Long wavelength photons in the red and near infrared region of the spectrum are poorly absorbed in thin film silicon cells, due to their long absorption lengths. Advanced light trapping methods are necessary to harvest these photons. The basic physical mechanisms underlying the enhanced light trapping in thin film solar cells using periodic back reflectors include strong diffraction coupled with light concentration. These will be contrasted with the scattering mechanisms involved in randomly textured back reflectors, which are commonly used for light trapping. A special class of conformal solar cells with plasmonic nano-pillar back reflectors will be described, that generates absorption beyond the classical 4n2 limit (the Lambertian limit) averaged over the entire wavelength range for nc-Si:H. The absorption beyond the classical limit exists for common 1 micron thick nc-Si:H cells, and is further enhanced for non-normal light. Predicted currents exceed 31 mA/cm2 for nc-Si:H. The nano-pillars are tapered into conical protrusions that enhance plasmonic effects. Such conformal nc-Si:H solar cells with the same device architecture were grown on periodic nano-hole, periodic nano-pillar substrates and compared with randomly textured substrates, formed by annealing Ag/ZnO or etched Ag/ZnO. The periodic back reflector solar cells with nano-pillars demonstrated higher quantum efficiency and higher photo-currents that were 1 mA/cm2higher than those for the randomly textured back reflectors. Losses within the experimental solar architectures are discussed.

Three-dimensional (3-D) optical modeling based on Finite Element Method of single, double, and triple junction thin-film silicon solar cells is presented. The combination of front periodic gratings with optimal geometrical parameters and rear ZnO/Ag reflector constitutes an efficient light trapping scheme for solar cells in superstrate (pin) configuration. The application of optimized trapezoidal 1-D and 2-D gratings resulted in 25.5% (1-D case) and 32.5% (2-D case) increase in photo-current density with respect to the flat solar cell. The application of inverted pyramidal 2-D gratings in double and triple junction silicon solar cells with very thin absorber layers resulted in a photo-current density > 11 mA/cm2 and > 9 mA/cm2, respectively.

In this study we compare light trapping in hydrogenated amorphous silicon (a-Si:H) solar cells deposited directly onto polycarbonate (PC) at low temperature (< 130°C). To that end, we embossed PC substrates with 400 nm and 10 μm square based pyramids to induce light trapping based on diffraction and on geometric effects. As a comparison, we deposited a-Si:H cells on flat glass substrates and on Asahi U-type TCO glass. The cells on PC generate current densities comparable (slightly higher) than cells on Asahi TCO glass, but suffer from a slightly lower Voc, resulting in cells with an initial efficiency of 6.8% and 7.4% on sub-micron pyramid and micro-pyramid structured PC substrates respectively, compared to 7.6% for cells on Asahi. This shows great potential for a-Si:H cells deposited directly onto cheap plastics.

We correlated the texture morphology and the solar cell properties by measuring the distribution in the texture morphology. As a result, the short-circuit current ISC was approximated across various types of substrates by the standard texture height. Furthermore, we investigated the texture morphology from the point of view of the electrical effects. With regard to this point, the open-circuit voltage VOC was correlated to the steepest texture angle. Therefore, we consider that the both of the ISC and the VOC can be improved by controlling the distribution in the texture morphology.

We present the fabrication and characterization of low-temperature ambipolar thin-film transistors (TFTs) based on hydrogenated amorphous silicon-germanium (a-SiGe:H) as active layer. Inverted staggered a-SiGe:H TFTs were fabricated on Corning glass. Spin-on glass silicon dioxide was used as gate dielectric to improve the quality of the dielectric-semiconductor interface. For positive gate bias the transfer characteristic showed n-type TFT behavior, while for negative gate bias p-type behavior was observed. The n-type region exhibits subthreshold slope of 0.45 V/decade while the p-type region shows a subthreshold slope of 0.49 V/decade.

This paper reports on light filtering devices based on a-SiC:H tandem pi´n/pin heterostructures. The spectral sensitivity is analyzed. Steady state optical bias with different wavelengths, are applied from each front and back sides and the photocurrent is measured. Results show that it is possible to control the sensitivity of the device and to tune a specific wavelength range by combining radiations with complementary light penetration depths. The transfer characteristics effects due to changes in the front and back optical bias wavelength are discussed.

Input red, green and blue pulsed communication channels are transmitted together, each one in a specific bit sequence and the multiplex signal is analyzed. By superimposing appropriate background and depending on the channel/background wavelength combinations, the device behaves as a long- or a short- pass filter, producing signal attenuation, or as an amplifier, producing signal gain. A physical model is presented to support the filter properties of the device.

