Search Documents

By M. Cohen, R. T. Howard

The purpose of this paper is to direct attention to the lower part of the austenite transformation diagram, or TTT curves, where considerable uncertainty still exists as to the blending of the bainite and martensite reactions. The dissimilarities between these two low temperature transforrnations have been summarized recently by Troiano and Greninger.' Of particular importance to the present work is the fact that martensite formation, unlike bainite formation, is not suppressed by the fastest quenching rates attainable. However, most of the investigations in this direction have been concerned primarily with the start of the martensite transformation (Mg) during cooling, rather than with the quantitative aspects of the subsequent reaction on either further cooling or isothermal holding. Several years ago, Cohen2 suggested a form of transformation diagram which reconciled, at least schematically, the isothermal characteristics of the bainite transformation with the insuppressible nature of the martensite transformation. In essence, the diagram consisted of a family of horizontal lines (representing the progress of the austenite-martensite reaction during quenching) which were interrupted in point of time by the bainite family of C-curves extending down from higher temperatures. This picture received later support by metal-lographic3 and dilatometric4,5 studies, but the techniques employed were not sufficiently critical to substantiate some of the details. Elmendorf6 and Grange and Stewardi have shown that the Greninger-Troiano metallographic technique8 may be used not only for M, determinations, but for studying the course of further austenite decomposition. In fact, the latter investigators have observed the amount of martensite formed as a function of temperature below M, in fourteen commercial steels. In the present paper, the Greninger-Troiano technique has been combined with the lineal method of Howard and Cohen9 to provide a unique, quantitative treatment of the austenite-martensite reaction. Furthermore, isothermal transformations have been similarly measured on holding at temperatures above and below M., thus permitting a fairly detailed presentation of the lower part of the conventional austenite transformation diagram. Experimental Details Materials The steels, listed in Table I, were selected for study because they exhibit a satisfactory range of M. temperatures and have not received as much attention as the lower carbon grades. They were prepared as 30 lb induction furnace heats in the laboratories of the Vanadium-Alloys

Jan 1, 1949

By M. Cohen, R. T. Howard

The purpose of this paper is to direct attention to the lower part of the austenite transformation diagram, or TTT curves, where considerable uncertainty still exists as to the blending of the bainite and martensite reactions. The dissimilarities between these two low temperature transforrnations have been summarized recently by Troiano and Greninger.' Of particular importance to the present work is the fact that martensite formation, unlike bainite formation, is not suppressed by the fastest quenching rates attainable. However, most of the investigations in this direction have been concerned primarily with the start of the martensite transformation (Mg) during cooling, rather than with the quantitative aspects of the subsequent reaction on either further cooling or isothermal holding. Several years ago, Cohen2 suggested a form of transformation diagram which reconciled, at least schematically, the isothermal characteristics of the bainite transformation with the insuppressible nature of the martensite transformation. In essence, the diagram consisted of a family of horizontal lines (representing the progress of the austenite-martensite reaction during quenching) which were interrupted in point of time by the bainite family of C-curves extending down from higher temperatures. This picture received later support by metal-lographic3 and dilatometric4,5 studies, but the techniques employed were not sufficiently critical to substantiate some of the details. Elmendorf6 and Grange and Stewardi have shown that the Greninger-Troiano metallographic technique8 may be used not only for M, determinations, but for studying the course of further austenite decomposition. In fact, the latter investigators have observed the amount of martensite formed as a function of temperature below M, in fourteen commercial steels. In the present paper, the Greninger-Troiano technique has been combined with the lineal method of Howard and Cohen9 to provide a unique, quantitative treatment of the austenite-martensite reaction. Furthermore, isothermal transformations have been similarly measured on holding at temperatures above and below M., thus permitting a fairly detailed presentation of the lower part of the conventional austenite transformation diagram. Experimental Details Materials The steels, listed in Table I, were selected for study because they exhibit a satisfactory range of M. temperatures and have not received as much attention as the lower carbon grades. They were prepared as 30 lb induction furnace heats in the laboratories of the Vanadium-Alloys

Jan 1, 1949

By Clarence Zener

The heat-treatment of steels will not pass from the stage of an art into that of a science until the mechanism of the phase transformations associated therewith is thoroughly understood. Such an understanding will involve two distinct types of inquiry: (I) the direction of the transformation, (2) the speed of the transformation. The first type of inquiry, which involves only the equilibrium relations between the various phases in steel, is the subject of the present and of a succeeding article; the second type, which involves the kinetics of transformations, will be the subject of a later article. In order that a theory of equilibrium relations may give useful information, certain experimental data must be put into the theory. It is customary for these data to be of a thermal nature. In the system discussed herein—steel—such thermal data are very scarce. In medium-alloy steels another type of data might possibly be used. In these steels the concentration of solute is so low that it may be possible to use the approximation of dilute solutions, an approximation that leads to linear relationships. The unknown functions, or constants, may then be determined in some cases from the binary iron-alloy system and in other cases from the ternary iron-carbon-alloy system. From the functions, or constants, so determined may then be computed the equilibrium relations in more complex medium-alloy steels. In the present article this second approach is used, the necessary experimental data 'being obtained from the simple binary and ternary equilibrium diagrams. Fundamental Equations In this article are derived the equilibrium relations between the gamma, alpha and cementite phases. In these three phases all important alloying elements except carbon form substitutional solutions, each alloying atom occupying a potential site for an iron atom. Carbon, on the other hand, forms an interstitial solution. In the face-centered cubic gamma phase, the interstitial positions are at the center of the unit cube and at the centers of the cube edges. There are therefore four interstitial positions per unit cell, or one per iron atom. In the body-centered cubic alpha phase, the interstitial positions are probably at the face centers, and also at the centers of the edges, which are crystallographically equivalent to the face centers. There are therefore six interstitial positions per unit cube, or three per iron atom. In cementite, (Fe, Alloy)3C, the carbon atoms may. be regarded as filling all the lattice positions of a given type. The introduction of the fundamental the of equilibrium has presented the author with a quandary. These equations are formally so similar to other equations that have been derived1,2 for related systems that they may appear to some readers

