June, 1940
COMMUNICATIONS TO TRE EDITOR
1629
that the substance surpasses phthiocol by far in converted in part into the comparatively inactive vitamin K activity. The observation that the phthiocol. MEMORIAL LABORATORY L.F. FIESER pure oxide [Fieser, Campbell, Fry and Gates, CONVERSE UNIVERSITY THISJOURNAL, 61, 3216 (1939)l is fully effective HARVARD CAMBRIDGE, MASS. in the chick assay at a dosage of 5 y prompted RESEARCH LABORATORIES, M. TISHLER the investigation of other oxides of 2-alkyl and MERCKAND Co., INC.,AND MERCKINSTITUTE FOR THERAPEUTIC RESEARCH 2,3-dialkyl 1,4-naphthoquinones, and i t was RAHWAY, w.L.SAMPSON NEWJERSEY found that the hydrogen peroxide procedure RECEIVED APRIL20, 1940 [Fieser, et al., loc. cit.] provides a convenient route to a number of substances of both types. Farne- MISCIBnITY OF CARBON DIOXIDE AND WATER UNDER HIGH PRESSURE sylnaphthoquinone oxide [found: C, 78.88; H, Sir: 8.161 and phytylnaphthoquinone oxide [found: C, 79.91; H, 9.761 were obtained as nearly colorWiebe and Gaddy [THISJOURNAL, 62, 815 less oils showing, respectively, very weak and (1940)l report the solubility of carbon dioxide in weak (500 7) antihemorrhagic activity. These water and make the following statement: “The substances are cleaved by alkali to a mixture of compositions of the gas and liquid phases a t 12’ 2-hydroxy-l,4-naphthoquinoneand its 3-alkyl and 600 atrn. were identical.” My understandderivative. Crystalline oxides were obtained ing of the behavior of carbon dioxide-water sysfrom 2,3-dimethyl-l,4-naphthoquinone[m. p. tems is that compression a t 12’ to about 47.7 104-104.5’, found: C, 71.26; H, 5.09; active atm. will change the phases from a gas (predomia t 25 71 and 2-methyl-3-cinnamyl-1,4-naphtho-nantly carbon dioxide) and liquid (mostly water) quinone [m. p. 85-86’, found: C, 78.94; H, 5.471. to a gas (mostly carbon dioxide), a liquid (mostly Vitamin K1 was converted in nearly quantita- carbon dioxide), and a liquid (mostly water). tive yield into the 2,3-oxide, which was obtained Further attempts to compress the system will as an almost colorless oil [found: C, 79.85; H, cause the gas phase to change to a liquid (mostly 9.691. Any uncertainty as to the structure is carbon dioxide) phase and leave a t its disappeareliminated by the observation that the absorption ance two liquids, one predominantly water and the spectrum corresponds closely with that of 2,3- other predominantly carbon dioxide. dimethyl-1,4-naphthoquinoneoxide. The K1 oxTheir data show that compression of the two ide shows antihemorrhagic activity of about the phases increases the carbon dioxide content of the same order as the vitamin (1.5 7) and gives no water phase only from 0.03 mole fraction to less purple-blue color with alcoholic alkali. The than 0.04 mole fraction. Lowry and Erickson properties of the oxide are of interest in connection [ibid., 49, 2729 (1927)] indicated that the water with the reports of Fernholz, Ansbacher and co- concentration in the liquid carbon dioxide was workers [THISJOURNAL, 61, 1613 (1939); Proc. much lower than the above concentration of carSOC.Exptl. Biol. Med., 42, 655 (1939)l stating bon dioxide in the water phase. Wiebe and that they have isolated from alfalfa concentrates Gaddy [ibid., 61, 315 (1939)] state the mutual a nearly colorless substance of high potency which solubility of water and carbon dioxide as liquiddoes not give the Dam-Karrer color test charac- liquid system is affected by pressure to only a teristic of vitamins K1 and Kz. However, in our slight extent. hands, there is no great difference in the ratio of These statements appear contradictory to the the effective dose of the oxide and of vitamin K1 above quotation of identical phases. Any critical when assayed by the six and eighteen hour method. phenomena in this region would be of the liquidWe have found that the oxides of vitamin K1 liquid critical solution type and not of the type and of methylnaphthoquinone can be reduced reported by Kuenen [Commun. Phys. Lab. Univ. smoothly t o vitamin K1 hydroquinone and methyl- Leiden, 4 (1892)l for the carbon dioxide-methyl naphthohydroquinone with sodium hydrosulfite chloride system. To become mutually soluble, in aqueous alcohol, even a t room temperature. it is normally expected that the two phases apThis lends plausibility to the hypothesis that the proach each other in composition. A considerahigh potency of the oxides is due to a reduction in tion of the probable behavior of these two phases the organism to the corresponding quinones or upon compression indicates a possible explanation hydroquinones; possibly the 2-methyl oxide is of the above contradictions. At 10 to 20°,
1630
Vol. 62
COMMUNICATIONS TO THE EDITOR
carbon dioxide is relatively close to its critical temperature and is more compressible than water. At low pressures the predominantly water phase is the more dense but the increase of pressure to about 200 atm. and about 500 atm. a t 10 and 20°, respectively, would cause the two immiscible fluids to reach the same density, I know no reason to believe that the same density makes the two phases miscible but the formation of a relatively uniform suspension of the two phases might occur. Unless the possibility of this reversal in the positions of the two phases in the equilibrium cell were recognized, experimental results might indicate a critical state had been reached. Details of the manipulation of the equilibrium cell and quantitative results a t the high pressures in the reported region of complete miscibility should show whether the proposed explanation of the unusual results is tenable.
