Volume 31, Issue 7, 01 July 1960

Diagrams Representing States of Operation of a General Thermocouple
View Description Hide DescriptionThe state of operation of a thermocouple of which (a) the bars have an arbitrary shape; (b) the properties of the materials are arbitrary functions of temperature; and (c) the composition is, under certain restrictions, inhomogeneous and anisotropic depends on three independent parameters: the current I and the temperatures T _{1}, T _{2} of the junctions. If T _{2} is kept constant, operating characteristics, such as curves of constant output power or efficiency, can be plotted in an I, (T _{2}−T _{1}) diagram. The existence of regions of generation of electricity and of cooling is proved. These regions are investigated. Possible generalization and reduction of the diagram are discussed. As an illustrative example, the cooling region of a general couple with temperature independent properties is dealt with.

Time Discrimination in Solid‐State Infrared Quantum Counters
View Description Hide DescriptionThe paper discusses various considerations in the practical realization of the infrared solid‐state quantum counter recently proposed by Bloembergen. It is suggested that a high degree of discrimination between the quantum to be counted, and the optical pumping flux, could be attained by a time‐sequence switching system; it is estimated that an isolation of ∼160 db might be required. The proposal demands that at least one of the levels in the crystal have a long lifetime, and a simple expression is derived for the overall efficiency of the detector in this case. An additional gain in efficiency and isolation would result if ``double pumping'' were used, with detection in the ultraviolet. In conclusion a suggested layout of components is presented incorporating the switching scheme discussed herein.

Power Flow and Stored Energy in Thin Electron Beams
View Description Hide DescriptionThe kinetic and electromagnetic components of ac power and stored energy are evaluated for space‐charge waves along thin drifting beams of simple geometry. It is found that: (a) when a modulated beam is decelerated, it radiates power into the surrounding space; (b) when both fast and slow waves are excited by a common source, the real kinetic power varies periodically with distance, exchanging energy with the electromagnetic field; and (c) when a finite‐area beam is current modulated with zero ac velocity, the total power need not be zero.
The energy transport velocity of a space‐charge wave is shown to equal its group velocity, when the time‐average stored energy is properly evaluated. A small portion of the stored kinetic energy propagates, together with the field energy, as an electromechanical wave along the beam. The larger part of the kinetic energy, which can be positive or negative, is transported by the motion of the beam itself.

Uniform Silicon p‐n Junctions. I. Broad Area Breakdown
View Description Hide DescriptionSmall area siliconp‐n junctions have been made which are free from exposed edges and dislocations passing through the space‐charge region. It is believed that the space‐charge regions of these junctions more closely resemble plane parallel geometries than any studied similarly hitherto. The avalanche breakdown phenomena in these uniform junctions are shown to be drastically different from those occuring in junctions that contain many dislocations. A comparison is made between the uniform junctions and one that is similar except that it possesses two breakdown‐inducing sites, probably dislocations. In the latter junction the reverse characteristic shows two slightly separated rapid increases in current which coincide, biaswise, with the formation of two isolated light‐emitting microplasmas, the occurrence of characteristic microplasma noise, and two singularities in the charge‐multiplicationcharacteristics. The uniform junctions show no such phenomena at intermediate voltages, breakdown occurring at a voltage roughly twice that at which the microplasmas form and which was previously thought typical for the given material resistivity. The light emission pattern accompanying breakdown in the uniform junctions appears more diffuse (giving rise to the term—macroplasma) than in nonuniform junctions where it normally appears as an array of intense local spots (microplasmas). It is concluded that microplasmas are not a necessary accompaniment of avalanche breakdown in silicon, but that they tend to occur where there are field or lattice inhomogeneities.

