Search

The frequency dependent capacitance-voltage (C-V) and conductance-voltage (G/ω-V) characteristics of Schottky barrier diodes (SBDs) with Cu contacts on Si doped GaAsN epilayers with (100) and (311)A/B orientations have been investigated in the frequency range from 20 kHz to 1 MHz at room temperature. C, G/ω and the deduced series resistance (Rs) show strong dependences on the applied frequency in the forward bias region, which is closely correlated to the frequency-dependent response of interface states (Nss). In GaAsN SBDs with all three growth orientations, the increasing N composition is found to increase the peak value of capacitance and enhance its dependence on frequency, which thus implies a general rule that increasing N incorporation causes an increase in Nss. The increasing extent of Nss due to N incorporation, however, differs a lot for different growth orientations as analyzed by using Hill-Coleman method. It is revealed that (311)B is the promising growth orientation to suppress the Nss generation over a wider N composition range in GaAsN Schottky devices. The reduced formation probability of non-substitutional N due to the efficient N incorporation on the (311)B plane is considered to be responsible for the observations.

In this paper, we propose a new method of using H2 supply in the atmosphere to increase Sn concentration in a Ge1−xSnx film epitaxially grown on a Ge substrate using MOCVD (metal organic chemical vapor deposition). H2 supplied in the atmosphere accelerates decomposition of Sn precursors and suppresses surface migration of Sn atoms during epitaxial growth of a Ge1−xSnx film. The proposed method is new and fundamentally different from the existing methods that increase Sn concentration through either crystallizing α-Ge1−xSnx or lowering growth temperature. The proposed method uses H2 supply in the atmosphere to increase Sn concentration. In order to show the effectiveness of the proposed method, we conducted experiments with varying ratios of supplying H2 in the atmosphere and epitaxially grew a Ge1−xSnx film on a Ge substrate using MOCVD. MO precursors that we used in our experiments (tertiarybutylgermane and tetraethyl tin) are new and safe. In our experiments, we observed that Sn concentration increased with H2 supply during growth, while maintaining a high growth rate of a Ge1−xSnx film.

We evaluated the ITO/p-type a-Si interface for Si heterojunction solar cells using XPS and TEM. It was found that ITO/a-Si interface which is 20 nm from the surface is oxidized by using non-destructive and non-contact HAXPES. The ITO/a-Si interface was oxidized during ITO film deposition by RPD technique, which leads to an increase in contact resistance. The chemical bonding states at the ITO/a-Si interface was modified by post deposition annealing (PDA). The TEM and the EDX mapping images revealed the interdiffusion of Si and Sn resulting in the interface roughness enhancement and the possible In precipitation in the a-Si layer.

Double layers of In 2 Se 3 and GaAs were grown on Si(111) 4o off substrate by molecular beam epitaxy (MBE). After the MBE growth, the GaAs layer transfer onto flexible PMMA sheet was attempted by the mechanical cleavage by hand or by the thermal stress induced by liquid nitrogen dipping. The GaAs layer was successfully transferred onto PMMA sheet from 2-inch Si substrate by the both method. The 2-inch Si substrate remained without any damage after the thermal stress peeling, though Si substrate was broken after the peeling by hand. The applying homogeneous stress between the over layer and the substrate is necessary. After the peeling, In 2 Se 3 layer remained on Si substrate, indicating that the GaAs layer was peeled via the van der Waals interface in the layered In 2 Se 3 . This epitaxial liftoff (ELO) technique enables the dramatically cost reduction of thin film flexible III-V solar cells than the usual selective etching technique by aqueous acid.

Komatsu NTC developed ultra-thin wafer slicing with low kerf-loss by multi diamond-wire saw. We modified slicing conditions and diamond-wire specifications to keep the straightness of wire for the fine pitch slicing and established 150 μm pitch slicing technology. The as-cut wafer thickness is 90 μm, and the kerf-loss is 60 μm. As a result, the highly flexible ultrathin wafer can be obtained with well-controlled surface crystallinity. The PERT cell fabrication trial using ultra-thin wafers showed that the relative deference of cell efficiency between 166 μm thick and 84 μm thick solar cells was just only 1.2%.

