\ 1 1969 AN ORNL=3373 UC=-80 — Reactor Technology TI1D-4500 (18th ed.) MAGTER THERMAL ANALYSIS AND GRADIENT QUENCHING APPARATUS AND TECHNIQUES FOR THE INVESTIGATION OF FUSED SALT PHASE EQUILIBRIA H. A. Friedman G. M. Hebert R. E. Thoma OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. DISCLAIMER Portions of this document may be illegible In electronic image products. Images are produced from the best available original document. Printed in USA. Price: $0.75 Available from the Office of Technical Services U. S. Department of Commerce Washington 25, D, C. LEGAL NOTICE This report was prepared as an account of Government sponsored work. Neither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Makes any warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this report. As used in the above, '‘person acting on behalf of the Commission’’ includes any employee or contractor of the Commission, or employee of such contracter, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, any information pursuant to his employment or contract with the Commission, or his employment with such contractor. ORNL-3373 Contract No. W-7405-eng-26 REACTOR CHEMISTRY DIVISION THERMAL ANALYSIS AND GRADIENT QUENCHING APPARATUS AND TECHNIQUES FOR THE INVESTIGATION OF FUSED SALT PHASE EQUILIBRIA H. A. Friedman, G. M. Hebert, and R. E. Thoma DATE ISSUED JBH & - 1963 OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION ~ THIS PAGE WAS INTENTIONALLY LEFT BLANK lf“ i -~ o) iii CONTENTS Abstract . . . . . . . o . 0 . . Introduction e e e e e e e e e Methods . . . . .+ « ¢« ¢ « « .+ o . Direct Thermal Analysis « v e e . Normal Procedure . . . . . . . Special Procedure . . . . . Quenching Techniques . . . . . . Preparation of Samples . . . Preparation of Quench Tubes Tubes for Non-volatile Salts Tubes for Volatile Salts . . Quench Furnaces . . « .« + .« . Furnaces with Stationary Thermocouples Furnaces with Traveling Thermocouples Accuracy and Precision of Measurement Acknowledgment . . . . . . . . . References . « ¢« o« o o o o+ o o o & Page 28 29 THERMAL ANALYSIS AND GRADIENT QUENCHING APPARATUS AND TECHNIQUES FOR THE INVESTIGATION OF FUSED SALT PHASE EQUILIBRIA H. A. Friedman, G. M. Hebert, and R. E. Thoma ABSTRACT A detailed description is presented of appara- tus and methods used at ORNL for determination of high temperature equilibrium phase relationships in condensed systems of molten salts. Principal empha- sis is given to experimental techniques required for investigation of non-volatile hygroscopic fluor- ides. Equilibrium phase behavior is elucidated by thec combined results of experiments in which measure- ments are made of the thermal effects occurring on melting and freezing polycomponent mixtures, and others in which unequivocal identification of solid phases formed during crystallization is obtained. Apparatus devised at ORNL for use in preparation, purification, equilibration, and handling of mater - ials for application in fluoride'phase studies is described in detail. The methods and techniques described are unique in providing such large quan- tities of phase data that phase diagrams of complex systems may be constructed in a relatively short time. —2- INTRODUCTION The advent of molten salts in nuclear reactor technology as fuels, converter-breeder blankets, heat transfer fluids, and reprocessing media for spent fuel elements has necessi- tated a large number of phase equilibrium investigations. Although many experimental methods have been applied in stud- ies of phase equilibria at elevated temperatures,1 e.g., through measurements of thermal expansion, magnetic proper- ties, viscosity, thermodynamic properties and crystallization equilibria, only the latter two of these methods are suited for rapidly acquiring the large number of data needed in constructing complex phase diagrams. These two general methods have therefore bheen applied for several years to in- vestigations of molten salt phase equilibria at ORNL. Adapt- ations of experimental methods to specific problems obviously require consideration of the most annoying properties‘of the‘. " materials to be studied and moditfication of the methods to per- mit investigation of the materials despite their intransigence. Molten halidés at elevated temperatures possess an impressive list of these characteristics. It