Index of content:
Volume 29, Issue 12, 01 December 1958
29(1958); http://dx.doi.org/10.1063/1.1723012View Description Hide Description
The shape and extent of space charge regions in semiconductor surfaces are briefly discussed. Curves describing the potential as a function of distance inside the semiconductor are obtained by numerical integration of the Poisson equation. The curves apply to all semiconductors. They show the effects of different resistivities and the increased depth of the space charge regions expected in pure III‐V compounds over those encountered in Ge and Si.
29(1958); http://dx.doi.org/10.1063/1.1723013View Description Hide Description
It is shown that the reduction of a metal oxide such as nickel oxide by hydrogen gas is strongly influenced by nuclear irradiation. The initial induction period is shortened and the subsequent reduction accelerated. The effects anneal out at higher temperatures. A probable mechanism of this acceleration is suggested.
29(1958); http://dx.doi.org/10.1063/1.1723014View Description Hide Description
The emf of a galvanic cell which consisted of a stressed metal electrode and an unstressed metal electrode was measured. The experimental data showed that for low torsional stresses there was a parabolic relationship between the emf and the torsional stress.
It was observed that the stressed electrode was more anodic with respect to the unstressed reference electrode in agreement with the majority of the previous investigations.
Experimental data of emf and torsional stresses for B & S gauge 12, 14, and 16 hard‐drawn copper wires are presented.
29(1958); http://dx.doi.org/10.1063/1.1723015View Description Hide Description
The occurrence of dislocation sites along grain boundaries is shown by transmission electron microscopy. Small thin sections of aluminum for these studies were obtained by electrolyticoxidation, with subsequent dissolution of the oxide. Dislocation sites along grain fragments in cold worked aluminum have been observed which, by Cottrell's formula, indicate grain orientation differences greater than 2° of arc. This method of examination makes possible the observation and analysis of dislocations in metals where boundaries are at other than extremely low angles of crystal misorientations. Also, it has been found useful for following the movement of dislocations within grains of the material and their migration to grain boundaries.
29(1958); http://dx.doi.org/10.1063/1.1723016View Description Hide Description
The dielectric relaxation spectra of a system of lithiumborosilicate glasses have been investigated over the frequency range 0.01 cps to 500 kc and temperatures between 25°–350°C. The spectra of the samples exhibit well‐defined peaks in the dielectric loss factor and dispersion in the dielectric constant. The activation energies associated both with the lithium ion mean relaxation frequency and the lithium ion dc conductivity (zero frequency conductivity) are constant with temperature; their difference for each sample appears to be an indication of the distribution of activation energies of the sample. The experimental data show that an analytical function proposed by K. W. Wagner adequately describes the distribution of relaxation times associated with the lithium ion relaxation process in the glass network. The distribution originates from a spread in activation energies for low lithium concentrations and a spread in the period of vibration of the lithium ions about their equilibrium position for higher concentrations.
29(1958); http://dx.doi.org/10.1063/1.1723017View Description Hide Description
By modulating a beam strongly with a high‐level signal, called the ``pump'' signal, a low‐level space charge wave of another frequency imposed on the beam can be parametrically amplified. With the pump frequency twice the signal frequency, 13‐db gain per plasma wavelength has been observed in a movable cavity experiment. Over a 10‐in. length of beam the signal has been observed to increase 41 db. When the imposed signal wave's frequency is not exactly half the pump frequency, the beam generates another wave called the ``idler'' such that ωsignal+ωidler=ωpump with both signal and idler growing exponentially with distance along the beam. In another experiment with the pump frequency lower than the signal frequency, the signal was observed to increase by 30 db over a 9.2 in. length of beam.
29(1958); http://dx.doi.org/10.1063/1.1723018View Description Hide Description
A technique of controlling the shape of detonation wave fronts in high explosives by inert wave control inserts was applied to generate discrete ultrahigh‐velocity pellets. Tests to determine the most suitable pellet shape as well as the optimum charge configuration are described, and velocities up to 7600 m/sec were realized for 0.95‐g aluminum pellets. The mechanism whereby pellets are accelerated by ``shaped'' waves is discussed, and the conclusion is reached that a simple model based upon the transmission of shock from the detonation wave to the pellet is not applicable.