Jan 1, 1947

By Clarence Zener

The heat-treatment of steels will not pass from the stage of an art into that of a science until the mechanism of the phase transformations associated therewith is thoroughly understood. Such an understanding will involve two distinct types of inquiry: (I) the direction of the transformation, (2) the speed of the transformation. The first type of inquiry, which involves only the equilibrium relations between the various phases in steel, is the subject of the present and of a succeeding article; the second type, which involves the kinetics of transformations, will be the subject of a later article. In order that a theory of equilibrium relations may give useful information, certain experimental data must be put into the theory. It is customary for these data to be of a thermal nature. In the system discussed herein—steel—such thermal data are very scarce. In medium-alloy steels another type of data might possibly be used. In these steels the concentration of solute is so low that it may be possible to use the approximation of dilute solutions, an approximation that leads to linear relationships. The unknown functions, or constants, may then be determined in some cases from the binary iron-alloy system and in other cases from the ternary iron-carbon-alloy system. From the functions, or constants, so determined may then be computed the equilibrium relations in more complex medium-alloy steels. In the present article this second approach is used, the necessary experimental data 'being obtained from the simple binary and ternary equilibrium diagrams. Fundamental Equations In this article are derived the equilibrium relations between the gamma, alpha and cementite phases. In these three phases all important alloying elements except carbon form substitutional solutions, each alloying atom occupying a potential site for an iron atom. Carbon, on the other hand, forms an interstitial solution. In the face-centered cubic gamma phase, the interstitial positions are at the center of the unit cube and at the centers of the cube edges. There are therefore four interstitial positions per unit cell, or one per iron atom. In the body-centered cubic alpha phase, the interstitial positions are probably at the face centers, and also at the centers of the edges, which are crystallographically equivalent to the face centers. There are therefore six interstitial positions per unit cube, or three per iron atom. In cementite, (Fe, Alloy)3C, the carbon atoms may. be regarded as filling all the lattice positions of a given type. The introduction of the fundamental the of equilibrium has presented the author with a quandary. These equations are formally so similar to other equations that have been derived1,2 for related systems that they may appear to some readers

Jan 1, 1947

By W. H. Brandt

ThEre has been much progress in the last two decades in understanding the hardenability of steel. Roughly, the progress has been along two lines, which may be designated as empirical and fundamental. This empirical work has progressed to the point where steels may be manufactured to the desired characteristics quite closely and heat-treatment and application problems can be approached with considerable confidcnce. Particularly noteworthy among the empirical work is that of Grossmann,12. who has given formulas for calculating hardenability as a function of composition, and the establishment of the isothermal transformation curve by Davenport and Bain.13 The latter does not fall entirely in the class of an empirical study, for while it provides a means of presenting a tremendous amount of empirical data in the form of a single curve, it also outlines clearly the problems that face the fundamental investigator. The isothermal transformation curve shows clearly that the phenomena occurring in the neighborhood of the knee of the curve are those that limit the hardenability of steel. Above the knee of the curve, the product of decomposition of austenite is lamellar pearlite and fundamental progress has consisted chiefly of a detailed study of structure and formation of pearlite, largely at the hands of Mehl and his collaborators.'-5 They have shown that formation of pearlite proceeds by a process of nuclea-tion and growth and they measured both the nucleation and growth rate. However, there is still a wide gap between our empirical knowledge and fundamental knowledge, owing largely to the fact that there is no satisfactory theory of nucleation rate or of growth rate. This paper is being written in an attempt to Hl in a small part of the gap. A method was prcsented recently1' for calculating the velocity of edgewise growth of pearlite using an expression for the carbon concentration that satisfies the diffusion equation. The edgewise velocity of pearlite growth in a plain carbon eu-tectoid steel was calculated, using linear extrapolations of the ferrite and comentite solubilities. While the results are in fairly good agreement with the experimental facts, the calculated velocity of edgewise growth is smaller at 600°C. than G, the radial growth velocity of pearlite nodules, whereas one would expect it to be larger. In this paper the calculation will be repeated using a new extrapolation for the ferrite solubility, which gives better agreement between G and T', the calculated velocity of edgewise growth. A calculation will also be made of the effect of carbon content on velocity of edgewise growth, and the problems of calculating the effects of other alloying elements on growth rates will be discussed. It will be assumed that the mechanism of pearlite growth, established by Mehl and

