Library

Notes: Polarization Raman microprobe spectroscopy is used to study crystalline silicon heated to the melting point by a tightly focused cw laser beam, which is either fixed or scanned across the surface. By examining optical phonons in solid silicon, the real-time Raman spectrum monitors the progress of silicon flow during melting and the trench depth during melt-assisted etching. Raman peaks lie between 482 cm−1, which is the Raman shift for silicon uniformly heated to the melting point (1690 K), and ∼510 cm−1, which is the Raman shift for c-Si heated just to the melting point and probed by the same beam. During laser melting with a static laser, the Raman spectrum of scattered light with z(x,y)z¯ polarization has two peaks, while the z(x,x)z¯ spectrum has one peak. This shows that at the beginning of melting in vacuum by a static laser there is a central region with solid silicon floating in the melt, which is surrounded by hot solid material. Because of the flow of the molten semiconductor, the temperature profile changes, causing the Raman spectrum to change rapidly. Laser melting of c-Ge and thin films of silicon in vacuum is also studied, as is the melting of c-Si by a static laser in the presence of an inert buffer gas. The presence of an inert buffer is shown to affect the temperature profile very strongly during melting and also during laser heating at lower laser powers when no melting occurs. During scanned laser melting and etching, the Raman spectrum has a single peak using either polarization configuration. Raman analysis during melting of silicon by a scanning laser shows that the average temperature in the probed region is much higher when there is a gas-phase argon buffer present (and no etching) than when there is an etching mixture of argon/chlorine gas (and etching). Along with these real-time Raman measurements, the reflection of the incident laser was monitored, and post-process Raman analysis and profilometry were also performed to characterize surface changes due to melting and etching.

Notes: Photoluminescence near 4450 A(ring) (violet) and 5900 A(ring) (yellow) from ZnSe/Zn1−xMnxSe strained-layer superlattices (SLS) is studied as a function of applied hydrostatic pressure up to ∼90 kbar for x=0.23, 0.33, and 0.51 at 9 K. For each of the three SLSs, the peak energy of the near-band-gap violet photoluminescence (PL) from exciton recombination increases sublinearly with pressure, and closely follows that of ZnSe. The observed dependence is consistent with type-I band alignment, at least at pressures where the violet PL intensity remains strong. For the x=0.23 SLS, the energy of the biexciton PL increases faster than that of the exciton, which suggests decreased confinement. The yellow PL is due to the 4T1 →6A1 intraionic transition in Mn2+ and the nearly linear decrease of the peak energy of this signal with pressure is explained by crystal-field theory. In most cases examined, the intensities of the violet and/or yellow PL decrease abruptly above a certain pressure that increases with x from ∼65 to ∼90 kbar. For the different cases, this suggests either the occurrence of a phase transition in the ZnxMn1−xSe layers or a band crossing.

Notes: Cerium dioxide particles excited in air with sub-band-gap radiation emit very broad radiation in the visible spectrum above a threshold intensity that decreases with increasing ambient temperature. Concomitant with this emission is the near disappearance of the Stokes and anti-Stokes Raman scattering peaks. Both phenomena are reversible in air up to just above threshold, and are seen for nanoparticles and several-micron-diameter particles with particle diameter comparable to or smaller than the laser focus. Temperature estimates using the Stokes/anti-Stokes scattering intensity ratio suggest there is laser heating due to small intragap absorption and possible nonlinear processes, given the very slow thermal conduction. The broad emission in this loose powder may well be due to thermal emission, on the basis of spectral fitting of the high-energy part of the spectrum to a blackbody radiator at ∼1200–1400 °C, although luminescence from a new phase is a possibility. The sudden decrease in Raman scattering and increase in emission in air are consistent with a transition to a new, possibly luminescent, phase, as is the continued disappearance of the Raman peaks in forming gas when the laser power is reduced below the upstroke threshold. Oxygen point defects and their complexes may play an important role in many of these processes. © 2002 American Institute of Physics.

Notes: Optical emission (OE) actinometry has been used to measure the absolute densities of Cl2, Cl, Cl+, and Ar+ in a high-density inductively coupled (ICP) Cl2–Ar plasma at 18 mTorr as a function of the 13.56 MHz radio frequency (rf) power and Ar fraction. The fractional dissociation of Cl2 to Cl increases with rf power, with the dissociated fraction increasing from 78% to 96% at 600 W (10.6 W cm−2) as the Ar fraction increases from 1% to 78% due to an increase in electron temperature. Emission from Cl+* and Ar+* originates primarily from electron excitation of Cl+ and Ar+ (and not excitation of Cl and Ar), making actinometric determination of Cl+ and Ar+ densities feasible. For powers exceeding 600 W, the neutral (Cl2 and Cl) to ion (Cl+ and Ar+) flux ratio is found to be strongly dependent on Ar fraction, decreasing by a factor of ∼3.0 as the latter is increased from 13% to 78%. This dependence can be attributed mostly to the decrease in Cl density and relatively little to the small decrease in the total positive ion density from 1.8×1011 to 1.4×1011 cm−3, over the same range. OE spectroscopy is also used to estimate the rate constant for the dissociative excitation of Cl2 to the Cl (4p2D0J′=3/2,5/2) excited state with emission at 822.2 nm, yielding ∼10−13 cm3 s−1. © 2001 American Institute of Physics.

