2072
NOTES
rivalent platinum which is not converted into an aquo compound even by concentrated hydrochloric acid.
Summary 1. A new complex compound of quadrivalent manganese has been prepared which has the corn-
Vol, 5s
position KZ[Mn(CtO&(OR)2]*2H20. The barium and chloropentamminecobalti salts were also isolated. 2. The absorption spectrum was measured in the visible region. BUFFALO, NEW YORK RECEIVED JUNE 19, 1930
Sodium Fluorenone as a Dehydrating Agent1
a capillary capable of being broken and thereby opened by a magnetic hammer, carried four sideBY H. E.BENTAND H. M. IRWIN,JR. tubes. The first two wete empty and used as a check on the method. The third and fourth conThe most rapid method of drying an organic tained activated alumina and magnesium perliquid is to dissolve in it some compound which chlorate, respectively. The samples of alumina reacts rapidly with water. For this purpose soand magnesium perchlorate were prepared by dium benzophenone has been widely used alexposing them in a thermostat to one liter of air though this involves using an alkali metal, sodium-potassium alloy or some other material saturated with water vapor at 25", The quanwhich introduces considerable hazard into such tity of water absorbed by the sample was at least work. Fluorenone has the advantage over benzo- one hundred times as much as was necessary to phenone that it will react readily with dilute destroy the sodium fluorenone with which it came sodium amalgam to form an addition compound in contact, while still being small enough to give which has a deep green color and therefore like the very low vapor pressure characteristic of the benzophenone gives a definite indication by its dehydrating agent. The tubes containing the color of the absence of water. Dry solvent is solid dehydrating agents were evacuated for a short time before sealing the apparatus, and imthen obtained by distilling from this solution, The object of this investigation is t o compare mediately the capillary was broken in order to the efficiency of sodium fluorenone with the cus- prevent diffusion from one solid to the other. The tomary drying agents, magnesium perchlorate fluorenone was decolorized in the first empty tube and activated alumina, Since the compound on account of moisture on the glass and in the formed by the addition of sodium to fluorenone vapor state. In the second tube the solution reis more stable than that formed with benzophe- tained its color. In the third and fourth tubes none, it is important to be sure that the ketyl is the solution was immediately decolorized by the still sufficiently reactive to be an effective drying alumina and by the magnesium perchlorate, indicating that the vapor pressure of water was agent. A dilute solution of sodium fluorenone was pre- sufficiently high to be removed by the ether solupared for this work by introducing 0.025 g. of tion of sodium fluorenone. A duplicate run verifluorenone and evacuating with a mercury pump fied the conclusion from this experiment. The vapor pressure of water is given as 0.003 for one hour. Three to five cubic centimeters of a saturated amalgam (about 1%) was then added mm. for alumina and the water is given as unin vacuo and 50 cc. of dry ether. Shaking pro- weighable in 210 liters for magnesium perchloduced a solution which was sufficiently concen- rate.a If unweighable means less than 1 mg. this trated to react with at least one hundred times as would give a vapor pressure of less than 0.005 mm. much water as is retained by the glass on its sur- If we assume that in the presence of a small face.2 A tube, connected to the reaction flask by amount of water the ketyl is converted to fluorenone and fl~orenole,~ then we can estimate a mini( 1 ) This problem was studied as a reading period assignment in the Erst course in physical chemistry at Harvard University. (2) Bent and Lesnick, Tm5 Jourur~~, 67,1246 (1936).
(3) Yoe, Chsm. News, 180,340 (1936). (4) Bachman, THISJOURNAL,66,1179 (19%).
NOTES
Oct., 1936
2073
small ampoule, The flask could then be saturated again and the vapor a second time condensed and in this way any desired amount of material obtained. Since the sodium hydroxide is present as a solid The method was checked by determining the we may express the equilibrium constant by the vapor density and the molecular weight of triequation phenylmethane. The result was high by about K = (R)llt(RH#/S/(NaR)(H20) 20% which we assupe to be due to adsorption. If we assume, as a consequence of the complete As it was expected that difficulty would be endestruction of the green color of the solution, that countered with hexaphenylethane on account of the reaction goes at least 99.9% to completion, decomposition a procedure was adopted with tnphenylmethane which would avoid this difficulty then we obtain as far as possible. The ground joint to the pump K = 0.5 X 999 X 23/0.003= 3.8 X 106 was opened just enough to reduce the pressure to referring the vapor pressure of water to that of about three-fourths of the saturation pressure. pure liquid water as the standard state. This After the pressure had become constant the valves gives a value for A F of -9.0 kcal. These figures to the pump and to the capsule of sample were represent the minimum efficiency of sodium closed simultaneously and the valve to the refluorenone as a drying agent. In any normal ceiver opened for a time sufficiently long to conprocedure a large excess of the drying agent dense about half of the total amount of material would be used which would result in the removal in the ten-liter flask. The valve to the receiver of water until the pressure is at least as low as was then closed and the pressure again measured, lo-' mm. the difference between the initial and the final CONTRIBUTION BROM THE RECEIVED JULY30, 1936 pressure giving a measure of the amount of mateCHEMICAL LABORATORY rial condensed. This process was repeated about HARVARD UNIVERSITY ten times and the flask then cooled and the reCAMBRIDGE, MASS. ceiver cracked off. The amount of material was determined by weight, both by difference and by evaporation of the solution of the compound to dryness in a crucible. This method of operating The Vapor Density of Hexaphenylethaae between three-fourths and three-eighths saturaBY H. E. BENTAND E. S. EBERS tion would also be effective in avoiding condensaIn a recent publication the statement is made tion in the ten-liter flask or in removing adsorbed that hexaphenylethane probably does not dis- material from the surface of the glass during consociate in the vapor state.' In order to verify densation. When the method was applied to hexaphenylthis statement an attempt was made to measure ethane it was found that decomposition was too directly the vapor density of hexaphenylethane. rapid to permit even an approximate determinaThis measurement combined with the previous tion of the vapor density. In a flask of such a data on the vapor pressure would give the molarge size decomposition was so great that even lecular weight in the vapor state. with the pump operating the pressure remained Since the vapor pressure of hexaphenylethane several times that of hexaphenylethane for a space at 120' is only 0.004 mm. a very large volume at 120' (the valve to the solid of three hours would be necessary in order to produce a weighremaining open). By this time apparently all of able amount of compound. For this purpose a the ethane had decomposed for the pressure graduten-liter flask was attached to the fiber gage used in the previous research and to a capsule of hexa- ally decreased. However, the material in the phenylethane. Ground joints controlled by elec- flask was no longer hexaphenylethane but a detromagnets were so arranged as first to saturate composition product for on closing off the flask the ten-liter flask with the ethane and then, clos- the pressure remained perfectly constant. Under ing the ethane capsule, condense the vapor in a these conditions hexaphenylethane decomposes rapidly enough to be measured easily. We con(1) G . R . Cuthbertson and H. E. Bent, Tsrs JOURNAL, 68, 2000 (1036) cluded, therefore, that it is not possible to measmum value for the equilibrium constant for the reaction NaR HzO = NaOH */aR '/&lH,
+
+
+
