NOTES
May, 1959 We have studied the problem and summarize our results in this note. Experimental Matheson C. P. grade ethylene and Matheson nitric oxide (min. 99%) were degassed by freeze-pump technique and subjected to bulb-to-bulb distillation on a vacuum line. The only detectable impurities in the purified gases were 0.007% ethane in ethylene and 0.3% nitrous oxide in nitric oxide. The irradiation vessel was made of Pyrex glass and equipped with break-off seal and capillary constriction at opposite ends. Four fuel elements from a Materials Testing Reactor, shielded by 5.5 meters of water, were used as an irradiation source. All experimental data were obtained a t 22.5 ==! 1’ (the temperature of pool water). Gamma field intensity was obtained by ionization-chamber technique, calibrated with a cerious-ceric chemical dosimeter.2 Ionization chamber readings before and after each run were averaged. In a week of irradiation the 7-field intensity decreased 14%. After irradiations, gaseous products were analyzed by gas-liquid partition chromatography.
Results and Discussion Gamma irradiation of ethylene yielded Hz, CH4, C2H2,CzH6 and n-C4Hloas gaseous product^,^ with G-values independent of energy input rate. Gvalues were also independent of initial ethylene pressure (Table I) and surface-to-volume ratio of the glass reactors (Table 11). I n order to identify the presence of free radicals, nitric oxide was added to a standard run. The results summarized in Table I11 show that the formation of ethane was . completely inhibited by the presence of 5% nitric oxide. I n electroii bombardment of e t h ~ l e n ehy,~ drogen atoms and methyl radicals are formed by direct detachment and by ion-molecule reactions. Similar reactions could occur in yradiolysis. Hence, ethane probably is formed by free radical reactions, such as
+
753
TABLE 11 EFFECTOF GLASSSURFACE UPONTHE GASEOUS PRODUCTS
FORMATION IN THE RADIOLYSIS OF ETHYLENE Ethylene pressure, 120 cm.; energy input rate, 2 X loL9 e.v./g. hr. Surface area, vel. om. - 1
a
,----
CzHz
G (molecules per 100 e.v.)
Hz
C Ha
CzHe
0.8 3.0 2.2 0.06 0.3G .05 .56 5.70 3.2 1.9 Packed with glass beads ( 4 mm. in diameter).
n-C~Hlo
0.30 .48
TABLE I11 EFFECT OF NITRICOXIDEON THE GASEOUS PRODUCTS THE 7-IRRADIATION OF E T H Y L E N E ~ Ethylene pressure, 100 cm.; energy input rate, 2 X e.v./g. hr. Nitric oxide pressure (cm.)
G-
CaHa
(molecules per 100 e.v.)Hz CaHs
IN
n-CdHe
0.48 0.45 1.8 3.0 0.0 .00 b 1.8 3.6 5.0 .00 .40 1.7 10.0 3.9 .00 .37 1.9 4.1 20.0 .00 .41 1.7 4.1 30.0 a Due to the overlap of methane and nitric oxide peaks in chromatographs using silica gel columns, the effect of nitric oxide on the G-value for methane was not obtained. bGvalue not measured.
MISCIBILITY RELATIONS OF LIQUID HYDROGEN CYANIDE BY ALFREDW. FRANCIS Sncony IlIobil Oil Company, Inc. Research and Development Laboratory, Paulsboro, N . J. Received September l Q !1068
H C& +C2H6 C2Hs fI +C2H6
+
Published miscibility relations of hydrogen cyanide are meager, perhaps because of its toxicity. and Several handbooks indicate complete miscibility with water or alcohol; and distribution a t low CH3 CH3 +C2He concentrations is reported1-6 between benzene and The G-values for hydrogen, acetylene and n-butane water or aqueous solutions. Freezing curves of were not affected by nitric oxide. This suggests binary systems of hydrogen cyanide have been that the reactions resulting in these products do not four observed.6-* Its unique structure suggested studies involve free radicals. of binary and ternary solubilities. Hydrogen cyanide was distilled from a cylinder TABLE I and condensed in a large tube in an ice-bath. It G-VALUESOF GASEOUSPRODUCTS IN ?-IRRADIATION OF was kept a t 0” (its b.p. is 26’) in a glass stoppered ETHYLENE AT DIFFERENT ETHYLENE PRESSURES vessel. All operations were conducted in a hood, Energy input rate: 2 X 10’9 e.v./g. hr. and no odor of cyanide was noted during the inEthylene vestigation. -G (molecules per 100 e.v.)--pressure C2H2 Hz CHI CzHe n-CIHlo (om.) Samples of hydrogen cyanide were taken by a 1120.0 3.2 1.9 0.05 0.56 0.48 ml. graduated pipet connected to a ‘ ( p r ~ p i p e t t e , ” ~ 1.8 a .48 99.8 3.0 .45 and mixed \$ith other liquids in a small graduated .59 a 80.0 3.7 2.0 .04 glass stoppered tube which could be shaken for 60.0 3.7 2.0 .06 .67 .89 equilibrium studies. I n view of the small samples,
+
a
40.0 3.9 2.0 G-value not measured.
