Morphological and nanostructural features of porous silicon prepared by electrochemical etching

Porous layers were produced on a p-type (100) Si wafer by electrochemical anodic etching. The morphological, nanostructural and optical features of the porous Si were investigated as functions of the etching conditions. As the wafer resistivity was increased from 0.005 to 15 Ω·cm, the etched region exhibited ‘sponge’, ‘mountain’ and ‘column’-type morphologies. Among them, the sponge-type structured sample showed the largest surface area per unit volume. Silicon nanocrystallites, 2.0 to 5.3 nm in size, were confirmed in the porous layers. The photoluminescence peaks varied in the wavelength range of 615 to 722 nm. These changes in the maximum peak position were related to the size distribution of the Si crystallites in the porous silicon. The doping levels of the wafers significantly affect the size distribution of the Si crystallites as well as the light-emitting behavior of the etched Si, which contains nanoscale Si crystallites.     

Phosphorus diffusion gettering process of multicrystalline silicon using a sacrificial porous silicon layer

ABSTRACT: The aim of this work is to getter undesirable impurities from low cost multicrystalline silicon (mc-Si) wafers and then enhance their electronic properties. We used an efficient process which consists in applying phosphorus diffusion into a sacrificial porous silicon (PS) layer in which the gettered impurities have been trapped after the heat treatment. As we have expected, after removing the phosphorus-rich porous silicon (PS) layer, the electrical properties of the mc-Si wafers were significantly improved. The PS layers, realized on both sides of the mc-Si substrates, were formed by the stain-etching technique. The phosphorus treatment was achieved using a liquid POCl3-based source on both sides of the mc-Si wafers. The realized phosphorus/PS/Si/PS/phosphorus structures were annealed at a temperature ranging between 700 degreesC and 950 degreesC under an O2 controlled atmosphere, which allows phosphorus to diffuse throughout the PS layers and to getter eventual metal impurities towards the phosphorus doped PS layer. The effect of this gettering procedure was investigated by means of the internal quantum efficiency (IQE) and the dark current-voltage (I-V) characteristics. The minority carrier lifetime measurements were made using a WTC-120 photoconductance lifetime tester. The serial resistance and the shunt resistance carried out from the dark (I-V) curves confirm this gettering-related solar cell improvement. It has been shown that the photovoltaic parameters of the gettered silicon solar cells were improved as regard to the ungettered one, what proves the beneficial effect of this gettering process on the conversion efficiency of the multicrystalline silicon solar cells.     

Diffusion in porous silicon: effects on the reactivity of alkenes and electrochemistry of alkylated porous silicon

Alkenes are known to react with hydrogen-terminated silicon surfaces to produce robust organic monolayers that are attached to the surface via covalent SiC bonds. In this report we investigate the dependence of the rate of alkylation of porous silicon samples on the reaction time using photochemical initiation. The kinetics of the photochemical alkylation of hydrogen-terminated porous silicon by undec-1-ene in toluene were observed to be pseudo first order, however the apparent rate constant decreased as the concentration of undec-1-ene increased. This behaviour is opposite to what would be expected if the rate-limiting process was an elementary chemical reaction step involving the alkene. Instead, it suggests that transport of the alkene to reactive sites and in the correct orientation is the rate-limiting step. Comparison of the rates of alkylation of porous silicon by undec-1-ene and dimethoxytrityl (DMT)-undecenol is consistent with such an interpretation as the bulky DMT headgroup gives a lower rate of alkylation. The diffusion of some simple redox-active probe molecules in porous silicon was investigated using a scanning electrochemical microscope (SECM). The probe molecules are converted at diffusion-controlled rate at an inlaid disk ultramicroelectrode (UME) consisting of the cross-section of a microwire sealed in glass. If the microelectrode is placed a short distance above the porous silicon, the microelectrode current depends on kinetics of the electrochemical reactions at the porous silicon and the mass transport properties within the open thin layer cell formed by the microelectrode and the alkylated porous silicon. In order to differentiate the effects of finite heterogeneous kinetics at silicon from diffusion limitations, current-distance curves were fitted over a wide range of applied potentials (on the Si) and it was observed that the diffusion coefficient in the porous layer was strongly anisotropic. The measured diffusion rates are comparable to those in bulk water along the pores, but with negligible diffusion between pores. This indicates that few pore-pore interconnections exist in the porous silicon.     