Security imaging systems working with crystalline silicon CCD or CMOS detectors are not able to distinguish colorimetrically between a large number of dangerous chemical substances, for example whitish powders [1]. In order to offer an alternative to expensive and destructive chemical methods of analysis, we developed optimized hydrogenated amorphous silicon (a-Si:H) multicolor photodiodes with different spectral response characteristics for a reliable, fast, cheap and non-destructive identification of potentially dangerous substances. Experimental optical, C-V and I-V studies were performed to explore the effect of combining linear graded a‑SiC:H-/a‑SiGe:H layers with low-reflective ZnO:Al back-contacts. Typically, a-Si:H with profiled energy gaps can be found in tandem solar cells to optimize the collection of incoming photons [2,3]. We determined the absorption coefficients of a group of a-SiC:H and a-SiGe:H graded and non-graded layers to calculate the penetration depth of photons at different energies into the device structure. Knowing the indices of absorption, refraction and extinction, it is possible to engineer diodes in such a way that accumulations of charge carriers are generated precisely at varying device depths. Common chromium back reflectors avoid a sharp falling edge of the sensitivity towards longer wavelengths and lead to interference fringes in the spectral response [4]. By combining linear graded absorption zones and ZnO:Al back contacts, we designed an optimized device with a highly precise adjustment of the spectral sensitivity reaching from 420 nm to 560 nm and reduced interference fringes at a very low reverse bias voltage of maximum -2.5 V. Similar three terminal devices allow a shift from 440 nm to 630 nm, however, at a much higher reverse bias of -11 V at 560 nm [4]. Present research efforts concentrate on the development of fast and high dynamic front illumination device structures which ensure a continuous narrow-band shift of the spectral photosensitivity and an optimum adaption to a predetermined light source-/sample measurement configuration.

Fluorescence Resonance Energy Transfer (FRET) is a standard technique used in many medical and biological applications. It involves the detection of transient fluorescent signals coming from the different fluorescent proteins that work in the visible range of the spectrum. Common fluorescent emissions come from the cyan/yellow fluorophores that emit respectively, at 470 nm and 588 nm. In this paper we use optical filters based on multilayered a-SiC:H heterostructures to detect optical signals at these wavelengths. The advantage of this type of sensor is that it does not rely on mechanical parts; it is compact and cost effective. The transducer consists of two heterostructures based on a-SiC:H/a-Si:H optimized for the detection of the fluorescence emissions at wavelengths 470 nm (cyan) and 588 nm (yellow). Both front and back structures were designed to optimize the detection at these wavelengths. Results show that the device photocurrent signal measured under reverse bias and using appropriate steady state optical bias, allows the separate detection of the cyan and yellow fluorescence signals.

We analyze thermoelastic dissipation in composite silicon MEMS resonators that exhibit multiple mechanical and thermal modes with complex dynamics. Silicon resonators that are coated with thin films of silicon dioxide can have near-zero temperature coefficients of frequency, making them attractive for use as precision time references. The quality factor of MEMS resonators can be dominated by thermoelastic dissipation (TED), which is triggered by the relaxation of mechanically induced temperature gradients. Recently, Chandorkar et al. (2009) have shown an expression of TED based on entropy generation as a weighted sum of the modal solutions of the three-dimensional heat transfer equation. This expression was obtained for weak coupling between mechanical and thermal dynamics. Applying this same technique to a fully coupled solution to the dynamics, we show that the TED contribution of the dominant thermal modes can be inhibited in the presence of a thin silicon dioxide film. Reduction of the contribution from the dominant thermal mode is shown with increasing oxide. We studied the effects of varying oxide film thickness and beam length. The quality factor was simulated for each unique case and compared to multimode energy dissipation. Our results suggest with some variability, thin film oxide coating affects the thermal relaxation of the composite resonator in the direction of lower TED and increased quality factor.

Light effects on the performance of the a-Si:H PIN photodiode has been studied. The leakage current-voltage and capacitance-voltage curves under the red, green and blue light illuminations were measured. The apparent charge storage density in the negative voltage range was quantified from the capacitance-voltage curve; charges in the positive voltage range were estimated from the leakage current-voltage curve. A comparison of charge storage capacities of diodes with different intrinsic layer thicknesses is also presented. The diode under the long wavelength light illumination condition stored more charges than that under the short wavelength light illumination condition because the former could penetrate the intrinsic a-Si:H layer deeper than the latter could. The leakage current and charge storage capacity of the diode are determined by the generation of electron-hole pairs, the depletion of charges in the intrinsic layer, and the supply of charges from the electrodes. The number of incident photons is critical to the process.

