We measure a series of hot-wire (HW) amorphous silicon films grown with hydrogen contents cH varying between 0.5–17at.%. From constant photocurrent method (CPM) measurements and the steady-state photocarrier grating method (SSPG) we find good agreement with previous measurements on similar hot-wire films. Electron spin resonance measurements on the same samples, however, yield significantly higher spin densities than expected. A thickness series indicates a highly defective layer close to the substrate interface. We propose that this defective layer may be due to excessive out diffusion of hydrogen during growth at high temperatures, as seen by secondary ion mass spectroscopy. ESR measurements on light-degraded samples indicate an improved stabilityof samples with cH < 9at.%.

Computer modeling is used as a tool for determining current matching in hydrogenated amorphous silicon (a-Si:H) alloy tandem cells on textured substrates. The increasing complexity of a-Si:H based solar cells requires continuous extending and testing of the computer models which are used for their simulation. To take light scattering at the textured interfaces of the cell into account we developed a multi-rough-interface optical model GENPR02 which was used for calculating the absorption profiles in the solar cells. The results of a sensitivity study of the parameters of this optical model such as the scattering coefficients of the reflected and transmitted light and the dependence of scattered light on the in-going and out-going angle are presented. In order to simulate multi-junction solar cell as a complete device we implemented a novel model for tunnel/recombination junction (TRJ), which combines the trap assisted tunneling and enhanced carrier transport in the high field region of the TRJ.

The current matching conditions were determined both for a-Si:H and a-SiGe:H bottom cells, while the top cell was an a-Si:H cell. We investigated the influence of light scattering at the textured interfaces and of the thickness of the intrinsic layer of the bottom cell on the optimal ratio (i2/i1) between the thicknesses of the bottom (i2) and top (il) intrinsic layers in the current-matched cell. The results show that increasing amount of scattering at the textured interfaces leads to higher efficiencies and lower ratio (i2/i1) in the current-matched cell. The use of a-SiGe:H material in the bottom cell leads to higher efficiency and 3 to 4 times lower i2/i1 ratio than in case of a-Si:H/a-Si:H cells.

We have prepared hydrogenated macrocrystalline silicon germanium by plasma enhanced CVD of a mixture of silane and germane gas diluted with hydrogen. The growth conditions have been systematically controlled to obtain large (∼400Å) crystallites of silicon-germanium as observed using Raman scattering and x-ray diffraction. The dangling bond (germanium) density has been reduced to <5×1016 cm−3 at low substrate temperatures (∼150°C). The optical absorption spectra of the 50% Ge containing material is red-shifted compared to microcrystalline silicon, consistent with a reduction of the indirect optical gap to 0.9eV. Schottky type cells fabricated using Au on an n+ crystalline silicon substrate confirm that the long wavelength response is remarkably enhanced in this material.

An initial conversion efficiency of 14.6% has been achieved using amorphous silicon-based alloy in a spectrum splitting triple-junction structure. After 1000 hours of indoor one-sun light soaking at 50 °C, the stabilized efficiency is 13.0%. Both efficiencies are the highest reported to date for amorphous silicon alloy solar cells and have been independently confirmed by the National Renewable Energy Laboratory. The device was deposited onto a stainless steel substrate coated with textured silver/zinc oxide back reflector. The bottom and middle cells use amorphous silicon-germanium alloys, employing high hydrogen dilution in the gas mixture and bandgap profiling in the cell design. The top cell uses amorphous silicon alloy with high hydrogen dilution. Key factors leading to the achievement include a) improvement of the bottom cell that exhibits an AM1.5 efficiency of 10.4% and quantum efficiency of 45% at 850 nm; b) improvement of the tunnel junctions between the component cells by incorporating a novel multilayered structure with microcrystalline p and n layers; and c) improvement of transparent conductive oxide for enhancing the short wavelength response of the top cell.

A detailed study of the influence of different materials for the amorphous top cell on the stabilized efficiency of amorphous silicon/microcrystalline silicon (”micromorph”) tandem cells is presented. The authors investigate different amorphous i-layer materials which are applied in cells with varying thicknesses. It is shown that it is preferable to optimize the top cell in order to obtain a high current rather than a high voltage.

