Marked morphology-dependent electrical, electrochemical, and spectroelectrical properties of PbS nanostructured particulate films are the subject of the present communication. Examination of the potential-dependent absorption spectra on optically transparent conducting glass supports and of the voltage-dependent capacitances and photocurrents on platinum support have revealed substantial differences between equilateral-triangular PbS (PbS-I), right-angle-triangular PbS (PbS-II), and disk-shaped PbS (PbS-III) nanoparticulate films.

The optical absorption and photoluminescence properties of nanosize PbS clusters synthesized in inverse micelles are presented. Relative to bulk crystals, the bandgap of PbS nanoclusters is significantly blue shifted from the near-IR into the visible and near-UV region with decreasing cluster size. The optical absorption spectra for clusters with a diameter of ∼2.5-3 nm contain a molecular-like feature at 284 nm (4.37eV) which matches the absorption of diatomic PbS. The main photoluminescence band occurs at 520 nm and is due to recombination at surface states. The PLE spectrum of the 520 nm band has a minimum which coincides with the absorption peak. The absorption line is interpreted to arise from a pre-dissociative excitation of PbS molecules on the surface. The PLE band is interpreted to be an excitation of molecular-like levels in the interior of the cluster. Evidence from size, stoichiometry and surface alloy (Au) dependent properties supports this interpretation.

We previously reported the synthesis and some properties of heterostructure nanoparticles, i.e. CdS coated with PbS (CdS/PbS). We report further characterization and investigation of the infrared photoluminescence (PL) of the CdS and coated CdS/PbS nanoparticles. This allows us to provide a more stringent test of the quantum confinement model. We consider the energy shift of the PL peak and the variation of the PL intensity correlated with the predictions of the quantum confinement model. The experimental results can be explained by this model of coated semiconductor nanoparticles.

In the bulk state FeS2 and MoS2 are optically opaque, narrow bandgap semiconductors with no optical applications. We demonstrate that nanosize FeS2 and MoS2 have bandgaps that can be adjusted to the visible and even UV region of the spectrum by control of the cluster size. This opens up a host of applications of these materials as inexpensive solar photocatalysts. We demonstrate that the band-gap of both materials shifts to the blue with decreasing size but ceases shifting when a size of ∼ 3 nm (in the case of MoS2) is attained. We interpret this observation as a change from bulk quantum confinement of the hole-electron pair of a tiny semiconductor to a set of discrete molecular-like transitions more characteristic of a large molecule. Room temperature photoemission studies of these clusters demonstrate that, while photoemission shifts to the blue with increasing bandgap for large clusters, small clusters have photoemission exclusively from trapped sub-bandgap surface states. Chemical modification of the surface to introduce hole or electron traps can result in either an enhancement or a decrease in the photoluminescence. In addition, we report our results concerning chemical purification and preliminary surface characterization of MoS2 clusters by chromatography.

The structural, optical and related properties (i.e. photoluminescence, photoconductivity etc.) of some natural three- and lower-dimensional semiconductor systems based on metal halides are briefly reviewed and some new results are reported. A blue shift of the excitonic bands was observed by decreasing the dimensionality or the size of the materials active part. The results are similar to those obtained from conventional semiconductor systems (e.g. GaAs, CdS, PbI2) by decreasing artificially the dimensionality or decreasing the size.

AgBr nanocrystals with radii R ≈ 2.5 to 5 nm comparable with the exciton Bohr radius are produced in inverse micelles. As compared with bulk AgBr, the exciton emission exhibits a substantial blue shift and enhanced intensity due to the spatial confinement. Besides luminescence, first- and second-order resonance Raman scattering is discovered the occurrence of a zero-phonon process clearly revealing mixing of L with Г point states. From time-resolved measurements, exciton lifetimes of the order of 500 μs are found. They are close to the radiative lifetime in the bulk demonstrating that nonradiative processes are negligible in these nanocrystals. In agreement with state mixing, the total decay rate is found proportional to R−2. A small shift of the TO(L) Raman line with excitation photon energy is analyzed in terms of the wavevector dependent interaction of the quantized exciton states with the dispersive phonon.

Previous work on quantum confined AgBr microcrystals has been limited to either large (500-1000 Å diameter) aqueous samples or extremely small (60-100 Å diameter) inhomogeneously sized micellar samples. This has prevented an investigation of the effect of size restriction on the phonon assisted transitions of the indirect exciton. A non-aqueous preparation was developed, capable of providing more homogeneous size distributions of AgBr microcrystals in the 100-400 Å diameter range. Low temperature luminescence studies were performed to determine the changes in energy and intensity associated with these excitonic transitions. The radiative lifetimes of some of the excitonic processes were measured in order to obtain a more global perspective on the photophysics of AgBr.