In this study, the crystalline improvement of GaAs grown on metal (Ga or In) selenides films is studied. We have found that the inevitable bonding between Ga and Se on the top of In 2 Se 3 was observed. Further, the wettability of GaAs nucleus has been improved by irradiating Ga 2 Se 3 surface with As, Ga, and Se simultaneously. The twin-domain less than 2% for 1 μmGaAs grown on an In 2 Se 3 /Si (111)4° off-cut to [11-12] has been achieved by introducing Ga 2 Se 3 on In 2 Se 3 with the subsequent (As-Ga-Se) treatment.

We have investigated effects of different tunnel oxide layers on the performance and light stability of tunnel oxide/TiO 2 stack electron-selective contacts on n-type Si. From XPS analysis, we found that chemical properties of a SiO x interlayer at an interface between a tunnel oxide/TiO 2 stack electron-selective contact and an n-type Si significantly depend on stack materials and fabrication processes of tunnel oxide layers. We also found that although chemical SiO 2 and thermal atomic layer deposited AlOx tunnel layers show higher passivation performance, plasma-enhanced atomic layer deposited AlO x tunnel layers demonstrate light-enhanced τ eff with an ITO capping layer as well as the highest light stability without an ITO capping layer.

In this paper, we report on the fabrication of Ta-doped TiO2 (TiO2:Ta) electron-selective contacts and their improved passivation qualities. By using thermal atomic layer deposition, 3.5 nm-thick, undoped TiO2 films and 3.5 nm-thick TiO2:Ta films with Ta concentrations of 1.5, 2.8, and 5.5 at.% were formed at 150 °C on n-type crystalline silicon substrates with ∼1 nm-thick chemical SiO2 layers. Rutherford backscattering spectroscopic measurements confirmed the Ta concentrations in the films and uniform distributions of Ta atoms in the films. X-ray photoelectron spectrometry measurements implied that Ta atoms rather cause local structural changes than enter substitutionally into Ti sites. After annealing at 130 °C in air, the effective minority carrier lifetime (τ eff) of the TiO2:Ta-passivated samples were significantly higher than those of the samples passivated with undoped TiO2 films. The highest τ eff was obtained for the samples with a relatively low Ta-doping concentration of 1.5 at.%. On the other hand, the passivation quality was degraded by Ta doping with a too high concentration (for example, 18.4 at.%). These results indicate that moderate Ta doping improves the passivation quality of TiO2 thin films.

Si tandem solar cells are attractive for new applications such as photovoltaic-powered vehicles because of their high-efficiency and low-cost potential. In particular, III-V/Si tandem solar cells have higher efficiency potential compared to perovskite/Si and other Si tandem solar cells. Although the direct growth of III-V layers on Si is very attractive for cost reduction and simple processing potential, high-quality growth of the III-V thin-film layer on Si is necessary. The paper discusses the effectiveness of the low-temperature growth of III-V layer on Si substrates for realizing low-density dislocations on Si substrates. Low dislocation density of less than 3 × 105 cm−2 in GaAs-on-Si by low-temperature growth is demonstrated in this study. According to our analytical results, this low dislocation density shows high potential efficiency of more than 33% and 38% for III-V/Si 2-junction and 3-junction tandem solar cells, respectively.

The theoretical lateral current of the surface inversion layer in a crystalline silicon (cSi) surface for a p-aSi:H/i-aSi:H/cSi heterojunction (SHJ) solar cell was calculated using computer simulation and was compared with the experimental one to study defects/traps at the aSi:H/cSi interface and/or in the cSi surface and to detect the acceptor concentration (N a) in p-aSi:H. To experimentally extract the lateral surface inversion layer current, a field-effect transistor type test element group device was co-integrated with SHJ cells on the same wafer. From the correlation between the experimental and calculated lateral surface inversion layer current, the density of defects/traps (D it) at the aSi:H/cSi interface and/or in the cSi surface and the value of N a were extracted. The calculated lateral surface inversion layer current stayed unchanged for various minority carrier lifetimes in the substrate, suggesting that this method is not suffered from the variation in the material parameters in the substrate.

We demonstrate that the passivation quality of TiO2 electron-selective contacts can be improved by Ta doping. By using atomic layer deposition, 3.5-nm-thick, undoped TiO2 films and Ta-doped TiO2 (TiO2:Ta) films with Ta concentrations of 1.5, 2.8, and 5.5 at.% were formed on n-type crystalline silicon substrates with chemical SiO2 layers. After annealing at 130 °C, the effective minority-carrier lifetime (τeff) of the TiO2:Ta-passivated samples were significantly higher than that of the undoped-TiO2-passivated samples, and the highest τeff was obtained at a relatively low Ta-doping concentration of 1.5 at.%. These indicate that moderate Ta doping improves the passivation quality of TiO2 thin films.

