Volume 37, Issue 2, 01 February 1966
Index of content:
37(1966); http://dx.doi.org/10.1063/1.1708198View Description Hide Description
Moiré patterns produced by viewing a large number of spatially periodic light sources through an appropriate grid structure of similar periodicity are sometimes characterized by a spatial location away from the grid structures. This paper gives a simple geometrical obstruction theory of this effect. It also demonstrates that the results predicted are observable.
37(1966); http://dx.doi.org/10.1063/1.1708199View Description Hide Description
This paper describes an investigation of the origin of moiré patterns which occur in display‐type lowenergy electron diffraction apparatus. With both planar and spherical geometry, ``real'' multiple source moiré patterns have been observed. With planar geometry, quantitative comparison of experiment and theory for the average moiré fringe spacing and pattern location are shown to be in good agreement. Spherical geometry patterns can be understood qualitatively on the same basis.
37(1966); http://dx.doi.org/10.1063/1.1708200View Description Hide Description
Moiré patterns of the multiple source type can be used in the evaluation of line shape, line spacing, and grating plane separation of two nearly identical gratings. For line spacing analysis, the location of the MP indicates which of the gratings has the greater line spacing in a particular region. A difference in the average spacing of the order of one part in 107 between adjacent lines has been detected.
37(1966); http://dx.doi.org/10.1063/1.1708201View Description Hide Description
Radiative donor‐acceptor pair recombination between the deep‐donor oxygen and a shallow acceptor (zinc or cadmium) in p‐type crystals remains a dominant radiative recombination process in GaP at room temperature as it is at low temperatures. This mechanism is deduced from the spectral shift when one shallow acceptor is replaced by another, and from the dependence of photoluminescence intensities upon the simultaneous presence of both oxygen and the acceptor (zinc or cadmium). The minority carrier electron captured by the ionized donor is not readily released back to the conduction band thermally because the donor is quite deep. The thermal population of holes on the shallow acceptors is quite large even at room temperature, so that the trapped electrons can recombine with these holes. No radiative transitions between free holes and the trapped electrons are observed, implying that the capture cross section of a neutral oxygen donor for a free hole is <10−19 cm2. Since the widths of the pair bands involving oxygen are greater than the difference in binding energies between zinc and cadmium, the two bands overlap in crystals containing oxygen and both acceptors. The sum band exhibits no structure but its position depends on the zinc‐to‐cadmium ratio.
37(1966); http://dx.doi.org/10.1063/1.1708202View Description Hide Description
Photoluminescence spectra were obtained between 20° and 298°K for GaP crystals containing germanium alone or in combination with shallow acceptors (zinc or cadmium), or shallow donors (tellurium or selenium), or a deep donor (oxygen). Emission bands at 77°K at 1.942, 1.907, 1.823, 1.815, 1.38, and 1.59 eV were correlated, respectively, with the simultaneous presence of both germanium and zinc, germanium and cadmium, germanium and tellurium, germanium and selenium, germanium and oxygen, and germanium alone. The shift between the zinc‐germanium and the cadmium‐germanium bands was just the difference in binding energies between the two shallow acceptors, indicating that the emission was due to donor‐acceptor recombination between the shallow acceptor and a deep germanium donor level of binding energy ≤0.36 eV. Similarly the shift between the tellurium‐germanium and selenium‐germanium bands indicated pair recombination between the shallow donor and a deep germanium acceptor of binding energy ≤0.45 eV. The band associated with germanium alone corresponded to a pair band involving both donor and acceptor germanium, the maximum in the emission corresponding to a much smaller donor‐acceptor separation (18 Å) than for the bands involving shallow levels (52 Å) because the carriers are more tightly bound in the former case. The band at 1.38 eV was ascribed to a pair band between oxygen donors and germanium acceptors. Thermal quenching of the emission bands was predominantly due to thermal release of minority carriers after (efficient) capture by the ionized deeper level of the pair in the compensated samples.
