Volume 33, Issue 5, 01 May 1962
Index of content:
33(1962); http://dx.doi.org/10.1063/1.1728800View Description Hide Description
Following a report of ferroelectricity at room temperature in Na(Nb1−x V x )O3, crystals of pure NaNbO3 and Na(Nb1−x V x )O3 have been investigated electrically, optically, and by x‐ray diffraction. Attempts to introduce amounts of vanadium larger than x≈0.005 into the crystals were unsuccessful. The room temperature phase of pure NaNbO3 has been reported to be antiferroelectric, that of NaVO3 nonferroelectric but perhaps also antiferroelectric. The vanadium‐doped crystals appear to be antiferroelectric in the room temperature as grown state. A transition to a ferroelectric phase can be induced in both the vanadium‐doped and the pure crystals by a sufficiently high field (of the order of 50 kv/cm), called the transition field, applied normal to the orthorhombic b axis. In one type of crystal plate, when the field‐induced transition into the ferroelectric phase occurs, a crystallographic reorientation is also observed. Some crystals revert to the nonferroelectric phase when the transition field is removed.
All of the crystals investigated while in the ferroelectric phase were found to have the length of the b axis twice the length of the simple perovskite cube, whereas in the antiferroelectric phase, the length of the b axis is four times that of the simple perovskite cube.
In so far as the phenomena studied in the present investigation are concerned, it can be stated that the vanadium‐doped crystals behave in most respects like pure NaNbO3 crystals. In particular, no evidence was found for an unbalanced antidipole configuration, i.e., ``ferrielectric'' structure, proposed by another investigator in an earlier paper. Further differences between the content of that paper and the present results are pointed out.
33(1962); http://dx.doi.org/10.1063/1.1728801View Description Hide Description
The elastic displacement field around a uniformly moving edge dislocation has been obtained for the case where the dislocation glides on a (111) plane in a 〈110〉 direction in an anisotropic face‐centered‐cubic crystal. An equation is obtained for the velocity (the Rayleigh wave velocity) at which the shear stress on the slip plane of a moving dislocation is zero. Dislocations of like sign moving at velocities faster than this velocity attract rather than repel each other. It is concluded that when anisotropy is small and the elastic constant c 11 is smaller than c 12+2c 44 (the more commonly occurring relationship between the elastic constants) the Rayleigh wave velocity is increased above its value in an isotropic crystal and therefore, the extent of the velocity region where dislocations exhibit an anomalous behavior is decreased. For slightly anisotropic crystals the limiting velocity of dislocationmotion is the slower of the two shear velocities in the 〈110〉 direction.
33(1962); http://dx.doi.org/10.1063/1.1728802View Description Hide Description
Calculations have been made of the shear stress produced by an edge dislocation moving on its slip plane in aluminum,copper,silver,gold,lead, and nickel. It is found that, except in the case of aluminum, the anomalous velocity range in which dislocations of like sign attract one another is very small. It is concluded that the effect of anisotropy in decreasing the extent of the anomalous velocity range is much greater in face‐centered‐cubic than in body‐centered‐cubic metals.
33(1962); http://dx.doi.org/10.1063/1.1728803View Description Hide Description
Heat transfer intensity and current density distributions at the anode of high current arcs in predominantly inert gas atmospheres were determined experimentally. Measurements were made for stable, axially symmetric arcs having a small diameter refractory metal cathode and a plane, cooled copperanode. The experimental method consisted of splitting the anode,measuring the heat flux and the current to one of the sections as a function of arc position relative to the splitting plane, and calculating therefrom the distribution functions. The work encompassed the effects of gas pressure (195–790 mm Hg), cathode geometry and material (tungsten and W‐1% ThO2), electrode separation (1.6–12.7 mm), arc current (100–300 amp), gas composition (argon, helium, and diatomic gas‐argon mixtures), and localized constriction of the plasma column. Peak heat transfer intensities ranged from 1.0 to 20 kw/cm2 and peak current densities from 100 to 2500 amp/cm2. The heat and current distribution curves were of similar shape and were generally sharper than Gaussian. The heat distributions were partially resolved according to transfer mechanism. An upper limit was set on the average kinetic energy of electrons arriving on axis at the anode. This was used to specify limits on gas heat transfer intensities for certain conditions. The space charge sheath at the anode was estimated to be 10−4 to 10−3 cm thick at the arc axis.