Jan 1, 1947

By W. H. Brandt

ThEre has been much progress in the last two decades in understanding the hardenability of steel. Roughly, the progress has been along two lines, which may be designated as empirical and fundamental. This empirical work has progressed to the point where steels may be manufactured to the desired characteristics quite closely and heat-treatment and application problems can be approached with considerable confidcnce. Particularly noteworthy among the empirical work is that of Grossmann,12. who has given formulas for calculating hardenability as a function of composition, and the establishment of the isothermal transformation curve by Davenport and Bain.13 The latter does not fall entirely in the class of an empirical study, for while it provides a means of presenting a tremendous amount of empirical data in the form of a single curve, it also outlines clearly the problems that face the fundamental investigator. The isothermal transformation curve shows clearly that the phenomena occurring in the neighborhood of the knee of the curve are those that limit the hardenability of steel. Above the knee of the curve, the product of decomposition of austenite is lamellar pearlite and fundamental progress has consisted chiefly of a detailed study of structure and formation of pearlite, largely at the hands of Mehl and his collaborators.'-5 They have shown that formation of pearlite proceeds by a process of nuclea-tion and growth and they measured both the nucleation and growth rate. However, there is still a wide gap between our empirical knowledge and fundamental knowledge, owing largely to the fact that there is no satisfactory theory of nucleation rate or of growth rate. This paper is being written in an attempt to Hl in a small part of the gap. A method was prcsented recently1' for calculating the velocity of edgewise growth of pearlite using an expression for the carbon concentration that satisfies the diffusion equation. The edgewise velocity of pearlite growth in a plain carbon eu-tectoid steel was calculated, using linear extrapolations of the ferrite and comentite solubilities. While the results are in fairly good agreement with the experimental facts, the calculated velocity of edgewise growth is smaller at 600°C. than G, the radial growth velocity of pearlite nodules, whereas one would expect it to be larger. In this paper the calculation will be repeated using a new extrapolation for the ferrite solubility, which gives better agreement between G and T', the calculated velocity of edgewise growth. A calculation will also be made of the effect of carbon content on velocity of edgewise growth, and the problems of calculating the effects of other alloying elements on growth rates will be discussed. It will be assumed that the mechanism of pearlite growth, established by Mehl and

Jan 1, 1947

By E. V. Potter, H. C. Lukens

In general, during all electrowinning processes, large volumes of gas are liberated at the cathodes of the electrolytic cells. Most of this gas escapes from the electrolyte, but much of it may be absorbed or entrapped in the metal being plated. In many cases, this is not objectionable, but since it is liberated when the metal is heated, it is a potential source of trouble where the metal has to be heated or melted. Since this gas is principally hydrogen, in escaping from a melt it may cause minor explosions and scattering of the molten metal, while in other cases it has been reported to make the metal rise in the molds when cast. Therefore such metals would be more generally useful if all or most of the gas were removed before they were used. In an earlier investigation1 the gas content of electrolytic manganese was determined and methods found for removing most of it without detrimentally affecting the properties of the metal. In this paper results of a similar investigation on electrolytic chromium are presented. Apparatus The apparatus used in these investigations is essentially the same as that used in the manganese work1 and is shown in Fig I. It consists of sample chamber A, vacuum or pressure gauge B, volume gauge C, vacuum pump D, hydrogen container E, auxiliary gas chamber F, and furnace G. The pressure in the system is measured by the mercury manometer B. Changes in gas volume are measured by volume gauge C which is essentially another mercury manometer in which the volume of the mercury column is large and can be varied by raising or lowering the well H. In this apparatus the tube had a volume of 100 cc for a 60 cm length. Larger changes in gas volume can be accommodated by auxiliary containers attached to the system at F. The sample is heated by furnace G, the temperature of which is regulated within 5°C by an automatic controller. The only change in the apparatus was the sample container which was made smaller because of the chromium metal's containing more gas than the manganese metal and a smaller sample was used. Method The measurements were made as follows: The complete system when cold was evacuated and refilled with H2 until all the air was removed. It was then refilled with H2 to the desired pressure, as indicated on manometer B. The temperature was then raised to the desired value and as gas was either liberated from or absorbed by the sample, the height of well H was changed to alter the volume of the system so as to keep the pressure constant. In determining rates of liberation of gas, volume readings were taken at convenient tirne intervals depending on the rapidity with which the gas was liberated. In determining total volumes, the specimen was held at constant temperature and pressure until

Jan 1, 1949

By H. H. Uhlig

A study of passivity in chromium-iron alloys holds considerable interest, both because of the present-clay practical importance of the stainless steels, and because of the scientific importance attached to the mechanism of their corrosion resistance. Upon the discovery of the passive chromium-iron alloys in 1912, it seemed logical to attribute the observed phenomenal corrosion resistance to a thin, continuous, protective reaction product on the surface. The product was thought to be Cr2O3 or a similar oxide containing both chromium and iron. This mechanism seemed likely in view of Michael Faraday's conjecture that passive iron is not reactive in concentrated nitric acid because of an oxide on the metal surf ace. Since 1912, data have accumulated that permit a further exploration into the nature of passivity and provide a new approach to the problem. A summary of the facts lends considerable weight to a mechanism based not on physical protection of an oxide or similar compound, although such compounds may be present on the surface, but to a change in tendency of surface metal atoms to react chemically. This change in reaction tendency is ascribed to chemisorbed films that effectively satisfy residual valence forces of the metal surface, but do not appreciably disturb the positions of metal atoms in the lattice. In other words, formation of a new lattice such as that of an oxide is not a necessary condition for this type of passivity. The decreased reaction tendency of the metal by reason of chemisorbed films can thereby account for low reaction rates of the stainless steels in many media, as well as a noble galvanic potential as is observed. Substances capable of adsorbing on the stainless steels, and hence accompanying the property of passivity, are typically oxygen, nitrites or chromates, but are not in any sense limited to these. Other elements and compounds can also chemisorb, and therefore may also play a major part in the mechanism of passivity in specific media. The mechanism of passivity based on chemisorption or activated adsorption begins with the interpretation Langmuir proposed for his experiments concerning adsorbed oxygen on tungsten.' He found that adsorbed oxygen was tightly bound, remaining on the tungsten surface above temperatures at which WO3 rapidly volatilizes. The oxygen in such a state did not readily react with hydrogen even at

