The x-ray diffraction from very thin buried semiconductor layers (Angstrom level) and from quantum wires are reviewed. The structural properties like layer thickness, chemical composition, lattice strain and the onset of strain relaxation are investigated analyzing the experimental data by using a computer simulation based on a dynamical diffraction model. The experimental diffraction patterns on quantum wires exhibit satellite peaks which are analyzed by using a kinematical diffraction model. An elastic strain relaxation of the quantum wires which results in an orthorhombic unit cell deformation is observed.

The need for higher resolution in the study of materials has led to the development of sharper probes for imaging of the geometrical and surface structures. This development has led to the Scanning Tunnelling Microscope (STM), and variations thereof such as the atomic force microscope. Simultaneously there is a strong need for the possibility of making spectroscopic investigations of nm-structures. In this paper we describe results from spectral analysis of photons emitted as a consequence of tunnel-injection of minority carriers in semiconductor bulk or Quantum-Well (QW) samples. We denote this techtiique Scanning Tunneling Luminescence (STL), which is yet a new type of tunneling microscopy. Two different types of STL-experiments will be described, differing in the type of tip being used. In the first experiment we use a metallic tip to inject electrons into p-type JnP, where the charge carriers recombine, partly radiatve1y. In the second experiment we employ a tip of a large band-gap semiconductor, p-type GaP, from which electrons tunnel into an n-type InP/GaInAs/InP QW sample. We report on the first observation of such minority carrier injection where the sharply defined carrier (hole) energy distribution of the emitting tip is employed for almost monoenergetic injection and on the observation at higher bias levels of how electrons are also being injected from the InP surface into the tip where they recombine radiatively.

The interaction between P and GaAs(100) surface has been studied by XPS, UPS, HREELS and LEED. The results show that, at room temperature, P is adsorbed on the GaAs surface to form amorphous P thin film. There is less than one monolayer of P atoms bonded to the Ga atoms of the substrate at the interface. The amorphous P overlayer covered on the top of GaAs results in 0.2eV lowering of the GaAs surface barrier. The thermal annealing at 100°C-300°C will cause most of the amorphous P desorbed, with some randomly distributed P-clusters left on the surface. High temperature annealing will make all the remaining P atoms interact with the substrate to form Ga-P bonds. The exchange reaction between deposited P and the substrate will take place successively to form GaAsP thin film when P is deposited on GaAs substrate at higher temperatures. -This film is suggested to be a promising passivating film for GaAs surface.

A fundamental study has been conducted on Fe78B13Si9 metallic thin film in order to understand the change of the properties and structures of the film as a function of Cr and NiMo additions. The seperate additions of Cr and Ni-Mo was successful in a way two new metallic thin films were produced. The composition of the new films are Fe77Cr2B1Si5 and Fe74Ni4 Mo3B17Si2. Therefore, the study was focused on evaluating the physical and magnetic properties of the films with respect to the Cr, and Ni-Mo additions. Furthermore, characterizations of the internal and surface structures of the films have been conducted by a Transmission Electron Microscope (TEM ) and a Scanning Tunneling Microscope (STM), respectively. A comparison between the internal and surface structures of the films was carried out on both amorphous and crystallined forms. As a result, a correlation between the properties and structures of the films is established.

We review our photoluminescence results concerning the structural disorder of the interfaces in GaAs/AlAs quantum well structures. In the highest quality samples structural disorder exists as monolayer—high islands. We show the types of possible luminescence spectra which can occur in quantum wells in which the bottom and top interfaces have differently-sized islands. It is shown how luminescence spectra are sensitive to both large and small island structures —whether they occur on the same interface or occur separately on the top and bottom interfaces.

We describe a polarization modulation technique employing a photoelastic modulator and present its implementation in the experimental setup of a photoluminescence (PL) and a photoluminescence excitation (PLE) experiment. In the PL experiment the technique is used to analyze the emitted light with respect to its polarization, whereas for the PLE the polarization of the exciting light is modulated, probing the polarization dependence of the absorption of the light. Since the modulation of the light is restricted to the polarization, the polarization dependence can be measured simultaneously with the PL or PLE intensity. The versatility and the sensitivity of the technique is exemplified by presenting results of polarized PL and PLE obtained on quantum wire samples grown on the vicinal (100) surface of GaAs by molecular beam epitaxy (MBE) that show a considerable anisotropy in the linear polarization for both the PL and PLE at low temperatures.

