Micro-Raman mapping on layers for crystalline silicon thin-film solar cells

Micro-Raman mappings have been used for characterization of our layers system developed for thin-film silicon solar cells. For the cubic SiC barrier layer a preferential orientation of the grains in 〈1 1 1〉 direction normal to the substrate was revealed. A high density of stacking faults resulted in the splitting of transversal optical (TO)-phonon modes, usually only observed in several non-cubic SiC polytypes. Within the silicon layers, which were obtained by zone melting recrystallization (ZMR) and subsequent epitaxial growth, a high residual stress of about 625 MPa was measured near the boundary towards the SiC layer. Outside of this boundary no residual stress could be detected, in spite of commonly found twin boundaries. Thus the main origin of residual stress in the silicon layers is due to the different expansion coefficients of the respective layers, while grain boundaries have no dominant effect.     

Preparation of microcrystalline silicon solar cells on microcrystalline silicon carbide window layers grown with HWCVD at low temperature

N-type microcrystalline silicon carbide layers prepared by hot-wire chemical vapor deposition were used as window layers for microcrystalline silicon n–i–p solar cells. The microcrystalline silicon intrinsic and p-layers of the solar cells were prepared with plasma-enhanced chemical vapor deposition at a very high frequency. Amorphous silicon incubation layers were observed at the initial stages of the growth of the microcrystalline silicon intrinsic layer under conditions close to the transition from microcrystalline to amorphous silicon growth. ‘Seed layers’ were developed to improve the nucleation and growth of microcrystalline silicon on the microcrystalline silicon carbide layers. Raman scattering measurement demonstrates that an incorporation of a ‘seed layer’ can drastically increase the crystalline volume fraction of the total absorber layer. Accordingly, the solar cell performance is improved. The correlation between the cell performance and the structural property of the absorber layer is discussed. By optimizing the deposition process, a high short-circuit current density of 26.7 mA/cm² was achieved with an absorber layer thickness of 1 μm, which led to a cell efficiency of 9.2%.     

Toward the fast deposition of highly crystallized microcrystalline silicon films with low defect density for Si thin-film solar cells

In this study, employing a high-density, low-temperature SiH₄–H₂ mixture microwave plasma, we investigate the influence of source gas supply configuration on deposition rate and structural properties of microcrystalline silicon (μc-Si) films, and demonstrate the plasma parameters for fast deposition of highly crystallized μc-Si films with low defect density. A fast deposition rate of 65 Å/s has been achieved for a SiH₄ concentration of 67% diluted in H₂ with a high Raman crystallinity of X_c > 65% and a low defect density of (1–2) × 10¹⁶ cm⁻³ by adjusting source gas supply configuration and plasma conditions. A sufficient supply of deposition precursors, such as SiH₃, as well as atomic hydrogen H on film growing surface is effective for the high-rate synthesis of highly crystallized μc-Si films, for the reduction in defect density, and for the improvement in film homogeneity and compactability. A preliminary result of p–i–n structure μc-Si thin-film solar cells using the resulting μc-Si films as an intrinsic absorption layer is presented.     

Electron and hole transport in microcrystalline silicon solar cells studied by time-of-flight photocurrent spectroscopy

A photocurrent time-of-flight study of carrier transport in microcrystalline silicon pin diodes prepared over a range of crystallinities is presented. Electron and hole drift mobilities at a crystalline volume fraction >0.35 are typically 3.8 and 1.3 cm²/(V s) respectively at 300 K and a thickness to electric field ratio of 1.8 × 10⁻⁷ cm²/V. A factor of five enhancement in hole mobility over amorphous silicon persists at a crystalline volume fraction as low as 0.1. Current decays are dispersive and mobilities are thermally activated, although detailed field-dependence is still under investigation. Evidence for a sharp fall in the density of states at 0.13 eV above the valence band edge is presented. Similarities in behaviour with certain amorphous and polymorphous silicon samples are identified.     

Microstructures of high-growth-rate (up to 8.3 nm/s) microcrystalline silicon photovoltaic layers and their influence on the photovoltaic performance of thin-film solar cells

Microstructures of microcrystalline silicon (μc-Si) deposited at a high-growth-rate have been investigated in order to apply to the photovoltaic i-layer. μc-Si films were prepared by very-high-frequency (100 MHz) plasma-enhanced chemical vapor deposition at 180 °C. High growth rates of 3.3–8.3 nm/s have been achieved utilizing high deposition pressures up to 24 Torr and large input powers. Applying μc-Si to n–i–p junction solar cells, as the optimum result in this experimental series, a conversion efficiency of 6.30% (J_SC: 22.1 mA/cm², V_OC: 0.470 V, and FF: 60.7%) has been achieved employing the i-layer deposited at 8.1 nm/s. Raman scattering and X-ray diffraction measurements revealed the crystalline volume fraction of around 50% with the (2 2 0) crystallographic preferential orientation, respectively. The cross-sectional transmission electron microscope image shows densely columnar structure grown directly on the underlying n-layer. These structural features are basically in good agreement those of low-growth-rate μc-Si used for a high efficiency solar cell as previously reported, implying advantages of the use of high pressures with regard to providing the photovoltaic i-layers. Finally, the implication is discussed from the photovoltaic performance as a function of the crystalline volume fraction of i-layer, and current problems in improving the photovoltaic performance are extracted.     

