T H E THERMAL DECOMPOSITION O F GASEOUS ETHYLENE IODIDE T. IREDBLE
AND
L. W. 0. MARTIN
Department of Chemistry, T h e University of S y d n e y , N e w South Wales, Australia Received M a r c h
$4,
19%
Ethylene iodide decomposes readily, even at moderate temperatures and in the absence of oxygen, into ethylene and iodine. This thermal decomposition has been studied recently by several investigators. Pollissar ( 5 ) used carbon tetrachloride as a solvent and fousld that iodine was a catalyst for the reaction, the rate being proportional to the product of the ethylene iodide concentration and the square root of the iodine concentration. He concluded that iodine atoms were the real catalytic agents, and he devised some equations to satisfy the kinetics of the reaction. Schumacher and Wiig (6) made a photochemical study of the influence of iodine atoms on the decomposition. In 1928 Mooney and Ludlam (4) published a paper dealing with the dissociation pressure of ethylene iodide, the gaseous equilibria of the gases ethylene iodide, iodine, and ethylene at different temperatures in the presence of solid ethylene iodide and iodine, and a few experiments on the velocity of decomposition of gaseous ethylene iodide, and of the substance dissolved in carbon tetrachloride, in which they concluded that iodine played the part of a catalyst. Mooney (4) recently carried out some experiments on the synthesis of ethylene iodide from ethylene and iodine, and found that the reaction took place on the surface of the solid iodine, and that the glass surface of the reaction vessel was not, in itself, catalytic unless covered with a layer of alcohol. A satisfactory investigation of the rate of decomposition of the purely gaseous substance does not seem to have been undertaken. This paper is a record of some experiments we have carried out under these conditions, and of our own conclusions, which differ in some respects from those of previous workers. We found, at the outset, that ethylene iodide which was in the slightest degree impure was quite useless for quantitative work. We prepared the pure substance by passing ethylene into an alcoholic solution of iodine containing powdered glass, which was a very efficient catalyst, though not all varieties of glass were found suitable for this purpose. The substance was recrystallized several times from alcohol, was of a pure white appearance, and melted sharply at 82°C. The reaction was carried out in a glass vessel (figure 1) and the change in 365
366
T. IREDALE AND L. W. 0. MARTIN
pressure recorded on a simple glass manometer. It will be seen from the figure that one side of this could be opened continuously to the vacuum system if necessary. The manometric liquid was a very concentrated solution (at 65°C.) of orthophosphoric acid, obtained by partially dehydrating the ordinary commercial phosphoric acid. It was heated under the vacuum until dissolved gases were removed. After opening TI (figure 1) to the vacuum system, and closing Tz for a time, no perceptible pressure difference was detected. We carried out some preliminary tests and found that neither iodine vapor nor ethylene were appreciably soluble in this liquid. It was the only one of a number we experimented with which satisfactorily fulfilled
FIG. 1. GLASSREACTION VESSEL
these conditions. The manometer was read from outside the thermostat, in which both it and the reaction vessel were immersed, by means of a cathetometer. The reaction vessel was cleaned in the usual way with chromic acid, etc., but in some cases we used hydrofluoric acid. These modes of treating the glass surface have a marked influence on the reaction rate. No attempt was made to bake out the vessel at a temperature much higher than that of the reaction itself. The taps TI and Tz were lubricated with graphite and metaphosphoric acid, and in some experiments we discarded TI and sealed off directly. The decomposition was studied at 65°C. and 75°C. The procedure was this: A small quantity of ethylene iodide was introduced through the top
THERMAL DECOMPOSITION
367
OF GASEOUS ETHYLENE IODIDE
of the reaction vessel, which was then sealed up and evacuated with a Hyvac pump. The solid ethylene iodide remained a t the bottom of the vessel and was prevented from vaporizing completely by cooling with a freezing mixture of ice and salt. The final pressure under these condiTABLE 1 T h e decomposition of ethylene iodide at 66OC. Experiment no. 9. Induction period pronounced TlMB
