Showing papers on "Silicon published in 1968"
TL;DR: In this article, piezoelectroreflectance in Ge, Si and GaAs was studied for uniaxial stress effects on electronic energy bands in three different energy bands.
Abstract: Piezoelectroreflectance in Ge, Si and GaAs studied for uniaxial stress effects on electronic energy bands
594 citations
394 citations
TL;DR: In this paper, the orientation dependence of the backscattered yield of 1.0-MeV helium ions has been used to investigate the lattice characteristics of silicon and germanium implanted at room temperature with 40-...
Abstract: The orientation dependence of the backscattered yield of 1.0-MeV helium ions has been used to investigate the lattice characteristics of silicon and germanium implanted at room temperature with 40-...
255 citations
TL;DR: In this article, the ion dose needed to form amorphous silicon as a function of the temperature of the silicon during bombardment was studied and it was shown that the bombarded regions exhibit various hues until eventually when they become completely amomorphous they appear "milky" due to the Rayleigh scattering of light.
Abstract: Thin films of single crystals of both n-type and p-type silicon contain disordered zones ∼ 50 A in diameter, discernible in the electron microscope after bombardment with ∼ 1013 Ne+ ions cm−2. As the dose increases these zones become more numerous until eventually (≳ 1014 ions cm−2) they overlap, creating a continuous surface layer which electron diffraction shows to be amorphous silicon. The individual zones disappear on annealing between 400 and 500°C, and at ∼630°C the amorphous layer recrystallizes epitaxially upon the underlying silicon, leaving an array of dislocation loops and dipoles. The bombarded regions exhibit various hues until eventually when they become completely amorphous they appear ‘milky’ due to the Rayleigh scattering of light. This appearance has been used as a convenient method to study the ion dose needed to form amorphous silicon as a function of the temperature of the silicon during bombardment.
202 citations
TL;DR: The surface recombination and generation process in silicon and germanium is studied theoretically and experimentally in this article, showing that the two surface systems are quite similar and that the recombination-generative centers are uniformly distributed in energy in the central region of the forbidden gap.
195 citations
TL;DR: In this paper, the chemical shifts for 29Si in seven series of molecules of the type XnSiY4−n have been measured where Y is an alkyl group and X varies widely in electronegativity.
Abstract: Chemical shifts for 29Si in seven series of molecules of the type XnSiY4−n have been measured where Y is an alkyl group and X varies widely in electronegativity. A considerable amount of proton and...
176 citations
TL;DR: In this article, the channeling of ions with through thin (0.24-1.5 μ) silicon samples has been studied for ion energies ranging from about 100 keV to 500 V.
Abstract: The channeling of ions with through thin (0.24-1.5 μ) silicon samples has been studied for ion energies ranging from about 100 keV to 500 keV. The effects of radiation damage and sample misorientat...
174 citations
TL;DR: The Hall mobility of holes in silicon p-type inversion layers has been measured as a function of gate voltage (perpendicular electric field), inversion layer orientation (100, (110), and (111) surfaces), direction of current flow within an inversion, and temperature as discussed by the authors.
Abstract: The Hall mobility of holes in silicon p‐type inversion layers has been measured as a function of gate voltage (perpendicular electric field), inversion layer orientation [(100), (110), and (111) surfaces], direction of current flow within an inversion layer, and temperature. It has been shown that hole mobility in silicon inversion layers depends not only on the crystalline orientation of the inverted surface, but also on the azimuthal direction of current flow within the inversion layer. Thus, on the (110) silicon surface at room temperature, the inversion‐layer hole mobility is 40% higher in the[l10] direction than in the [001] direction. Room‐temperature piezoresistance tensors have been experimentally determined for inversion layers on the (100), (110), and (111) surfaces of silicon. It is found that, in general, the piezoresistance coefficients are not the same as those for the same directions in bulk silicon and that they depend on the orientation of the surface. The existence of these anomalous effects can be understood in terms of quantization of the carrier wavefunction in the surface channel, which is narrow compared with the carrier wavelength in bulk silicon. This quantization tends to depopulate that part of the Brillouin zone within which k⊥, the component of wavevector perpendicular to the surface, is small. If the dependence of energy on k is not quadratic, as is true for the valence band of silicon, the effective masses for conduction in the plane of the surface can be complicated functions of k⊥. The mass anisotropies estimated from the cyclotron resonance parameters for the valence band of bulk silicon are in qualitative agreement with the experimental mobility values.
