I 5 3 TRICHLOROMETHYL RADICALS THE KINETICS OF THE INTERACTION OF TRICHLOROMETHYL RADICALS WITH CYCLOHEXENE BY H, W, MELVILLE, J. C. ROBB AND R. C. TUTTON Received 2nd February, 1951 This paper outlines the technique adopted in obtaining kinetic data for the photochemically induced reaction between trichlorobromo methane and cyclo- hexene with the formation of I-trichloro-methyl 2-bromo-cyclohexene. The standard methods which have been previously developed in the investigations of high polymer reactions have been modified to suit this reaction. The reaction exhibits chain properties and will be specially useful in measuring the absolute reactivities of olefinic or unsaturated compounds in general towards a specific and unambiguous free radical, CCl,. It also gives valuable information concerning reactions of the type responsible for the phenomena of transfer in the polymerization of vinyl compounds in a way that is not possible by other means .Preliminary results indicate that the reaction CCl, + C6H,, proceeds with a velocity constant k , = 2-56 x loa 1. moles-l sec.-l a t 30' C and a termination velocity constant zk4 = 1.0 x 108 1. moles-lsec.-I, being that for z CC1, -+ C,Cl,. The overall energy of activation is 3-4 kcal./mole and it is indicated that the energy of activation for termination of the chains is also low. Quantitative radical chemistry has now progressed to such a stage that not only is the mechanism of certain reactions thoroughly known but the velocity coefficients of the constituent radical reactions have been determined with reasonable accuracy.In order to build up a general picture of radical reactivity both towards molecules and radicals it is essential to explore the possibilities in reactions other than those involving polymerization and oxidation. One type of reaction which appears to be suitable is the addition of simple organic halogen compounds to ethylenic molecules. Kharasch and his co-workers have investigatedH. W. MELVILLE, J. C. ROBB AND R. C. TUTTON 155 the chemistry of a number of reactions of this kind and have shown that they take place to form well-defined products, that they can be initiated easily photochemically and proceed by a radical chain mechanism.1 A typical example is the addition of trichlorobromomethane to cyclohexene. The terminal product has the structure CH2-CH2-CHCCI I I CH2-CH2-CHBr.It is suggested that initial fission of the halomethane is into Br + CCl,, that the trichloromethyl radicals attack cyclohexene, and the radical then formed abstracts a bromine atom from another molecule of bromo- trichloromethane. The CCl, radical starts the process and the cycle comes to an end by the interactions of the hydrocarbon and trichloromethyl radicals. It should be possible by so arranging the conditions of re- actions that the two rate determining steps would be CCl, + cyclohexene and CCl, + CCl,. A comprehensive investigation of these reactions is of importance for many reasons. Radical double bond interactions are the rate deter- mining step in addition polymerization. By experiments on copolymer- ization systems it is possible to measure such interactions in which the attacking radical is kept fixed in structure and its absolute reactivity towards a whole series of monomers investigated.Br . CCl, is further a good transfer agent in polymerization and the rate determining step is the debromination by a polymer radical. Nothing, however, is known of the rate of addition of the CCl, radical to start the growth of a new polymer radical. Recent experiments in the addition of atomic hydrogen to olefines have thrown some light on the factors that affect chemical reactivity in this system.2 It is important to explore other systems in which the nature of the attacking radical is varied. Radical-radical reactions are of great interest since although in general little energy of activation is needed in these encounters, quite highly specific factors enter into such processes.Further data are required to see how structure affects the interaction coefficients in this kind of reaction. The present paper is specifically concerned with the elucidation of the precise mechanism of the addition of trichlorobromomethane to cyclo- hexene initiated photochemically, together with the determination of the velocity coefficients of the constituent radical reactions involved. Towards the end of this investigation the authors became aware of a similar investigation being carried out on the interaction of CCl , radicals with methallyl acetate., Theoretical.