Ellipsometry characterization of a-Si:H layers for thin-film solar cells

In order to determine microscopic structures of hydrogenated amorphous silicon (a-Si:H) layers incorporated in a-Si:H-based thin-film solar cells, the spectroscopic ellipsometry (SE) analysis of a-Si:H layers prepared by plasma-enhanced chemical vapor deposition has been performed. In particular, we have characterized the a-Si:H layers by applying a new dielectric function model that allows the evaluation of the SiH₂ bond densities in a-Si:H networks. This model is based on our finding that the a-Si:H dielectric functions in the visible/ultraviolet region vary systematically with the formation of SiH₂-clustered microvoids. We have applied this model to estimate the SiH₂ content in a-Si:H layers fabricated on glass substrates, on which the characterization of the SiH₂ bonding is generally difficult. The validity of the SE analysis has been confirmed from the direct characterization of the SiH_n local structures using infrared ellipsometry.     

Crystallinity of the mixed phase silicon thin films by Raman spectroscopy

Raman spectra of the mixed phase silicon films were studied for a sample with transition from amorphous to fully microcrystalline structure using four excitation wavelengths (325, 514.5, 632.8 and 785 nm). Factor analysis showed the presence of two and only two spectrally independent components in the spectra within the range from 250 to 750 cm^?1 for all four excitation wavelengths. The 785 nm excitation was found optimal for crystallinity evaluation and by comparison with surface crystallinity obtained by atomic force microscopy, we have estimated the ratio of integrated Raman cross-sections of microcrystalline and amorphous silicon at this wavelength as y = 0.88 ± 0.05.     

Design and application of dielectric distributed Bragg back reflector in thin-film silicon solar cells

A dielectric distributed Bragg reflector (DBR) formed by four pairs of hydrogenated amorphous silicon/silicon nitride layers is used as the back reflector in thin-film silicon solar cells. The DBR was designed to perform in a broad wavelength range with the peak reflectance at 600 nm. The DBR was fabricated at low substrate temperature (172 °C) and applied at the rear side of flat and textured amorphous silicon single-junction solar cells in both superstrate (pin) and substrate (nip) configurations. The spectral response and electrical I–V characteristics were measured. Solar cells with optimized DBR exhibit an enhanced external quantum efficiency in the long wavelength range and the electrical performance is comparable to solar cells having conventional Ag back reflector.     

Comparison of photocurrent spectra measured by FTPS and CPM for amorphous silicon layers and solar cells

Fourier Transform Photocurrent Spectroscopy (FTPS) has been recently introduced as a fast and highly sensitive method for the evaluation of the optical absorption coefficient of photoconductive thin films such as microcrystalline silicon layers. This contribution represents the first study of FTPS utilization for amorphous silicon layers and cells. FTPS spectra are compared with results of Constant Photocurrent Method (CPM) and Dual Beam Photoconductivity (DBP) measured at different chopping frequencies. We will concentrate to highlight the appropriate measuring conditions and evaluation procedures for correct data interpretation. Moreover, we will present our novel approach for the interference free determination of absorption coefficients of thin films grown on transparent substrates, which is mainly important for very thin layers where broad interference fringes do not allow correct evaluation of parameters such as a slope of the Urbach tail and the defect density.     

Raman spectroscopy study of the structure of gallium nitride epitaxial layers of different orientations

The composition nonstoichiometry and structural quality of undoped gallium nitride layers grown by the hydride vapor phase epitaxy on sapphire substrates of different orientations have been estimated using Raman spectroscopy. It is found for the first time that the peak position of the phonon mode E₂(high) in the Raman spectra of gallium nitride films at a wave vector of 572 cm^–1 depends on the initial orientation of sapphire substrate and is low-frequency shifted when passing from Ga-polar to partially N-polar orientation. Additional modes are found in the spectra of GaN layers grown on substrates with m and r orientations. It is shown that a decrease in the composition deviation from stoichiometry, caused by reducing the HCl flow through the gallium source during the growth of GaN layers, leads to an increase in the phonon-mode intensity in the Raman spectrum.     