We have grown Si/Sii_Ger multiple quantum wells (r 8%) lattice-matched to silicon with well thicknesses between 3 and 20 nni using UHV-CVD. The sample parameters were obtained accurately with high-resolution X-ray diffraction ( rocking curves) and transmission electron microscopy. From an analysis of the band-edge related photoluminescence energies we find a blue-shift due to confinement for thin wells.

Photoconductivity (PC), photoluminescence (PL), and photoreflectance (PR) have been carried out on In_0.52Al_0.48As/In_xGa_1-xAs single quantum wells, in lattice matched and lattice mismatched composition. The unstrained (x_In equals 0.53) and the strained (x_In equals 0.60) samples have been grown by molecular beam epitaxy (MBE), with well thicknesses of 5 nm and 25 nm. Low temperature PL measurements have shown a narrow full width at half maximum (FWHM) for the unstrained samples, indicating a very good interface quality. In strained samples a broadening on the FWHM has been found, indicating a small degradation of the structure quality with the introduction of strain. With the PC and PR measurements we have been able to observe transitions between electron and heavy holes levels (E_iH_i) up to i equals 5 and also between the first electron and light holes levels (E₁L₁). We have then calculated the theoretical values of these transitions by solving the Schroedinger equation in a finite square well, using an envelope function approximation, an effective mass approximation, and including the effects of strain on the band structure and on the effective mass. For the lattice matched composition, the best fit is obtained for conduction band offset (Delta) Ec equals 0.50 +/- 0.05 eV, in agreement with the literature. For example, with x_In equals 0.60 composition the best fit is obtained for (Delta) Ec equals 0.55 +/- 0.05 eV, in agreement with theory which predicts that (Delta) Ec increases with indium content.

The low-temperature photoluminescence (PL) and photoluminescence excitation (PLE) spectra characteristics of Al_0.48In_0.52As have been studied under high pressure from 1 bar up to 92 kbar. We have obtained, for the first time, the (Gamma) -(Chi) crossover critical pressure P_c (approximately 52.5 +/- 0.5 kbar), the linear pressure coefficients (alpha) ^(Gamma ) and (alpha) ^(Chi ) (7.9 meV/kbar and -2.9 meV/kbar, respectively) at helium temperature. By measuring temperature and excitation intensity dependences of the PL spectra together with the PLE spectra, we have demonstrated that the low-temperature luminescence of the Al_0.48In_0.52As is not excitonic but due to (D degree(s), A degree(s))transitions with a relatively deep acceptor of 68 meV, which occurs in both the direct- and indirect-band gap. We suggest that the shallow donor ground state associated with the (Chi) - and (Gamma) -conduction bands seem to be tied quite rigidly to these conduction bands. Variations in the donor binding energies with the pressure and the direct-indirect crossover seem to be minor.

Raman spectroscopy is used to study the incorporation of Si in (delta) -doped GaAs layers via light scattering by local vibrational modes. This includes the analysis of Si monolayers embedded in GaAs. Raman scattering by excitations of the two-dimensional electron gas in (delta) -doped GaAs gives information on the subbands formed by the space charge induced potential well. In addition, the effect of additional lateral confinement on these subbands can be studied.

