TRANSITORY PRODUCTS IN THE REACTION OF ETHYLENE WITH OZOXE
Jan., 1960
of the current plateaus is about 15%. However, one can see that the thickness does definitely change with wire diameter and that the values do not depend markedly on viscosity or temperature. The points a t room temperature are the most reliable since each is the average of about five determina-
9
tions, while each one at the other temperatures is the average of only two determinations. Acknowledgment.-The authors wish to express their thanks to Professor Karl F. Herzfeld of the Department of Physics, The Catholic University of America, for very helpful discussions.
TRANSITO'RY PRODUCTS IS THE GAS PHASE REACTIOS OF ETIIYLESE K I T H OZOKE BY XRTHUR
E. HEATH, SAMUEL J. BROADWELL, LOWELLG.
W A Y S E AND 1'ht-L
P. 31.4DER
Los Angeles County Air Pollution Control District, Research Division, Los Angeles, Cal. Recezied FebTuaTy 63,1959
In a study of the rates of reactions between ozone and various olefins a t low concentrations in air, Cadle and Schadt' sought but did not find evidence for the existence of transitory reaction products in the ethylene-ozone system. We have observed such products recently by means of their infrared absorption in a one-meter path. The evidence is definitive for the existence of a t least two such substances and one or more others may be involved. Reactions were carried out in a darkened Vycor cell, one meter long, placed in the beam of a doublemonochromator recording spectrophotometer. Partial pressures ranged from 700 to 3500 microatm. for ethylene and from 500 to 1000 microatm. for ozone. Some forty experiments in which the reagents (diluted with oxygen to about one atmosphere) were intimately mixed by simultaneous introduction in to the cell through a single stopcock consistently confirmed the order of magnitude of the rate reported by Cadle and Schadt for the disappearance of ethylene. Within five to ten minutes the concentration of ethylene or of ozone became substantially stationary, the other reagent having disappeared from the system. In two tests, spectra recorded within five minutes after mixing showed strong development of a peak a t 8.6 p, which was much reduced in subsequent spectra. Six tests in which the absorption a t 8.6 p was continuously recorded showed that this peak reached a maximum within 30 to 60 seconds after mixing; it then subsided, with a half-time of perhaps five to ten minutes. Formic acid, recognized by its triplet band centered at 9.0 p, accumulated much more slowly. When the reactants mere introduced into the cell by a technique which avoided rapid mixing, there was strong development of peaks which did not belong to formic acid nor to the compound absorbing at 8.6 p , which we call Intermediate I. These new peaks subsided after several hours, furnishing clear evidence of the presence of other transitory compounds in the system. In these tests the cell (volume, 1.4 1.) was filled to atmospheric pressure with oxygen containing about 0.1% ozone. A few ml. of ethylene then was injected into the cell by means of a hypodermic needle and syringe through a serological stopper near (1) R.Cadle and C:. Schadt, J . Am. Chem. Soc., 74, 6002 (1952).
one end. Under these conditions, reaction rates were limited by diffusion, so that the time available for observing possible transitory products was prolonged; about an hour was required for the substantial disappearance of one reactant. Figure 1 illustrates changes observed in the spectrum of one diffusion-limited reacting system containing excess ethylene. After two hours in the dark the spectrum (dotted line) showed no traces of
J
I
I
Fig. 1.-Near infrared absorption spectra of reacting system of ethylene and ozone in air, with diffusional mixing. Dotted line shows spectrum after 2 hours; solid line, after 18 hours. Ethylene appears in region 9.9 to 11.2 microns; formic acid a t 5.6, 5.7, 8.2 and (triplet) a t 9 microns; transitory compounds a t 5.5 to 5.7, 9 and 10 microns.
the ozone peaks (9.5 to 9.7 p ) , but absorption peaks due to products different from Intermediate I were very prominent a t 5.6 to 5.7, 9.0 and 10.0 p. After 18 hours in the dark, the 10-p peak had largely subsided, while typical formic acid peaks were well developed a t 5.6, 5.7, 8.2 and 9.0 p . The changes in the interim can only have been caused by the disappearance of one or more transitory products which decompose or react as formic acid accumulates. The absorption a t 10 p is especially intriguing; it has been observed clearly in only three instances, although in several other spectra its presence appears likely. Comparison of our spectra with published spectra for pure compounds2-4 reveals that ethylene
J. BRAUNSTEIX ASD M. BLANDER
10
oxide and ethylene ozonide were substantially absent from the reacting mixtures, and that the bands a t 8.2, 9.0 and 10.0p were not due to formaldehyde, acetaldehyde or acetic acid. Performic acid has main peaks at 9.0 and 5.6 p4and thus seems a likely constituent in the diffusion-limited systems. It is worth noting that the possible presence of these transitory compounds in smog-laden atmospheres could provide an explanation for dis(2) R. H. Pierson, A. N. Fletcher and Et. St. c. Gantz, Ind. Eng. Chcm., 28, 1218 (1956). ( 3 ) D. Garvin and C. Schubert, THISJOURNAL,60,807 (1956). (4) P. A. Giguere and A. W. Olmos, Can. J . Chem., S O , 821 (1952).
T’ol. 64
crepancies between chemical6 and spectroscopic6 estimates of formic acid concentrations in the atmosphere; that is, one or more of these intermediates might be chemically indistinguishable from formic acid by the usual analyses, while its absorptions would differ from those of formic acid and perhaps obscure them. This research has been discontinued, (5) P. P. Mader, G. E. Cann and L. Palmer, Plant Physiology, 80, 318 (1955). (6) W.E. Scott, E. H. Stephens, P. L. Hanst and R. C. Doerr, Paper presented at meeting of American Petroleum Institute’s Division of Refining, Philadelphia, Pa., May 14. 1957.
THE THERMODYNAMICS OF DILUTE SOLUTIOSS OF AgNO, AND KCI I N MOLTEN KNOs FROM ELECTROMOTIVE FORCE MEASUREMENTS. 111. TEMPERATURE VARIATION3 OF THE ACTIVITY COEFFICIENTS BY
J. B R A U S S T E I N ’ AND M.
BLSNDER
Oak Ridge National Laboratory,z P . 0. Box Y , Oak Ridge, Tennessee Received March 86, 1959
The quasi-lattice model of the molten reciprocal salt systems A+, B+, C- and D- previously proposed3 leads to the relaX/p(l - X)lz-’]/ND-, if A + and C- ions tion for the activity coefficient of the component AD TAD = [(I - X ) z [ l have an extra-coulombic interaction, where 2 is the lattice coordination number, p = e-AE/RT, A E is the energy of formation of a mole of A+-C- ion pairs and X is the fraction of positions adjacent to A + ions occupied by C- ions and can be calculated in terms of 2, aE and the concentrations of the ions. The theory should be valid for low vnhies of N.kg+and X . Measurements of the activities of AgN03 were made in the molten salt concentration cell
+
for dilute solutions of Ag+ and C1- ions a t 385, 402 and 423’. KC1 in solution lowered the activity coefficient of AgNO8, the lowering being larger the larger the concentration of C1- ions and the smaller the concentration of Ag+ ions. A comparison of the ex erimental data and the theory a t low values of N A ~(50.720 + X 10-3) and X (