be zero in all cases. Thus in every type of exchange studied certain bonds are broken and others are formed, but the latter are always identical with the former except for slight steric effects, For AH = 0, K = eAS0lR,and As0 for random distribution should be measured by the relative external symmetry numbers of the molecules involved in the equilibrium; e. g., in the type reaction RaM 12
cr =
AS: =
R
+ R:M
2RzR;M 12 2 123 In - = 7.11 units or K = 36 2‘
(1)
where is the external symmetry number. By the nature of the reactions studied, the only remaining appreciable contributions t o As0 would lie in the effect of the redistribution of mass on the translational entropies, and on the redistribution of principal moments of inertia on the possible rotational entropies. Based on six specific reDEPARTMENT OF CHEMICAL ENGIXEERING UNIVERSITY OF MICHIGAN distributions from systems typified by the leftANNARBOR,MICHIGAN D. L. KATZ hand side of equation 1, the former of the two RECEIVED APRIL18, 1940 mentioned contributions (i.e., mass redistribution) averages 0.076 entropy unit (a heat effect of about 27 small calories a t 350’K.); and, assuming REDISTRIBUTION REACTIONS stretched molecules and free rotation, the average Sir: of the upper limit of the second contribution to Recently Calingaert and co-workers’ have deA 9 mentioned above is, for the same six reactions, scribed a “hitherto unrecognized type of intermolecular exchange of organic radicals” which 0.86 entropy units (a heat effect of about 300 small they term “redistribution reactions.” These re- calories a t 350OK.). It is therefore not surprising that the controlling actions are characterized by equilibrium constants factor in determining the equilibrium state for which are independent of temperature through the such reactions is the value of A$, or randomness temperature range employed, and which have of distribution. values that agree with the idea of a random disOF MISSOURI tribution of the exchanging radicals to within the UNIVERSITY COLUMBIA, MISSOURI ALLENE. STEARN precision of the data obtained. RECEIVED MARCH11, 1940 I n view of the large amount of experimental work done by these authors already, it seems justifiable to point out that, while such reactions have THE ISOMERIZATION EQUILIBRIUM OF %-BUTANE AND ;-BUTANE AND THE THIRD LAW OF not hitherto been recognized as a type, their existTHERMODYNAMICS ence need not surprise one; and, furthermore, Sir: given any of the reactions so far studied, the Recently complete measurements in this Laboraequilibrium results obtained could have been pretory of the thermal properties of the two butanes dicted about as closely as the experiments justify from 11OK. to their respective boiling points furthe idea of random exchange of radicals. nish values for the entropies of the gases a t their The significance of these reactions lies, first, in normal boiling points. For n-butane S&,660~. = the fact that, to the precision to which a modified 72.05 * 0.2 e. u.; for i-butane S&1,440~, = 67.54 * Redgrove2 rule would apply, values of AH would 0.2 e. u. These values, together with available (1) Calingaert and Beatty, THISJOURNAL, 61, 2748 (1939); heat capacity data on the gas, yield a value of Calingaert, Beatty and Neal, ibid., 61, 2755 (1939); Calingaert and Soroos, ibid., 61,2758 (1939); Calingaert Beatty and Hess, i b i d . , AS,& = -3.7 =t0.3 e. u. for the reaction 61, 3300 (1939). (2) Redgrove, Chem. News, 116, 37 (1917).
n-butane(g) --+ i-butane(g)