We present a systematic study of the optical and structural properties of InGaAs grown on InP:Fe by molecular beam epitaxy (MBE) and metalorganic molecular beam epitaxy (MOMBE). The properties of these films were measured using photoluminescence, photoreflectance, and micro-Raman spectroscopy and correlated to double-crystal x-ray diffractometry. Lineshape analysis of low temperature Fourier transform photoluminescence (FTPL) allowed the identification of four optical transitions; bound exciton, donor-to-acceptor pair, and two bands which are impurity or defect related. The band-gap energy to XRD composition, relationship demonstrates films off the lattice match composition are under biaxial strain. Room temperature photoreflectance (PR) with a complex Airy functional model were used to directly yield band-gap energies (light and heavy hole). The complex Airy lineshapes were applied to both intermediate electric field (Franz-Keldysh oscillations) and low field PR spectra illustrating that the complex Airy analysis represents a generalized treatment. We correlate the band-gap energies from the PR spectral fits to those determined from PL measurements. Micro-Raman spectroscopy was performed in the <100> and <011> backscattering directions to identify four phonon modes; InAs-like TO (226 cm^-1), InAs-like LO (233 cm^-1), GaAs-like TO (255 cm^-1), and GaAs-like LO (270 cm^-1), and one alloy disorder mode R^* (244 cm^-1) for InGaAs on InP. For all five Raman features, a linear relationship between the Raman frequency and composition was determined for near lattice matched conditions (0.04 < 1 - x < 0.52).

The pseudomorphic GaAs/In_xGa_1-xAs structures are of particular interest because of the high value of the In_xGa_1-xAs electron mobility and the good electron confinement in the ternary alloy. Raman scattering experiments were carried out to measure the optical lattice modes of a series of GaAs/In_xGa_1-xAs strained- layer superlattices grown by molecular beam epitaxy on (001) surface of GaAs substrates. The frequency shifts between commensurate and incommensurate layers give a quantitative determination of strain in each type of the strained layers. For this purpose, we have also measured the Raman phonon frequencies of bulk In_xGa_1-1As alloy samples upon a large scale of composition. Double crystal x-ray rocking curve (XRC) data on superlattices are compared to those of Raman experiments. This allows a more valid estimation of the obtained results.

Raman scattering has been used to probe the effects of both chemical and plasma etching on the photosensitive area of GaAs planar photoconductors, and also to examine the stress at a Si₃N₄ film/n-doped GaAs interface. Electrical performance on the photoconductors has been measured and then correlated to the Raman results. As expected, increasing the number of processes leads to slightly poorer electrical performance.

The use of Raman spectroscopy to study interdiffusion in SiGe/Si superlattices and multilayers has been considered. We have modeled the interdiffusion process using an error function concentration profile and modeled the Raman scattering spectrum as an integration over this profile. The calculation has been compared to experimental data. The Raman spectra were calculated by appropriately averaging over the diffused interface. We find good agreement between the calculated and measured spectra. The Raman spectra are shown to be sensitive to the interfacial profile.

Raman scattering and low temperature photoluminescence (PL) have been used to characterize conventional ion beam (CIB) and focused ion beam (FIB) implanted multiple quantum well (MQW) structures. Lattice damage induced by both CIB and FIB implantation is studied by analyzing the peak position and the lineshape of the Raman spectra. CIB and FIB induced lattice damage was similar at lower doses, but for the highest dose the FIB induced significantly larger damage. Following implantation and rapid thermal annealing, all samples are compositionally mixed. PL spectra suggest that CIB and FIB implanted samples with the lowest dose behave in a manner similar to a lightly compensated semiconductor.

We present results of a new contactless mode of electroreflectance (ER) which utilizes a condenser-like system. One electrode consists of a transparent conductive coating on a transparent substrate which is separated from the sample surface by a thin layer of air. We have measured the contactless ER (CER) spectra from a number of materials including semi- insulating bulk GaAs; epitaxial In_0.15Ga_0.85As; bulk Hg_0.8Cd_0.2Te (80 K and 300 K), a GaAs structure with a large, uniform electric field; and a GaAs/Ga_1-xAl_xAs (x approximately equals 0.2) coupled double quantum well. Some measurements were performed up to 500 K. The phase of the CER signal yields information about the nature of the band bending at the surface of bulk or epitaxial material. The relative merits of CER and photoreflectance are discussed.