A simple optical and electrical model is presented which allows one to predict the optimum optical bandgap of the i-layer material of the top cell and to determine its optimum thickness for maximum stabilized efficiency. By means of this model it is shown that the optimum top cell for a total cell current of 26 mA/cm2 should contain an about 2500 Å thick i-layer with an optical gap Eo4 ≃ 1.8 eV. The model furthermore shows that it is desirable to obtain slightly bottom-limited conditions after degradation in order to maximize the output power.

A stabilized efficiency of 10.7 % for a micromorph tandem cell has been confirmed by an independent measurement (FhG-ISE). Another such cell yields a stabilized efficiency of 11.2 % as measured in our laboratory.

We have extended previous real time spectroscopie ellipsometry (RTSE) capabilities in order to investigate the effects of H2-plasma treatment of i-type hydrogenated amorphous silicon (a-Si:H) on the deposition of the overlying p-type microcrystalline silicon (μc-Si:H:B)) in the formation of an n-i-p solar cell structure. In this study, we compare in detail the nucleation and growth of p-layers by plasma-enhanced chemical vapor deposition (PECVD) from SiH4 highly diluted in H2 on the surfaces of untreated and H2-plasma treated a-Si:H i-layers. We find that for intended single-phase μc-Si:H:B p-layer PECVD under optimum conditions on an untreated i-layer surface, a wide gap (∼2.0 eV Taue gap) amorphous layer nucleates and grows in the first ∼150 Å. This layer develops uniformly to a bulk thickness of ∼150 Å, but gradually acquires a crystalline structure for thicknesses greater than the desired p-layer thickness (200 Å). In contrast, for p-layer PECVD under identical conditions on the H2-plasma treated i-layer, high-density crystalline nuclei form immediately. This conclusion is drawn on the basis of the unique optical properties of the bulk p-layer that develops on the surface of the H2-plasma treated i-layer. Specifically, an absorption onset near ∼2.5 eV is observed for a 48 Å fully-coalesced p-layer, as measured by RTSE at 200°C. For this μc-Si:H:B p-layer, the optical gap decreases by ∼0.15 eV with increasing thickness from 50 to 200 Å. This effect is attributed to a reduction in the quantum confinement energy with an increase in the average crystallite size in the film.

An initial conversion efficiency of 13.5% has been obtained on a triple-junction triple-bandgap device fabricated in a large-area deposition reactor capable of producing one-square-foot modules. The intrinsic layer of the top cell is a wide bandgap amorphous silicon alloy. The middle and bottom cells employ high quality amorphous silicon-germanium alloy. The high efficiency of the triple-junction cell is attributed to the relative reduction of the optical loss in the top tunnel junction and the improvement in the quality of the middle and bottom component cells. Triple-junction devices with initial efficiency of 13.3% have shown saturation at 11.6% after light soaking. Modules of aperture area 909cm2 have been fabricated using an assembly process similar to the one being currently used in our manufacturing line. The module design consists of onelarge-area, high-current monolithic multijunction device. The status of the small-area devices andmodules is described

Space-charge-limited currents have been examined in a wide variety of n-i-n devices. If the devices were completely symmetric, the current-voltage characteristics would be identical for positive and negative bias, but in several devices differences between the two polarities were observed. In order to understand in which part of the device these differences originate, the influence of the contacts and interfaces on the JV characteristics were examined by using different metal top contacts, different n-layers and different i-layers. Ag and Al top contacts gave minor differences between the polarities, whereas with Cr contacts no differences were observed. Incorporation of a defect layer in the i-layer results in asymmetric JV curves. We have observed a small asymmetry in an experimental device, and a large asymmetry using AMPS modeling. N-i-n devices appear to be a sensitive probe for interface defects.

We present measurements of the photocapacitance in hydrogenated amorphous silicon (a-Si:H) Schottky barrier diodes under reverse bias. A calculation relating photocapacitance to hole drift mobility measurements is also presented; the calculation incorporates the prominent dispersion effect for holes in a-Si:H usually attributed to valence bandtail trapping. The calculation accounts quantitatively for the magnitude and voltage-dependence of the photocapacitance.