29(1958); http://dx.doi.org/10.1063/1.1723019View Description Hide Description
An analytical model, which is applicable to static fatigue of lime glass, has been extended to account for dynamic fatigue of the same glass. The model successfully predicts the room temperature strain rate sensitivity of the failure process in lime glass and indicates a method by which the stress concentration relationship, applicable to microscopic flaws on glass surfaces, may be obtained by experiment. Resulting experiments showed that the stress concentration relationship proposed by Inglis is valid.
29(1958); http://dx.doi.org/10.1063/1.1723020View Description Hide Description
An arc cathode spot in a transverse magnetic field may move either in the Amperian (forward) direction, or in the opposite (retrograde) direction. Experiments are described which show that forward continuous motion is caused by magnetic fields at the cathode surface, as is the case with retrograde motion, and that both of these motions depend upon the material and surface condition of the cathode. Forward cathode velocity both when continuous and when moving in a stepping process is independent of arc current over a wide range.
The theories so far suggested for retrograde motion do not appear to be able to account for the dependence of retrograde cathode movement on the nature of the cathode surface.
29(1958); http://dx.doi.org/10.1063/1.1723021View Description Hide Description
Thermometric observations have been made of the phenomena accompanying the stabilization and the recombination of free nitrogen atoms in a solid molecular nitrogen matrix at liquid helium temperatures. Condensation from the gas phase may occur at temperatures considerably above the 4.2°K coolant temperature as the result of the normal change in heat content and the heat of atom recombination. Estimates, made from the latter effects, of the free atom concentration in the gas phase were as high as 4 atom %. Qualitative confirmation of the presence of an energetic species stabilized in the solid phase was made by detection of heat release upon warmup. From the latter effect a minimum nitrogen atom concentration in the solid of 0.03 atom % was estimated. For the conditions used the initiation of atom recombination occurs at a temperature of about 9°K. The upper limit of about 36°K for stabilization coincides with the known thermal transition in solid nitrogen. The data suggest that at least above 9°K diffusion of nitrogen atoms occurs throughout the solid matrix during deposition from the gas phase. If recombination during warmup is diffusion‐controlled, unusual dependence upon temperature is indicated.
29(1958); http://dx.doi.org/10.1063/1.1723022View Description Hide Description
The operation of a pulsed solid‐state Maser consists essentially of alternate periods of spin inversion and stimulated spin emission. Inversion can be accomplished by a short driving pulse or an adiabatic rapid passage. In the emission phase a weak signal field causes the spin system to spiral back coherently and radiate. The emissive phase can be treated by equations which for finite relaxation times can not be solved in closed form, but which are amenable to machine computation. Such numerical solutions giving the transverse moment, output field, output power, and power gain as a function of time were computed for various relaxation times, input fields, and values of the oscillation parameter. The results are interpreted with respect to applications of both regenerative and superregenerative modes of operation.
29(1958); http://dx.doi.org/10.1063/1.1723023View Description Hide Description
Dislocations introduced into single crystals of germaniumgrown by the Czochralski technique have been studied. It was found that most of the dislocations do not originate during growth, but are produced after growth as a result of thermal shock. Thermal stresses resulting from nonuniform temperature gradients during growth seem to have little or no effect on the dislocation density and distribution.
29(1958); http://dx.doi.org/10.1063/1.1723024View Description Hide Description
The range of applicability for the solution of the linearized diffusionequation is discussed for a particular case—heating of a homogeneous tube by internal energy generation. The range is defined by comparisons of series solutions of the linear and nonlinear diffusionequations (constant material propertiesvs variable material properties). Two sources of errors in the constant‐property solution are evaluated and a criterion for the limits of applicability of the linear solution is derived.