173 citations
TL;DR: The average energy expended for electron-hole pair generation in silicon and germanium lithium-drifted detectors by gamma rays, electrons, and alpha particles has been measured as a function of temperature as discussed by the authors.
170 citations
TL;DR: In this paper, the forward and reverse I-V characteristics of Schottky barrier diodes with PtSi contacts and diffused guard rings were investigated and the forward current was shown to follow the expression I f = I s exp (qV/nkT) over eight orders of magnitude in current with I s = 10−12 A and n = 1.02.
Abstract: Metal-semiconductor diodes with near-ideal forward and reverse I-V characteristics have been fabricated using PtSi contacts and diffused guard rings. Typically, for a device with an area of 2.5 × 10−6 cm2 made on an n-type (111) oriented, 0.35 ohm-cm silicon epitaxial substrate, the forward current follows the expression I f = I s exp (qV/nkT) over eight orders of magnitude in current with I s = 10−12 A and n = 1.02. The reverse breakdown is sharp and occurs at the theoretical breakdown voltage of p+ n silicon junctions of the same n-type doping. The premature breakdown observed in nearly all previous Schottky barrier diodes has been shown to be caused by electrode sharp-edge effects. Besides giving sharp breakdown voltage, the guard ring also eliminates anomalously high leakage currents, yet still retains the fast recovery time characteristic common to other Schottky barriers. Typically, the recovery time measured at 10 ma is less than 0.1 ns, the resolution of the measurement.
162 citations
TL;DR: In this article, it was shown that the Schottky barrier height depends on the particular method of surface preparation, and subsequently shows a slow change with time, reaching a steady value over a period of days or even weeks.
Abstract: Measurements have been made of the height of Schottky barriers obtained by evaporating metal films on to n-type silicon. In the case of surfaces prepared by chemical methods, the height of the barrier initially depends on the particular method of surface preparation, and subsequently shows a slow change with time, reaching a steady value over a period of days or even weeks. This final value is independent of the method of surface preparation, but depends on the choice of metal. For junctions made by evaporation on to silicon cleaved in an ultra-high vacuum, the barrier height shows no ageing and is substantially independent of the metal. These observations can be explained in terms of the existence of a thin oxide layer on the chemically prepared surfaces, together with the assumption that the density of surface states is about two orders of magnitude lower on the chemically prepared surfaces than on the cleaved surfaces.
TL;DR: In this article, the Hall theory of the pin-rectifier is extended by considering the diffusion currents in the heavily doped regions, and a general solution with low-level injection in the exterior p - and n -regions and with high-level injections in the middle region is derived.
Abstract: The Hall theory of the pin -rectifier is extended by considering the diffusion currents in the heavily doped regions, and a general solution with low-level injection in the exterior p - and n -regions and with high-level injection in the middle region is derived. At small current densities, this general solution passes into the Hall solution: for high current densities an approximate quadratic I–V dependence is found. The dependence of the forward characteristic within this range upon the properties of the heavily doped regions and of the middle region is discussed in detail.
IBM1
TL;DR: In this paper, a mathematical analysis on the measurement of an impurity atom distribution in silicon by the differential capacitance technique is presented, and it is shown that the measurement is applicable only in regions of semiconductor material exhibiting charge neutrality.
Abstract: A mathematical analysis is presented on the measurement of an impurity atom distribution in silicon by the differential capacitance technique. This analysis shows some inherent errors that can arise when the technique is applied to material containing a small impurity atom density. An important conclusion is that the differential capacitance measurement establishes the distribution of majority carriers, rather than the distribution of impurity atoms; therefore this measurement technique is applicable only in regions of semiconductor material exhibiting charge neutrality.