The efficiency potential of chalcopyrite and kesterite solar cells including CIGSe (CuInGaSe2), CIGS (CuInGaS2), CZTS (Cu2ZnSnS4) and CZTSSe [Cu2ZnSn(S,Se )4] solar cells is discussed based on external radiative efficiency (ERE), open-circuit voltage loss, fill factor loss, non-radiative recombination and resistance loss. CIGSe cells achieve efficiency potential of 26.8% and 27.5% by improving the ERE from around 1% to 10% and 20%, respectively. CIGS and CZTS(Se) cells achieve the efficiency potential of 25% and 22%, respectively, by improvement in ERE from around 1 × 10−4% to 3%–5%. The effects of non-radiative recombination and resistance loss upon the properties of wide-bandgap CIGSe, CIGS and CZTS(Se) cells are discussed. In the case of wide-bandgap CIGSe cells, lattice mismatching between the buffer layer and CIGSe active layer and deep-level defects are thought to originate from non-radiative recombination loss. CIGS and CZTS(Se) cells are shown to have lower ERE and higher resistance loss compared to that of CIGSe cells.

The growth orientation dependence of N incorporation on traditional (100) and high-index GaAs (311)A/B planes has been investigated by the first principle calculation and experimental measurement. Due to the electronegativity and atomic radius differences between As and N atoms, the Ga N bond is much shorter than the corresponding Ga As bond as N occupies As atom site. The optimization to energetically preferred construction leads to the inward moving of N atom, while the displacement extent of three surface models exhibits considerable differences. More importantly, the theoretical result reveals a lower formation energy of N doping for N@(311)B GaAs surface and a larger one for N@(311)A GaAs surface as compared with the traditional (100) GaAs plane. The first principle investigations thus suggest the enhanced N incorporation in (311)B GaAsN, which is in good accordance with the secondary ion mass spectrometry (SIMS) measurement results of GaAsN layers by chemical beam epitaxy (CBE).

The paper presents a quantitative approach to the investigation and comparison of the material qualities of III-V on silicon (III-V/Si) solar cells by using external radiative efficiencies. We use this analysis to predict the limiting efficiencies and evaluate the criteria of material quality in order to achieve high efficiency III-V/Si solar cells. This result yields several implications for the design of high efficiency III-V/Si solar cells.

In this paper, the Chemical Beam Epitaxy approach to GaAsN material growth has been investigated. Photoluminescence and Hall effect measurements were performed to clarify the influence of defects on the transport properties, those of holes in particular. The PL intensity of near-band emission (1.35 eV-peak) for GaAs0.996N0.004 at low temperature ( 1) indicated that the N atoms were alloyed in a disordered arrangement. As a result, the corresponding alloy scattering mechanism was enhanced, contributing to the hole mobility decrease. However, compared to p-type GaAsN grown by MBE or MOCVD, our CBE-grown GaAsN showed a higher value of hole mobility.

The contact behavior of Cu on n-type GaAsN, grown on (100) and (311)A/B GaAs substrates by chemical beam epitaxy, has been investigated by current–voltage (I–V) and capacitance–voltage (C–V) measurements. Both N incorporation and growth orientation are found to influence the electrical properties of Cu/GaAsN Schottky diodes. The increasing N composition leads to an increase in the ideality factor and Schottky barrier height (C–V), whereas a decrease in the carrier concentration in samples with all adopted growth orientations. Compared with (100) and (311)B samples, the Schottky diodes constructed on (311)A GaAsN epilayer exhibit better performance with ideality factor close to 1 and higher barrier height. The interfacial chemical reaction between metal layer and surface atoms on the growth surface is believed to account for the observations.