-The most suitable method for producing CCl radicals in a controlled fashion is by the direct photochemical decomposition of trichlorobrornomethane CCI , .Br which absorbs light of wavelength around 3650 A, which has sufficient energy to rupture the C-Br bond, (I) CC1,Br + hv -+ CCI, + Br I Initiation The reaction mechanism is as follows : 1 Kharasch, et al., J . Amer. Chem. SOG., 1947. 69, 1105 ; J . Org. Chem., 1949, 14979s 239,537- Melville and Robb, Proc. Roy. SOC. A , 1949, 196, 445, 466, 479, 494 ; 1950, 202. 18T. 3 Matheson and Halter (private communication).TRICHLOROMETHYL RADICALS (4) 2 CCl, +- C,CI, 2 c> CCl, 3 \ / k5 I Termination k6 I By arranging the experimental conditions such that the unsaturate con- centration is low, reaction (2) can be made rate controlling, in which case a normal kinetic analysis leads to the expression If conditions are such that low CCI, .Br concentrations are employed and reaction ( 3 ) becomes rate determining, assuming long chains, By the application of standard techniques, developed to deal with polymerization and chain reactions, the kinetic constants in eqn. ( z ) , (3), (4) and ( 5 ) can be evaluated. The remaining constant for reaction (6) can be obtained by working in a region of intermediate concentration such that all three termination processes occur and, having previously determined the others, it is a relatively simple task to evaluate k6. The bromine atom produced in the initiation step is used up, presum- ably in addition to cyclohexene. If this radical is capable of propagating a chain, it means that two chains are started for each molecule of CCl, .Br decomposed. This does not affect the picture since the inhibitor tech- nique used later measures the number of chains starting, irrespective of the nature of the initiating radical atom. The technique adopted for following the progress of reaction was by dilatometry. This necessitates knowledge of the densities of initial and final components in the system, but in this case, direct measurement of the density of the product is somewhat difficult and liable to introduce errors of considerable magnitude. The method adopted was to make use of the fact that due to utilization of unsaturate, the reaction rate decreases with time. Let xco be the meniscus drop for roo yo reaction (at t = a), xt be the meniscus drop for a time t, Y o be the amount of unsaturate in the reaction vessel at time t = 0, Yt be the amount of unsaturate in the reaction vessel at time t.Then dxjdt = C’Yt, . . (3) where C’ is a constant. Now xm is proportional to Y o and xt is proportional to the amount of unsaturate utilized between t = o and t = t, therefore the amount of unsaturate remaining in the reaction system at time t is proportional to (xm - xt). where c is a proportionality constant. Substituting in eqn. ( 3 ) , dx/dt = G ( X ~ - ~ t ) , . * (4) - In (xa - xt) = ct + const. . - ( 5 ) Integrating eqn. (4), When t = 0, xt = o and so the constant of integration is - In xa, Hence orH. W. MELVILLE, J. C. ROBB AND R. C. TUTTON 157 By taking the values for xt at two times tl and t,, * (7) . xh - xtl = xin e-cti - xtle-&.The values of t,, t,, xtl, xh, are known and by plotting the function on the right-hand side of eqn. (7) against arbitrarily chosen values of C, a curve is obtained from which the value of c, which gives the right hand side function equal to (xtl - xt,) is obtained. This value for G is substituted back into eqn. (6) and a value for xco is obtained. This gives the relation- ship between extent of reaction and percentage contraction of the re- acting mixture. The kinetic constants are determined as follows. Measurement of the overall rate gives the value of R2(I/2k4)* and the temperature coefficient of the overall rate gives the overall activation energy, where Measurement of the rate of chain starting I is accomplished by means of an inhibitor technique, described more fully below, thus giving a value for k2(2k4)-*.By means of experiments done with a rotating sector producing intermittent illumination, the value of 2k4 is obtained, thus leading to the individual values of k2 and k4. This investigation has been restricted to conditions where reaction (2) is rate determining. Experimental Apparatus and ~ateria~S.