37(1966); http://dx.doi.org/10.1063/1.1708203View Description Hide Description
Evaporated silicon oxidefilms can be made with reproducible electrical conductivity when the evaporation rate, the oxygen pressure, the substrate temperature during the evaporation, and the subsequent heat treatment are controlled.
In the high electric field region the conductivity follows a Schottky emission law and it is shown to be bulk limited. The conductivity is also shown to be a function of the degree of oxidation, and a variation in conductivity of a factor of 109 has been accomplished by varying the composition from SiO to SiO x , where x is between 1.5 and 2. At an electric field of 250 V/μ the variation is from 10−7 (Ω·cm)−1 to 10−16 (Ω·cm)−1.
A model for the conduction mechanism is proposed, where the rate‐limiting process is assumed to be Schottky emission from small silicon islands into a silica structure in which the silicon islands are imbedded.
37(1966); http://dx.doi.org/10.1063/1.1708204View Description Hide Description
The threshold energy for carrier multiplication Ei in Ge, Si, and GaAs is calculated theoretically. The theoretical values of Ei are obtained from a consideration of the semiconductor energy band structure along with energy and crystal momentum conservation in the ionization process. Calculated values are found to be consistent with experimental data.
37(1966); http://dx.doi.org/10.1063/1.1708205View Description Hide Description
In recent years, a number of power converters utilizing the pyroelectric effect have been reported in the literature, and attempts at computing their efficiency have been made. In this paper the maximum efficiency for power converters made with nonferroelectric pyroelectrics, with ferroelectricpyroelectrics having a first‐order ferroelectric‐paraelectric transition, and with ferroelectricpyroelectrics having a second‐order transition is computed. The last two cases have been evaluated starting from Devonshire's equations, which describe the properties of ferroelectrics better than other equations used in previous calculations of efficiency.
Magnetization of Synthetic Filamentary Superconductors. A. The Dependence of Flux Jumping on Temperature and Magnetic Field Sweep Rate37(1966); http://dx.doi.org/10.1063/1.1708206View Description Hide Description
Synthetic high‐field superconductors have been prepared by impregnating porous glasses of various pore size with indium, using pressures up to 70 000 psi. The magnetic fieldHfi at which the first flux jump occurred in the initial magnetization of the sample was measured as a function of temperatureT and magnetic field sweep rate dH/dt. Values of dH/dt between 24 and 375 Oe/sec were used. For a given value of dH/dt, Hfi increased initially as T was lowered, reached a maximum, and finally fell with T. For a given T, Hfi was reduced as dH/dt increased.
For large values of dH/dt, Hfi was almost independent of T and dH/dt. Hfi was reproducible to 50 Oe from cycle to cycle except for T <1.5°K, where the behavior became more erratic. After the sample was warmed to room temperature,Hfi was reduced. The samples showed hysteresis in the magnetization. Flux jumping occurred around the hysteresis loop; the values of field at which flux jumps occurred were reproducible from cycle to cycle. The field interval Δ between flux jumps had a temperature dependence similar to Hfi ; Δ is slightly reduced for large values of the applied magnetic field. The size of the flux jumps were also reduced for large values of magnetic field. For small values of T, the flux jumping behavior became erratic. Measurements of the critical field and transition temperature have been made using a low‐frequency mutual‐inductance technique. The critical field, that is the magnetic field for the normal‐superconducting transition, for a sample with a 65‐Å pore diameter was 25 770 Oe at 2.67°K. At a reduced temperature of 0.6°K, samples with pore diameters of 65, 117, and 250 Å have critical fields of 156, 102, and 39 times the critical field of bulk indium (176 Oe). The transition temperatures for the samples were 4.24°, 4.04°, and 3.61°K, respectively, compared with 3.408°K for bulk indium. The over‐all behavior of these samples was similar to inhomogeneous type II superconductors.
