Mar. , 1956
NOTES
379
The gold chloride probably also exists mainly in the dichloroether as HAuC14, or as has been suggested for indium bromide,I2 as the associated ionpair (H+)(AuCI4-). The crystalline chloroaurates show a planar structure for the AuC14- ionlSand if this exists in such an associated ion pair, the activated complex with HC1 should be in a form favorable for ligand exchange. The role of associated water in such molecules in organic solution is uncertain. The work of Gibson and Colles‘4 indicates that for the bromoaurates water is essential for dissolution in dry diethyl ether, since they find that a mixture of AuBra, HBr arid ether is immiscible in the absence of water, but addition of water causes HAuBr4 to dissolve in the organic phase exothermically. This suggests that the species present in organic soluwhere the symbols represent the moles of each tions has the formula HAu&.nH20 as has been species concerned, and hence the ratio gives the shown to be the case for iron16in dichloroether. Rich and Taube4 have pointed out that for the equilibrium distribution of activity. The results of C1--AuC14- exchange in aqueous solution the two experiments are given below. presence of reductants in the solution can enhance HC1 HAuClr % the rate of exchange. It seems unlikely that in the Run (mmoles) (mmoles) exchange highly purified dichloroether there would be an 2.1 0.040 0.0051 3.1 appreciable concentration of reducing entities and 2.3 0.071 0.0116 20.5 in fact, any impurities produced in this solvent These values show that for low concentrations this would be expected to be oxidants. effect is negligible. The increase in exchange with Acknowledgment.-The senior author acknowlincreasing concentration suggests that there is probably not so much an induced exchange but edges a fellowship under the M.I.T. Foreign rather a fairly rapid isotopic exchange between HC1 Students Summer Project. The experimental work was supported by the U. S. Atomic Energy Comand the resin-AuCL species in these experiments. The exchange between HC1 and the two chlorine mission. (12) L. A. Woodward and P. T. Bill, J . Chem. SOC.(London), I699 atoms of the solvent was briefly investigated to ensure that this could not affect the results. A (1955). (13) A. F. Wells, “Structural Inorganic Chemistry,” Oxford Unisolution of labeled HCl in dichloroether was al- versity Press, 1945, p. 290. lowed to stand for 5 hours, after which l ml. was (14) C. 8. Gibson and W. M. Colles, J . Chem. SOC.(London), 2407 shaken with two 15-ml. portions of water and cen- (1931). (15) J. Axelrod and E. H. Swift, J . Am. Chem. SOC.,62, 33 (1940). trifuged. The nonextracted activity represented only 5% of that calculated for total exchange. For the much shorter time intervals between HCl absorption and the actual exchange experiments THE KINETICS OF T H E CS2-NO REACTION (normally about 30 minutes or less), this effect is AND THE MECHANISM OF LIGHT negligible. EMISSION I N THE EXPLOSIVE Discussion REACTION Taking account of the results presented above, BY WALTER ROTH AND THEODORE H. RAUTENBERG one can set an upper limit of approximately 10 seconds for the half-time of the HCl-HAuC1, exchange Qeneral Electric Research Laboratory, Schenectady, New York in the 0.01f concentration region. Received October 8, 1066 It has been emphasized elsewheregJo that the rate of any reaction between two ions of the same There has been a recent renewal of interest in charge should decrease as the dielectric constant D flames supported by nitrogen Van of the solvent decreases. For dichloroether, D = Liempt and de Vriend have studied the composition 21.2, whilst for water, D = 80. Comparison of explosion limits and the light output of the reaction the present results with those of Rich and Taube4 