Potential-induced degradation (PID)-test results of modules fabricated from the rear- and front-emitter silicon heterojunction (SHJ) solar cells were compared to clarify the influence of the emitter position of SHJ photovoltaic (PV) cell modules on their PID behaviors. The PID tests were performed by applying a bias of −2000 V to the shorted interconnector ribbons from the front surface of the cover glass, at 85 °C. In the initial stage, both modules showed the same degradation characterized by a reduction in the short-circuit current density (J_). After the firststage degradation, the rear-emitter SHJ PV modules exhibited subsequent degradation characterized by a significant reduction in the J_ and open-circuit voltage (V_), due to the enhancement of the minority-carrier recombination in the front surface region of the n-type crystalline silicon base. The front-emitter SHJ PV modules, on the other hand, showed a reduction in the fill factor (FF), in addition to moderate reductions in J_ and V_. The FF reduction of the front-emitter SHJ PV modules is considered to be caused by the enhancement of the recombination in the front surface region of the n-type crystalline-silicon base as the region corresponds to the pn junction interface of the front-emitter configuration. The moderate reductions in both J_ and V_ may be due to further progression of the first-stage degradation. These findings are essential for understanding the mechanism of PID in SHJ PV cell modules.

This research investigates the cause of lifetime reduction properties of a crystalline defect layer introduced by the plasma process such as reactive plasma deposition (RPD). The plasma irradiation damage to silicon substrate with the different oxygen and carbon concentrations were evaluated. Minority carrier lifetime of the silicon substrate after the RPD process has been significantly reduced by plasma irradiation. Furthermore, photoluminescence (PL) spectroscopy revealed that the cause of the lifetime degradation on the silicon substrate is Ci–Oi defect generation originated in the plasma irradiation during the RPD process.

GaAs were grown by the migration-enhanced epitaxy (MEE) method on vicinal (001) Si substrates oriented 4 ° off towards [110] at low temperature. Investigations of crystalline and optical properties of the grown GaAs/Si were carried out by the methods of X-ray diffraction (XRD) and low-temperature photoluminescence (PL). The possible cause for tilting in the GaAs grown at low-temperature is discussed.

We have investigated the effects of chemical rounding on the performance of bifacial n-PERT Si cells with a random-pyramid textured p+-emitter. From experimental results, we found that solar-weighted reflectanceR SW of pyramid textured p+-emitters passivated with an ${AlO}_{x}/{SiN}_{x}$:H stack increases and short circuit current densityJ sc of bifacial n-PERT Si cells with an ${AlO}_{x}/{SiN}_{x}$:H stack-passivated textured p+-emitter consequently decreases as a function of chemical rounding time. However, the loss inJ sc is fully offset by the increased fill factor FF, and the efficiency of the 200-sec-rounded cells is consequently enhanced, compared to the non-rounded cells. In addition, the deposition of an indium tin oxide layer on the non-rounded and rounded $p^{+}-$ emitters passivated with an ${AlO}_{x}/{SiN}_{x}$:H stack can reduce reflectance largely in the near-infrared region. Based on the experimental results, although we would need further studies, it could be expected that cell efficiency of a bifacial ${n}$-PERT Si bottom cell and its monolithic tandem cell with a perovskite top cell would be improved without additional complicated and costly processes if chemical rounding and boron doping processes are properly optimized.

We investigated the damage using hard X-ray photoemission spectroscopy (HAXPES), and found that the a-Si surface was oxidized due to the energetic oxygen ambient during the deposition. For the conventional and advanced sputter, the SiO peak intensity at the ITO/a-Si interface did not change by post deposition annealing, but for RPD the peak intensity decreased. From the O 1s and In 3d spectra, it is considered the oxygen at the ITO/a-Si interface moved to the oxygen deficient in ITO. We also verified the damaged oxide was partially recovered by the post deposition annealing process.

Lifetime degradation caused by oxygen precipitates combined with metal contamination in Cz-Si for solar cells was evaluated for the samples with different interstitial oxygen concentrations using PL imaging and IR–LST. It was confirmed that phosphorus diffusion gettering is effective for the samples with low oxygen concentration, however the lifetime never recovered for the samples with high oxygen concentration. The enhanced oxygen precipitation process made it more difficult to recover from the metallic contamination. We believe the high density and large size oxygen precipitates are harmful for the lifetime in combination with metal impurities.

Effects of an ITO capping layer on the interface workfuncion of MoO x was studied for an ITO/MoO x /SiO2/Si structure device from the viewpoint of the behavior of In atoms. The workfunction of MoO x at the MoO x /SiO 2 interface, which were extracted by capacitance-voltage (C-V) analysis, was largely different from the bulk MoO x film in its value and post-deposition forming gas annealing (PDFGA) temperature dependency. The interface workfunctoin decreased by 0.13 and 0.35 eV by 200 and 400 °C PDFGA, respectively, while bulk workfunction decreased by only 0.1 eV by 400 °C PDFGA. A secondary ion mass spectroscopy (SIMS) analysis showed that In atoms were penetrated from the ITO layer into the MoO x layer by PDFGA. The concentration of In near the MoO x /SiO 2 interface increased from ∼1019 cm−3 for the as-grown ITO sample to ∼1020 cm−3 for the sample with 400 °C PDFGA. The increase in the In concentration near the MoO x /Japan>SiO 2 interface is a possible cause to decrease the interface workfunction.