Uniform Silicon p‐n Junctions. II. Ionization Rates for Electrons
View Description Hide DescriptionCharge multiplication as a function of reverse bias has been studied in a number of uniform siliconp‐n junctions (junctions free from defects which promote local avalanche breakdown sites). From the multiplication characteristics new data have been derived for the field dependence of the ionization rate (α) for electrons. As previously found in junctions containing dislocations, α obeys the empirical relation α=α_{∞} exp (−b/E), where α_{∞} and b are constants and E is the field. Thus, this law is not simply a consequence of distortions to the junction introduced by dislocations or other microplasma‐inducing defects.
The ionization rates and breakdown voltages for the uniform junctions can be made consistent with older data (for nonuniform junctions) if the junction fields, as determined from capacitancemeasurements, in the new junctions are multiplied by 0.65. The most obvious interpretation of this experimental fact is that breakdown is made relatively easier at defects because the field is actually higher by a factor of 1.55. Possible causes of such an enhancement are discussed together with alternative hypotheses for the different ionization rates.

Propagation Mechanism of Germanium Dendrites
View Description Hide DescriptionThe role of the twin planes in the rapid dendritic propagation of germanium crystals is examined on the basis of corner nucleation. A model is proposed, and it is shown that at least two twin planes must be present for continued easy propagation in <211> directions, in good agreement with experimental observations. The absence of the dendrite with only one twin is explained.

Monte Carlo Calculation of Molecular Flow Rates through a Cylindrical Elbow and Pipes of Other Shapes
View Description Hide DescriptionA method is devised for the calculation of molecular flow rates through pipes where the mean free path for intermolecular collisions is large compared to the dimensions of the pipes. Results of the calculation are given for a straight cylindrical pipe, a cylindrical elbow, the annulus between two concentric cylinders, a straight cylindrical pipe with restricted openings, and a straight cylindrical pipe with restricted openings and a plate to block the direct beam between the openings.

Transport Equation for the Spectral Density of a Multiple‐Scattered Electromagnetic Field
View Description Hide DescriptionThe use of a first Born approximation is open to question when the path length is greater than a mean free path in the scattering region. It is therefore of interest to develop a transport equation capable of predicting the spectral density for such cases. The general theory presented in this paper is applied to the case of multiple scattering by dielectric noise. It is independent of models for the dielectric fluctuations. A method of solution is developed for the case of forward scattering. The particular case of a monochromatic plane wave incident of a half space is discussed in detail. The results are applied to a numerical example in the troposphere.

Small Angle Scattering Investigation of Ludox Mixtures
View Description Hide DescriptionSmall angle scattering measurements have been made on distributions of particles prepared by mixing two silicasols of about 90A and 190A mean particle diameter in varying proportions. The scattering at small angles arises predominantly from the larger particles, whereas the smaller ones control the scattering at large angles, as predicted by theory. Attempts to determine the size distribution by the graphical method of Jellinek et al. were unsuccessful, but Porod's result relating the scattering at large angles to the total surface area was verified. Bounds for the largest and smallest sizes were estimated from the experimental data by means of simple calculations based on the properties of the scattering function.

Charged‐Particle Orbits in Varying Magnetic Fields
View Description Hide DescriptionA solution for the paraxial orbits of charged particles in varying magnetic fields is given. The magnetic field is assumed to be azimuthally symmetric but is allowed to vary arbitrarily with time or axial distance. The instantaneous guiding center and radius vector of the orbit are given in terms of the initial guiding center and radius vector and two parameters which are determined from the solution of a first‐order differential equation containing the magnetic‐field variation. One exact and two approximate constants of the motion are evaluated and discussed.