TL;DR: In this paper, a replication technique for electron microscopy has been developed which is capable of detecting 3 A steps on surfaces, which is used to observe the growth of silicon on a (111) silicon surface in ultrahigh vacuum.
Abstract: A replication technique for electron microscopy has been developed which is capable of detecting 3 A steps on surfaces. This technique was used to observe the growth of silicon on a (111) silicon surface in ultrahigh vacuum. On a substrate surface that is ``clean'' by current standards of silicon surface study, three‐dimensional growth centers developed. However, on a cleaner surface consisting of a continuous layer of freshly deposited silicon, growth proceeded by a step‐motion mechanism. By the use of a simple model to calculate the critical supersaturation ratio and the equilibrium adatom concentration, and from the observed absence of nucleation on atomically flat terraces between steps, a surface self‐diffusion coefficient of 10−3 cm3/sec at 800°C was estimated. Straight steps which propagate in the 〈112〉 direction were observed, in contradiction of the simple bond model of step edges which predicts straight steps that propagate in the〈112〉 direction. Step arrays and step separation were observed ...
TL;DR: In this article, the detailed morphology and crystallography of silicon in aluminium-silicon alloys are described and the various structures are rationalized in terms of the variables, composition, freezing rate and temperature gradient.
Abstract: Aluminium-silicon alloys in the composition range 12 to 16 wt.% silicon have been frozen unidirectionally over a range of rates from ca . 0.3 to 30 μ m/s and with imposed temperature gradients ranging from 0.35 to ca . 40°C/mm. The detailed morphology and crystallography of silicon in these alloys are described and the various structures are rationalized in terms of the variables, composition, freezing rate and temperature gradient. Three growth processes are distinguished in which (A) massive silicon crystals grow at a planar aluminium front by a relatively long range diffusion process; (B) short range diffusion occurs at the growth front between silicon crystals which develop a preferred fibre texture; and (C) silicon crystals are more or less heavily twinned and grow by steady or fluctuating short-range diffusion processes. There do not appear to be any preferred epitaxial orientations between the phases. Similarities are noted between the microstructures of eutectic alloys in the systems Al-Si and Ag-Si, and it is shown that in chilled or ‘modified’ alloys the silicon occurs in an irregular fibrous form rather than as isolated globules. The mechanisms are discussed for (I) the growth kinetics of silicon, (II) the transition from a long-range to a short-range diffusion process, (III) steady-state, and (IV) non-steady-state growth processes, and (V) repeated nucleation of silicon crystals from the liquid.
TL;DR: The experimentally observed distributions of surface state density versus energy is correlated with the existence of coulombic centers in the oxide as mentioned in this paper, where single charges give rise to peaks in surface state densities close to the band edges, deeper levels are introduced by two charges in close proximity.
Abstract: The experimentally observed distributions of surface state density versus energy is correlated with the existence of coulombic centers in the oxide. Single charges give rise to peaks in surface state density close to the band edges, deeper levels are introduced by two charges in close proximity.
TL;DR: In this article, the production and annealing behavior of divacancy and the $A$ center in fission-neutron-irradiated silicon was studied by infrared absorption, using the 1.8-, 3.9-, and 12-\ensuremath{\mu} bands.
Abstract: The production and annealing behavior of the divacancy and the $A$ center in fission-neutron-irradiated silicon was studied by infrared absorption, using the 1.8-, 3.9-, and 12-\ensuremath{\mu} bands. The production rate of the divacancy was found to be high, about 5.7 ${\mathrm{cm}}^{\ensuremath{-}1}$, and to be enhanced by the presence of boron (\ensuremath{\sim}2\ifmmode\times\else\texttimes\fi{}${10}^{17}$ atoms per ${\mathrm{cm}}^{3}$), but not by the presence of oxygen (\ensuremath{\sim}1\ifmmode\times\else\texttimes\fi{}${10}^{18}$ atoms per ${\mathrm{cm}}^{3}$). The annealing of divacancies in neutron-irradiated Si required an activation energy of 1.25 eV, as in electron-irradiated Si, indicating that most of the divacancies were removed by diffusion to sinks. The annealing results also indicate that the local defect concentrations in the damaged regions can be as high as \ensuremath{\sim}${10}^{20}$ defects ${\mathrm{cm}}^{\ensuremath{-}3}$, in which the divacancies still retain their individual properties as far as their infrared absorption and annealing properties are concerned. The production rate of the $A$ center was found to be extremely low. The near-edge absorption band was also studied. About 95% of the near-edge band disappeared upon annealing in the same broad temperature range as did the divacancies. From these results, it was concluded that the majority of the total volume in localized damage regions produced by the fission-neutron irradiation of silicon is rich in divacancies and is still crystalline.