The properties of the acceptor states in GaAsN grown by chemical beam epitaxy (CBE) are studied by analyzing their charges based on the Poole–Frenkel model. Deep level transient spectroscopy (DLTS) shows two acceptor levels at 0.11 and 0.19 eV above the valence band maximum. The emission rates of carriers from these states are enhanced with increasing the electric field during the DLTS measurement, which indicates that the energies required for the emission are decreased. By analyzing this field-enhanced emission process, the polarizabilities of the levels at 0.11 and 0.19 eV are found to be −1 (±0.1) and −2 (±0.1), respectively. In addition, these states have almost the same concentration. Therefore, we conclude that they originate from the same defect, acting as a double acceptor in GaAsN film grown by CBE.

The N–H related acceptor defects in GaAsN grown by chemical beam epitaxy (CBE) are studied by hydrogen isotopes, H and D. When the films are grown by a conventional arsenic source, deep level transient spectroscopy (DLTS) reveals two energy levels at 0.11 and 0.19 eV above the valence band. These levels were considered to act as a double acceptor in the literature. When the films are grown by a deuterated arsenic source, new signals appear in DLTS spectra at 0.15 and 0.23 eV. This indicates that the new signals are originated from D-related defects. The energy differences between 0.15 and 0.11 eV, and that between 0.23 and 0.19 eV are same (0.04 eV). Although these energy levels become deeper with increasing the growth temperature, the energy differences are almost constant independent of the growth condition. In addition, the intensity ratios of the peaks at 0.15 (0.23) eV to that at 0.11 (0.19) eV have a good correlation with the isotopic concentration ratio of D to H in the grown films. Therefore, we conclude that the energy differences and intensity ratios of the DLTS peaks occur due to the structural change from N–H to N−D in the same type of defect, and that this acceptor is an N–H related defect.

Combining Si PV module technology with an efficient III- V CPVs top PV into a tandem device is an attractive approach to enhance the efficiency of PV modules. Here, a method called III- V/Si hybrid configurations for enhancing the Si module efficiency is placing III- V static CPVs at the edge gap of each Si wafer rather than stacking in the top of Si. The prototype PV module shows the enhancement of efficiency. The cost evaluations of the III- V lSi hybrid configurations compared with Si module is discussed under various presumed cost ratio of 111-V/Si PVs.

The behavior of oxygen precipitates under solar cell fabrication processes and the effect on device performance were investigated using TEM observation. Samples were prepared with different carbon concentration and with two sets of growth conditions. The number of precipitates correlates monotonically with the carbon concentration. When the initial carbon concentration is high, the cell efficiency is improved by decreasing the carbon concentration. When the initial carbon concentration is reduced to smaller than 1016 cm−3, the oxygen precipitate grows largely and introduces dislocation. Precipitates grown in a plate form introduce dislocations to the surroundings at a high density, while dislocation density is relatively small around the precipitates polyhedral-grown on the basis of an octahedron. Under the growth conditions for introducing a plate-like precipitate, the cell efficiency was not improved even if the carbon concentration was reduced.

When our novel Ag paste is used, the high solar cell performance is obtained. A thick SiON layer is formed between Ag line and Si substrate after the firing process. To understand the effect of this layer on the electrical properties, the current-voltage analysis of the contact is carried out. The obtained results indicate that there might be locally contacted areas between Ag and Si, and these regions are distributed in laterally, which determined the carrier transportation at the contact. This SiON layer with local contacts prevents the Ag diffusion into Si crystal during the firing process, and decreases the minority carrier recombination, resulting in the relative high performances.

The high quality and thin Si wafer technology for the future higher conversion efficiency and lower cost crystalline silicon solar cells are realized. The high minority carrier lifetimes even after the processes are obtained by controlling the Czochralski growth condition, which prevents the interstitial oxygen segregation enhanced by the substitutional carbon. The Cz ingot is sliced by the advanced diamond multi-wire saw technology and the thin wafers with relatively thin damaged layer and 100μm kerf-loss are realized. The thin wafer with low kerf-loss decreases the wafer cost and improve the cell performance. The thin wafer solar cells are fabricated using the PERC processes, and these technologies are evaluated from the view point of photovoltaic.