The photoreflectance lineshape of the 11H exciton peak in an In_0.21Ga_0.79As/GaAs single quantum well was measured from 10 K to 500 K. In contrast to the GaAs/GaAlAs system our 10 K data exhibited no `wings' and can be fit equally well by the first-derivative of a Gaussian or Lorentzian profile. The former is probably relevant since the well material is an alloy. These results indicate that care must be taken in associating spectra without pronounced `wings' with the Lorentzian profile. The temperature dependence of the 11H linewidth, which has been fit to a Bose-Einstein expression, yields important information about the quality of the material and interface. The variation of the energy of 11H with temperature agrees with that of bulk material.

Room temperature photoreflectance (PR) has been used to study a series of AlGaAs/GaAs heterostructures, which have been shown by electrical measurements to contain a two- dimensional electron gas (2-DEG), with varying carrier concentrations, mobilities, and structures. Oscillatory features in the PR spectra observed above the GaAs bandgap energy cannot be simply interpreted as Franz-Keldysh oscillations (FKO).

This paper describes the application of a novel approach in electric field modulation spectroscopy, the differential photoreflectance (DPR), to study buried GaAs/AlGaAs heterostructures. We show that in most complicated device structures. DPR can provide selective photoreflectance (PR) response from layers buried within thin multilayer structures. Such responses are often superimposed on one another in conventional PR measurements. DPR measurement is achieved through alternative modulations from two laser pumps with different penetration depths, so that the modulation has a gradient with respect to the depth within the sample. The application of this technique is demonstrated for a two-dimensional electron gas (2DEG) of a modulation doped heterojunction in comparison with conventional PR. In one case the heterojunction of interest was buried under two highly doped GaAs and AlGaAs layers 40 nm thick. We show that this heterojunction is barely distinguishable in a PR measurement. Nevertheless, at room temperature DPR shows distinct peaked signals that correspond to the previously reported PR from a two-dimensional electron gas.

We extend our previous measurements of the lineshape of GaAs at the E₁ transition (2.9 eV). This study covers the combined effect of temperature and carrier concentration along with a discussion of the effect of the electric field intensity and the field inhomogeneity within a depth of 20 nm from the surface. A systematic study of changes in the lineshape of the above bandgap transition, E₁ as a function of temperature (80 - 400 K) and carrier concentration (CC) (2 - 200 X 10¹⁶cm^-3umps with different penetration depths, so that the modulation has a gradient with respect to the depth within the sample. The application of this technique is demonstrated for a two-dimensional electron gas (2DEG) of a modulation doped heterojunction in comparison with conventional PR. In one case the heterojunction of interest was buried under two highly doped GaAs and AlGaAs layers 40 nm thick. We show that this heterojunction is barely distinguishable in a PR measurement. Nevertheless, at room temperatutric field on the photoreflectance lineshape is discussed. The observed effect may be applied as an optical measurement of the electric field and the carrier concentration within a depth of about 17 nm from the surface/interface.

We report the first photoreflectance measurement of strain-induced piezoelectric field in a (111)B InGaAs/GaAs structure. The InGaAs quantum well was pseudomorphically grown in the undoped regions of a GaAs undoped-heavily doped structure. Four structures, two each with the same layer structures but different orientation, (111)B and (100), were used in this study. The electric fields in the undoped GaAs region were measured by Franz-Keldysh oscillations in photoreflectance. All the samples have a surface barrier height of about 0.7 eV. However, the measured electric field is 30% stronger in the (111)B sample compared to the (100) sample. We attribute this difference to the strain induced electric field in the (111)B sample. The piezoelectric field in (111)B strained In_0.15Ga_0.85As obtained in this measurement is 2.2 +/- 0.5 X 10⁵ V/cm, which agrees very well with theory.

The Fermi level position in low temperature (LT) GaAs is studied by photoreflectance (PR). The experiments show that the Fermi level in both the as-grown and the annealed LT-GaAs is firmly pinned, however, the pinning position occurs at different energies: 0.47 eV below the conduction band edge for the as-grown samples and 0.65 eV below the conduction band edge for the annealed samples. The pinning in the as-grown LT-GaAs is the result of a high degree of charge compensation of deep levels, while the pinning in the annealed LT-GaAs is due to the depletion of carriers by the Schottky barrier at the metallic As precipitates. From the measured Fermi level and ionization ratio of As antisites, the (0/+) donor level of the As antisite is found to be at E_c - 0.57 eV.