Jan 1, 1949

By R. T. Webber, R. B. Owen

The vapor phase deposition of molybdenum and tungsten from the hexacar-bonyls has been thoroughly investigated by Lander and Germerl, and shown to yield well-bonded coatings on a variety of hot metallic surfaces. They also demonstrated that chromium coatings2'3 could be similarly deposited, but did not investigate this metal extensively. Recently, conditions have been described by which this compound can be prepared in quantities sufficient for laboratory scale plating experiments, and it is the purpose of the present communication to outline the results of preliminary experiments upon the plating of chromium from its carbonyl. Apparatus and Technique The apparatus used in these experiments is shown in Fig I. At one end, connected by a short length of rubber tubing, is a Cenco Hypervac 20 pumping continuously. At the opposite end is a needle valve which allows a small stream of purified hydrogen to flow through the apparatus. Connected in series are a trap T1, maintained at the temperature of dry ice and acetone, the plating chamber, and two traps T² and T3 containing crystals of the carbonyl. T² is kept in a constant temperature bath (not shown), while T³ is at room temperature. Stopcock S6 allows the hydrogen flow to by-pass the carbonyl traps. A McLeod gauge is connected so that it can measure the pressure at the pump through Se, or the pressure at the plating chamber through S9 and S6. The standard (12 liter) volume connected by Sb is used to determine the volume included by S4, S6, SB, Slot the needle valve and the McLeod gauge. Knowledge of this volume allows calibration of the flow of hydrogen. The apparatus* is built entirely of pyrex, except for the needle valve, the rubber connection to the Hypervac, and the cover of the plating chamber. The plating chamber is water-jacketed, with a large brass top allowing insertion of the specimen to be plated. Set into the brass top is a ground glass stopper through which thermocouple leads are sealed. The thermocouple wires (pure Pt, and 87 pct Pt-13 pct Rh) serve the double purpose of measuring the specimen temperature and suspending it in the chamber. Surrounding the chamber is the coil of an induction heater, H. The apparatus is designed to allow variation of the following conditions: (I) the specimen temperature, from room temperature to about 1000°C, (2) the carbonyl vapor pressure, from 40 to 220 microns, and (3) the hydrogen flow. In several runs, 0.8 to 1.8 PCt of hydrogen sulfide was introuced through a needle valve that replaced S4. The vapor pressure of the carbonyl is

Jan 1, 1949

By J. J. Lander, L. H. Germer

Molybdenum and tungsten have desirable corrosion and temperature resistant properties, but the metals in bulk are expensive and their fabrication is difficult. Such considerations led to a search for a method of plating them upon base metals. This search was successful and led furthermore to a number of developments which could not have been foreseen. Neither pure tungsten nor pure molybdenum has been electroplated successfully, although the literature of electroplating is liberally sprinkled with the description of unsuccessful attempts. The process of plating these metals by thermal reduction of their halides has achieved onlv limited usefulness because of technical limitations and because of limited range of product quality. Plating by thermal decomposition of the carbonyls does not suffer from limitations of the same sort and the properties of the plate are readily controllable. This paper describes the plating of molybdenum and tungsten, and of compounds and alloys of these metals, by pyrolysis of their carbonyl vapors. Chromium also has been plated in this manner, but the experiments upon chromium have been less extensive and the results much less satisfactory. Probably the most important of the products which we have obtained are the alloys with carbon, and the carbides of the metals. Either molybdenum or tungsten can be deposited upon a base metal as a coherent and adherent coating. By adjusting the plating condition the carbon content of the coating can be made to have any value from substantially zero to about 35 atomic pct. The hardness is dependent primarily upon the amount of carbon in the plate and can be regulated over a very wide range, some coatings containing considerable carbon being harder than sapphire. In addition to products already known this method of plating has yielded new carbides of molybdenum and tungsten and alloys supersaturated with carbon. These new products are possible because, in the process of plating, carbonyl can be decomposed at temperatures as low as about 150°C, and thus in the presence of other substances reactions occur which could perhaps not be produced in any other way. In addition to carbides and alloys with carbon, sulphides and oxides of molybdenum and tungsten have been produced and studied. There are certain other advantages of the carbonyl method of plating. It is not required that the base have good electrical conductivity as in electrochemical processes, and the base need not have the temperature stability required in plating from a halide. Thus low melting glasses, as well as other ceramics, can be plated almost as readily as metals. In particular, the plating of glass with