We applied aluminum-induced crystallization (AIC) as a technique to realize heavily doped p-type polycrystalline Si layer at low temperature for the rear side of tunnel oxide passivated contact concept solar cell. Systematic temperature variation of AIC revealed that annealing at a temperature slightly lower than the eutectic temperature (577°C) is effective to improve solar cell performance. This is explained by reduced resistivity for low contact resistance while avoiding meltinz of AI-Si.

By stacking a III - V cell to the Si cell, it is expected to reach to more than 30 % of efficiency. However, the contribution of the Si bottom cell may be reduced to 5 to 6 % in maximum. Another problem is that the III-V cell cost is still expensive and high concentration without relying on tracking is expected. For solving these two problems, a new configuration that places small III-V multi-junction solar cells at four corners of a single-crystalline Si solar cell was invented.

Adding SiCl4 to the atmosphere during the growth of a crystal suppressed crystalline defects and impurities from the edge of silicon ingots. Crystalline silicon ingots with seed crystal were grown by using the unidirectional solidification technique. Most of the grain boundaries were inactive Σ3 in the ingot with the SiCl4 injection. There were no small-angle grain boundaries at the top and edge of ingots. The carbon concentration with SiCl4 was decreased to less than half of that without the SiCl4 injection. Carbon precipitation did not occur even at the surface of ingots and the surface of grown ingots had a metallic luster.

A reactive-plasma deposition (RPD) process damage is evaluated for an indium-tin-oxide (ITO)/SiO2/Si structure in terms of the recombination properties at the interface. To calculate the surface recombination velocity (SRV), the defect density at the SiO2/Si interface (Dit), the effective total charge (Qtot) which is the sum of the charges in the SiO2 and the trapped charge at the SiO2/Si interface, and the workfunction of ITO at the ITO/SiO2 interface (ϕITO), are extracted by capacitance-voltage (C-V) analysis as a function of a postdeposition annealing (PDA) temperature. The Dit and Qtot significantly increase by the RPD process to 1.4×1012 cm-2eV-1 and 1.5×1012/q cm-2, respectively, and decrease by the PDA at 200 °C in N2 ambient to 1.4×1011 cm-2eV-1 and 1.1×1011/q cm-2, respectively. The ϕITO also varies from 4.80 eV to 4.35 eV by the PDA. The correlation is examined between the SRV values calculated based on the extended SRH model using the extracted Dit, Qtot, and ϕITO values [SRV(CV)], and the SRV ones obtained from the minority carrier lifetime measurements [SRV(LT)]. It is found that multiple sets of capture cross-sections were required for SRV(CV) values to coincide with SRV(LT) ones. The effect of variation in the Qtot and ϕITO on the band bending is almost canceled out in the present PDA condition. It is thus considered that the increase in Dit is the main cause to decrease the minority carrier lifetime.

In situ three-dimensional X-ray reciprocal space mapping ( in situ 3D-RSM) was employed for studying molecular beam epitaxial (MBE) growth of InGaAs multilayer structures on GaAs(0 0 1). Measuring the symmetric 004 diffraction allowed us to separately obtain film properties of individual layers and to track the real-time evolution of both residual strain and lattice tilting. In two-layer growth of InGaAs, significant plastic relaxation was observed during the upper layer growth, and its critical thickness was experimentally determined. At the same thickness, it was found that the direction of lattice tilting drastically changed. We discuss these features based on the Dunstan model and confirm that strain relaxation in the multilayer structure is induced by two kinds of dislocation motion (dislocation multiplication and the generation of dislocation half-loops).