The ellipsometric technique has been around since the end of the last century. However, until the introduction of computers in the 1960s mainly nulling techniques were used. These were accurate, but extremely slow, and data analysis capabilities were limited. The use of personal computers permits rapid data acquisition and analysis. Thus the emphasis is now on ellipsometry for materials research and analysis. Ellipsometry is a unique tool in that it is totally non-destructive and can be performed in any ambient, including air, vacuum, high pressure, or liquid. In ellipsometry a polarized, collimated beam of monochromatic light is reflected at a preselected angle of incidence from the surface of a material, and the polarization state of the reflected beam is determined.1'2 For each wavelength and angle of incidence, one pair of numbers is obtained. These can be used to determine at most two materials parameters; such as the real and imaginary parts of the complex dielectric functions. Frequently, however, two or more variables are correlated and cannot be uniquely determined. This is an important caveat for ellipsometry: one always obtains numbers from the experiment, but they may not be correct. It is extremely important for the user to understand the basis of regression analysis, and the possibility for correlated variables when using ellipsometry for any new materials system.3 To minimize this possibility, multiple angles of incidence and wavelength are used to maximize sensitivity and uniquely determine parameter values. In general, ellipsometry determines the real and imaginary part of the optical dielectric function or equivalently the index of refraction and extinction coefficient. For filmed materials systems, one can potentially determine film thickness (with monolayer sensitivity), void fraction (density), alloy fraction, interfacial and/or surface roughness, and other microstructural properties.4 To do these sophisticated levels of microstructural materials analysis requires spectroscopic and variable angle measurements, in order to minimize correlation, increase sensitivity, and maximize total information obtained

Recent applications of spectroscopic phase modulated ellipsometry, from UV to IR, to the study of the growth of plasma deposited thin film semiconductors like amorphous (a-Si:H) and microcrystalline ((mu) c-Si) silicon are reviewed. The high sensitivity of this technique is emphasized. In the UV range, the ability of kinetic ellipsometry, with fast time resolution, to study the complex growth mechanism of (mu) c-Si is illustrated. In particular, the importance of hydrogen etching during (mu) c-Si growth is evidenced. In the IR, the hydrogen incorporation during a-Si:H growth can be precisely investigated. Photoelectronic quality a- Si:H films grow beneath a hydrogen rich overlayer (1-3 monolayers thick) containing SiH₂, the hydrogen being bonded as SiH in the bulk material.

Recent applications of spectroscopic phase modulated ellipsometry, from UV to IR, to the study of the growth of plasma deposited thin film semiconductors like amorphous (a-Si:H) and microcrystalline ((mu) c-Si) silicon are reviewed. The high sensitivity of this technique is emphasized. In the UV range, the ability of kinetic ellipsometry, with fast time resolution, to study the complex growth mechanism of (mu) c-Si is illustrated. In particular, the importance of hydrogen etching during (mu) c-Si growth is evidenced. In the IR, the hydrogen incorporation during a-Si:H growth can be precisely investigated. Photoelectronic quality a- Si:H films grow beneath a hydrogen rich overlayer (1-3 monolayers thick) containing SiH₂, the hydrogen being bonded as SiH in the bulk material. 2.2 +/- 0.5 X 10⁵ V/cm, which agrees very well with theory. thick. We show that this heterojunction is barely distinguishable in a PR measurement. Nevertheless, at room temperatutric field on the photoreflectance lineshape is discussed. The observed effect may be applied as an optical measurement of the electric field and the carrier concentration within a depth of about 17 nm from the surface/interface. phonon modes; InAs-like TO (226 cm^-1), InAs-like LO (233 cm^-1), GaAs-like TO (255 cm^-1), and GaAs-like LO (270 cm^-1), and one alloy disorder mode R^* (244 cm^-1) for InGaAs on InP. For all five Raman features stoichiometry, in terms of the RD response and the materials properties, which was recently demonstrated for VCE growth. Finally, we compare the quality and the character of real-time RD-detected growth oscillations as obtained for CBE, VCE, and MOVPE growth.