-TRICHLOROBROMOMETHANE was prepared by allowing anhydrous aluminium bromide to stand in contact with carbon tetra- chloride a t room temperature for 3 days.4 The resulting mixture was filtered and the filtrate washed with aqueous potassium carbonate and with water, dried in the usual manner and fractionally distilled under reduced pressure, the fraction boiling a t 49'9O C a t I I G mm. pressure being collected and stored in a dark bottle.CYCLOHEXENE was obtained commercially and purified by distillation, the fraction boiling a t 33" C a t 80 mm. piessure being collected. It was found to be necessary to treat the cyclohexene before distillation by refluxing with copper stearate.5 This treatment removes the peroxide formed on standing in contact with air. Failure to remove this peroxide leads to the development of pure thermal reactions, which complicate the kinetic analysis. TETRACHLORO-0-BENZOQUINONE was prepared by passing chlorine into a hot solution of catechol in acetic acid.6 After two final crystallizations from anhydrous di-isopropyl ether, the dark red crystalline solid had a melting point of 130' C . The size of dilatometer used throughout these investigations had a volume of about 10 ml.and a capillary cross-section of about 0-02 sq. cm. Under the conditions of illumination employed, these dimensions made i t possible to follow the meniscus fall readily with the aid of a cathetometer and to measure ac- curately the extent of reaction. The dilatometer was mounted rigidly and reproducibly in a thermostat controlled by a conventional mercury-toluene regulator to better than f IO-, "C. The source of active radiation was a 125-W Osira lamp, the radiation being filtered through 5 mm. soft glass and about 15 cm. water. In the sector ex- periments, the radiation also passed through 5 mm. of Perspex and a soft glass lens system, The complete thermostat was enclosed in a light tight box, fitted with suitable apertures for irradiation of the reaction vessel and for observation of the meniscus level.The sector used in the intermittent illuminaticn experiments was constructed by blackening a 180' sector of a 16-in. Perspex wheel driven by a D.C. motor, either direct or via a 15/1 reduction gear box. The sector chopped the light The density of the product was d425'C = 2.005 g./ml. JVesper and Rollefson, J . Amer. Chem. Soc.. 1934, 56, 1456. George and Robertson, Trans. Faraday SOL, 1946, 42, 217. Zincke, Ber., 1887, 20, 1776.TRICHLOROMETHYL RADICALS beam a t the focus produced by a suitable lens and could be used to give flash times of from 1/20 sec. up to about 3 sec. For longer flash times, a manually operated shutter was empIoyed. All experiments were carried out in absence of air, since oxygen was found to retard the reaction markedly.The techniques for the vacuum manipulation of materials in this kind of investigation are well established and need no further explanation. The outgassing of the liquids used was achieved by repeated freezing and pumping until no more non-condensible gas was evolved. Results Extinction coefficient for CCl, . Br.-This has been experimentally determined and is shown as a function of wavelength in Fig. I. Since the light used was largely a t A = 3650 A, the molar extinction coefficient is I = 5-0 x 10'~. The amount of light absorbed by the reaction vessel containing pure CC1,. Br, of thickness about I cm., is therefore about 6 yo of that entering the system, so that the photochemical initiation takes place uniformly throughout the reaction vessel.0 -1 '% -2 3700 3{GOd 3400 3600 3800 FIG. I .-Extinction coefficient for CCl,. Br. I = I, IOUCd ; E = extension coefficient, c = concentration in moles/litre, d = thickness of sample in cm. Dependence on light intensity.