33(1962); http://dx.doi.org/10.1063/1.1728804View Description Hide Description
Charge multiplication arising from carriers injected into p‐n junctions in gallium phosphide has been studied as a function of applied bias. From these and capacitance studies, the carrier ionization rate as a function of field has been determined. It is found to be very similar, both qualitatively and quantitatively, to the behavior in silicon. In particular, the mean free path for hot carriers between collisions involving optical phonon emission is of the order of 100 A. An additional result obtained from simple analysis of the photoresponse and capacitance behavior of the junctions is that the dominant carrier lifetime is 10−10 sec.
33(1962); http://dx.doi.org/10.1063/1.1728805View Description Hide Description
During the drive phase of the operation of a flux‐limited multipath core the flux change produced in the driven branch is the irreversible flux available to that branch plus a certain amount of reversible flux. Most of this flux will return through the shortest path in that part of the magnetic circuit of the core which constitutes the load for the driven branch. It is shown, however, that if the irreversible flux of some branch of this path saturates, the additional flux from the driven branch will produce an irreversible flux change in the next shortest path. This flux change reappears as noise during the reset phase. This fact leads to the surprising result that under some circumstances a greater flux change occurs in the longer branch of the load structure of a three‐rung laddic than in the shorter. This source of noise may in many cases be circumvented by reducing the cross section of the driven branch. These ideas are applied to explain the existence of optimum hold currents in the general laddic. A modification of the geometry of the three‐hole memory core which improves the signal‐to‐noise ratio in inhibited‐flux operation is described.
33(1962); http://dx.doi.org/10.1063/1.1728806View Description Hide Description
The diffusion of copper in polycrystallineAg2Se was determined from measurement of the rate of change of the resistivity profile of a partially copper‐plated ingot of the semiconductor material. The copper concentration profile that was determined from the resistivity profile, was found to agree very satisfactorily with a theoretical concentration curve computed from a normalized complementary error function. The room temperature diffusion coefficient of copper in Ag2Se was computed from these data and was found to be 4.7×10−8 cm2 sec−1.
33(1962); http://dx.doi.org/10.1063/1.1728807View Description Hide Description
The rapid evaporation or condensation of a vapor bubble when out of equilibrium compared to the slow dissolving or growth of an air bubble results in quite different propagation properties of sound waves through water containing one or the other types of bubbles. Adapting a method developed by Foldy in his treatment of the air bubble case, we derive an expression for the propagation number for sound waves in a ``vapor bubbly'' medium. As an application of the theory it is shown that as few as 10/cm3 vapor bubbles of radius about 10−3 cm will (for ambient pressure about one atmosphere) attenuate an ultrasonic (f=̇3·104 cps) sound wave by a factor of 10 in 10 cm. Suggestions are made concerning the use of these waves for the early detection of shut‐down bubbles following a power burst in a reactor.