Jan 1, 1949

By E. A. Gulbransen, W. S. Wysong

The behavior of molybdenum and its surface oxides in oxidizing and reducing gas atmospheres and in high vacua at elevated temperatures is a question of scientific and technical importance. The use of molybdenum as a metal or as an important component in alloys in oxidizing atmospheres at elevated temperatures has been limited because of the unprotec-tive nature of its oxide film and the volatility of this oxide at moderate temperatures. This communication will present results of a vacuum microbalance studyl, 2 of the following problems: (I) the oxidation kinetics in the temperature range of 250°to 450°C; (2) the reduction with pure hydrogen of thin oxide films formed on molybdenum; (3) the volatility of the thin oxide films; and (4) the vacuum oxidation of molybdenum at high temperatures. Molybdenum is an interesting metal to study for the following reasons: (I) although the oxide-to-metal-volume ratio is greater than one, the metal is not considered protective; (2) the metal oxidizes readily at temperatures of 2 50°C and higher; (3) the oxides of the metal are reduced at temperatures as low as 500°C by pure hydrogen; (4) the oxide Moo3 melts at 795°C; and (5) the oxides of molybdenum are volatile at moderate temperatures. The following variables are studied in the oxidation kinetics: time, temperature, pressure, surface preparation, surface area, concentration of inert gas in the lattice, cycling procedures in temperature, vacuum effect and high temperature vacuum oxidation. The general interpretation of many of these factors has been previously discussed. Literature The oxidation of molybdenum has not been studied critically. Although many studies have probably been made on the kinetics of the oxidation process, none has been reported in the literature available to us. The crystal structures found in the oxidation of molybdenum have been studied by Hickman and Gulbransen." Their results showed that Moo2 is the oxide which is stable in contact with the metallic substrate throughout the temperature range 300 to 700°C. As the film thickens in the low temperature range, MoO3 becomes predominant on the surface. Above 400°C, Moos is no longer observed, MoO2 being the only oxide found. Hagg and Magneli6 have shown from an X ray study that the a-phase, orthor-hombic Moo3 is stable up to 1050°C. Two other phases, ß and ß with compositions close to MoO2.90 are also obtained. The formula for the 0 phase is not given but that of the ß phase is defined as Mo9O26 The 0' phase is found to be built upon a monoclinic unit cell. A ?-phase is isolated from a preparation MoO2.75 heated for 17 hrs at 670°C. This phase is

Jan 1, 1949

By B. M. Harra, F. G. McCutcheon, O&apos

Until about 1930, the universal practice of horizontal-retort zinc smelters in the United States, as far as the writers are aware, was to operate the retort furnaces on a 24-hr cycle; that is, the retorts were charged in the early morning of one day, and in the early morning of the following day the residues were cleaned out, the retorts recharged, and a new cycle started. During the business depression of the early thirties, zinc production fell off to the extent that zinc smelters had idle retort capacity. Rather than continue to run some of the retort furnaces on a 24-hr cycle and shut down the remainder, a few companies decided to try out a longer cycle. Although the output of zinc per retort per day is lower, the longer cycle permits a larger charge per retort and improved zinc recovery. Advantages of the longer cycle were soon evident, and the practice spread rapidly. Various lengths of cycle were used, typical examples being 28-hr, 32-hr, 36-hr, and 48-hr. The 48-hr cycle fits in best with the standard routine of smelter practice that was built up on the basis of the 24-hr cycle; that is, the retort furnace "maneuver" always comes at the same time of day, and can be performed in the cool hours of the early morning. Under some conditions, particularly in some of the older plants, where the retorts are smaller than in newer plants, a 48-hr cycle is longer than necessary, and some intermediate length of cycle may theoretically be most economical. Use of an intermediate cycle, however, means that the maneuver comes at different times on consecutive days, and that sometimes it comes during the hottest part of the day. The work on the retort furnaces is extremely hot and arduous at best, and it is much worse, of course, if it is done during the heat of a summer day. When labor was plentiful and tractable, many plants were able to operate on cycles of intermediate length, but the workmen did not like the working hours. Under present labor conditions, it is hardly practicable to operate with a cycle of intermediate length, so the choice has narrowed down to the alternatives of 24-hr or 48-hr cycles. In any discussion of economics, too comprehensive generalization is not warranted, because local conditions affecting economics vary greatly. The best economic practice at one plant at any given time may be uneconomical at another plant, or even at the same plant at another time. To determine the length of cycle that is the better practice at a given plant at a given time requires a weighing of the advantages and disadvantages of each cycle, taking into consideration all the local conditions. some of the factors that must be considered in making the comparison are discussed briefly in the following paragraphs.

Jan 1, 1949

By L. W. Strock

Metallurgists handling lead and zinc ores have long been familiar with the spectrograph as a routine analytical tool, as its earliest regular use by American industry was in controlling impurities of zinc metal and its alloys. This application has been extended to other metal industries on an extensive scale, accompanied by refinements of procedure to such an extent that a large part of some metals produced, especially magnesium, aluminum and steel, are checked only by spectrographic methods for conformity to specifications. The spectrographic analysis of ores and other poorly conducting powdered solids presents difficulties not encountered with metallic samples, which can be arced or sparked as self-electrodes or can be readily dissolved, so that their solutions can be analyzed, as in the American Society for Testing Materials standard methods for zinc and zinc-base alloys. Spectra of powdered solid samples can be prepared by using carbon electrodes provided with a crater of suitable size, into which the sample is packed. Ore samples and many other powdered solid materials usually analyzed in the chemical and metallurgical laboratory are more complex and varied in composition than samples of metals produced in large tonnages. 'lore preliminary work therefore is usually necessary in devising a spectrographic method for analyzing a given type of powdered sample than for analyzing metal samples, in order to use that method with confidence. The howledge that the standard samples and unknowns produce equivalent line intensities at equal concentration is the fundamental basis of any quantitative spectrographic method. On the other hand, powdered samples have an advantage over metals in that any desired number of chemical compounds can be mixed dry in any desired ratio to produce a synthetic standard. Expcrience has shown, however, that spectra of elements in finely divided compounds introduced into a standard in preparing a synthetic ore sample do not always behave in the electric arc in a manner that produces spectral line intensities similar to those observed at the same concentration in a natural mineral. In ore samples minor elements frequently are contained in the crystal lattice of larger particles of an abundant ore mineral. Their dissociation and volatilization from the sample into the arc then proceeds differently in the standard and sample, with resultant differences in intensity. During the past 10 years I have devised and applied several quantitative spectrographic methods for determining various elements in rocks and minerals in connection with studies on their geochemistry,' but the incentive has gradually changed since the summer 0. 1940 to the specific practical problem of searching for domestic sources of rare and stratcgic metals. The