There is significant interest in optical sensors whose fabrication process is fully compatible with existing flat panel display thin film transistor (TFT) technology. Here, we report a field-effect phototransistor with a channel comprising a thin nanocrystalline silicon (nc-Si:H) transport layer and a thicker hydrogenated amorphous silicon (a-Si:H) absorption layer. The implementation of nc-Si:H layer improves device stability in comparison with a-Si:H phototransistors, resulting in reduced threshold voltage shift. Semiconductor and dielectric layers were deposited by radio-frequency plasma enhanced chemical vapor deposition at 280°C. The device characterization included the dark and light transfer characteristics, spectral-response and dynamic measurements. The external quantum efficiency was measured as a function of incident photon flux at different biasing conditions. The phototransistor with channel length of 24 microns and photosensitive area of 1.4 mm2shows an off-current of about 1 pA, and photo-conductive gain up to 200 at low incident intensities. Thus, the results demonstrate the feasibility of the phototransistor for low light level detection.

Fluorescence detection is more advantageous than electrochemical detection in terms of high sensitivity, multiplexed detection capability and isolation from analyte. Integration of fluorescence detection, however, is much more difficult. First, it would require heterogeneous integration of various optical components including an excitation source, an optical filter, a lens, a mirror and a detector. Second, most of integrated fluorescence detectors, even though not fully integrated, suffer from high limit of detection (LOD) compared to conventional optical system that consists of discrete optical components. We have reduced laser light scattering in an integrated hydrogenated amorphous Si (a-Si:H) fluorescence detector, significantly improving a limit-of-detection (LOD). The detection platform comprises a microlens and the annular fluorescence detector where a thick SiO2/Ta2O5 multilayer optical interference filter is monolithically integrated on an a-Si:H pin photodiode. With a microfluidic capillary electrophoresis (CE) device mounted on the platform, the integrated system is demonstrated to separate DNA restriction fragment digests with high speed, high sensitivity and high separation efficiency, implying single molecular DNA detection when combined with polymerase chain reaction (PCR). We are now working towards integration of an excitation source to fabricate heterogeneously integrated laser-induced fluorescence detection (LIF) device that would be comprised of an InGaN laser diode, microlenses and the integrated a-Si:H fluorescence detector.

This work presents the development of three-terminal photodiodes which have their origin in bipolar pinipip-diode-structures [1,2]. The idea is to develop an unipolar diode and to contact a buried p-layer by an interior TCO-anode (transparent conductive oxide) instead of nip-stacking. The simulation of the optical material properties shows promising results. At first, both parts of the diode-structure were produced separately. The manufactured bottom (Cr/nip/TCO) and top (TCO/pip/TCO) parts were measured optically and electrically. These measurements are required to simulate the SR of the total-diode. Finally, a Cr/nip-a-Si:H/TCO/pip-a-Si:H/TCO multi-layer stack was deposited. The measured SR of the integrated diode validates the simulated data. The SR maximum shift amounts to 100nm, from 540nm by contacting the interior anode, to 640nm at the top anode. Furthermore, the curves are clearly split and do not enclose each other. The presented approach, with additional bandgap engineering, promises good prospects to improve color separation compared to currently existing detectors and should lead to a tunable multi-spectral photodiodes for high quality color recognition. Such a diode can be used in photonic devices, e.g. for safety and security applications.

Stacked layered pin a-SiC/a-Si devices based on a filter design are approached from a reconfigurable point of view. This paper shows that a double SiC/Si pin photodiode can be de-composed into two photonic active filters changeable in function. Reconfiguration is provided by optical control signals to the optoelectronic front and back pin building blocks. Depending on the wavelength and irradiation side of the external optical bias the device acts either as a short- and a long- pass band filter or as a band-stop filter, amplifying or rejecting a specific wavelength range. Particular attention is given to the amplification coefficient weights, which allow taking into account the wavelength background effects. We illustrate these effects in detail and discuss the filters transfer function characteristics. We present examples of filters and we propose a reconfigurable device for directed optical logic. An algorithm to decode the information is presented. An optoelectronic model supports the optoelectronic logic architecture.

In this paper, we present an example of a new generation of laser-crystallization approaches that can crystallize Si films for large active-matrix displays at drastically increased effective crystallization rates. The particular scheme presented in this paper is referred to as the advanced excimer-laser-annealing (AELA) method, and it can be readily configured for manufacturing large OLED TVs using various available and field-proven technical components. As in ELA, it is mostly a partial-/near-complete-melting-regime-based crystallization approach; AELA can, however, eventually achieve greater than one order of magnitude increase in the effective rate of crystallization over that of the conventional ELA technique utilizing the same laser source. We discuss in this paper how and why such a dramatic increase can be attained, and some strategical and technological benefits and options that can be entertained regarding, and as a result of the availability of, the AELA technique.