-Table I shows the constancy of the intensity exponent n obtained by using an intensity screen having a transmission factor of 0.418. The constancy of the exponent indicates that termination throughout the range of experimental conditions is accurately bimolecular. TABLE I Temp. O C CCla . Br/CqH,O molar ratio Intensity I Exponent I O / I I O / I I5/I 2O/I I O / I * I O / I 0.48 0.53 0.50 0.52 0.49 0.52 * This exponent was measured after inhibition for IOO min. by tetrachloro-o-benzoquinone.H. W.MELVILLE, J. C. ROBB AND R. C. TUTTON 159 Rate dependence on cyclohexene concentration.-Table I1 shows how the rate is accurately proportional to the concentration of cyclohexene, thus establishing the proposed kinetic analysis. TABLE I1 Temp. (o c) Temp. (" C) 25 Rate of removd of cyclohexene moles 1.-1 Sec-1 30'7 40 *The standar Relative amount of cyclohexene* 1 0.76 0.52 0.76 0.55 I Relative rate 1 0.75 0-52 I 0.78 0.54 concentration emdoved was I , CC1,. Br/C,H,, = IO/I molar. Measurement of absolute rate .-A typical experimental run is shown in Fig. 2 in which the fall of the meniscus is plotted against time. It will be seen that the rate falls off with time as the cyclohexene is used up. By applying the mathematical treatment outlined above in order to obtain the rate of re- action in terms of the amount of cyclohexene used up per unit time, the curve FIG.2.-Fall of meniscus as a function of time. shown in Fig. 3 is obtained, being a plot of the function xt, . e-ell - xtl. e-cl, against c. The value of xtt - xtl chosen corresponded to 3-535 and for a value of c = 0.0053, the above function has this value. Substitution of this value for c in eqn. (6) enables a value for x, to be obtained from which the curve in Fig. 2 is drawn. The curve fits the experimental points well. In this way, the rate of removal of cyclohexene can be computed and for three temperatures, the rate is shown in Table 111. In the usual manner, by plotting log rate against r / T , the overall activation energy is obtained. This plot is shown in Fig.4, The overall energy of activation measured from the slope is 3-4 kcal. /mole. TABLE I11 I 1-66 x I O ~ 1-84 x 10-6 49'2 2'00 x 10-6I 60 TRICHLOROMETHYL RADICALS Measurement of rate of initiation.-It has been shown that the com- pound tetrachloro-o-benzoquinone is a particularly efficient inhibitor for re- actions involving trichloromethyl radicals. It has, unfortunately, a high extinction coefficient, the absorption spectrum being shown in Fig. 5. C -004 -006 - 008 -3-49 I I I t FIG. 3.-Curve for obtaining absolute rate of reaction. f = xb e-di - xtl e-clo. 31 32 33 I FIG. 4.-Dependance of overall reaction rate on temperature. In Fig. 6 are shown two typical inhibited runs with an uninhibited one f a comparison. It will be seen that there is a retardation of the subsequent rate by the products of the inhibited period of the reaction and i t is found that the degree of retardation does in fact increase with increasing amounts of inhibitor.On account of the fairly strong absorption by inhibitor, measurement of the overall duration of the inhibition period would not be a true measure of the rate of initiation. The technique adopted was to use low concentrations of inhibitor such that its absorption of 3650 A is weak, and t o measure the rate of inhibitor removal photometrically. This was done by removing the dilatometer from the thermostat at various intervals and placing i t in the cell-holder of a Unicam Diffraction Grating Visible Spectrometer. The wavelength used wasH. W. MELVILLE, J. C. ROBB AND R.C. TUTTON 161 5000 A and the instrument was calibrated by using solutions containing known amounts of tetrachloro-o-benzoquinone. Two such runs are shown in Fig. 7. Removal is quite linear over a considerable period near the end of the induction period and kom this slope, the rate of initiation can be calculated. In this way, the rate of initiation is found to be 1-17 x 10-6 mole 1.-l sec.-1 and is sensibly inde- pendent of temperature over the range of temperatures required. The overall rate of reaction is k I& given by 2 (C,,Hlo). Knowing the overall rate I , and (C,,Hlo), the ratio k,/(2k4)4 can be obtained and is given in Table IV. To obtain the absolute values of k2 and k, it is necessary t o measure another characteristic property of the reaction, namely the lifetime of the radical involved in the termination step.