is the purpose of this report to furnish detailed descriptions of the practical pro- cedures which have found spepial application at ORNL.for in- vestigations of molten salt phase equilibria. Phase equilibrium diagrams are generally derived from two kinds of experiments, those from which deductions are made from measurements of thermal effects.occurring in heating and cooling curves, and those which permit a direct o, -3 or indirect identification of the numbers and compositions of phases occurring at all temperature-composition points. Commonly, fused salt diagrams are based on information from cooling curves. Changes in slope of the temperature of the sample, when plotted as a function of time, reflect phase changes which occur on cooling. This technique is generally adequate for determining all except the steep- est liquidus curves; steep curves represent small changes in saturation concentrations with temperature and hence small heat effects. Cooling curves also provide information on the solidus and subsolidus phase changes, but are prone to give misleading indications because of the impossibility of maintaining equilibrium during cooling. Phase transitions inferred from cooling curves are Qerified by quenching of equilibrium samples and an identification of the phases by crystallographic examination with microscopic and X-ray diffraction techniques. Chief interest.in salt phase equilibria has focused on the fluorides and chlorides of the actinide elements and on low melting solvents for these fissile and fertile materials. Such salts vary'widely with respect to their hygroscopic character. It is necessary to employ experimental techniques which maintain a genuinely anhydrous environment for these salts. Though the tri- and tetrafluorides of the actinides, rare earths and zirconium are not hygroscopic, they are easily hydrolyzed at elevated temperatures. It is necessary, therefore, if the sample e under examination is to be free from extraneous phases due to the presence of oxides or oxyfluorides, to remove all water and to protect the heated sample from contact with air. METHODS Direct Thermal Analysis Two techniques have been developed for obtaining thermal- analysis data. These have evolved as a '"'mormal" procedure, employed for mixtures whiéh are known to be non-hygroscopic, and a "special' procedure, employed for hygroscopic mixtures. - Normal Procedure A convenient means of accommodating four samples in graph- ite or nickel crucibles is shown in Figure (1). The graphite crucibles, 5 1/2" high, 1 5/8" o.d., with a wall thickness of 1/8", are Tabricated from high density graphite; the nickel crucibles, 5 1/2" high, are constructed from 1 1/2" tubing with a 1/16" wall thickness. A graphite disc, approximatoly the size of the internal diameter of the crucible and with two holes to admit the stirrer and thermocouple well, may be inserted into the crucible on top of the sample if thc major ingredient has an appreciable vapor pressure. The disc re- duces the volatilization of the sample by floating on the melt thus decreasing the liquid surface area. An annealed copper gasket between the flange of the reactor vessel and the 1id acts as a seal when the assembly is fastened together with three clamps. Nickel stirrers of 1/8" diameter shanks and 14" in length are inserted into the melts through closely fitting sleeves "x W UNCLASSIFIED ORNL-LR-DWG 622!13R CRIVE BELT (COILED SPRING}) BUSHING (GRAPHITE) /ALIGNER SUPPORT ROD (Al . (I, BT P T PP ¢ ' oo 5 | 4 _ I ik, Lk : % 1? GAS OUTLET ~ GAS INLET | — THERMOCOUPLE VENT PLUG (GRAPHITE ) —a 1 7— SWAGELOK FITTINGS VENT \\ BUSHING { GRAPHITE) = VRANNIANN s o e e [ | 3 LT X J i LS “Il' .'h A L | flt%: i ; il m\\\&;& %%m g \ LID (Ni) R, =) VJZ% : ¢ Al | e o, pH s f il Y3 UNEAY: : 15 ! GASKET {Cu) CLAMP {TOOL STEEL ! * i g W CRUCIBLE (GRAPHITE / OR Ni ) [} 2 § VESSEL (Ni) L . ! S$TIRRING ROD (Ni) g1 ¥ I T :' Y | B THERMOCOUPLE WELL (Ni) . N l' \ ' { T ' ] L et H “l @_ Br b ey i ) INCHES I T T T T N I L 0 i 2 3 4 Fig. 1. Reactor Vessel. —b— of graphite in the 1id of the assembly (Figure 2). A holder capable of fine adjustments aligns the top of the stirrers (Figure 3). The melts are protected from the atmosphere by maintaining a small positive pressure of helium,'purified by passage through a liquid nitrogen-charcoal