between CSz and N0.486 They have reported shows a decrease by a factor of 20 or more in the luminous efficiencies as high as 83 lumens/watt. half-time for the exchange for similar concentra- In addition, they have estimated an ignition temtions. It seems likely that this apparent contra- perature of 1900” from experiments with melting diction with theory is due t o reaction of uncharged wires. In the course of determinations of P-T species in the organic solution. For HC1 in dichlo- explosion limits for the CS2-N0 system, we have roether conductance measurements show that ( 1 ) E. Bartholome and H. Sschae. Z . Eleklrochem., 68, 326 (1949). this acid is only slightly ionized; the ionization (2) H. Behrens and F. Rossler, Naturwissenschaflen, 86, 218 (1949). (3) W. a. Parker and H. G. Wolfhard. Fourth Symposium on Comconstant“ for ion-pair formation for millimolar bustion, 420-8 (1952); Fifth Symposium on Combustion 718-28 solutions is of the order of 10-7. The specific activities in HC1 and HAuCl were identical within experimental error. By comparison with aliquots of the original HC1* solution, it was shown that a complete activity balance was being achieved. Certain experiments were carried out to investigate the possibility of an exchange induced during the separation procedure. In these experiments, a solution of HAuC14i n dichloroether was run onto the column, which was then washed with the pure solvent. A 1-ml. aliquot of labeled HCl solution in the ether was then passed down the column and eluted in the usual manner. The percentage exchange occurring was calculated from the following eauation activity on resin x100 yo exchange = C1 C1 activity eluted 4 X HAuCL
(1954).
(9) G. Scatchard, Cham h s . , 10, 229 (1932). (10) D. Pemhanski. J . Aim. PLUS., 60, 634 (1953).
(11) A. M. Poakamer, private communication.
(4) J. A. M. van Liempt and J. A. de Vriend, Rec. lrav. chim., 52, 160 (1933). (5) J. A. M. van Liempt and J. A. de Vriend, ibid., 52, 549 (1933).
NOTES
380
Vol. 60
been able to ignite mixtures in a 5-cm. diameter of Lindsay and Bromley? Deviations of calculated spherical silica vessel immersed in a furnace at from experimental results are well within the temperaturesaslow as 775" and, in general, ourigni- errors iiivolved in the experiments and in the caltion temperatures have been of the order of 1000" culations of mixture thermal conductivities. , lower than those estimated by van Liempt and de TABLE I Vriend. The explosion limits are of the thermal type and can be expressed approximately by log EFFECT OF TH0 ADDITION OF INERT GASES ON THE EXPLOSION (PIT) = A/T B6 a t pressures above about 20 LIMITPRESSURE IN CM. IIg OF CS2 3N0 Lower value is calcd. Upper value ia exptl. cm. for mixture composition ratios, (NO)/(CS2), 10.0 from 1 to 8. The slope, A , decreases sharply a t 20.0 40.0 4.8 %A 0 the lower prernures, perhaps indicating an increasT,"C. ing importance of surface reactions a t these pres... ... 37.8 825 35.2 36.1 sures. In a separate series of experiments, a num38.8 36.9 ber of short pieces of silica tubing were put into the ... ... 850 27.2 ... ... vessel such that the surface to volume ratio was 24.0 24.6 ... 23.2 875 21.7 doubled while the diameter of the vessel was not 26.5 ... 23.8 22.5 20.4 ... seriously altered. A comparison of the results is 19.6 19.0 900 18.0 19.8 shown in Fig. 1. It can be seen that the effect of 20.0 ... 18.7 16.7 increased surface area is to increase the limit pres17.2 22.6 15.5 925 15.3 16.8 18.6 23.5 sure a t pressures below about 30 cm. 16.1
-
+
+
950
13.6
...
lo00
12.1
12.5 12.7 5.0
50
% H e ----f 0 T,"C. 850 27.2 35
E 15
875
21.7
900
18.0
925
15.3
950
13.6
lo00
12.1
... 21.8 23.8 19.7 19.2 16.2 16.5 14.1 14.7 12.8 13.0
14.7 15.0 12.5 13.4 10.0
15.2 16.6 13.3 14.8
19.2 20.8
...