37(1966); http://dx.doi.org/10.1063/1.1708207View Description Hide Description
The photoemission properties of the (100) surface of lead sulfide prepared by cleaving, cleaving and annealing, and ion bombarding and annealing have been investigated in the region within about 2 eV of the threshold. For the cleaved and annealed, or ion bombarded and annealedsurface, the photothreshold is 3.8±0.1 eV giving an electron affinity of 3.4±0.1 eV. This value of electron affinity is lower than the 4.6 eV value previously reported and is found to be primarily dependent on surface preparation; the surface prepared in vacuum by cleaving and annealing or ion bombarding and annealing gives the lower threshold. Surfaces prepared by cleaving in air and then mounting in the vacuum apparatus have a higher threshold and lower yield than those prepared in vacuum. Subsequent anneal and/or ion bombardment of a surface prepared by cleaving in vacuum lowers the threshold and increases the yield even more significantly. Exposure of the freshly cleaved surfaces to oxygen at room temperature indicated a sticking coefficient of less than 10−4.
37(1966); http://dx.doi.org/10.1063/1.1708208View Description Hide Description
The author previously reported a variety of coincidence lattices for oxygen chemisorbed on single‐crystal faces of platinum by low‐energy electron diffraction. The purpose of the present work is to show that theoretical calculations confirm the stability of such structures. For the case of two one‐dimensional lattices it is shown that both a coulomb and a cosine interaction potential do not lead to any significant lowering of the energy for coincidence lattices. However, a variety of other potentials do favor the formation of coincidence lattices when the lattice parameters are in the ratio of small whole numbers. Since the Coulomb interaction leads to a potential which is virtually indistinguishable from the cosine potential for cases of interest, it is concluded that deviations from the cosine potential favor the formation of coincidence lattices. It is further concluded that in real cases such deviations exist.
37(1966); http://dx.doi.org/10.1063/1.1708209View Description Hide Description
The segmented‐rod ruby laser structure has been operated as a giant pulse laser by use of a Kerr cell to provide the required Q switching. The giant pulse output, peak power of 106 W/cm2 and pulse width of 30 nsec, has been analyzed by use of Fabry‐Perot etalons. The number of oscillating modes is shown to be similar to that previously observed in pulsed operation. Good agreement is found between the measured and computer‐calculated frequency differences. The computer calculations presented indicate that the best performance can be expected when the two laser rod segments are of approximately equal length.
37(1966); http://dx.doi.org/10.1063/1.1708210View Description Hide Description
Attenuation in single crystals of silicon was measured as a function of frequency with a previously described technique. It was confirmed that, as in the case of aluminum, this dependence, in the range of frequencies from 20 to 110 Mc/sec, is of the form α=α0 exp (Cf), where α is attenuation,f is frequency, and α0 and C are constants. Heat‐treating experiments indicate that the form in which oxygen is present in the material has a pronounced effect on the energy loss.Multiple scattering is proposed as a mechanism of loss in silicon and aluminumsingle crystals. An empirical model of multiple scattering is discussed.
It is confirmed that dislocations in silicon, at room temperature, do not contribute to the observed energy loss. Also it is proposed that the accepted relation for the evaluation of thermoelastic losses should be modified.
Observations of Dislocations in Ammonium Dihydrogen Phosphate: Production of Dislocation‐Free Crystals37(1966); http://dx.doi.org/10.1063/1.1708211View Description Hide Description
The results of etch‐pit and diffraction topographic studies on ammonium dihydrogen phosphate (ADP) are reported. These studies were carried out on specimens sectioned by abrasive wheel and string‐saw from large single crystals grown from aqueous solution. The two characterization methods are shown to be in approximate one‐to‐one correspondence in the case of sections produced by abrasive sawing. Absence of both etch pits and topographic images in sections made by string‐sawing of material well removed from the seed region is taken to indicate that this material is dislocation‐free. Preliminary results on the plasticity of ADP are reported.