On the Kinetics and Mechanism of the Precipitation of Lithium from Germanium
View Description Hide DescriptionThe kinetics of precipitation of lithium from germanium have been investigated and the following variables have been studied: (a) crystal source; (b) preannealing treatments prior to the introduction of lithium; (c) the effect of selective impurity additions, namely, oxygen and copper. The precipitation process can be described as following the law given below: .
Upon saturation at elevated temperatures, (about 425°C), n is found to be for the first part of the process and 1 for the later part of the process. Upon saturation at low temperatures, however, n is found to be ½. It is suggested that this results from the nonuniform distribution of lithium in the crystal prior to precipitation.
Nucleus number densities varying by orders of magnitude were obtained from different crystals. It was found that a preannealing treatment at 900°C was sufficient to increase the nuclei number densities for all crystals and the differences tend to disappear upon the preannealing treatment.
Doping with oxygen tended to increase very markedly the number density of nuclei whereas doping with small amounts of copper (several orders of magnitude less than lithium present) decreased the number density of nuclei by about an order of magnitude. Interpretation of these results suggests that an impurity, consisting either of particles or oxygen‐vacancy pairs, is catalyzing the precipitation process.

Beam Refrigeration by Means of Large Magnetic Fields
View Description Hide DescriptionThe resistive loading which an electron beam produces in an adjacent structure can be made to exhibit a very low noise temperature. This is achieved by coupling to the fast cyclotron wave in a large magnetic field; the noise temperature at a given signal frequency is shown to be equal to the cathode temperature times the ratio of signal frequency to cyclotron frequency.
An experiment is described in which this ratio is The coupling structure interacts with the fast cyclotron wave but rejects the slow cyclotron wave. A noise temperature of 186°K is measured.
In conclusion, it is shown that the large magnetic field required for beam cooling need not extend throughout the tube.

Growth of Atomically Flat Surfaces on Germanium Dendrites
View Description Hide DescriptionThis note presents a theory of the formation of extended flat surfaces on germanium dendrites. It is suggested that part of the liquid‐solid interface is a supercooled {111} surface, on which new atomic planes are occasionally nucleated. Thermal limitations on the nucleation and propagation of these monolayers are discussed. When such a growing layer reaches the melt surface it creates a step on the solid. It is proposed that the meniscus momentarily sticks to the corner of this step. Subsequent planes nucleated during this sticking period result in a step several atomic layers high, in accord with experiment. Elementary estimates of corner energy and amount of meniscus sticking yield results consistent with experiment.

Parametric Diode Figure of Merit and Optimization
View Description Hide DescriptionThe noise figure and gain expressions for the diode parametric amplifier are presented and discussed which include the effects of diode losses. From these relationships, a new diode figure of merit is defined as follows: f_{D}=a_{n} / 4πa _{0} ^{2} C _{0} R_{D} , where a_{n} is the normalized Fourier coefficient of the time dependent (pumped) capacitance;C _{0}, the diode capacitance at the operating bias point; and R_{D} , the equivalent series diode resistance. It is shown that this figure of merit is directly applicable in comparing diodes of various types as well as in optimizing their respective design for low‐noise amplifier use. Further, it is indicated how this figure of merit can be directly employed by the circuit designer to predict an amplifiers' noise figure performance including the effects of choice of bias, pump swing, and signal and idle frequency for a given diode.
A complete evaluation of the proposed figure of merit is made for the abrupt junction diode by expressing f_{D} in terms of device design parameters and examining as a function of bias and base doping. It is concluded, if no constraints are considered, that the maximum value of f_{D} is obtained for a bias approximately equal to half the breakdown voltage with full pump swing and that f_{D} (max) increases with decreasing base doping from 10^{18} to 3·10^{15} atoms/cc. It is further concluded that the maximum possible value of a _{1}/a _{0} ^{2} for the inverse square root of voltage capacitance law is approximately 0.557 such that the greatest value f_{D} (max) can attain is 20% of f_{c} , the diode cutoff frequency. Some possible limitations on this evaluation are discussed including the effects of diode deviation from model, pump power or diode dissipation restrictions, nonsinusoidal pumping waveform, and ambient temperature.