29(1958); http://dx.doi.org/10.1063/1.1723025View Description Hide Description
A low‐temperature dislocation creep model of the exhaustion type is proposed to account for the temperature independent activation energies which are found in recent experiments. Logarithmic creep can be obtained from the model if a Gaussian distribution function of Frank‐Read sources is assumed. The theory is applicable under easy glide conditions but should not be used when extensive double glide occurs. The theory can be extended into the liquid helium temperature range with the aid of Glen‐Mott quantum‐mechanical dislocation tunneling.
29(1958); http://dx.doi.org/10.1063/1.1723026View Description Hide Description
A dimensional change that simulates magnetostriction can be expected when a porous or spongy ferromagnetic body is placed in a magnetic field. Such an effect has been observed. Polycrystalline cobalt with a density less than 50% of normal developed a strain of −139×10−6, whereas cobalt of normal density developed a strain of −28×10−6. This finding appears to open the way for the development of materials with enhanced values of strain in a magnetic field.
29(1958); http://dx.doi.org/10.1063/1.1723027View Description Hide Description
The design and analysis of a bimodal cavity for the observation of microwaveFaraday rotation is presented. An equivalent circuit of lumped elements is developed and the coupling between degenerate cavity modes is expressed in terms of the elements of the susceptibilitytensor of the material producing the rotation. The theory is checked against experimental results with a paramagnetic salt and substantial agreement is obtained. A cavity of this type when used in conjunction with superheterodyne detection appears to provide a high sensitivity spectrometer for the observation of magnetic resonance.
29(1958); http://dx.doi.org/10.1063/1.1723028View Description Hide Description
In work reported earlier, the average length‐to‐diameter ratios determined magnetically did not agree with those observed under the electron microscope; this was thought due to magnetic interactions. This paper describes further shape determination measurements on γ‐Fe2O3 samples of increasing dilution by the original method and by two analogous methods described by Wohlfarth. The results show a slight increase of mean length‐to‐diameter ratio with dilution and a considerable broadening of the distribution toward large length‐to‐diameter ratios. The different methods of measurement described do not yield the expected congruent curves at very low concentration. This indicates a deviation from Stoner‐Wohlfarth rotation in these particles. The details of the mathematical methods are also described.
29(1958); http://dx.doi.org/10.1063/1.1723029View Description Hide Description
Noise measurements have been performed on oxide‐coated cathodes at frequencies of 8 Mc and of 30 cycles. At 8 Mc, where thermal noise and shot noise predominate, the measurements show thermal noise at high cathode temperatures (pore conduction region) and at low cathode temperatures (grain conduction region) with a pronounced noise peak caused by shot noise in the temperature range where pore conduction switches over to grain conduction.
Noise measurements at 30 cycles where flicker noise (with a 1/fnoise spectrum) predominates, indicate that the pore noise corresponds to the flicker noise expected for a semiconductor with a voltage‐independent conductivity, whereas the grain noise corresponds to the flicker noise expected for a semiconductor with voltage‐dependent conductivity. The measurements also show that the pores are inherently noisier than the grains at low frequencies. The results agree with a qualitative theoretical analysis.
29(1958); http://dx.doi.org/10.1063/1.1723030View Description Hide Description
Alternating electric fields can modify the time distribution of luminescent recombination in sulfide‐type phosphors with the expenditure of less electrical energy than is contained in the fluctuating component of the emitted visible radiation. Within the range of excitation and electrical signal input conditions where this is true, the phosphor sample may be regarded as an amplifier deriving its power from incident radiation, which can convert an applied electrical signal into an output light signal with energy gain and frequency fidelity.
29(1958); http://dx.doi.org/10.1063/1.1723031View Description Hide Description
A nonuniform magnetic configuration for rapid rotational flux reversal in square‐loop ferrite cores is analyzed. It is shown that the demagnetizing effects due to the boundary of the toroid are, to a good approximation, negligible for toroids with cross‐sectional radii as small as 2.5×10−2 cm. A basic assumption of an earlier rotational model of flux reversal was that these demagnetizing fields could be neglected. The predictions of this modified model at high drives are similar to those of the earlier model. The possibility of correlating H c (the intercept of the extrapolated linear region of the inverse switching time versus drive field plot) with more basic parameters through a model of this type is indicated.