Notes: Three-dimensional arrays of organically passivated CdSe nanocrystals were investigated under hydrostatic pressure using photoluminescence (PL) and absorption spectroscopies. Interdot separations were varied coarsely by varying the organic ligand on the nanocrystal and finely by applying hydrostatic pressure. The PL and absorption spectra of solutions and arrays of CdSe nanocrystals capped by either tri-n-octylphosphine oxide or tri-n-butylphosphine oxide are the same up to 60 kbar, which suggests that they exhibit no interdot coupling since the interdot separations in the solutions (∼50 nm) are much greater than those in the arrays ((approximately-less-than)1 nm). While the variation with pressure is roughly that expected from the increase in band gap energy of bulk CdSe with pressure and the increase in confinement energies of electrons and holes with increased pressure, there is still a significant difference in the energy of the PL peak and the first exciton in absorption (the Stokes shift) for both these solutions and arrays that increases with pressure. This is attributed mostly to increased vibrational relaxation due to the movement of nuclei in the excited state. In contrast, there is a distinct difference between the pressure dependence of CdSe/pyridine dots in solution and arrays; the increase of the energy of the first exciton peak in absorption with pressure becomes markedly slower above about 30 kbar in CdSe/pyridine arrays, and is lower than that in the corresponding solution by ∼50 meV at 50 kbar and ∼70 meV at 60 kbar. Experiments with CdSe/shell/pyridine dots, with large electron and hole barriers, cast doubt on the mechanism of interdot electron and/or hole tunneling leading to a decrease in electron and/or hole confinement energy. Also, interdot tunneling of single carriers may be inhibited by the charge separation energy. The differences in the dielectric medium surrounding each dot in the solution and array explain their different absorption exciton energies at ambient pressure, but not the changes at elevated pressure. The observed loss of much of the pyridine ligands during array drying could be very significant, and contact between pyridine-capped dots at elevated pressure may be important. © 2001 American Institute of Physics.

Notes: UV-induced modifications in undoped metalorganic chemical vapor deposition grown GaN on sapphire are observed from 9 to 160 K. The photoluminescence intensities of bound excitons (3.476, 3.482 eV), the yellow band (2.2 eV) and the blue band (2.9 eV) change with time when a fresh sample is irradiated by 325 nm (He–Cd laser). The free exciton peak at 3.488 eV is unchanged by laser irradiation. Initially the blue and donor-bound exciton emission degrade rapidly and the yellow luminescence increases, each at the same rate. Later, the yellow luminescence degrades and the donor-bound exciton emission increases very slowly, at the same rate. Mechanisms are proposed that may explain the luminescence pathways and defects involved. © 1999 American Institute of Physics.

Notes: Raman scattering was used to study optical phonons in a Si12Ge4 strained-layer superlattice on c-Si(001) that was subjected to hydrostatic pressure at room temperature. The change of phonon frequency with pressure, dω/dP, for the principal quasi-confined LO mode in the Ge layers, is found to be significantly smaller than that for bulk crystalline Ge. This difference is shown to be due to the tuning of biaxial strain in the Ge layers and the pressure response of the confined mode as hydrostatic pressure is varied. Both strain and confinement make comparable contributions to dω/dP for the Ge layers in the superlattice examined here.

Notes: Thin films of Ge-Si alloys were deposited by 193 nm photolysis of GeH4/Si2H6 gas mixtures using an ArF laser. For substrate temperatures below 350 °C, deposition occurred only with the laser present, while for temperatures above 400 °C, film growth was little influenced by laser photolysis and resembled conventional chemical vapor deposition (CVD). The Si/Ge ratio in the films was about three times the PSi2H6/PGeH4 ratio of reactant partial pressures for deposition in either the laser photolysis or the CVD regime. This result indicates that there is strong cross chemistry between silicon and germanium-bearing species in the gas phase. Film stoichiometry was measured by Auger analysis and Raman spectroscopy, with both methods leading to the same film composition.

Notes: Micron-dimension structures of germanium-silicon alloys of various compositions are formed by several direct-laser writing techniques, including pyrolytic deposition of silicon from silane on laser-melted germanium substrates and codeposition from silane/germane mixtures on different absorbing substrates. In situ composition analysis of these polycrystalline alloy microstructures is performed by Raman microprobe analysis. The measured Raman shifts and widths of the laser-deposited alloys are found to be in better agreement with published Raman data on germanium-silicon alloy films than with the Raman data on alloy bulk solids. In the codeposition of alloys, the decomposition of germane to form Ge is observed to be about six times faster than the decomposition of silane to form Si, independent of the silane/germane ratio, laser power, and substrate type.