Gasification reactivity of charcoal with CO2. Part I: Conversion and structural phenomena

The gasification reaction of fir charcoal with CO₂ was studied by isothermal thermogravimetric analysis under kinetic control. The derived reaction rate (r=dX/dt) as a function of the converted carbon mass (X) was compared with random pore model predictions and found to be much higher at elevated conversion levels than predicted by theory. Similar enhanced reaction rate behaviour was evidenced after removing the natural alkali catalyst from the charcoal by acid washing, suggesting that with untreated charcoal the late reaction rate contribution stems from both, catalytic and additional structure effects. Literature attributes the unpredicted late reaction rate behaviour to the disintegration of the porous char particle into small fragments, which, in line with percolation theory predictions, seems to occur only after a critical conversion level has been reached. However, our gasification data reveal a gradual rise in the charcoal reactivity thereafter, suggesting a breaking up (embrittlement) of the solid phase accompanied by the exposure of fresh surface area from fracturing. The original random pore model derivation given by Bhatia and Perlmutter is extended to account also for these peculiarities and the resulting kinetic relation described our reaction rate data well over the entire conversion range.     

Explicit numerical simulation of suspension flow with deposition in porous media: influence of local flow field variation on deposition processes predicted by trajectory methods

A new explicit numerical simulation (ENS) method based on lattice-gas automata (LGA) is introduced here for the flow of solid/fluid suspensions with deposition in porous media. The ENS method explicitly resolves the dynamics of the individual solid particles and the suspending fluid in the domain defined by the pore walls and solid particle surfaces. After describing the new method, it is applied to the study of solid/fluid suspension flow with deposition in a constriction and in a model random porous solid. This study clearly demonstrates that the deposits strongly influence the local flow fields, which in turn affect the deposition process indicating that this interplay should be modelled if accurate results are desired from trajectory methods.     

Modelling of catalyst particle skin effects using a 3-D pore network model and quantitive microscopy

The pore structure of fragments of steam reforming catalyst ring-shaped pellets was characterised so as to simulate a larger particle using larger pore networks comprising millions of pore segments. The pore size distribution generated for a 3-D stochastic network was deduced from the experimental mercury penetration curve. The penetration processes which take place within the pore network at different pressures can be visualised by randomly slicing irregular 3-D pore networks to create 2-D “Virtual reality slice” images. The experimental visualisation technique of low-melting alloy impregnation was applied to the larger ring catalyst fragments at several pressure levels in order to investigate the effect of size, the fragment geometry, and a suspected skin effect on the mercury penetration curve for these particles. The experimental scanning electron microscopy images, obtained from polished sections of a large 5-mm ring fragment, can be compared directly with the 2-D virtual slice images from randomised 3-D pore networks for direct quantification of the pore structure using image analysis techniques.     

Pore structure of carbons coated on ceramic particles

Pore structure of carbon coated on ceramic particles by carbonization of precursor in a powder mixture at 900 °C was studied by focusing on the effects of substrate ceramics (MgO, TiO₂ and various phases of Al₂O₃) and of carbon precursor (poly(vinyl alcohol) (PVA), poly(ethylene terephthalate) (PET) and hydroxyl propyl cellulose (HPC)). By dissolving substrate MgO particles, carbon coated was found to have a high BET surface area, more than 1000 m²/g, which was almost the same as the value estimated from apparent surface area measured on carbon-coated MgO particles under the assumption of zero surface area of the substrate. The carbon separated was found to be rich in micropores from the analyses by DFT method and α_s plot. The dependence of the BET surface area on the amount of carbon coated on TiO₂ with a high surface area was the same for three carbon precursors, although the carbon yields from the precursors were slightly different. Porous Al₂O₃ substrates, γ-Al₂O₃as-received and that formed from Al(OH)₃ during carbonization, gave a high BET surface area, but dense Al₂O₃, α-Al₂O₃, gave a low surface area.     