TL;DR: In this paper, the surface effects on metal-silicon contacts have been studied in detail using gate-controlled Schottky-barrier diode structures and it was shown that the generation current in the depletion region often constitutes a significant part of the total reverse current and cannot in general be neglected.
Abstract: Surface effects on metal‐silicon contacts have been studied in detail using gate‐controlled Schottky‐barrier diode structures. The ``excess'' forward and reverse currents at small and moderate bases, respectively, and low breakdown voltages usually associated with metal‐silicon contacts are clearly shown to be due to high fields near the corner region when the surface is accumulated. These currents can be eliminated and breakdown voltage increased by depleting or inverting the surface. By studying the diode characteristics when the surface is inverted, it is shown that the generation current in the depletion region often constitutes a significant part of the total reverse current and cannot in general be neglected.
TL;DR: In this article, two special open-tube furnace systems, respectively containing all fused silica and all alumina furnace area components, and in which P is supplied through the reaction of wet H2 with AlP to form PH3, have been constructed for the growth of GaP crystals from Ga solution.
Abstract: Two special open‐tube furnace systems, respectively containing all fused silica and all alumina furnace‐area components, and in which P is supplied through the reaction of wet H2 with AlP to form PH3, have been constructed for the growth of GaP crystals from Ga solution. Careful attention has been paid to the purification of the PH3 and H2 transport gas. Unwanted contamination from the flow tubes and furnace area has been minimized. The crystals are then suitable for the controlled study of the optical properties of the impurities C and Si, persistent residual impurities in GaP crystals grown under less stringent conditions. The residual concentrations of S and N have also been dramatically reduced in this system. Clearly defined chemical evidence indicates that carbon rather than silicon is the shallow acceptor (ionization energy EA ∼48 meV) in the residual green donor‐acceptor pair luminescence spectrum in GaP. Analysis of sharp line donor‐acceptor pair transitions observed at the high‐energy tail of a broad red luminescence band (peak energy ∼1.96 eV) characteristic of silicon‐doped crystals indicates that silicon is a deep acceptor (EA∼204 meV) on P lattice sites and a shallow donor (ED∼80 meV) on Ga lattice sites in GaP. Sharp line green spectra observed for recombinations at C–Si and Zn–Si pairs confirm this value of (ED)Si. The shallow green pair transitions involving Si donors are strongly coupled to ∼17.5 meV and ∼29 meV low‐energy phonons, unlike pair recombinations involving P site donors. This marked difference in phonon coupling invalidates judgment of the relative values of ED for S, Te, and Si donors from the relative positions of the peak intensities of the green or red pair spectra involving these donors. Small shifts in the no‐phonon discrete pair lines between crystals doped with Si28 and Si30 confirm the suggested role of Si in these spectra.
TL;DR: The electrical properties of vapordeposited silicon nitride and silicon oxide films on silicon have been investigated in this article, showing that the silicon oxide properties are more similar to those of thermal oxides.
Abstract: The electrical properties of vapor‐deposited silicon nitride and silicon oxide films on silicon have been investigated. The silicon nitride films were produced by the reaction at 950°C, while the oxides were prepared at 400°C using the reaction. The properties were compared with those of thermal oxides prepared in dry O2 at 1200°C. As contrasted to the thermal oxides, the silicon nitride films are characterized by polarization and room temperature trapping instabilities, relatively high conductance, and high surface state charge densities. The vapor‐deposited oxides tend to resemble the nitrides in those properties which are associated with the silicon‐ dielectric interface, but the bulk properties are more like those of thermal oxides.