trap, in the as- sembly. Leakage of helium through the graphite bushings on the 1lid prevents diffusion of air to the molten mixture. Pow- er %S supplied to the stirrer through the bclt of coilcd spring (Figure 1); slippage of the belt prevents possible dam- age to the drive motors. Temperatures are measured with Chromel-Alumel thermocouples in a thin walled (10 mil) nickel thermocouple well immersed in the melt. The e.m.f.'s are measured using Minneapolis Honeywell "Electronik" Recorders that are frequently calibrated with a potentiometer. To remove oxide and water vapor, 10 grams of ammonium . bifluoride* are added to each crucible followed by the sample. . The crucibles are then loaded into the reactor vessel and the reactor is assembled and placed into a 5" pot furnace. The stirrers are aligned and the furnace heated until the ammonium bifluoride becomes molten, 120-225°C. These low temperatures are maintained for at least one hour before heating to elevated temperatures. Fuming of ammonium bifluoride occurs until ap- proximately 550°C. When evolution of the fumes is no longer *Ammonium pifluoride will react with a number of oxides, in- cluding those of uranium, zirconium, yttrium, aluminum, beryl- lium, cobalt, iron, vanadium, cerium, and chromium (valence 6) to form fluorides. Nickel and chromium (valence 3) oxides will not be fluorinated with ammonium bifluoride. (Private communication from B.J. Sturm, Reactor Chemistry Division, ORNL. ) UNCLASSIFIED ORNL-LR-DWG 62215 VENT HOLE FOR THERMOCOUPLE (4} T e GAS INLE LID (Ni) ~ T~ GAS QUTLET [ Fig. 2. Reactor Vessel Top. UNCLASSIFIED ORNL-LR-DWG 62214 ALIGNER .SUPPORT ROD (Al) - ADJUSTMENT SCREW Fig. 3. Stirrer Rod Aligner. -9- observed, the escape vent (Figures 1 and 2) is closed with a graphite plug, Fiberfrax insulation is packed on the top of the reactor and the stirrer motors are started. One of sev- eral electric timers can start or stop any part of the equip- ment mechanically. A mineral oil bubbler located in the gas exit line is used to check for a positive pressure. The furnace is cooled after the temperature has reached approximately 100°C above the highest estimated liquidus of any sample and the ingredients of each sample have melted and mixed. The rate of cooling is regulated by controlling the voltage to the furnace with an auto-transformer. The heating and cooling cyclcs are usually repeated with the cool- ing rate varied to verify Llhe thermal data. It is possible to magnify the thermal effect by increasing the sample size and by decreasing the cooling rate. An approximately optimal choice of sample size appearé to be 50 g. This sizé choice is a compromise of desirably larger sample sizes with the convenience of employing laboratory scale equipment. Suit- able cooling rates for such sample sizes are 3-49C/ min. A complete heating and cooling cycle requires about 6 hours. Special Procedure To obtain equilibrium cooling curves and quench data in systems of hygroscopic salts all manipulations except weigh- ing the starting materials are performed in a vacuum dry-box. -10- Samples are purified in the same manner as described in the other procedure and melted 1in a 5" pot furnace set into the floor of the dry-box. Ammonium bifluoride fumes are exhausted through alnickel funnel (Figure 4). 1In use, the funnel is placed over the furnace well and sealed to the well with the teflon gasket. Fumes evolved during the ammonium bifluoride purification step are pulled by a vacuum pump in turn through the funnel, a rubber hose, and 3/4" diameter copper tubing in- to a soda lime trap and into a - sulfuric acid trap. The fun- nel is placed in its rack at the back of the dry-box when fuming is complete, as indicated b& the temperature of a thermocouple fitting into a well in the nickel funnel; the stirring mechanism, which also contains the thermocouple wells, is then positiqned over the heating'well after it has been removed from its holder (Figure 5). One thermocouple well and one stirring rod areinserted into each crucible. A single motor rotates the four stirrers which have slip clutches to permit the motor to revolve without damage when the melts freeze, Temperatures are measured and recorded using Chromel- Alumel thermocouples inserted into the dry-box through Conax fittings*. An atmosphere of érgon gas, dried