(2k4P TABLE IV 30 40 50 0.0256 0.0308 0.0365 20 I 40 I 3600 4400 5200 6000 6600 I 1 FIG. 5.-Extinction coefficient for tetrachloro- o-benzoquinone as a function of wavelength. 200 & FIG. 6.-Inhibition by tetrachloro-o-benzoquinone concentration of inhibition. A : 5.5 x I O - ~ moles/l. B : 8.0 x 10-8 moles/l. Measurement of lifetime of the CCl, radical.-This was done using the For these experiments, the rate of Fig. 8 shows the curve Burnett and now wellestablished sector technique.? initiation was reduced t o 7-16 x 10-9 mole 1.-1 sec.-l. Melville, Proc. Roy. SOC. A , 1947, 189, 456. Chapman, Briers and Walters, J. Chem. SOG., 1926, 562. RI 62 TRICHLOROMETHYL RADICALS obtained by plotting log I&t against R,/R,, where t is the duration of the flash, R, the sectored rate and R, the unsectored.The value of log (2k4)* is obtained by measuring the displacement of this experimental curve from the theoretical.* The curve drawn in Fig. 8 is that for a value of 2kq = 1.0 x 108 1. mole-l sec.-l. Furthermore, under these conditions, the lifetime of the radical is found to be t = 1-18 sec. It will be Seen that a t the two temperatures a t which sector experiments have been carried out, there is no significant change in the position of the curve. This means that only a low energy of activation can be ascribed to the termination process. Further work is proceeding in an attempt to obtain a value for the termination activation energy. FIG. 7.-Inhibition concentration as a function of time.FIG. &--Effect of intermittent illumination on the rate of reaction. Conclusion.-The kinetic values obtained are tabulated in Table V. TABLE V 2-56 x 102 1. mole-l sec.-l 3-08 x 102 ), ), 1'0 x I08 ) J J , 3.4 kcal. /mole low 6-96 x 105 -1.0 x 108- 3.65 x Io2 JB > J 8 Dickinson, Photochemistry of Gases, by W. A. Noyes and P. A. Leighton (Reinhold, New York, 1941)~ p. 207.H. W. MELVILLE, J. C. ROBB AND R . C. TUTTON 163 The low energy of activation for the addition of CCl, to cyclohexene is in general accord with the high reactivity of this radical as, for example, indicated by the fact that CC1, is an effective transfer agent in polymeriz- ation. The temperature independent factor is perhaps surprisingly low for a relatively simple addition of this kind.When, however, the attack- ing radical is a hydrogen atom the velocity coefficient increases to 6 x IO* 1.mole-l sec.-l at 1 5 O C.a The present measurements indicate a low energy of activation for termination but again the temperature in- dependent factor is considerably less than the normal value as happens in polymerization and in interactions in oxidation processes. The general conclusion from the results is that the velocity coefficients of these radical are controlled to a large extent by factors which affect the temperature independent factors rather than the energies of activation. More refined experiments are necessary to measure the termination constants with sufficient precision while it is hoped, by conducting similar experiments with IO/I molar excess of cyclohexene, to produce further constants for reactions (3) and (5). By working in the intermediate regions, the values for reaction (6) should be available. This paper serves to outline, very briefly, how study of systems such as that already described, yields information which can be obtained in no other way concerning the absolute reactivities of unsaturated com- pounds towards a specific radical. One of us (J. C . R.) is indebted to the Department of Scientific and Industrial Research for a Senior Award (1948-50), during the latter part of the tenure of which this work was carried out, and to the University of Birmingham for the award of an I.C.I. Fellowship (1950-51). Another of us (R. C. T.) is indebted to the Department of Scientific and Industrial Research for a maintenance allowance. We are also indebted to Messrs. Imperial Chemical Industries for the loan of the spectrophotometer. Chemistry Department, The University, Edgbaston, Birmingham, I 5.