.10
.85
.89
(1) E. Collinson and A. J. Swallow, Chem. Revs., 56, 482 (1956). (2) J. Weiss, Nucleonics, 10, No. 6, 28 (1952).
(3) Negative results of Hayward and Bretton to identify these gaseous products may be due to the ineffectiveness of their analytical technique, viz., copper oxide reduction method. (See J. C. Hayward, Jr., and R. H. Bretton, Chem. Eng. Progr. Symposium Seriea, 50, 78 (1954)). (4) F. H. Field, J. L. Franklin and F. W. Lampe, J. A m . Chem. Sac., 79, 2419 (1957).
(1) A. Hantnsch and F. Sebaldt, 2. physik. Chem., 80, 258 (1899). (2) A. Hantzsch and A. Vagt, ibid., 8 8 , 705 (1901). (3) P. Gross and K. Schwarz, Monatsh. Chem., 55, 287 (1930). (4) P. Grossand M. Iser,iMd., 55,329 (1930). ( 5 ) M . Randall and J. 0. Halford, J. A m . Chem. Sac., 52, 192 (1930). (6) “Solubilities of Inorganic and Metal-Organic Compounds,” A. Seidell, Ed., D. Van Nostrand Co., New York, N. Y.,1940, PP.
569-70.
(7) J. E. Coates and N. H. Hartshorne. J. Chenz. Soc.. G57 (1931). (8) A. L. Peiker and C. C. Coffin, Can. J . Research, 8 , 114 (1933). (9) Instrumentation Associates, New York 23,N. Y .
754
NOTES
Vol. 63
.
Fig. 1.-Ternary
systems of hydrogen cyanide.
solubilities are only approximate. Liquids (including condeiised gases) found miscible with hydrogen cyanide were matched for ternary systems with other liquids known to be immiscible with HCN, and vice versa. This gives a better idea af miscibility relations than do binary solubilities alone. A gaseous reagent was condensed into a cooled thick-walled glass tube containing the other components. The tube was sealed, weighed, shaken, then cooled, reopened, more reagent added and resealed, etc. Table I lists 23 liquids found completely miscible with hydrogen cyanide a t room temperature. The “codes” are letters suggestive of the names, referring to graphs in Fig. 1. Table I1 lists 13 liquids not miscible with hydrogen cyanide, with estimates of the two solubilities. From their structures many more substances could be added to each of these groups with reasonable certainty. Thus, most liquid organic oxygen compounds up to Clo and aromatic hydrocarbons up to Cs are com-
TABLE I LIQUIDSCOMPLETELY MISCIBLEWITH HYDROGEN CYANIDE (AT 25’) Code
Solvent
Ac A B BA C CE CF Cr DA D EG FA
Acetonitrile Aniline Benzene n-Butyl alcohol Chlorex p-Chloroethanol Chloroform ni-Cresol Decyl alcohol Dioxane Ethylene glycol Formic Acid
Code
F HBr HCI M N NM Pe S W X Xp
Solvent
Furfural Hydrogen bromide Hydrogen chloride Met,lianol Nitrobenzene Nitroinethane Propylene Sulfur dioxide Water m- and p-Xylenes p-Xylene
pletely miscible with hydrogen c,yanide and would be included in Table I. Olefins above CS and all paraffins and iiaphtthenes are probably only par-
.