Jan 1, 1949

By H. L. Gilbert, C. T. Anderson, W. J. Kroll

In a previous paper,' the use of a split graphite tube resistor as a heater element for high-temperature furnaces has been described. The principal advantages of this type of construction are: I. The quadruple resistance or voltage compared with the top-bottom clamp resistor furnace. 2. Free expansion of the resistor. 3. The ease of protection from oxidation by air. 4. The firmness by which it may be held in the compact split clamp. This construction appeared to be well adapted for the fusion of high melting metals which require fusion in a vacuum. Zirconium metal had previously been melted at the Albany Laboratory of the Bureau of Mines in a vacuum arc furnace on a water-cooled plate, using a movable tungsten top electrode.% When larger furnace capacities were desired, this type of construction proved to have its limitations as the largest tungsten rods available were about 7/8 in. in diameter. The supply of direct current of high amperage for producing the desired arc characteristics was also not available. Considerable difficulty was also found in compensating the atmospheric pressure on the flcxible tubing of the movable electrodc, at the same time keeping its movement sufficiently free to enable the striking of an arc on any part of the metal briquet. Among the various materials used for the electrodes in conducting experiments in the vacuum arc furnace was graphite. It was found that zirconium fused in a graphite cup with a carbon arc was contaminated to only 0.2 pct carbon. It was therefore decsed to melt zirconium directly in graphite crucibles. A high-frequency furnace was first used in the attempt to melt the zirconium in graphite crucibles. The high-frequency heating coil was placed outside a silica tube which was evacuated by means of an oil diffusion pump backed up by a mechanical pump. The graphite crucible was thoroughly degassed at 19oo"C before use. Considerable difficulty was experienced in operating at such temperatures in vacuo because, as shown by Zintl,3 carbon vapor formed and reacted with silica accortling to the reaction: SiO² + C = SiO + CO. The silicon monoxide which is volatile at these high temperatures reacted with the zirconium, thereby contaminating it with silicon and oxygen. The presence of carbon monoxide was detected by means of a gas discharge tube placed in the vacuum system as indicated by the pale blue glow of the tube as soon as carbon vapor started to react with the hot silica. The reaction between the carbon vapor and hot silica was avoided by the use of split cylindrical molybdenum shields which were inserted between the graphite crucible

Jan 1, 1949

By B. D. Cullity

This investigation of bismuth-selenium alloys was made in an attempt to find a suitable material for use in thermoelectric generators. These devices are simply thermocouple circuits designed for the efficient production of electrical energy from thermal energy, which may in turn be derived directly from solar radiation or from the combustion of fuels. The requirements of a material for use in a thermoelectric generator are obviously quite different from those of a material used in a temperature-measuring thermocouple. Naturally, for power generating purposes, the thermoelectric power (voltage produced per degree difference in temperature between the junctions) must be as high as possible. Furthermore, in order to obtain reasonably high power outputs, the materials used in the generator must have an electrical conductivity which is as high as possible. Finally, in order to minimize heat lost by conduction to the cold junction, the thermal conductivity must be as low as possible. The relation between these several factors and the efficiency of conversion of thermal into electrical energy—the thermoelectric efficiency—has been discussed by Telkesl in a recent review of the subject. She gives the thermoelectric efficiency as: W o Efficiency = -Wo ______________i____________ TR-TC+ O2TR-Tc) where W, = power output in external load (watts) Wi = rate of heat input to hot junction (watts) TE,T, = temperature of hot and cold junctions (oA) K1,Kz = thermal conductivity of the two thermocouple materials (watts/cm°C) p,,pz = electrical resistivity of the two thermocouple materials (ohm cm) 9 = thermoelectric power (volts0/C) For many materials, some of the desired properties (for example, high electrical and low thermal conductivity) are almost mutually exclusive; the problem thus becomes one of finding a material which has the optimum combination of pfoperties. Bismuth-selenium Alloys work on the equilibrium diagram of the bismuth-selenium alloys has been reviewed by Hanson who gives the diagram reproduced in There are two intermediate phases in this system, BiSe and Bi²Se3. The existence of an alpha and beta modification of BiSe has not been demonstrated satisfactorily and the writer found no evidence of it in the present work. BiSe and Bi2Se3 have, like bismuth, a rhombohedra1 crystal