37(1966); http://dx.doi.org/10.1063/1.1708212View Description Hide Description
The defects in hydrothermally grown synthetic quartz have been studied by x‐ray diffractiontopography.Linear defects which are apparently associated with the growth of the crystal are the predominant defects in z‐growth material. These defects which originate at inclusions or other defect sites on the seed surface are believed to be areas of high impurity content. If there are dislocations associated with the defects, their effect on the diffracted x‐ray intensity is affected by a surrounding cloud of impurity atoms. Defect densities range from 10 to 104 lines per square cm and the crystals with high densities of defects exhibit high acoustic loss at 5 Mc/sec and 300°K. The defects observed in synthetic quartz by etch, γ‐ray coloration, and polarized light are compared to those observed by x‐ray diffractiontopography.
37(1966); http://dx.doi.org/10.1063/1.1708213View Description Hide Description
A collecting factor for ions created at a given distance is computed on the basis of a simplified model and under the assumption of a Maxwell distribution of the initial velocities of ions. The factor is independent of the mass of the ion, but is a function of the temperature and takes values differing substantially from unity for usual dimensions of the collector.
The effect of the collector dimensions on the x‐ray‐limited pressure may be analyzed by using an approximate expression for the integral collecting factor.
Calculation shows that for very fine collecting wires the collecting factor is affected even by the extremely small momentum transferred to the helium ion from the electron during the ionizing collision. Experiments have confirmed the foregoing considerations, thus opening the way to the measurement of such momenta.
Calculations of the Conductivity of a Medium Containing Cylindrical Inclusions by the Method of Orthogonalized Particular Solutions37(1966); http://dx.doi.org/10.1063/1.1708214View Description Hide Description
Calculations of the conductivity of an infinite medium containing a square array of perfectly conducting cylindrical inclusions were carried out using the method of orthogonalized particular solutions. The results agree with those reported by Keller and Sachs using the method of finite differences.
37(1966); http://dx.doi.org/10.1063/1.1708215View Description Hide Description
A magnesium‐rich phase and a cadmium‐rich phase have been found to coexist for 25 and 30 at.% cadmium over temperature ranges of about 10 and 2 C°, respectively. With changing temperature, x‐ray diffractometer tracings of the (004) and (110) reflections showed a pair of peaks associated with each phase, one increasing in intensity while the other decreased. The appearance or disappearance of one pair of peaks established the temperature boundaries of the two‐phase region. Lattice parameters, calculated from the positions of these peaks, changed with temperature in a nonrectilinear manner which indicated maximum changes in composition of 2% and 1% for the 25% and 30% alloys, respectively. The sudden increase or decrease in the intensity of the (101) superlattice line confirmed the upper boundary of the region. Relative amounts of the phases, as determined from areas of the peaks, agreed with the relative amounts determined by applying the ``lever law'' to the revised phase diagram. These results suggest an ordinary first‐order phase transformation. Moore and Raynor reported two phases of the same composition for Mg3Cd. However, Kamath, Craig, and Wallace found phases of different composition for MgCd3.
37(1966); http://dx.doi.org/10.1063/1.1708216View Description Hide Description
X‐ray production by proton bombardment may be used to determine the surface density of thin films. It is desirable that the energy loss of the protons in the film be small and that the self‐absorption of the produced x‐ray be negligible. To illustrate the method, thin surface films of aluminum are discussed. The aluminum x‐ray yield (x rays/proton) as a function of surface density is presented for the range 0 to 115 μg/cm2 and at an incident proton energy of 100 keV. Under the conditions of a 10‐μA proton current incident on a thick target, a typical detection system would record 10 000 counts/sec with a counter signal‐to‐noise ratio of 104. At a signal‐to‐noise ratio as low as unity, the calculated surface density is 0.005 μg/cm2 or approximately 1014 atoms/cm2. Details of the calculation are presented, along with a table of eleven elements to which the method is currently applicable.
37(1966); http://dx.doi.org/10.1063/1.1708217View Description Hide Description
Direct observations of the growth of ice on, and perpendicular to, a water‐air interface are described, in an experiment reproducing the conditions of formation of natural lake ice. The final, preferred crystal orientation is affected by the initial orientations and by selective wedging cut of grains of certain orientations, but often the most important factor is grain boundary migration. Boundary migration is driven in part by interfacial energy and probably in part by stress. The origin of the stress is not known. Preferred intergrain orientations are found.