33(1962); http://dx.doi.org/10.1063/1.1728808View Description Hide Description
Diffusion coefficients of Al26 and Mn54 in aluminum have been determined between 450° and 650°C. Low specific activities of the isotopes necessitated use of a thick layer technique. The exact solution to Fick's second law for the appropriate boundary conditions was used in treating the data. Temperature dependence of the diffusion coefficients may be expressed by the following equations:
33(1962); http://dx.doi.org/10.1063/1.1728809View Description Hide Description
The effect of adding the rare‐earth elements to cobalt,iron, and nickel has been investigated. Antiferromagnetic coupling was found to exist in the cobalt and iron rare‐earth compounds. The moment per formula unit of the compound Co5Y is increased when Sm, Pr, and Nd are substituted for yttrium and decreased when Tm, Gd, Tb, Er, Dy, and Ho are substituted. This is explained on the basis that although the spin moments of cobalt and rare earths are coupled antiferromagnetically in these compounds, the orbital moments add to the cobalt moment in the compounds containing Sm, Pr, and Nd, but they subtract from the cobalt moment in the compounds containing Tm, Tb, Er, Dy, and Ho. For the iron compounds investigated, the moment of the compound Fe4Y was found only to decrease when Tm, Er, Ho, and Dy were substituted for yttrium. The nickel compounds behaved entirely differently from the cobalt and iron compounds since very little or no antiferromagnetic coupling was observed in these compounds.
33(1962); http://dx.doi.org/10.1063/1.1728810View Description Hide Description
Measurements of the currents leaving thin films of uranium dioxide undergoing a known rate of fission have determined the secondary electron yield due to fission fragments emerging from the films. This yield was found to vary from about 570 to 300 electrons per fragment over a film thickness from 0.1 to 3 microns. The ratio of charge carried by the secondary electrons emitted from the surface to the charge carried by fission fragments was also determined. This ratio varies slowly with film thickness, from 28 for a 0.1‐micron coating to about 21 for a 3‐micron coating. The behavior of the charge ratio and secondary electron yield with coating thickness is in qualitative agreement with theory.
33(1962); http://dx.doi.org/10.1063/1.1728811View Description Hide Description
A review is given of the energy band structures in a number of relatively well understood semiconductors, including Si and Ge, the Group III–V compounds, the IV–VI compounds (lead salts), CdAs2, and Bi2Te3, and of the experimental methods which have been used in determining these band structures. This review is directed particularly toward the interests of those engaged in applied research and development of semiconductor devices which depend more or less specifically on band structure properties for their operation. It is also meant to provide a single source of the latest available energy band information for those engaged in basic research on semiconductors and for students.
Ferroelectric Harmonic Generator and the Large‐Signal Microwave Characteristics of a Ferroelectric Ceramic33(1962); http://dx.doi.org/10.1063/1.1728812View Description Hide Description
At microwave frequencies, ferroelectric ceramics behave as nonlinear dielectrics and exhibit large signal scalar nonlinear characteristics in both the dielectric constant and the rf conductivity. The nonlinear behavior of a ceramic of 73% BaTiO3‐27% SrTiO3 has been used to construct a harmonic generator operating from 3 kMc to 9 kMc. An efficiency of 8.5% for a peak input power of 2200 w was attained. In addition, a new measurement technique has been developed for measuring the large signal properties of this ceramic as well as for other nonlinear materials. This technique makes explicit use of the anharmonic response of a resonant system that contains a nonlinear element. The analysis of both the harmonic generator and the nonlinear resonant system takes into account nonlinear reactance as well as nonlinear resistance. For temperatures about 30°C above the Curie temperature (the Curie temperature for the ceramic composed of 73% BaTiO3‐27% SrTiO3 is approximately 20°C) the large signal dielectric constant and rf conductivity were found to be given by the expressions ε′(|E|)=ε0[2200−7.5×10−11|E |2] farad/m, and σ(|E|) =ωε0[160+29×10−12|E|2] mhos/m, where |E| is in volts per meter. These results were obtained at 3 kMc and are valid up to field strengths of 20 kv/cm.
