the molecular composition of sodium iodide vapor from molecular

THE MOLECULAR COMPOSITION OF SODIUM IODIDE VAPOR FROM. MOLECULAR WEIGHT MEASUREMENTS. BY SHELDON. DATZ AND WILLIAM T...
0 downloads 0 Views 416KB Size
938

SHELDON DATZA N D WILLIAM T. SMITH, JR.

that, in the presence of oxygen, glycine reacts only with the OH radicals. It is generally agreed13 that G(0H) for a-particles is of the order of 0.1-0.3, which is much too low to account for the observed yields. This proposed reaction may well account for most of the effect produced by X-rays but it appears that other reactions are of considerable importance with a-particle irradiation. These are probably re(13) A. 0. Allen, R a d i a t i o n Research, 1, 85 (1954).

Vol. 63

actions within the track itself. In fact we are rapidly coming to the conclusion in this Laboratory that the formation of HCHO even with X-rays may be almost entirely in the ‘%purs)’or regions of high local concentration of radicals and excited molecules. A study of the initial yield of all the products ns a function of glycine concentration and particularly studies in the presence of oxygen with He ions or a-particles, should be most rewarding.

THE MOLECULAR COMPOSITION OF SODIUM IODIDE VAPOR FROM MOLECULAR WEIGHT MEASUREMENTS BY SHELDON DATZAND WILLIAMT. SMITH,JR. Contribution from the Chemistry Division, Oak Ridge National Laboratory,’ Oak Ridge, Tenn., and the Deparlment of Chemistry University of Tennessee, Knoxville, Tenn. , Received F e b r u a r y 16, 1969

A technique has been developed for the determination of the molecular weight of vapors a t high temperatures, by measurement of the pressure at constant vapor density, the pressure measuring device being a molten gold manometer. For the dissociation of Na212into NaI measurements from 1175 to 1350’K. indicate a AEo of 40.2kcal. mole-’ and a AS0 of 27.0 e.u. a t 1260°K.

Introduction Considerable effort has been put forth in recent years in the determination of molecular association in alkali halide vapors. Of the experimental methods employed, the most satifactory has been that of effusive velocity distribution analysis devised by Miller and Kusch. 4,3 Mass spectroscopic investigation~~-’have yielded much valuable information, but uncertainties in ionization cross sections and modes of dissociation on electron impact preclude quantitative measurements of the degree of association. Infrared absorption studies of alkali halide vapors have indicated the presence of dimer,s but quantitative measurements of amounts present are a t best very difficult. Since even the most precise of the above methods is highly complex in both execution and interpretation, the present research was undertaken to provide a more simple and direct method of measurement, and to extend the measurements over a wide range of temperature and pressure. In this work, the association equilibrium of sodium iodide vapor was studied by the measurement of molecular weight as a function of temperature. The molecular weight was determined by measurement of the absolute pressure exerted by a known weight of completely vaporized salt contained in an isothermal bulb of known volume. The choice of fused silica for use in fabrication of the (1) Operated for the United States Atomic Energy Commission by the Union Carbide Corporation. (2) R. C. Miller and P. Kusch, J. Chem. P h y s . , 26, 860 (1956); 27, 981 (1957),hereinafter referred to as MK. (3) M.Eisenstadt, G.M. Rothberg and P. Kusch, J. Chem. Phys., 29, 797 (1958). (4) L.Friedman, ibzd., 23, 477 (1955). (5) T. A. Milne, H. M. Klien and D. D. Cubicciotti, ibid., 28, 718 f 1958). (6) R. C. Schoonmaker and R. F. Porter, ibid., 30, 283 (1959); 29, 1070 (1958). (7) J. Berkowitz and W. A. Chupka, ibid., 29, 653 (1958). (8) W. Klemperer, Chemistry Dept., Harvard Univ., private communication.

bulb was dictated by the requirements of high temperature dimensional stability, non-reactivity with halide vapors and the ability to be degassed and sealed off a t high temperatures. The pressure measuring device must have its sensing element at a temperature higher than the condensation point of the gas, and must be accurate to 10.1 mm. Initially a vitreous silica Bourdon sickle gauge was used, but it was found that a t 1000° the rapid diffusion of argon (used to balance the gauge pressure) through the thin membrane of the gauge caused large errors in pressure measurement. The gauge used in the measurements reported here was a balanced manometer in which the manometric fluid was molten gold. The two arms of the manometer could be observed through a window in the furnace and Dhe external pressure of argon required to balance the bulb pressure was read from an external mercury manometer. The choice of molten gold for the manometric fluid was made on the basis of the temperature range over which it is a liquid, its lack of chemical reactivity and its low vapor pressure over the required temperature range. Experimental Apparatus.-The bulb was fabricated from 80 mm. i.d. 82 mm. 0.d. clear silica tubing with end plates cut from 3.2 mm. thick silica plates made slightly convex to resist high temperature distortion. The bulb is 16 cm. in length and has a volume of ca. 800 ml. The pressure transmitting tube, which connects the top of the bulb to the top of the manometer which is mounted vertically above it, is 8 mm. i.d. and 30 cm. in length. The manometer tube is 10 mm. i.d. and contains 130 g. of gold (7.6 cc. when molten) giving a height of 3 cm. in each arm. Additional mechanical support of the bulb is provided by two 4 mm. silica rods which brid e the manometer. Additional 6 mm. tubes are sealed to t i e top of the bulb for sample loading and pumping. The pressure balancing arm is connected to a Pyrex pressure manifold through a graded seal located away from the high temperature region. Pressure in the manifold is adjusted by manipulation of two vacuum stopcocks, one connected to a vacuum system and the other to a bulb containing argon at a pressure slightly above the maximum anticipated pressure. Flow