Hematite Fe2O3, is a semiconductor, and its electrical properties are highly sensitive to the preparation methods and purity. Its precursors can be hydrated ferric oxides; however, these are usually obtained from poorly defined ferric gels, which are obtained by hydrolysis of an aqueous solution of a ferric salt by a base. We have designed a novel synthetic route to ferric hydroxide, by reaction of a peroxo-compound with an aqueous solution of a ferrous salt, which involves simultaneous oxidation of Fe(ll) to Fe(lll), and hydrolysis, in the same reaction process. The two kinds of ferric hydroxide are highly different, however, both give hematite by dehydration/recrystallization, however, the way this occurs for each is different.

The formation of porous silicon (PS) by electrochemical dissolution of bulk Si is described by a new model involving quantum mechanical calculations of the tunneling probability of holes through small crystallites (< 60 Å) into the electrolyte. This tunneling probability shows oscillations as a function of crystallite size. The presented model calculations are in agreement to the microstructure of p-PS — deduced from Raman measurements — as a function of etching parameters and substrate doping level.

Understanding the evolution of porous silicon (PS) layers at the early stages of growth is important for determining the mechanism of PS film growth and controlling the film properties. We have used X-ray reflectivity (XRR) to determine the evolution of layer thickness and interfacial roughness during the growth of thin PS layers (< 200 nm) prepared by electrochemical anodization. The porous layer grows at a constant rate for films as thin as 15 nm indicating a very short incubation period during which the surface may be electropolished before the PS structure begins to form. Interface roughness measurements indicate that the top surface of the film remains relatively smooth during growth while the roughness of the PS/silicon interface increases only slightly with film thickness. The XRR results are compared with results obtained from the same films by cross-sectional transmission electron microscopy (XTEM), atomic force microscopy (AFM) and gravimetry.

Porous silicon superlattices (PS-SL) were investigated by transmission electron microscopy (TEM), scanning electron microscopy (SEM), spectroscopic ellipsometry and reflectance spectroscopy. Type I superlattices were formed by periodically changing the current density during the etch process. The quality of the interface between layers of different porosity depends on the quality of the silicon substrate. Optical components such as Bragg reflectors or Fabry-Perot filters were designed using the optical data of single porous layers. A good long term stability of the layers is achieved by using thermal oxidation steps. Type II superlattices were formed on substrates with layers of alternating doping level. The more complex layer structure of these superlattices is explained by the selectivity of the etch process on the doping level.

Using special electrochemical etching and lift-off steps, we have fabricated large-area freestanding porous silicon films in the thickness range from 0.1 μm to 50 μm. Their transmission is near 100% in the near infrared which is indicative of very high porosity/low index of refraction films. These optically flat and homogeneous films exhibit no surface and bulk scattering, despite the fact that they did not undergo supercritical drying. The relationship between the absorption coefficient, the luminescence spectrum, and the chemical and structural properties is examined as a function of preparation and post-treatment conditions. Because of their superior optical properties, these films are suitable for many device applications.

The formation of porous silicon layers was examined with respect to crystal orientation and the presence of water. Unlike the etching of (111) silicon in aqueous HF solutions, no pores were formed in HF-MeCN. The etching of (111) silicon in HF-MeCN resulted in the formation of triangular pits defined by the {111} planes. A mechanism for the etching is proposed where steric hindrance of the surface terminated hydrogens induces bond strain and enhances the chemical reactivity. The mechanism also accounts for the formation of pores in (100) silicon, and the formation of a highly branched microporous structure when silicon is etched in an aqueous solution.

A new type of HF solution, HF-acetonitrile (MeCN), has been employed to produce 10-30 μm thick porous silicon (P-Si) layers by photoelectrochemical etching of different types of Si wafers, Si(100), Si(111) and polycrystalline Si, with different resistivities. A combined optical, surface and nuclear microscopic assessment of these P-Si layers was performed using photoluminescence (PL), Raman scattering, X-ray photoelectron spectroscopy (XPS) and Rutherford backscattering spectroscopy (RBS). The PL emission intensities, Raman line shapes and structural features are strongly dependent on the properties of the substrates such as the crystallinity and resistivity of the Si wafers used for forming P-Si. With increasing resisitivity of the Si(100) wafers, the resulting P-Si layers show a slight blue-shift of their visible light emission peak energy, an up-shift of the peak position and a narrowing of the band width of the dominant Raman band, and a decrease in the amount of residual elemental Si on the surface. Those Si(l 11) wafers, etched in HF-MeCN, showed no porous structures and no visible light emission.