Hole mobilities and the physics of amorphous silicon solar cells

The effects of low hole mobilities in the intrinsic layer of pin solar cells are illustrated using general computer modeling; in these models electron mobilities are assumed to be much larger than hole values. The models reveal that a low hole mobility can be the most important photocarrier transport parameter in determining the output power of the cell, and that the effects of recombination parameters are much weaker. Recent hole drift-mobility measurements in a-Si:H are compared. While hole drift mobilities in intrinsic a-Si:H are now up to tenfold larger than two decades ago, even with recent materials a-Si:H cells are low-mobility cells. Computer modeling of solar cells with parameters that are consistent with drift-mobility measurements give a good account for the published initial power output of cells from United Solar Ovonic Corp.; deep levels (dangling bonds) in the intrinsic layer were not included in this calculation. Light-soaking creates a sufficient density of dangling bonds to lower the power from cells below the mobility limit, but in contemporary a-Si:H solar cells degradation is not large. We discuss the speculation that light-soaking is ‘self-limiting’ in such cells.     

On the development of single and multijunction solar cells with hot-wire CVD deposited active layers

We present an overview of the scientific challenges and achievements during the development of thin film silicon based single and multijunction solar cells with hot-wire chemical vapor deposition (HWCVD) of the active silicon layers. The highlights discussed include the development of Ag/ZnO coatings with a proper roughness and morphology for optimal light trapping in single and multijunction thin film silicon solar cells, studies of the structural defects created by a rough substrate surface and their influence on the performance of nc-Si:H n–i–p single junction solar cells, and studies of the phase change during the growth of nc-Si:H by HWCVD and the use of a ‘reverse’ H₂ profiling technique to achieve nc-Si:H single junction n–i–p cells with high performance. Thus far, the best AM1.5 efficiency reached for n–i–p cells on stainless-steel with HWCVD i-layers is 8.6% for single junction nc-Si:H solar cells and 10.9% for triple junction solar cells. The opportunities for further improvement of cell efficiency are also discussed. We conclude that the uniqueness of HWCVD and of the i-layers deposited with this technique require some adjustments in the strategy for optimization of single or multijunction solar cells, such as using a reverse H₂ profiling technique for the deposition of nc-Si:H i-layers. However, the output performance of solar cells with HWCVD deposited i-layers is close to those with i-layers deposited by other techniques. The difference between the best nc-Si:H n–i–p cells obtained so far in our lab and the reported best n–i–p cells with PECVD i-layers can be mainly attributed to the differences in the rough substrates and to the use of rather thin i-layers.     

In-process monitoring and control of doping gas concentration during vapor phase silicon epitaxial growth by using a flame photometric detector

An in-process monitoring and control method of the doping gas concentration during epitaxial growth of Si was developed. A flame photometric detector (FPD) can be used as a monitor for the PH₃ and B₂H₆ dopant concentrations in the injected doping gases. A combination of this dopant monitor with an automatic control system of the silicon source (SiHCl₃) gas concentration using an infrared spectrophotometer as a monitor, makes possible an automatic in-process control of the concentrations of dopant and of silicon source gas supplied to the reactor. The present system provides an accurate and reproducible control of impurity concentrations in Si epitaxial layers. Good correlation between the monitored signal (or the doping gas concentration) and the impurity concentration incorporated into the growth layers was confirmed for PH₃ (n-type) and B₂H₆ (p-type) doping. For the B₂H₆ doping, a divergence from the linear relationship between the doping gas concentration and the impurity concentration in the layers was observed in the level of acceptor concentration below about 10¹⁵ atoms/cm³. The transient response of the present system was measured by growing epitaxial layers with increasing and decreasing step-changes in the dopant gas flow during continuous deposition of the layers. Some interesting, but complicated, transient responses of impurity concentration in the growth layer were observed. The responses are different between the PH₃ doping and the B₂H₆ doping, and also different between increasing and decreasing steps especially for the B₂H₆ doping.     

Improving the performance of amorphous and crystalline silicon heterojunction solar cells by monitoring surface passivation

The influence of thermal annealing on the crystalline silicon surface passivating properties of selected amorphous silicon containing layer stacks (including intrinsic and doped films), as well as the correlation with silicon heterojunction solar cell performance has been investigated. All samples have been isochronally annealed for 1 h in an N₂ ambient at temperatures between 150 °C and 300 °C in incremental steps of 15 °C. For intrinsic films and intrinsic/n-type stacks, an improvement in passivation quality is observed up to 255 °C and 270 °C, respectively, and a deterioration at higher temperatures. For intrinsic/n-type a-Si:H layer stacks, a maximum minority carrier lifetime of 13.3 ms at an injection level of 10¹⁵ cm^? 3 has been measured. In contrast, for intrinsic/p-type a-Si:H layer stacks, a deterioration in passivation is observed upon annealing over the whole temperature range. Comparing the lifetime values and trends for the different layer stacks to the performance of the corresponding cells, it is inferred that the intrinsic/p-layer stack is limiting device performance. Furthermore, thermal annealing of p-type layers should be avoided entirely. We therefore propose an adapted processing sequence, leading to a substantial improvement in efficiency to 16.7%, well above the efficiency of 15.8% obtained with the ‘standard’ processing sequence.     