Jan 1, 1949

By J. L. Tokich, A. R. Troiano

Since 1921, when Hull1 discovered that cobalt can exist in the face-centered cubic and hexagonal close-packed modifications, the transitions that occur in cobalt have been extensively studied. It is generally agreed that on heating the hexagonal lattice changes to the cubic structure at ternperatures ranging from 400° to 500°C although considerable doubt exists as to the exact temperature. Data concerning the temperature of the transformation on cooling are much less concordant and their interpretation has not been satisfactory. The study is further complicated by the fact that many investigators have reported the existence of a high temperature allo-tropic change in the range 850 to 1020°C, at which the cubic lattice reverts to the hexagonal modification. The transition on cooling, which has been reported as Occurring at various temperatures from approximately 400 to 100°C, is of particular interest. Troiano and Greninger2 have recently suggested and considered the possibility Of allotropic transformations occurring in pure metals by a shear-type, diffusionless (martensite) transformation. The case of cobalt was specifically discussed and certain observations of previous investigators evaluated in terms Of this type of Of Table 1 briefly outlines the methods and results Of previous investigators Of the transformation temperatures of cobalt. Materials and Methods The major portion of this investigatioll was performed with cobalt sheet containing a total of 0.2 pet impurities with carbon present to the extent of o,I pct. After melting in an induction furnace under an atmosphere of hydrogen, the carbon con. tent analyzed 0.03 pct. Some observations were repeated with hydrogen reduced cobalt compresses analyzing 99.9 pet pure. No significant variations in results were observed employing the as-received sheet, the remelted stock, or the cobalt compresses. Filings for X ray diffraction analysis were sealed in evacuated glass ampoules. Pyrex glass tubing was used for temperatures below (600°C and for higher temperatures clear fused quartz was satisfactory. After appropriate heat treatment at various ternperatures, the specimen ampoules were furnace cooled, air cooled or quenched by crushing under distilled water. This allowed a range of cooling velocities from less than one degree C per set to greater than 1,0oo°C per sec. X ray diffraction patterns were taken with a Debye camera and filtered characteristic cobalt K-alpha radiation. The of of the existing phases was roughly approximated by visual examina-

Jan 1, 1949

By M. E. Wadsworth, T. L. MacKay

Sintered UO, samples were leached in sulfuric acid solutions of various concentrations. A pressurized system was used so that it was possible to investigate the kinetics of the reaction to 270°C with oxygen overpressures as high as 900 psi. The rate was observed to be a function of the concentration of hydrogen ions and directly proportional to the partial pressure of oxygen. Evidences are presented which indicate that a UO2 surface site reacts with a molecule of water to form a hydroxyl complex which in turn can dissociate with the characteristics of a weak acid. A rate determining step has been proposed which involves the reaction between an oxygen molecule and the hydroxyl complex on the surface of UO,. ThE 2 principal methods for uranium dissolution are carbonate and acid leaching. The sulfuric acid leach is the more popular and is used for the treatment of the majority of the ores of Africa, Canada, and the United States. Low recoveries in basic leach circuits led investigators at the University of British columbia1 to study leaching of uranium ores in pressurized vessels. Early in the study of dissolution of uranium it was found that only uranium in the hexava-lent state could be leached in acid or basic circuits. Therefore, the use of oxygen over-pressure in an autoclave offered an interesting approach to solving the low recovery problems of carbonate leaching. UO, was used in this study because it could be obtained in high purity and also because it is representative of the most refractory of the primary uranium minerals. A pressurized system was used to provide a means whereby important temperature and pressure parameters could be varied for the evaluation of the kinetic processes. The mechanism for the dissolution of UO2 should be similar for any of the uraninite type minerals. At the present all of the kinetic studies that have been conducted have been in carbonate media. Peters and Halpern2 carried out a kinetic study of the leaching of pitchblende. Their specimens were pitchblende ores selected on the basis of high grade and homogeneity. Pearson and wadsworth3 conducted a kinetic study of the dissolution of UO2 in carbonate solutions with results very similar to those obtained by Peters and Halpern. EXPERIMENTAL The UO2 as received* was found by spectroscopic *Mallinckrodt reagent, supplied by the Atomic Energy Commission. analysis to have a purity of 99.94 pct 0.03. It was ground in a mechanical agate mortar and screened through a 400-mesh sieve. Thin flat disks of UO2 were prepared by pressing the sized powder in a specially constructed die at a total pressure of approximately 25 tons per sq in. The pressed disks were approximately 0.25 cm thick, 1.6 cm diam, and 3.5 g in weight. As pressed, the samples were approximately 65 pct of theoretical density. It was essential that a binder such as poly-vinyl alcohol be added before pressing to prevent formation of cracks when fired. It was found by trial and error that pressed disks sintered close to 1870°C in a hydrogen atmosphere, reach densities between 91 and 91.5 pct of the theoretical density. At this density the samples had no measurable porosity based upon a 2-hr emersion in boiling water. An X-ray examination of these sintered UO2 disks showed them to be identical with the original UO2 received. Reaction of UO2 with the alundum furnace core was prevented by placing molybdenum between the samples and the core according to the method of Corwin and Eyerly.4 Leaching studies were carried out in a specially designed autoclave, the details of which have been presented e1sewhere.6 The main features of the equipment as applied to a kinetic study are: 1) solution samples may be removed during the course of a single run; and 2) temperature and agitation are carefully controlled. Two-ml samples removed from the autoclave during the course of a run were analyzed for U3O8 content with a Beckman model DK-2 spectrophotometer by the method of

Jan 1, 1959

By J. Halpern, D. R. McKay

SOME of the most important hydrometallurgical developments in recent years relate to processes involving the oxidation of sulfide minerals in aque- ous suspension at elevated temperatures by air or oxygen under pressure. Among these is the acid pressure-leach process which has been developed, particularly for the treatment of pyritic uranium ores.1-5 In this process, an aqueous suspension of the ore is digested in an autoclave, in which an overpressure of air or oxygen is maintained, at temperatures above 100°C. Oxidation of the pyrite (FeS,)