Morphology development of 10‐μm scale polymer particles prepared by SPG emulsification and suspension polymerization

Classicalparticle morphologies, core‐shell, hemisphere, sandwich, and so on, were all reproducible by starting from ca. 10‐μm uniform droplets composed of monomers, initiator, solvents, and polymer, and polymerizing them by subsequent suspension polymerization. SPG (Shirasu porous glass) membrane was employed to form uniform size droplets having the coefficient of variation (CV) around 10%. Styrene (ST) and acrylic monomers were used as monomers, and their polymers were dissolved in the droplets to investigate the development of phase separation. When hydrophilic methyl methacrylate (MMA) was polymerized in the droplets with a mixed solvent consisting of hydrophilic hexanol (HA) and hydrophobic benzene and hexadecane (HD), the resulting morphology shifted from hemisphere to sandwich and eventually to PMMA/solvent core‐shell with increasing hydrophilicity of the mixed solvent. The sandwich was converted to the core‐shell after several weeks elapsed. As styrene was added to MMA, the morphology shifted from hemisphere core/solvent shell to raspberry core/solvent shell as the fraction of ST increased. The domain of the mixed solvent in the raspberry core was reduced with increasing the hydrophilicity of the mixed solvent. All these morphologies were eventually converted to the copolymer core/solvent shell. When a mixed monomer of styrene and MMA dissolving polystyrene (PS) was polymerized, the resulting morphology shifted from salami to core‐shell with increasing the MMA fraction in the comonomer. The salami particles were then swollen with toluene, and after the swelling, toluene was removed under the different temperature and pressure. The final particle morphology converted to the core‐shell with a milder rate of toluene removal which was predicted from the thermodynamic model. When styrene and cyclohexyl acrylate (CHA), a pair with widely different reactivity ratios, were copolymerized, salami morphologies, with tiny CHA‐rich domains dispersed in the matrix, were obtained even at a higher fraction of CHA in comonomer. Effects of glass transition temperature of the polymers, molecular weight, and the composition of copolymers were taken in consideration whenever the final morphologies were discussed. By these experiments, the authors tried to demonstrate an advantage of using large uniform spheres for the particle morphology studies. SPG emulsification technique was a potential tool because of its free formulation of the droplets, and the subsequent polymerization could undergo without the breakup or coalescence of the droplets. © 2001 John Wiley & Sons, Inc. J Appl Polym Sci 79: 2200–2220, 2001     

Modeling of Multiphase Transport during Drying of Honeycomb Ceramic Substrates

Multiphase transport model to simulate drying of honeycomb ceramic substrates in a conventional (hot air) drier is developed. Heat and moisture transport in the honeycomb walls as well as channels is modeled. The model predictions are validated against experiments done for drying of cylinder-shaped substrates by comparing histories and axial profiles of moisture loss and point temperature histories at various locations. Drying experiments are performed at two different values of air temperature, 103°C and 137°C, at a relative humidity value of 5%. Sensitivity analysis reveals that the drying process is controlled by heat and water vapor transport. External heat transfer is the dominant resistance mechanism for energy transport, whereas internal convection and binary diffusion dominate the resistance to vapor transport.     

CONVECTIVE DRYING OF THIN AND DEEP BEDS OF GRAIN

The object of this paper is the experimental and theoretical investigation of heat and mass transfer during drying of packed beds of grain. A deep bed of grain was regarded as a series of thin beds. Analytical expressions for the thin bed drying rate were obtained by defining the air parameters at the grain surface in the falling rate period of drying and using the results of drying experiments. The paper also contains a simulation model for drying deep beds of grain, consisting of four partial differential equations based on energy and mass balances in a bed element. The system of equations was solved using finite difference techniques and a digital computer. A comparison between numerical solutions and experimental results is illustrated.