A new method to form high quality crystalline silicon thin films on cheap glass substrates is developed using a single pass of a line-focus cw diode laser in air. The laser process results in the formation of large high-quality crystals as they grow laterally in the scan direction – seeded by the previously crystallised region. Grains 10 μm in thickness, up to millimetres in length and hundreds of microns in width have been grown with virtually zero detectable intragrain defects. Another mode is found which results in much smaller crystals grown by partial melting. The dominant grain boundaries identified are Σ3 <111> 60° twins. Hall mobilities as high as 470 cm2/Vs have been recorded. A diffused emitter is used to create a p-n junction at the rear of the films which produces open-circuit voltages as high as 539 mV.

We have investigated the effect of varying the film thickness on the surface orientation texturing in polycrystalline Si films obtained via mixed-phase solidification (MPS) of initially a-Si precursor films on SiO2. It is found that, for a given number of MPS exposure cycles, the degree of (100)-surface texturing is reduced as the film thickness is increased. We discuss how this trend can be accounted for by the previously proposed thermodynamic model of MPS, wherein a decreasing local solid/liquid interface curvature with increasing film thickness is identified as the primary cause for decreasing the influence which anisotropic solid-Si/SiO2 interfacial energies have on the survivability of the grains. This, in turn, leads to other factors becoming more significant in determining the grains that survive the MPS cycle, thereby reducing the degree of (100)-surface texturing in the resulting films.

Amorphous (a-Si) and polycrystalline silicon(c-Si) films have been obtained by low pressure chemical vapor deposition (LPCVD) in fixed silane flow at low pressure (200 mtorr) with variable growth temperature. Measurement of residual stress of polysilicon films growth between 570°C to 620°C was reported. Residual stress of polycrystalline silicon depends on film microstructure. The poly-silicon microstructure is a strong function of LPCVD growth temperature and pressure. In this work, Raman Scattering and X-ray diffraction (XRD) were used to study the film structure and composition. The multilayer optical model of Spectroscopic ellipsometry (SE) was used to crosscheck the crystallinity fraction. The surface roughness was identified by Atomic force microscopy (AFM) and SE. The residual stress changed from compressive to tensile and back to compressive for deposition temperature between 570°C and 620°C. Film’s c-Si fraction increased as a function of deposition temperature. The roughness surface was found at deposition temperature of 580°C. For deposition temperature larger then 580°C, all films shown (200) texture. The (220) grain size increased from 18.6 nm to 25 nm when deposition temperature increased from 580°C to 620°C. The film residual stress change can be explained by grain structure, surface stress and volume stress. At deposition temperature from 580°C to 587°C, grain is equi-axes type and volume stress dominate which cause the tensile stress. For temperature higher then 587°C inverse conical grain formed from oxide interface to surface and surface stress dominate cause the stress back to compressive. The columnar structure formed when deposition temperature > 600°C, grain growth push the compressive stress decrease again.

The polycrystalline n+/intrinsic silicon thin film stacks with various original intrinsic amorphous silicon layer thicknesses were formed using the multiple pulsed rapid thermal annealing process with the Ni-induced crystallization mechanism. The thick polycrystalline silicon stack was prepared by repeated steps of 1) amorphous silicon thin film deposition, 2) solution oxidation, 3) dehydrogenation, 4) pulsed rapid thermal annealing, and 5) oxide stripping. The poly-Si film properties, such as the grain size, orientation, and volume fraction of the crystalline phase, were related to the original intrinsic silicon film thickness and the total thermal budget. This process is effective in preparing the high volume fraction polycrystalline silicon thin film, which is important for low-cost thin-film solar cells, electronic and optoelectronic devices.

Layer transfer and simultaneous activation of phosphorus atoms in Si films induced by semiconductor diode laser (SDL) irradiation have been investigated. Phosphorus-doped a-Si films supported by columns on a starting substrate (quartz) and a counter substrate (glass) were closely contacted face-to-face, and an 812 nm light from a SDL was irradiated to the a-Si films from the backside of the starting substrate. After SDL irradiation, 20μm wide and 1000μm long Si films were transferred to the counter substrate and were crystallized simultaneously. From optical microscope images, it was confirmed that the original form was completely maintained after the film transfer. The electrical conductivity of transferred Si film was as high as 708 S/cm. Hall measurement of the films revealed very high electron concentration of 9.5×1020 cm-3, which indicated efficient doping is achieved by the laser transfer technique.