Jan 1, 1959

By F. A. Forward, A. H. Ross, J. Halpern

Application of the carbonate leaching process to carno-tite ores has often been limited by the difficulty of obtaining simultaneous high recoveries of uranium and vanadium. The chemical aspects of this problem are discussed and the effect of prior roasting procedures on this discovery is described. IN treating carnotite ores by carbonate leaching it is often necessary to subject the ore to a prior roast. Among the advantages that may result from roasting are: 1) improvement in settling and filtering characteristics of the ore; 2) destruction of carbonaceous impurities that would otherwise contaminate the leach solution and consume carbonate reagent; and 3) substantial increase in extraction of vanadium, which is usually low when the un-roasted ore is leached. Recovery of vanadium may be an important factor in processing carnotite ores, and the fact that vanadium is usually recovered in good yield by acid leaching has often been an economic advantage over results obtained with carbonate leaching. It has been found that response of an ore to carbonate leaching, following roasting, depends primarily on its composition and on the conditions of the prior roast treatment, particularly with regard to the roasting temperature, the atmosphere maintained above the charge during roasting and cooling, and the presence of certain reagents. These factors, especially the last, virtually determine the extractions of uranium and vanadium attainable by carbonate leaching, as well as the reagent consumption on leaching, whereas adjustment of the leaching conditions has only a limited influence. These variables were examined systematically to elucidate roasting and carbonate leaching of carnotite ores and to establish a treatment yielding high recoveries of both uranium and vanadium. Tests were conducted on samples of four different Colorado Plateau ores, described in Table I. Roasting Procedure: The roasting tests were conducted in an electrically heated furnace in which the temperature was controlled to ±15°C. The ore (usually a 500-g charge) was ground to —100 mesh, mixed with the desired reagent, and spread in a silica tray to a bed depth of about 1/2 in. During the roast the charge was rabbled at intervals of 15 min. To provide free access of air the furnace was usually ventilated; however, in a few experiments a controlled atmosphere such as carbon dioxide or hydrogen was maintained above the charge. Leaching Procedure: The pulp, generally comprising 500 g of ore and a 2000-cc solution containing 80 g per liter of Na2CO8 and 20 g per liter of NaHCO3, was agitated in an autoclave at 120°C, under 30 psi 02. Samples were withdrawn periodically, and the solution and ore residue were analyzed. Under the conditions employed, maximum attainable extractions were generally achieved in 4

Jan 1, 1959

By J. J. Stokes

Literature on the electrolysis of aluminum from cryolite melts and on the structure of these melts is surveyed critically. Data on density, freezing point, and other properties are reviewed. Theories of the electrolysis are examined in the light of these data. Two theories are presented which account equally well for the observations. ALTHOUGH the production of aluminum is a large-scale industrial process, consuming approximately 3 pet of all of the electric current generated in this country, the mechanism of the Hall Process remains a 70-year-old mystery. It is a remarkable fact that the aluminum industry, which was established largely through research and owes much of its growth to research, still employs as its basic process the one developed by Charles Martin Hall while he was still a young man working in a woodshed. There are, of course, modifications in detail and a tremendous expansion of the size of the units used, but basically the process employs carbon anodes and a carbon-lined iron vessel as a cathode for the electrolysis of alumina dissolved in molten salt. Of the almost infinite variety of salts available, the relatively minor Greenland mineral, cryolite, 3NaF.A1F3, is still the only one which gives good performance in the electrolysis. In the early days, the process was thought of as a simple one. The dissolved alumina dissociated and, during the electrolysis, the aluminum was carried to the cathode where it was discharged and collected in a pool of metal. The oxygen was carried to the carbon anode where it reacted to form primarily carbon dioxide together with some carbon monoxide. About the only virtue that can be claimed for this hypothesis is that of simplicity. The actual mechanisms must be tremendously more complex. This paper is devoted to a rather narrow part of the whole subject, namely, what ions are present in the bath, and by what mechanism the current is carried. The complexity of the literature on this system is demonstrated in Table I, which shows the ions postulated as being present during the process. This system is not easy to investigate. Cryolite melts at about 1000°C, so that all the investigations are at relatively high temperatures. Molten cryolite fumes in air, particularly if excess aluminum fluoride is present. The cryolite changes composition on heating at 1000°C, principally by reaction with moisture in the air, but apparently also by reaction with oxygen. Cryolite reacts with most of the customary materials of construction, and those that are not attacked by cryolite seem inevitably to react with molten aluminum. Graphite is the only common material that is inert to the electrolytic bath. Pure cryolite can be held in platinum. There is hope that some of the newer refractories will stand up in bath, and their development for this purpose would be a major advance in the field. The techniques to be discussed, then, are not the most advanced ones. Rather, they are examples of old standardized methods applied under difficult conditions. The principal methods that shed real light on the mechanism are density, electrical conductivity, freezing-point lowering, transport number, viscosity measurements, and reactivity of bath with carbon dioxide. The first measurement to be discussed is density. This procedure is relatively straightforward. A hollow platinum bob is suspended by fine platinum wire from an analytical balance into the melt being investigated. Density1-4 of the system sodium fluoride-aluminum fluoride is shown in Fig. 1. There is a maximum in the density near the cryolite composition, but actually the maximum is on the NaF-rich side of cryolite, close to the composition NaF - Na3AlF8. Density measurements are also useful in giving a clue to the nature of cryolite-alumina melts.1, 4 On the first addition of alumina, density decreases, Fig. 2, although the density of alumina, either solid or liquid, is higher than that of cryolite. These density results indicate that the systems are not simple. A great deal of effort has been spent in measuring the electrical conductivity of molten cryolite and of cryolite with various additives. Conductivity measurements are quite easy for fused salts that can be measured in a dip cell, but no successful dip cell has been made available yet for cryolite. Recourse is had

Jan 1, 1959