In previous work we reported that porous silicon (PS) films formed using a dilute HF:HNO3 chemical etch on polycrystalline, implant damaged single crystal, or amorphous starting material have luminescent characteristics that differ from PS fabricated on single crystal silicon1. Polycrystalline and implant damaged porous silicon exhibits brighter luminescence compared to single crystal silicon etched under identical conditions. No photoluminescence is detected from the porous amorphous silicon. In this work these effects are examined using HF:NaNO2 solutions with freely available NO2. The accelerated etching effects from work damage are reduced, and the PS from polycrystalline and implant damaged silicon luminesce with the same intensity as the PS from single crystal silicon. Again, etched amorphous silicon does not luminesce. TEM and EDX porosity measurements are used to determine the differences in structure and etching characteristics between the luminescent and non-luminescent materials.

Electrochemical anodization in the transition regime, between porous silicon formation region and electropolishing region, of monocrystalline silicon was investigated. Using this process bright and stable photoluminescence could be obtained on a very large range of substrate resistivities: p=12-0.005 Ωcm for p-type silicon and p = 20-0.001 Ωcm for n-type substrates.

Photoluminescence spectra, Fourier Transform IR (FTIR) absorbance and X-Ray Diffraction (XRD) measurements are reported. Investigations showed that anodic silicon suboxide was formed on the surface. The porous structure obtained in the transition regime is suggested to consist of silicon crystallites built inside an anodic oxide.

Porous silicon samples prepared in the dark under "gentle" etching conditions clearly demonstrate effects of quantum confinement, such as a correlation of the photoluminescence peak energy with the downshift of the Raman line from 521 cm−1 for bulk silicon, and a blue shift in the remaining weak photoluminescence after thermal annealing. On the other hand, samples prepared under illumination as well as those heavily etched in the dark, though luminesce brightly, show no significant effects of quantum confinement, suggesting a different dominant mechanism for the observed luminescence.

Photoluminescence (PL) of porous silicon after rapid thermal oxidation was studied using excitation energies of 5.0 and 6.4 eV. The emission spectra in both cases are dominated by a broad blue band centered at 430-450 nm and a much weaker red band positioned at 680-720 nm. Using atomic force microscopy (AFM), we have found a correlation between blue PL intensity and increase in feature size caused by progressive oxidation. The blue luminescence, found in porous Si only after oxidation, is identical, with respect to spectrum and fast decay, to that of thermally grown oxide on crystalline Si. Based on these findings, we conclude that the blue luminescence originates from SiO2 rather than from Si nanocrystals embedded in the oxide matrix. The red luminescence, on the contrary, has a different origin. The intensity of red band is a function of oxidation temperature and at RTP above 1050 °C, the red band is completely eliminated. The red luminescence characteristics are discussed in relation to the possible mechanism.

We have demonstrated a strong, room-temperature, 1.54 μm emission from erbium-implanted at 190 keV into red-emitting porous silicon. Luminescence data showed that the intensity of infrared (IR) emission from Er implanted porous Si annealed at ≤ 650°C, was a few orders of magnitude stronger than Er implanted quartz produced under identical conditions, and was almost comparable to IR emission from In0.53Ga0.47As material which is used for commercial IR light-emitting diodes (LEDs).

The strong IR emission (much higher than Er in quartz) and the weak temperature dependency of Er in porous Si, which is similar to Er3+ in wide-bandgap semiconductors, suggests that Er is not in SiO2 or Si with bulk properties but, may be confined in Si light-emitting nanostructures. Porous Si is a good substrate for rare earth elements because: 1) a high concentration of optically active Er3+ can be obtained by implanting at about 200 keV, 2) porous Si and bulk Si are transparent to 1.54 μm emission therefore, device fabrication is simplified, and 3) although the external quantum efficiency of visible light from porous Si is compromised because of self-absorption, it can be used to pump Er3+.

Proper surface passivation is critical for achieving stable, efficient PL from light-emitting porous silicon (LEPSi). As-anodized LEPSi is passivated by hydrogen which desorbs at a temperature as low as 400 °C. For device purposes, it is necessary that porous Si can tolerate at least 450 °C for post anodization annealing/metallization steps. We have established that, if the hydrogen at the surface is substituted by oxygen, the resulting Si-Ox passivation is significantly more stable. One way of achieving this is to implant low energy/low dose oxygen to form a thin coating of SiO2 on the surface. Post implantation FTIR data report the absence of Si-H peaks. XPS data indicate the formation of nearly stoichiometric SiO2 at the surface. Similar results were achieved by implanting with nitrogen to form Si3N4. As an alternative to implantation, we have deposited thin capping layers of SiO2, Si3N4 and SiC by plasma-enhanced chemical vapor deposition (PECVD) which resulted in a similar degree of passivation. Wafers were pre-treated at 400 °C to remove hydrogen from the surface. Finally, we carried out a low-pressure CVD (LPCVD) oxide deposition on LEPSi. Post implantation/CVD annealing was done at temperatures up to 600 °C. In most cases, little or no change was observed in the resultant PL intensities.