Properties of light-induced degradation and the electronic properties of nanocrystalline silicon solar cells grown under functionally graded hydrogen dilutions

The electronic properties of hydrogenated nanocrystalline silicon (nc-Si:H) were studied using drive-level capacitance profiling (DLCP) to obtain defect density profiles as well as transient photocapacitance (TPC) and transient photocurrent (TPI) spectroscopies to study the spectra of defect related optical transitions. These measurements were performed on a series of n–i–p solar cell devices with intrinsic layer thickness of roughly 1 μm. The nc-Si:H intrinsic layers were deposited using RF or MVHF glow discharge with various hydrogen dilution profiles predominantly on specular stainless steel substrates (SS/n⁺/i nc-Si:H/p⁺/ITO), but also on textured back reflectors (SS/Ag/ZnO/n⁺/i nc-Si:H/p⁺/ITO) in some cases. Crystallite fractions were estimated using Raman spectroscopy. The electronic properties determined by our measurements could be correlated with variations in structural device parameters and with the degree of hydrogen dilution profiling during growth. We also found, depending on the growth conditions, that the devices exhibited markedly different behaviors after prolonged light exposure (100 h using light at 610 nm and 500 mW/cm² intensity). We discuss one specific microscopic mechanism that may be responsible for the light-induced changes that we have observed.     

Mass transport effect within the boundary layers in solution crystal growth: Holographic study of KTP and KDP crystal growth

Using holographic phase-contrast interferometric microphotography, we have carried out real-time investigations of the mass transport processes taking place during the high temperature solution growth of KTiOPO₄ (KTP) crystal and low-temperature solution growth of KH₂PO₄ (KDP) crystal. Our experiments demonstrate that a mere diffusion boundary layer is not existing. The mass transport process within the boundary layers is a result of the coupled effect of diffusion and convection actions, no matter whether it is high-temperature or low-temperature solution growth. Under free convection state, the influence of bulk supersaturation on the thickness of solute boundary layer exists in the two different regions. The solute concentration distribution within the layer is an exponential function of the position.     

Investigation of the growth processes from vapor phase of silicon carbide epitaxial layers: Thermodynamic analysis of the high temperature processes in the system Si‐C‐H‐Cl

In the present work are presented the results of the thermodynamic analysis of the interaction processes in the system Si‐C‐H‐Cl in the temperature interval 1000‐3000 K. The equilibrium pressures of the components in the system Si‐C‐H‐Cl with taking account the formation of the condensed phases C, Si and SiC have been determined. The optimal conditions giving the maximum yield of silicon carbide by pyrolysis of mixture of volatile compounds of carbon and silicon have been defined. The thermodynamic analysis of the examined system showed that the increasing of the content of hydrogen in the initial mixture allows to decrease the optimal temperature for obtaining of silicon carbide by the method of pyrolysis and essentially to increase its maximum possible yield. (© 2008 WILEY ‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Liquid phase epitaxial growth and characterization of thin In1−xGaxAsyP1–y layers lattice-matched to InP

Growth studies enabled the deposition of In_0.71Ga_0.29As_0.68P_0.32 single quantum well structures with InP or In_0.88Ga_0.12As_0.26P_0.74 confinement layers lattice-matched to (001) InP by liquid phase epitaxy (LPE). Well widths in the order of 50–100 Å have been achieved using a conventional step cooling technique. The physical characterization has demonstrated the capability of the employed method to produce multilayered heterostructures which display confined particle states; quantum mechanically induced blue-shifts of the low temperature PL-emission up to 125 meV were measured. A remarkable reduction of the FWHM values of the shifted PL peaks was attained by optimization of the growth conditions.     

Structural characterization of the glassy phase in majolica glazes by Raman spectroscopy: A comparison between Renaissance samples and replica processed at different temperatures

Tin-opacified lead glaze, prepared according to Renaissance recipes, has been fired at different temperature from 300 °C to 990 °C, and investigated by Raman scattering. A chemometric treatment and a systematic curve-fitting procedure have been applied in the range of 700–1250 cm⁻¹ in order to monitor quantitatively the structural changes of the silicate network that occurred with firing. The results obtained on model glazes are compared with Raman spectra collected on various Renaissance potteries. This method is suggested for non-invasive surface analysis of ancient glazes aimed at the characterization of processing techniques.