Differential reflectance (DR) spectroscopy is similar to other optical modulation techniques in so far as the resulting spectra exhibit sharp derivative-like lineshapes at photon energies corresponding to the critical point transitions. DR signals originate from inhomogeneities on or below the semiconductor surface. These inhomogeneities may be intrinsic, such as fluctuations in surface field, layer thickness, alloy composition, or externally induced, such as ion implantation, hydrogenation, etc. The DR spectra of semiconductor layer structures may be used to determine mole fraction, doping concentration, critical point energies, etc., much like photoreflectance (PR). In many cases, DR spectra have better signal-to-noise ratio than that of the PR spectra. In this report we discuss the application of DR in the study of doping inhomogeneities in GaAs as well as the use of DR to determine damage profiles in ion implanted GaAs.

Dielectric functions of strained In_xGa_1-xAs (x approximately 0.1, 0.2, 0.24) measured by spectroscopic ellipsometry are presented for the first time. Critical point transition energies obtained from differential spectra are correlated with photoreflectance measurements to provide information on layer thickness, composition, and strain.

The results of the first picosecond time-resolved measurements performed using a free- electron laser are reported. These results, obtained on an a-Si_0.5Ge_0.5:H film, are compared to those we have obtained on a-Si:H using conventional lasers. The properties of free-electron lasers that make them attractive for time-resolved spectroscopy are also reviewed.

High excitation luminescence spectra from a homogeneous electron-hole system have been investigated in undoped strained In_xGa_1-xAs/GaAs and unstrained In_xGa_1-xAs/InP quantum wells in magnetic fields up to 12 T. Clear excitonic effects are observed in the optical spectra of neutral, dense magnetoplasmas in the In_0.53Ga_0.47As/InP quantum well at low temperatures. A direct evidence that electrons and holes from the N-th Landau levels bind into magnetoexcitons is the suppression of the density dependence of transition energies between the uppermost occupied (j_e equals j_h equals N) Landau levels in the range of filling factors N < (nu) /2 < N + 1. The effect disappears with increasing temperature. Good agreement between experimental results and theoretical calculations is achieved by including both the inter-Landau-level coupling and screening effects. For the dense (n > 10¹² cm^-2) magnetoplasma with electron temperature of the order of the cyclotron energy, the Landau level splitting is described within the framework of a simple plasma approximation. The renormalization of the reduced effective mass has been carefully measured.

We have performed far infrared cyclotron resonance absorption saturation experiments of both electrons (GaAs/AlGaAs MQW structure) and holes (InGaAs/GaAs <111> strained- layer superlattice) with the Free Electron Laser at UC Santa Barbara, Calif. The Landau level lifetime is found to be laser power dependent at high powers, and is longer for electrons than for holes. Reasons are discussed. Proper analysis of the transmission data is crucial for strongly absorbing multi-layer structures to obtain meaningful results. Limitation of the 3-level model used to determine the lifetimes are also discussed.

Vacuum electroreflectance (VER), a new method for performing electroreflectance on a free standing surface, is introduced. In VER, the surface of the sample is in vacuum and does not come in contact with a dielectric or a metal. The sample is interferometrically aligned, and precise positioning is obtained by piezoelectric motors. Cooling to 80 K is accomplished by the Joule-Thompson effect, and temperatures up to 400 K are achieved by resistive heating. Comparison of data obtained at room temperature by VER, photoreflectance (PR), and electrolyte electroreflectance (EER), and at low temperature by VER ad PR on the same GaAs sample is presented. Also, the VER data is compared to the data obtained in air by the newly introduced contactless electroreflectance (CER) technique.

A new modulation spectroscopy, microwave modulated photoluminescence (MMPL), is described. Application of the technique to a well characterized semiconductor system, InP:Zn, in which the radiative recombination processes are not understood, allows interpretation of the resulting spectra. In the ordered ternary Ga_0.52In_0.48P MMPL provides information about both carrier transport properties and the extent of ordering. When applied to new materials, MMPL can aid identifying radiative recombination processes.