31.1 30.1 24.8 25.3 20.2 21 .o 17.6 17.8 15.4 15.8 13.5 14.2
The thermal nature of the explosion limit has been confirmed by experiments with the addition of inert gases to the mixture (NO)/(CS*) = 3. In these experiments, the results of which are given in Table I, it has been found that for a given (km /ku)'/Sbu, where Pm temperature, P , = pg is the explosion limt pressure for the diluted mixture, P , for the undiluted mixture, p d is the partial pressure of the diluent, and k's are thermal diffusivities a t 273OK. The latter have been determined from thermal conductivities of the mixtures which were calculated by the method
Explosions are characterized by a bright blue flash of light and a decrease in pressure. Deposition of sulfur is observed and this increases as the mixtures are made richer in CS2. At temperatures slightly below the explosion limit a slow reaction has been observed. The rate of this reaction was measured a t 750" for the mixture (NO)/(CS2) = 3.00 and, despite the expectation that non-isothermal conditions would obtain, was found to fit a third-order rate law, second order with respect to NO and first order with respect to CS2, with remarkable precision. The measured rate conmm.-2 set.-'. Therstant was 7.7 i 0.2 X mal decomposition of NO8 could not have resulted in an error greater than 3% in the rate measurement. There is evidence which indicates that the reaction is non-chain, and of the third order, even in the explosive region. Since the reaction is thermally propagated, the rate of heat production must be equal to the rate of heat loss a t the explosion limit. An equivalent statement is that a critical rate of heat production must be attained for explosion to become self-propagating. Then, if the reaction is second order with respect to NO and first order with respect to CS2,it can be shown
(6) N. Semenoff, "Chemical Kinetics and Chain Reactions," Oxford Univ. Press, 1935. pp. 79-83; D. A. Frank-Kamenetzky, d e h Physicochim., 10, 365 (1939).
(7) A. L. Lindsay and L. A. Bromley, Ind. Eng. Chem.. 62. 1508 (1950). (8) H.Wise and M. F. Frech, J . Chem. Phya., '20, 22 (1952).
' 5O I
I
,
TEMPERATURE, 'C.
+
Fig. 1.-The effect of surface area on the CSs 3 N 0 explosion limit: A 5 cm. diameter vessel; 0,aame vessel with S/V doubled.
+
NOTES
Mar., 1956 that the composition-pressure explosion limit should have a minimum in pressure at a mole fraction of NO of 2/3. This is actually the position of the minimum indicated by the work of van Liempt and de Vriend4 and confirmed in our experiments. In addition, the 1/3 power dependence of pressure on thermal diffusivity indicates a n overall third-order reaction. The explosion limit for a reaction exhibiting third-order kinetics is more accurately represented by log(P3/T2) = A / T B where the slope A = (E/2.303R). From the average slope, we have calculated an activation energy of 70 f 5 kcal. This is in agreement with the heat of dissociation of an S-atom from CS2. The latter value is about 70-80 kcal.; a specific value connot be determined because of uncertainties in the heats of dissociation of CS and sublimation of carbon. If the reaction involved free radical chains, it would be expected that 0-atoms would participate. I n this event, evidence for the reaction NO 0 + NO2* should be detected. We have photographed spectra of CS2-NO and CS2-O2 flashes in the visible region, using a Hilger Medium Quartz Spectrograph with a slit width of 25 1.1 and Eastman 103-F plates. Ignition was accomplished with exploding 1 mil tungsten wires. These spectra exhibit identical feat!res. Two continua are evident, one from 4900 A. to the short wave length limit of the explosion vessel (3100 A.) and the other from 5500 to the long wave length limit of the plate (7000 A.). Superimposed on the continua are weak SO bands and stronger S2 bands. As the pressure is increased the bands are observed to decrease in intensity while the continua increase in intensity. At a sufficiently high pressure, the bands disappear entirely. Most important is the fact that neither of the continua can be ascribed to the reaction NO 0 + NO2* since this could not occur in the CS2-0 flashes. Furthermore, the region 4900is one of very low intensity, whereas the 5500 continuum of NOz* + NO2 hv is strong in this region. In addition, mass spectrometric analyses indicated that no NOz was present among the products even when NO was initially present in exce&. These facts rule out the reactions SO 0 + SO2*, CS 0 + COS*, and CO 0+ C02*, as emitters of the continua in CS2-NO flashes. Nevertheless, the continuum from 4900 A. to the violet was observed to go through an intensity maximum with variation in initial mixture composition at the same composition at which mass spectrometric analyses of products indicate a maximum in SO2 production for both CS2-NO and CSz-O2 flashes. This would indicate that an excited SOz is the emitter of the short wave length continuum. Furthermore, this continuum appears to be the same one observed by Gaydong in the 0 + SO2* SO2 after-glow. If the reaction SO occurs during CS2-O2 flashes, the energy of the reaction is enough to take the continuum down to about 1900 8. Formation of SO8 could not be responsible for light emission since this compound is never observed among the products.