u
NOTES
May, 1959 tially miscible with hydrogen cyanide and would be iiicluded in Table 11. TABLE I1 MUTUALSOLUBILITIES OF HYDROGEN CYANIDE. (AT25') Code
BB CD CT Cy De H Hx L MN 0 PF P T
Solvent
sec-Butylbenzene Carbon disulfide Carbon tetrachloride Cyclohexane Decalin n-Hexane 1-Hexene Lauryl alcohol a-Methylnaphthalene 1-Octene Perfluorodimethylcyclohexane Propane Tetralin
Solubility, w t . % Of HCN In HCN
30 5 5 5 3 2 10 20 5 5 2 5 5
30 2 2 5 3 2 2 10 20 5 5 2 2 5
Figure 1 presents 32 ternary systems on 28 graphs arranged concisely as in systems of liquid carbon dioxide.l0 I n each graph the top corner represents hydrogen cyanide. The other components are indicated by letters referring to the codes in the tables. The left hand component is the nonhydrocarbon, when present, or the more polar component. However, the usual measures of polarity are distorted since benzene and xylenes are miscible with hydrogen cyanide while cnrboii disulfide and carbon tetrachloride are not. The majority of the graphs have simple binodal curves on the right side, though the depths of the curves and the positions of the plait points vary. Five of the systems have binodal curves on the bottom and three on the left sides. Graphs 20 and 27 have isopycnicsl0J1 as represented by dashed tie lines indicating equal densities for the two layers. Other tie lines are shown only on Graph 1, based partly on references 1, 2, 0. These tie lines are solutropic (reversing in slope). This system also contains isooptics11*'2(not shown) for some tie lines, indicating equal refractive indices for the two layers. Tie lines in the other systems can be apprmimated from the positions of the plait points. Graph 2 is unusual in showing a binodal curve and a separate island curve, and so has three plait points. The island may be related to the complex noted by Peiker and Coffin.8 This island is larger than in the other published graph of this type, namely, the aqueous system of dioxane with hydrogen chloride. l 3 The system water-hydrogen 1x0mide-dioxane is certainly of the same type. HOWever, there is no island in the HBr-HCN system, Graph 3. Two graphs, 6 and 11, represent three systems each because those of each graph are identical within the accuracy of the observations. An actual extraction in the benzene-hexane system, graph 20, indicated a fair selectivity of hydro(20) A. (11) A. (12) A. (13) A.
W. Francis, THIBJOURNAL, 56, 1099 (1954). W. Francis, Ind. Ene. Chem., 46, 2789 (1953) W. Francis, THIEJOURNAL, 66, 510 (1952). W. Francis, ibid., 62, 579 (19581, Fig. 2.
755
gen cyanide for benzene. However, other less toxic substances show higher selectivities. Acknowledgment.-The author is indebted to his colleague, Dr. R. D. Offenhauer, who suggested the investigation, and prepared the sample of hydrogen cyanide. -
A SPECTROPHOTOMETRIC INVESTIGATION OF OUTER-SPHERE ASSOCIATION OF HEXAMMINECOBALT(II1) ION AND HALIDE ION1 BYEDWARD L. KING,JAMES H. ESPENSON A N D ROBERT E. VISCO
Department o/Chemislry, Uniuersitv W i s c o n s i n , Madison 6 , Wisconsin Receiued Seplcmbw 87, 1068
This paper presents the results of experiments which indicate that the values of &,, the equilibrium quotients for the reactions Co(NH3)6+3 + Sc0(x!&)6+3.xwith X- = Cl- and Br- are much lower than the values reported by Evans and Nancollas.2 Very sinall values of the absorbancy enhancement have been relied upon in this earlier work t o obtain values of The present study was made with the coiicentration of hexamminecobalt (111) ion held a t a lorn and constant vnlue and the concentration of halide ion varied up to 0.9 AI. The results obtained in the present study at X 250 mp for X- = C1- and X 2G0 mp for X- = Br- are presented in Fig. 1. It is seen that the absorbancy enhancement in each series of measurements is, within the experimental uncertainty, proportional to the concentrstion of halide ion over the ent.ire range of concentration. Curvature in a plot of the apparent molar absorbancy index of hexamminecobalt(III) ion (a = (log Io/I)/[CoIII]b, where b is the cell length and [CoIII] is the stoichionietric molar concentration of hexamminecobalt(II1) ion) versus [X-] sets in with increasing [X-] when Q1[X-] becomes appreciable compared to unity, as shown by the equation
where a0 and nl are the molar absorbancy indices of Co(NH3)e+ 3 and Co(NH3)fi+3.X-, respectively. If Ql[X-]
