C O ~ ~ M T J N I C . ~TO TIO T IN I ESEDITOR
4iSI
complex, and (2) an allyl complex in which one of the chlorines or an additional hydrogen has added to the ring. T h e latter possibility merits serious consideration in view of the recent finding t h a t butadiene-palladous chloride appears t o be an allyl c o i ~ i p l e x . ~However, the chlorine attached to carbon in these allyl complexes has been found to be labile and readily displaced by alkoxy. Y o such alkoxy compound has been found in the case of the gold complexes. LVhen an ethanolic solution of chloroauric acid is warmed with 1,3-cyclooctadiene in the presence of sodium carbonate, I is formed in place of 11. Compound I is therefore tentatively identified as an aurous and I1 as an auric complex of 1 , k y c l o o c t a diene. The formulation of I as an olefin complex is consistent with the identification of the infrared absorption a t 15:jO and 1,521 cm.-l as being due to a carbon double bond stretching frequency. Infrared absorptions in I and I1 at 3100 mi.-' are identified as being due to olefinic C H stretching frequencies. '41 €3 1. S h a w
C'hi.iii
GENERAL ELECTRIC
I>iii i l a n d o n , , 1190 i,l!J{i2),
CO.
-1. J.
RESEARCEI L A B O R A T O R Y
CHALK
S C H E S E C T A D Y , XE\V \-ORK
RECEIVED SEPTEMBER 21, 1061
Effect of Inert G a s Pressure and Solubility on Electrical Conductance in Fused Salts
Sir: L1-e wish to present a preliminary report on recent data resulting from a study of the mechanism of electrical conductance in pure fused salts. X ~jOO-ml.capacity Inconel metal bomb, Type A243HC3 of the Parr Instrument C o . , was fitted with two CON-iX thermocouple glands packed with a natural magnesium silicate. One of these glands contains a Chromel-Alumel thermocouple, and the other contains a pair of ?()-gauge platinum wires for conductance electrodes. Pyrex bomb liners were prepared with capillary conductance cells fixed rigidly within them. I n a typical experiment, such a liner-cell is charged with about 450 g. of dry Sax03 weighed t o the nearest 0.5 g. Argon is admitted at room temperature into the sealed. evacuated bomb, containing the liner-cell and its salt charge, to a desired pressure. The number of moles of Xr introduced is calculated with the aid of the compressibility factor chart of llaslan and Littinan' using the equilibrium values of pressure, room temperature, and the gas volume (known from the total volume of the bomb less the volumes of the liner-cell and its salt charge). The bomb is then heated in a furnace to a new temperature well above the melting point of NaNOrI. Conductance measurements are made on the fused salt a t equilibrium temperature and pressure. The number of moles of gaseous Ar still remaining is calculated from the new 1-alues of equilibrium temperature, pressure, ant1 gas \wluine ( t h e o l d volume corrected for change in density o f the salt a n d its compression) with the aid o f the s a m e coiiil)ressibility chart. Decrease i t i moles of y,aseous Ar is attributed to its solubility i n the melt, t h e latter then bein,y calculated. The gas is slowly evacuated from the system. and the latter is pumped 11 Ar from the melt. Kew conductance I
1;
D \ l , < , l a n i i n c l 7' \ I
Littm;in.
iirt!
iiu:
ih i m
,
46, I .
Vol
X(j
readings are then taken on the gas-free nielt a t the same temperature. First results for NaN03, using a capillary conductance cell of about lSLcni.-' cell constant, indicate a , j , i C / L decrease in specific conductance when the tnelt a t ; j ( i 9 c was subjected to ;W2 atm. of Ar. The conductance measurements showed excellent precision for both the gas-saturated and gas-free melt, yielding specific conductances of 1 . 1 6 ohm-' cni.-' for the melt under pressure, and 1.23 ohin-' cm.-' for the normal liquid, both a t 309". Maximum experimental error in these values is believed to be less than * O . O l ohn-I cm--I. The Henry's law constant determined for the solubility of Ar in fused K a N 0 3 under these conditions, based on two independent determinations, was 1 S . i X and I D . ( i X I(1-j mole of -ir (cm? of melt)-' atm.-I. 11-e believe t h a t the observed conductance decrease under pressure is due to loss of free volume available for transport in the melt. In the hole theory of fused salts, the volume increase on melting. A I ' , is attributed almost entirely to holes2 For h'aNOs this is 4.:32 cm.3 mole-'. or a 10.77; volume increase.3 The isothermal compressibility coefficient, $, for fused NaNOa is about 20.4 X 1O-l2 cm.2 dyne-' a t 3 T O c , based on interpolation of the data of Bockris and K i c h a r ~ l s . ~ Using these data in conjunction with a density of I.ST g. c i i i . - ~for NaSO?,a t 3(iL1°,5 we arrive a t an approximate upper limit loss of holes of about S.SYc due to compression b y 3N2 atm. of Xr alone. Thus, if hole density is the major factor controlling conductance of a pure fused salt,6 we would expect an approximate upper limit conductance decrease of fused N a X 0 3 of S.Sq2, under these conditions owing to compression alone. This is not irreconcilable with our observed 3.iC& decrease. On the other hand, the theoretical molar density of holes in a pure fused salt is given as4
where y is the liquid-vapor surface tension, R is Boltzmann's constant, and 1- is the absolute temperature. Using this equation with a value of y = 113.5 dynes c m - ' a t 3 G 9 O , ' in conjunction with the density of Nan'O;,, the volume density of holes is 4.8G X 10" holes crn.-:j. Our average Henry's law constant together with this figure re\-eals t h a t if A i ratoms dissolve by occupying existing holes, about S,GYc of the holes would be occupied and thereby removed from the conductance mechanism. This value is also not a t odds with the observed conductance decrease. Considered separately, the theoretical conductance decrease due to either compression of the melt or gas solubility alone is in line with observation. There can be no doubt that free volume is reduced by compression of the melt, but it may be doubtful if inert gas molecules dissolve by occupying existing holes. The positive enthalpies of solution observed b y Grimes, Smith, and \\'atsons may well indicate that solution occurs by the molecules 198 H 131