We demonstrate that the internal field of a thin a-Si:H pin solar cells can be measured using the transient-null-current method. This method was previously developed to measure the internal field profile in a-Si alloy Schottky barrier. The internal electric field profile was determined by measuring the forward-bias voltages that tune the transient photocurrents generated by a pulsed laser at a various wavelengths to zero. We adopt the same technique to a-Si:H p-i-n solar cells. In the case of p-i-n structure, we need to consider both space charge contributed by photogenerated carriers and carrier recombination which disturb the internal field. We use two critical conditions to minimize these effects. (1) To limit the contribution of photocarriers to space-charge distribution, the total charge collected is less than 10−10 C per pulse, and a repetition rate 1 Hz is used to ensure that the diode remains close to its equilibrium state. (2) The measuring time window is about 1 – 6 μs following the displacement current. Typically the RC constant of diode is < 1 μs and the rise time of the forward-bias recombination current is 6.0 × μs. We apply the signal average to process the forward-bias voltage. The error is within ± 0.05 V. With this technique we can study the effect of variety of structure design or processing on the device performance.

An experimental and numerical study of a-SiGe:H based solar cells with band gap graded i-layer in the shape of a ‘V’ is presented. The variation of the location of the band gap minimum has strong influence on the solar cell performance. Comparisons of experimental and simulated data of the dark IV-behavior, IV-curves under illumination and the quantum efficiency allow insights into the transport and recombination behavior within the solar cell. The simulations reveal that the position as well as the charge state of the defects determine the device characteristics.

Metal-Induced-Crystallization (MIC) by the contact of amorphous semiconductors with metals is one of the degradation factors in solar cells. This study has been made on the barrier properties of a ZnO layer between undoped a-SiGe:H and Al metallization films in the structure (001)Si/SiO2/a-SiGe:H/ZnO/Al. Plasma assisted CVD deposition was used to produce a-Si1.xGex:H (x=0 to 1) undoped films over thermally oxidized Si-wafers. There were covered with 500Å and 1000Å thick transparent conductive layers of ZnO. Al and then 1000Å thick films of Al. A set of Al-implanted a-Si, a-Ge, and a-Sio.5Geo.5 films on Si/SiO2 substrates was also prepared to study MIC in an amorphous system with dispersed Al. The structures were annealed in vacuum in the temperature range of 200°C to 400°C for lh. X-ray diffraction studies demonstrated the a-SiGe:H stability against crystallization under ZnO protection up to 400°C. Secondary Ion Mass Spectroscopy didn't reveal any noticeable redistribution of Al inside Al-implanted a-Si:H and a-Si0.5Ge0.5:H samples after annealing at 400°C for lh, but strong Al diffusion was seen in the a-Ge:H layer. Nevertheless, no MIC was observed in any of the Al-implanted a-materials.

Studies have been carried out on a-Si:H materials and corresponding solar cells fabricated with and without hydrogen dilution of silane by rf PECVD. The effect of hydrogen dilution on the growth kinetics and microstructures and their dependence on the substrate temperature have been studied. Hydrogen diluted a-Si:H materials and solar cells exhibit improved properties and higher stability to light induced changes. Distinct differences are found in the electron mobility lifetime (μτ) products and subgap absorption over a wide range of generation rates. Striking differences are also found in the kinetics of light induced degradation in both the materials and their corresponding solar cells. Direct correlations are presented between the degradation kinetics of p(a-SiC:H)/i(a-Si:H)/n(μc-Si) solar cells and those of thin film materials constituting the i-layers.

We study the effects of hydrogen dilution on the open circuit voltage of a-Si:H pin solar cells fabricated by rf glow discharge growth. We keep the p and n layers the same and only vary the i layer properties. A normal a-Si:H i layer, an H-diluted i layer, and a thin H-diluted layer inserted between p and normal i layer are selected for this study. We measure the JV characteristics and the internal electric field distribution using a transient-null-current technique both in annealed and light soaked states. We find that hydrogen dilution does stabilize the Voc either in a bulk H-diluted i layer or in a thin layer between p and normal i layer after 100 hours AMI sun light soaking. From dark IV measurement, both H-diluted cells show little change in current at voltage near Voc before and after light soaking; while the normal a-Si:H cell does show a noticeable change. Also the internal field measurements find a stronger electric field starting from p and i interface for both H-diluted cells compared to the normal a-Si:H cell. Furthermore, there are no measurable changes in the field profiles after 100 hour AMI light-soaking for both H-diluted and normal a-Si cells. All these suggest that hydrogen dilution increases the field strength near p and i interface, which is the key that leads to a more stable Voc of H-diluted cells.