Scattering of a Plane Wave by an Infinite Inhomogeneous, Dielectric Cylinder—an Application of the Born Approximation33(1962); http://dx.doi.org/10.1063/1.1728813View Description Hide Description
The scattering and absorption of an obliquely incident, plane electromagnetic wave by an infinite, cylindrically symmetrical region characterizable by a complex dielectric ``constant'' is determined by application of the Born approximation for the case of dielectric ``constant'' near unity. Differential‐scattering cross sections and absorption cross sections are derived as a function of angle of incidence, angle of polarization, and integrals over the inhomogeneous region. The expressions for the scattered fields are compared to the expansion of an exact solution for the homogeneous dielectric cylinder in the limit of dielectric constant near unity. The approximate expressions agree exactly. Of singular notability in this respect is the fact that polarization is altered in the Born approximation by the appearance of dyadic terms in the propagation equation in the region of inhomogeneity, while the exact solution requires the generation of cross‐polarized waves to satisfy boundary conditions.
33(1962); http://dx.doi.org/10.1063/1.1728814View Description Hide Description
The pulsed Townsend discharge technique for the determination of electron attachment coefficients of electronegative gases is examined for situations where the ion density is sufficiently high that space charge effects must be considered. The equations of motion for the ions are solved to yield a first‐order differential equation for a parametric function, in terms of which the current and boundary position may be expressed. For the particular case of interest, an initial distribution which decays exponentially from the cathode, an exact solution for the current is obtained by numerical evaluation of this function. Comparison with experiment shows excellent qualitative and fair quantitative agreement.
33(1962); http://dx.doi.org/10.1063/1.1728815View Description Hide Description
The ferroelectric and pyroelectric properties of mineral and synthetic colemanite were investigated. It is shown that the synthetic compound, prepared by the controlled dehydration of higher order calcium borates, has a much lower Curie temperature than mineral colemanite. The substitutional impurity concentration of strontium for calcium in natural colemanite is proposed to be primarily responsible for its higher Curie temperature. The occurrence of a double dielectric anomaly in some specimens of mineral colemanite is described and it is shown that the temperature at which the lower anomaly occurs is consistent with the ferroelectric transition observed in the synthetic compound while the upper anomaly is characteristic of the more common calcium‐strontium‐borate pentahydrates.
33(1962); http://dx.doi.org/10.1063/1.1728816View Description Hide Description
Fission fragment tracks observed in electron microscope studies of platinum and palladiumfilms are interpreted in terms of thermal processes involving melting and vaporization. The results require that the energy lost by the fragment be converted into thermal energy over the entire volume of the track in a time short enough that little of it is conducted away. A model in which free electrons disperse the energy over this volume is used to account for the width and structures of the tracks.
33(1962); http://dx.doi.org/10.1063/1.1728817View Description Hide Description
Amplitude dependent dislocation damping was measured in LiF. It was found that average dislocation velocities comparable to those present during plastic deformation occur under periodic stresses more than an order of magnitude below the macroscopic yield stress. This implies that the impediments to dislocation motion effective in plastic deformation are localized at widely separated points along the dislocation.
33(1962); http://dx.doi.org/10.1063/1.1728818View Description Hide Description
Trap depths and densities can be determined from the analysis of the dependence of space‐charge‐limited currents on applied electric field in insulators. Values obtained from space‐charge‐limited currentmeasurements are compared with those obtained from thermally stimulated currents and from photocurrent decay on the same crystal of cadmium sulfide. The results indicate that gross errors in trap depth and density can be made if it is simply assumed that the rapid rise in current toward the trap‐free curve with increasing applied voltage always occurs when the traps are filled. Although this may sometimes be true, it is also possible for such a rapid rise to result from field ionization of the traps, or even in certain cases from the onset of double injection. Even if the latter effects occur, however, proper analysis of the space‐charge‐limited current data permits an evaluation of trap depth and density which is completely consistent with independent measurements of thermally stimulated current and photocurrent decay.
33(1962); http://dx.doi.org/10.1063/1.1728819View Description Hide Description
The initial slope of the correlation function γ(r) is proportional to the specific surface of the scattering sample in principle, and is equal to zero in practice. In this paper, a simple model is investigated which illustrates the problem. Calculations are carried out to show how surface areas may be obtained from scattering data.