Defect characterization in molecular beam epitaxial (MBE) compositionally-graded InxGa1-xAs layers on GaAs substrates consisting different thickness of overshooting (OS) layers was carried out using cathodoluminescence (CL) and transmission electron microscopy (TEM). We found that the thickness of the OS layer influences not only stress but also lattice defects generated in a top InGaAs layer. While the top InGaAs layer with a thin OS layer is under compression and has mainly threading dislocations, the top layer with a thick OS layer is under tension and exhibits inhomogeneous strain associating with phase separation. We will discuss the mechanisms of defect generation and their in-plane distribution based on strain relaxation at the top and OS layers. (© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

The two N-H local vibrations in GaAsN grown by the chemical beam epitaxy have the same preferential orientation in the crystal. The bond orientations of the N-H wagging mode at 961 cm-1 and stretching mode at 2952 cm-1 tend to align along [1-10]. The peak intensities of the three major N-H local vibration modes were examined by the polarized Fourier transform infrared spectroscopy. The integrated absorption peaks at 961and 2952 cm-1 showed the two-fold rotational symmetry in (001) plane, while the one at 3098 cm-1 was almost constant within the accuracy of this measurement. The maximum peak intensities were observed when the IR polarization direction was [110] for the N-H wagging mode at 961 cm-1, and [1-10] for the N-H stretching mode at 2952 cm-1, which means that those bond orientations projected on (001) plane were the same. The result indicates that the N-H wagging mode at 961 cm-1 and stretching mode at 2952 cm-1 are originated from the same N-H bond. (© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

To get the optimized condition and ideal furnace structure, we have performed seed cast growth of mono-crystalline Si by using unidirectional solidification furnace. More than 20 ingots of 10 cm diameter and 10 cm height were grown under different growth conditions. The quality of ingots was characterized by using Fourier transform infrared spectroscopy (FTIR), infrared microscopy, scanning infrared polariscope (SIRP), X-ray topography, etc. We have realized reduction of carbon, residual strain and extended defects, which may contribute the increase of solar cell efficiency.

The dependence of solar cell parameters on i-aSi:H (non- or lightly-doped hydrogenated amorphous silicon) layer thickness in an aSi:H/cSi (crystalline silicon) heterojunction solar cell was analyzed using numerical simulation. By considering the quantum confinement effect at interfaces between i-aSi:H and cSi, experimental data which had not been explained by simulation could be successfully interpreted. The mechanism of an open-circuit voltage increase was visually presented by analyzing carrier distributions and quasi-Fermi levels near cSi surfaces and in the i-aSi:H layers. The optimized thicknesses of the i-aSi:H layers in both front and rear junctions were suggested to obtain the maximum conversion efficiency. The influences of the quantum confinement effect on the simulation results were discussed.

The effect of the tridimethylaminoarsenic (TDMAAs) source flow rate on the electronic signature as well as on the trapping density of electron traps in undoped n-type GaAsN grown by chemical beam epitaxy was investigated by deep level transient spectroscopy and isothermal capacitance transient spectroscopy methods. It was found that the densities of four electron traps, E1 to E4, localized at average energy depths of 0.05, 0.11, 0.16, and 0.39 eV, respectively below the bottom edge of the conduction band, significantly decrease with increasing TDMAAs up to 1.2 sccm. Nevertheless, the decrease of trapping density for these energy states showed quite different behaviors. This result implies that the excess of As atoms in the alloy could limit the formation of these electron traps. Furthermore, it indicates that the degradation of the optoelectronic properties of the alloy is not only caused by the addition of N atom to the lattice of GaAs. An interesting feature of this result is the possibility of recovering the lifetime of minority carriers in p-type GaAsN material, and, therefore, enhancing the conversion efficiency of GaAsN based solar cells. (© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

An approach of creating CMOS channel with GeSn is one of the post-scaling techniques and is widely considered to be promising to improve LSI performance. In this paper, we selected the metal-organic (MO) precursors to deposit GeSn and investigated their characteristics. We confirmed that it is possible to deposit GeSn on a substrate using the MOCVD technique.

Current transport mechanisms were investigated in a GaAsN homojunction solar cell (HJSC) grown by chemical beam epitaxy. At each temperature of measurement, the current-voltage (I–V) characteristics were found to deviate from the diffusion model. The fitting of these characteristics affirmed that the recombination current in the space charge region of the HJSC is mainly attributed to a localized state at 0.31 eV below the bottom edge of the conduction band of GaAsN. This recombination center was identified by deep level transient spectroscopy. It was previously confirmed to act as a nitrogen-related non-radiative recombination center and the split interstitial formed from one nitrogen atom and one arsenic atom in a single V-site (N-As) As was tentatively suggested to be its possible origin. The lifetime of electrons from the conduction band to its energy level was calculated to be around ∼0.20 ns, using the Shockley-Read-Hall model for trap-assisted recombination process. Furthermore, such order of magnitude was confirmed by time-resolved photoluminescence measurements.