An account is given for the conditions under which the collection efficiency in hydrogen ated amorphous silicon pin-diodes increases to values larger than 100 %. By specific bias illumination through the p-side bias generated photocarriers are collected under certain probe beam conditions of the collection efficiency measurement, leading to apparent large collection efficiencies. By numerical modelling we investigated the influence of the diode thickness, bias photon flux and probe absorption coefficient as well as applied voltage for possible sensor applications which may utilise this optical amplifying principle. The alternative with bias light through the n-side and probe light through the p-side is also explored. Collection efficiency values determined by the photogating of bias generated holes become only slightly larger than 100 % in contrast to the electron case where values in excess of 3000 % are presented.

The behaviour of a inicrocrystalline p-i-n junction locally illuminated with monochromatic radiation (incident power of 50 mW/cm2) is analysed by means of numerical experiences. The model used for the two-dimensional analysis of the transport properties of a μc-Si:H p-i-n photo-detector is based on the simultaneous solution of the continuity equations for holes and electrons together with the Poisson's equation. The solution is found on a rectangular domain, taking into account the dimension perpendicular to the junction plane and one on the parallel plane. The lateral effects occurring wimin the structure, due to the non-uniformity of the illumination, are outlined. The results we present show that the potential profile has a linear variation from the illuminated to the dark neutral region. The lateral components of the electric field and of the current density vectors reveal to be mainly localised inside the doped layers.

An integrated electrical-optical model has been used to simulate and examine ways of optimising the performance of double junction solar cells, where both the component cells have a-Si:H absorber layers of identical material quality. In the optical modelling part we take into account both specular interference effects; and diffused reflectances and transmittances due to interface roughness. The model simulates carrier transport in the junction between the two p-i-n subcells with the help of a thin heavily defective recombination layer (RL) having a reduced band gap.

Our results reveal that in order to simulate the current-voltage and the quantum efficiency (QE) characteristics of these cells, window losses and light-trapping effects need to be properly accounted for. Results indicate that the highest open-circuit voltage is attained when the majority carrier quasi-Fermi levels on either side of the RL coincide. Also for the highest multijunction cell efficiency the thicknesses of the component subcells are such that the electric field in both are fairly close to one another. Finally, the QE under AM 1.5 bias light at the maximum power point has been shown to be extremely sensitive to thickness variations of the component subcells and hence an useful tool for multijunction cell optimisation.

We have used a new numerical model and here present initial results on how texturing and backreflectors affect the maximum achievable short-circuit current densities in amorphous silicon solar cells.

In this work we report on the effect of current-induced degradation and annealing on p-i-n amorphous silicon solar cells. Current-voltage curves and capacitance measurements under forward bias have been used to monitor the current-induced changes as a function of time. We found that the recovery rate increases with the annealing current, while the stabilized value of efficiency decreases. Comparison of short circuit current and capacitance evolution suggests that defect kinetics in the electronic gap occurs in a different way during degradation and annealing. This behavior can be modeled assuming a faster annealing of defects closest to the extended band and a slower annealing of mid-gap defects.

We have deposited and analyzed solar cells from mixed SiCI2H2/SiH4 plasmas. Intrinsic and doped films were also grown to explain the cell results. The SiCI2H2 cells have lower fill factor and short circuit current density than the SiH4 cells due to the higher defect densities and lower photoconductivities of SiCI2H2/SiH4 material. Cell results showed low VQC if the n-layer was in contact with an i-layer deposited from SiCI2H2/SiH4. A pure silane buffer layer between the i- and n-layers was needed to attain the standard open circuit voltage. The reduced Voc results from a reduced conductivity of phosphorus-doped SiCI2H2/SiH4 material. However, comparison of doped films showed higher conductivities for boron-doped material grown from the SiCI2H2/SiH4 mixture than from pure silane.