+
+
4.
+
1.
+
+
+
+
(9) A.
0.Gaydon, Proc. Roy. &e. (London). A146, 901 (1934).
+
38 1
Third-order gas phase reactions involving NO are thought by some to proceed via the formation of a collision dimer of NO. If this dimer were to lose one 0-atom in reaction with CS2 or CS, there is the possibility that an NzO residue would remain. N20 is indeed observed among the products, but appears only when NO is initially in excess a s might be expected since N 2 0 can react further with CS2. The presence of N20as an intermediate suggests a possible reaction for the formation of excited SO2. This may be formed by SO( *X)
+ N20( '2)+SOz('Z) + Ne( 'Z)
where, if the N2 is produced in its ground electronic state, the Wigner-Witmer correlation rules'o indicate that the electronic state of the SO2 must be a2. Magnetic susceptibility measurementsL1on gaseous SO2, indicate that it is diamagnetic. This establishes its ground state as a singlet state. Radiation of the continuum would, in this case, involve a forbidden triplet-singlet transition. However, the selection rules may be weakened somewhat by a non-linearity of the SO2 analogous to the situation encountered in the C02 continuum observed in CO-02 explosions.12 I n addition, the heavy Satom would be expected to contribute to this weakening. From the heat of the SO-N20 reaction, we can set an upper limit for the position of the excited state of SO2 above the ground state of 110 kcal. if SO is in its ground state, and 220 kcal. if SO is in the excited state which results in radiation of the SO band system. I n both cases, the upper limit results.when it is assumed that all of the reaction energy is carried away by the SO2 and none by the N2. The long wave length continuum goes through an intensity maximum at a mixture composition which is different from the optimum for the SO2 contiriuum. It is felt that COS(311) is responsible The for this radiation by analogy with CO#II). reaction CS 0 + COS involving an 0-atom has been ruled out. Furthermore, this would have enough energy to take the continuum to 1900 A. The reaction CO S -+ COS* can, on eneTgetic grounds, result in a continuum to 4750 A. On the basis of arguments already discussed, the COS would have to be formed in a triplet excited state. COS has been found among the products of the reaction when CS2is initially in excess.
+
+
(10) E. Wigner and E. E. Witmer, 2. Ph,ysik, 61, 859 (1928). (11) P. Pascal, Compt. rend., 148, 413 (1909). (12) R. S. Mulliken, J . Chem. PAYS.. 8 , 720 (1935).
EFFECT OF IODINE ON THE RADIOLYSIS OF BENZENE' BY ROBERT H. SCHULER Contribution jrom the Chemislru Department Brookhaven National Laboratory, Upton, L. I., N: Y . Received October 4, 1866
Burton and Patrick2 have shown that the hydrogen yield from the radiolysis of benzene is little affected by the presence in solution of sub(1) Research performed under the auapices of the U. s. Atomic Energy Commieeion. (2) M. Burton and W.N. Patrick, J . ch6hstn. Phgs., PI. 1150 (1954).