June, 1959

MOLECULAR COMPOSITION OF NAI VAPORBY MOLECULAR WEIGHT MEASUREMENTS939

rates through the stopcocks are kept a t a convenient level for manipulation by plugging the bore entirely with paraffin and then piercing a hole with a 0.004 in. tungsten wire. Furnaces .-The heating system consists of three electric tube furnaces. The bulb furnace is a “Kanthal” wound “Marshall” tube furnace I6 in. long with a 3.5 in. i.d. Its temperature is controlled by a “Speedomax” controller with a 10 mv. span, using a chromel-alumel thermocouple as a sensing element. Gradients over the length of the bulb may be adjusted by shunting sections of the furnace winding. With this system, the temperature can be maintained constant to f 0 . 5 ’ for several hours t n d the gradients over the length of the bulb range from 1 at the lowest operating temperature to 5’ a t the highest. To facilitate loading, pumping and sealing, the furnace may be lowered 4.5 in. by a counterweight system. The manometer is heated by a “Kanthal” wound “Marshall” furnace 16 in. long and having a 2.5 in. bore. The furnace is fitted with two 1 in. diameter windows set at 180’ to each other and centered 8 in. from the end of the furnace. The temperature control of this furnace is accomplished with a“Micromax” controller using a chromel-alumel thermocouple. This furnace is mounted in a fixed position directly above the bulb furnace such that there is a ‘/z in. gap between furnaces. I n order to prevent condensation in the pressure transmitting tube, a third furnace was made of a MgO swaged “Kanthal” heater contained in a 1/8 in. Inconel sheath wrapped in the form of a helix and welded to a 2.5 in. i.d. Inconel tube 6 in. long. The tube is mounted firmly to the bottom of the manometer furnace such that three inches extend up into the upper furnace tube and 2.5 in. into the lower furnace. The furnace ends are sealed with fire brick and fire brick dust and the gap is lagged with “Fiberfrax” packing. Gold Manometer.-The purest gold commercially available contains considerable amounts of dissolved gas and small amounts of other impurities which tend to collect at the liquid surface. For this reason, it is further purified as follows. Gold wire is first de-greased and washed in hot HNOa. It is then melted in a silica tube under vacuum and pumped for several hours. After cooling, the cast ingot is removed and cleaned again in boiling “03. Once the gold has been melted in the manometer, the manometer temperature should not be dro!ped below 600’ since, if the temperature drops to ca. 500 , the silica will crack. This effect may be due to a slight wetting of the silica by the gold, although the markedly convex appearance of the gold meniscus suggests that the silica is not wet. The temperature a t which fracture occurs might be associated with the sharp decrease in coefficient of expansion of quartz occurring at 550’, although no such transition is observed in vitreous s i l i ~ a . ~ Sample Purity.-The samples were prepared from reagent grade NaI (99.95% pure) which was heated under vacuum to 500” and then melted and recrystallized under argon pressure.lO Optically clear crystals were selected and a single piece cleaved from the center of a crystal constituted a sample. All handling was done in a nitrogen filled PZOs dry box. Procedure.-Initially, the system is evacuated ( p = 10-6 mm.) and degassed a t 1000’ for 24 hours with an external atmosphere of argon supplied through a silica feed tribe. After cooling, the bulb furnace is lowered, argon is admitted to the bulb and the sample is introduced through the loading tube, which is then resealed. The bulb containing the sample then is evacuated and heated to 200” for several hours. The pumping tube is then sealed a t a point 1 in. away from the bulb. The bulb furnace is raised into posi; tion and the gap between furnaces filled with “Fiberfrax. The bulb then is heated in an argon atmosphere. Pressure readings are made by measuring with a cathetometer the argon pressure necessary to balance the gold manometer. Temperature measurements are made with calibrated Ptr 90% Pt 10% Rh thermocouples placed along the length of the bulb. Since it has been observed that some gas appears to diffuse into the bulb, especially a t high temperatures, the tempera(9) R . B. Sosman, “The Properties of Silica,” Chemical Catalog Co.. New York, N . Y., 1927. (10) J. W. Johnson, M . A. Bredig and Wm. T. Smith, Jr., J. Am. Chem. SOC.,11, 307 (1955).

ture of the bulb is dropped below the point where the salt vapor pressure is measurable and a determination of the residual gas pressure is made with the manometer. To change samples the vacuum is broken and the umping tube is reconnected to the pumping system. T i e old sample is then distilled out of the bulb to a cold portion of the pumping tube and the new sample introduced. Errors .-Errors in the measurement of sample weight and volume (after correction for thermal expansion) amount to less than 0.5%, which is negligible. Absolute temperature measurements are accurate to &lo (0.1%). Since a temperature difference of 4” will vary the equilibrium constant by 3%, a weighted average temperature of the bulb is used in the calculation of the equilibrium constants. The larger gradients existing along the pressure transmitting tube and the manometer itself do not significantly contribute to the error, since the volume involved is only 1.5% of the total. Pressure measurements were found to be reproducible to within 2~0.05mm., which introduces a maximum error of leyoin the measured pressure range (10 to 30 mm.). A precise knowledge of the density of the molten gold (ca. 17.0) is not necessary since a null method is used. The largest error introduced is in the correction of the pressure for permanent gas diffusion into the bulb. This pressure was measured a t 200°, corrected to the temperature of the observation and subt,racted from the total pressure observed. The leak rate varies greatly with temperature from 0.4 mm. per hour at 135OoI