by passing through mag- nesium perchlorate and dry ice-trichloroethylene traps, is maintained in the dry-box. The dry-box entrance chamber is * Made by Conax Corp., 2300 Walden Avenue, Buffalo 25, New York, UNCLASSIFIED PHOTO 562814 CONAX FITTINGS FOR THERMOCOUPLE 9/, 3/1-i~. Cu LINE TO D e sooyLive Rap if i | vfi. ; b GASKET o " NICKEL FJNNEL ! STIRRER MECHANISM | =P FURNACE Fig. 4. Vacaum Dry-Box. -12- PHOTO 36217 (o) w L n w < <) Q & 2 ————— T A Mechanism. Stirrer Ds Fig. -=]3= evacuated and refilled with dried argon after each transit through the cham ber. Metal plates cover the glove ports when the dry-box is evacuated and when the ports are not in use. The atmosphere is circulated inside the box over several trays of P,0; to absorb any moisture. An oxygen-free dry atmosphere must be maintained in the box. In usual operation with closed glove ports, the water content of the dry-box atmosphere can be maintained at about 20 ppm. Quenching Techniques Purified melts from the thermal analysis procedure may be further used by being equilibrated at and quenched from elevated temperatures to verify the transition temperatures and to observe the phases present at the transitions. Quench tubes containing 25-28 sample segments are equilibrated in gradient quench furnaces over pre-determined temperature ranges and then rapidly cooled. Methods used to interpret thermal gradient quenching data have been discussed in reports of fluoride phase investigationsz_4 and will not be treated here. Preparation of Samples Specimens to be equilibrated are obtained from either of the thermal analysis procedures, purified by a special prepar- ation described below, or prepared from pure components. The samples which have been purified in the normal thermal analysis procedure are transferred into a dry-box, ground with an elec- tric mortar and pestle to <100 mesh, bottled, removed from the ~14- dry-box and homogenized on a converted ball mill (Figure 6) for approximately 16 hours. These bottles, sealed with a coating of paraffin and beeswax, are clamped into place on the face plate. After mixing, the samples are returned to the dry-box and loaded into quench tubes. Hygroscopic samples, purified in the special thermal analysis procedure, are homocg- enized by hand mixing within the dry-box rather than external- 1¥a Preparation of quench tubes Tubes for non-volatile salts.- A rolling and crimping machine (Figure 7)5 has been constructed to insure equal sam- ple spacing in the quenching tube and to lessen the time re- quired for loading. A nickel tube 6-1/2" long, 0.10" in out- side diameter, and 0.010" in wall thickness which has been annealed for 1 hour in a H, atmosphere at 800°C or a dried platinum tube of similar size is rolled with the knurled wheel to flatten all but 3/8" at one end. The flattened tube has a void space 0.015" thick. The bottom of the tube is then sealed by welding. A sample is loaded by inserting the end of the sample tube into the shaft of a specially constructed funnel (Fig- ures 7 and 8). A small lip on the inside of the funnel shaft prevents over-insertion of the tube. The tube is tapped against a solid surface to insure complete filling, and then crimped with pliers 3/8" from the top. The upper 3/8" is cleared of powder, cleaned with a pipe cleaner and flattened Fig. 6. Mixer. UNC_ASSIFIED PHCTO 36190 o - b ‘;%&% — e S Fig. 7. Roller and Crimping Machine. UNCLASSIFIED PHOTO 30159 UNCLASSIFIED ORNL-LR—DWG 70688 L 8. Loading Funnel. Fig. T B with pliers. Care must be taken to see that this space is well cleaned, for a small amount of the sample lodged in the weld can prevent sealing of the tube. The tube is removed from the box, flattened in a vise, crimped with the crimp- ing wheel, and the end closed with a gas-oxygen torch. A piece of wet cleansing tissue held around the upper portion of the tube while welding prevents vaporization of the sample. Tubes for volatile salts.=Nickel or platlinum queunclh tubes prepared as discussed above are of little use for investi- gating systems containing one or more components which exert significant vapor pressure at elevated temperatures. An in- novation in the tube design was made to minimize expansion of crimped joints by volatile materials and migration of salts within the tubes. The sample is loaded into an unflatten-