TL;DR: In this paper, the concentration profiles of 32P ions implanted into silicon crystals have been studied over the energy range 10 −110 kV using an electromagnetic separator, where the beam directions near the major channel were studied.
Abstract: Concentration profiles of 32P ions implanted into silicon crystals have been studied over the energy range 10–110 keV, using an electromagnetic separator. Beam directions near the major channeling ...
TL;DR: In this article, the Coulomb interaction of the free carriers results in lower energy gap in the heavily doped emitter than in the rest of the transistor, and the difference in the energy gaps is experimentally determined from the activation energy difference of the emitter-current and the ideal component of the base Current.
Abstract: Theoretical treatments predict higher injection efficiency for double diffused silicon transistors than the experimentally observed values. This paper shows that the discrepancy can be partly explained by the difference in the effective energy gaps in the emitter and base regions. Coulomb interaction of the free carriers results in lower energy gap in the heavily doped emitter than in the rest of the transistor. The difference in the energy gaps is experimentally determined from the activation energy difference of the emitter-current and the ideal component of the base Current. It is concluded that too much doping in the emitter lowers the transistor gain, increases the temperature dependence of the gain, and results in a higher excess noise.
IBM1
TL;DR: A correlation between the lattice parameter and the carbon concentration in otherwise pure samples of silicon has been observed as discussed by the authors, and typical values of carbon concentration are given for some float-zoned silicon crystals and for a series of intentionally carbon-doped crystals.
Abstract: A correlation between the lattice parameter and the carbon concentration in otherwise pure samples of silicon has been observed. Carbon content has been determined by mass spectrometry, infrared, and lattice parameter measurements. Typical values of the carbon concentration are given for some float‐zoned silicon crystals and for a series of intentionally carbon‐doped crystals.
Patent•
08 Nov 1968TL;DR: In this article, the authors present a plan for the re-construction of the self-bonded SINTERED DIAMOND COMPACT with the help of agents from the INTERCONTINENTED PARTICLES.
Abstract: PROCESS INCLUDING PRELIMINARY CLEANSING AND ESSENTIAL PRECONDITIONING TREATMENT OF FINELY DIVIDED DIAMOND PARTICLES FOLLOWED BY COMPACTION OF THE PRECONDITIONED PARTICLES AT HIGH TEMPERATURES AND PRESSURES IN THE DIAMOND STABLE REGION TO PRODUCE DENSE SELF-BONDED SINTERED DIAMOND COMPACT. INCORPORATION OF FORON, SILICON OR BERYLLIUM AS SINTERING AID AGENTS WITH THE PRECONDITIONED PARTICLES STILL FURTHER PROMOTES SINTERING AND BONDING OF THE COMPACT.
TL;DR: The morphology of the solid-liquid interface and the contact angle configuration of the liquid alloy droplet determine the direction of growth of crystals prepared by the vapor-liquid-solid (VLS) technique.
Abstract: The morphology of the solid‐liquid interface and the contact angle configuration of the liquid alloy droplet determine the direction of growth of crystals prepared by the vapor‐liquid‐solid (VLS) technique. There are four different processes by which both growth kinks and branches can be formed. A change in solid‐liquid interface shape during VLS caused by a lateral temperature gradient results in the formation of growth kinks. Branches are formed if the alloy droplet ruptures during the kinking sequence. A sudden increase in temperature can cause an unstable contact angle configuration. The alloy droplet may run down the side faces of the growing crystal, leading to the formation of growth kinks or branches. A sudden decrease in temperature may cause "pinching off" of small droplets from the main droplet, giving rise to branches. Finally, the codeposition of liquid‐forming impurities may also lead to branch and kink formation. The proposed models have been verified experimentally for VLS growth of silicon and germanium. Crystalline defects, such as dislocations, are not essential for the branching and kinking process. It is shown that "growth shaping" during the VLS process is possible.