Use of the Omegatron in the Determination of Parameters Affecting Limiting Pressures in Vacuum Devices
View Description Hide DescriptionThe omegatron mass spectrometer has been used to determine residual gases in vacuum systems, including ion pump systems. The predominant residual gas in ion pumps was found to be methane, while the gases found in oil diffusion pump systems include water vapor and carbon monoxide. Analysis has been performed to determine the residual gases in vacuum tubes which are difficult to outgas, and conditions that lead to considerable hydrogen content are described. Examples of the use of the omegatron mass spectrometer as a tool in aiding solution of outgassing problems are given.

Low‐Temperature Transport Properties of Commercial Metals and Alloys. IV. Reactor Grade Be, Mo, and W
View Description Hide DescriptionThe thermal conductivity,electrical resistivity, Lorenz ratio, and thermoelectric power are given in the temperature range 4°–120°K for four reactor grade metals: beryllium with sample axis parallel to the pressing axis, beryllium with sample axis perpendicular to the pressing axis, molybdenum, and tungsten, the latter two doped with ThO_{2}. Within the given temperature range the molybdenum and tungsten show slight maxima in thermal conductivity, the beryllium does not. The various components of the electronic and lattice thermal conductivities and resistivities are resolved. The residual electrical resistivities are, respectively, 0.64, 1.01, 0.57, and 0.16×10^{−6} ohm‐cm.
The Lorenz ratios for molybdenum and tungsten show both maxima and minima and are always below the Sommerfeld value. The ratios for the beryllium samples have broad maxima and generally are above the Sommerfeld value. The thermoelectric powers of the samples with respect to pure copper are negative for the higher temperatures and positive for the lower temperatures.

Ultra‐High‐Speed Photographs Refuting ``Cohesion in Plasma''
View Description Hide DescriptionIn a recent issue of this journal [M. A. Cook and W. S. McEwan, J. Appl. Phys. 29, 1612 (1958)] and in a recent book [Melvin A. Cook, The Science of High Explosives (Rheinhold Publishing Corporation, New York, 1958), 1st ed., pp. 158, 420–426], a sequence of framing camera pictures was presented, and the sequence was interpreted as evidence that, as the result of the detonation of a quantity of explosive, a plasma was ejected into the atmosphere, and that the plasma exhibited a strong cohesive force. This paper presents pictures of an essentially identical subject, taken at three exposure times: the same exposure time used for the sequence in the references (4 μsec); 1/20th of that time; and 1/400th of that time. These pictures show that the phenomena are those of a shock wave, and that no new assumption of cohesive force is necessary to interpret the pictures. A simple analysis of the shock wave interpretation using the usual theory is presented to show that the interpretation is consistent with the known properties of air and the explosive used.

Magnetization of a Dilute Suspension of a Multidomain Ferromagnetic
View Description Hide DescriptionThe observed magnetization curve of a dilute suspension of carbonyl ironpowder, type E, is shown to be closely that one would derive from consideration of a dilute assembly of randomly oriented, single‐crystal, multidomain spheres. The deviations are in the directions expected from the effects of stress and non‐spherical clumping. The effects of packing and the applicability of various laws of approach to saturation are discussed.

Thermionic Properties of UC
View Description Hide DescriptionThermionic emission measurements using pulse techniques have been obtained from UC‐coated W filaments in the temperature range of 1200°–2100°K. From the slope and intercept of the Richardson plot, a value of φ=2.94 ev and A=33 amp/cm^{2} deg^{2} was indicated. There is evidence that the energy required for electrons to escape at low fields might be several tenths ev higher than the 2.94 ev value.

Criteria for the Choice of a Superconducting Bolometer
View Description Hide DescriptionThe usefulness of the superconductingbolometer was recognized a long time ago. All experimental data obtained up to now are far below the calculated value for a sensitivity of such a bolometer, because of the presence of the noise in the transition region of a superconductor. We have presented the correlation between the noise and the value of the surface boundary energy (interface between normal and superconducting laminas) and given on this basis the criteria for the choice of a superconductingbolometer. Attempt has been made to evaluate the noise on the basis of the Thomas‐Fermi approximation.