PRESSURE
minutes
mm. HzPOd
10 15 25 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 310 320
19.6 19.9 20.4 20.7 20.7 20.7 22.4 23.2 24.0 24.5 25.6 26.5 27.8 28.6 29.6 30.8 31.9 32.8 33.7 34.5 35.3 35.7 36.3 36.7 36.9 37.4 37.6 37.7 37.8 38.0 38.0
ka X 103
ki X 106
170 125 110 95 98 97 101 99 101 102 101 100 99 97 94 92 89 85 80 78
372 280 252 255 238 26 1 26 1 271 29 1 307 318 330 341 354 350 358 358 349 357 351 339 330
k2
x
106
869 914 919 1007 1023 1009 998 945
33 30 27 30 31 36 37 40 46 51 56 62 69 77 80
90 79 64 65 61 61 51 56 55 54 52 51 50 49 46 46 44 43 42 41 39 37
tions was not much more than the limit obtained by the pump, namely, 0.001 mm. of mercury. We tried solid carbon dioxide as a cooling agent, but found that eihylene iodide was too unstable at this low temperature; some decomposition took place as was apparent from the brown color developed by the ethylene iodide on the side adjacent to the cold glass
368
T. IREDALE AND L. W. 0. MARTIN
surface. It may be that the excessive refrigeration condenses some very volatile substance which acts as a catalyst for the decomposition, or there is another crystalline form of ethylene iodide which is unstable. Unless the ethylene iodide used is very pure, the reaction will not exhibit the special features now to be described. The reaction vessel was plunged into the thermostat at 65'C. and immediate readings of the manometer taken. As the ethylene iodide vaTABLE 2 The decomposition of ethglene iodide at 65°C. Experiment no. 13. Induction period pronounced TIME
PRESSURl
minutes
mm. HaPo4
10 15 20 25 30 35 45 60 75 90 105 120 135 165 180 195 210 225 285 330 360 390 405 420
(9.3) 9.1 9.2 9.2 9.2 9.2 9.8 10.1 10.8 11.2 11.8 12.2 12.5 13.3 13.7 14.1 14.4 14.6 15.2 15.4 15.7 15.9 15.9 15.9
(601 36 40 36 36 35 33 32 31 31 30 28 24 21 20 19
208 194 198 202 193 197 '201 206 207 202 194 165
174 145 130 122 109 96 92 88 84 79 65 56 53
808 813 797 762
porized and acquired the temperature of the thermostat, the pressure rose and remained steady for a while. No trace of iodine vapor could be detected in the vessel. The pressure began to increase again, and plotting the pressure against the time, we obtained the type of curve shown in figure 2 (curve A). Towards the end of the reaction a definite violet vapor filled the reaction vessel. If the ethylene iodide used was partially decomposed or if a small crystal of iodine was introduced into the reaction vessel before sealing up, the form of the curve was different (figure 2,
THERMAL DECOMPOSITION
OF GASEOUS ETHYLENE IODIDE
369
curve B), The pressure never remained steady a t any time. Undoubtedly the portion ab of the curve A indicates an induction period, and it seemed reasonable, a t first, to attribute this to the lack of the necessary catalytic agent, presumably iodine, which, however, soon begins to accumulate in the system, with consequent increase in the speed of decomposition. We would expect the reaction to be bimolecular and to follow an equation of the second order, involving either the iodine concentration or its square root. But this does not seem t o be the case. The results of a number of our experiments have been tabulated. We calculated the velocity con35
30
a,
2 25
E p’
10
L
50
150
100
zoo
250
300
Time
FIG. 2. DECOMPOSITION OF ETHYLENE IODIDE AT 65°C. Curve A, experiment no. 14; curve B, experiment no. 15
(dE)
stants ko, k l , kz (Id, kz and kz, for the zero order, monomolecular, bimolecular (catalyzed), and bimolecular (uncatalyzed), respectively. If a = initial concentration of ethylene iodide (in millimeters of phosphoric acid), x = concentration of iodine formed, and t = time in minutes, then dz - = - - d(a - 2) = K dt dt dX
-
dt
=
R(o - z), K
=
1 -
t
In -E a--2
KO,zero order)
(1)
( K = K 1 ,monomolecular)
(2)
(K
=
370
T. IREDALE AND L. W. 0. MARTIN
K z (Iz) bimolecular catalyzed)
(3)
(R= K z (2/G) bimolecular catalyzed)
(4)
(K
dx - = K(a dt
- x)2, K
1
x
=- t a(a - 2)
=
( K = Kz, bimolecular uncatalyzed)
(5)
The initial pressure was taken to be the pressure recorded during the induction period. The rate of decomposition is best represented by the zero order equation in the first stage, and after the pressure has fallen, by the monomolecular equation, and finally by the bimolecular equation (uncatalyzed). This is characteristic of a heterogeneous reaction, where strong adsorption takes place on the walls of the vessel. We verified the heterogeneous nature of the reaction by introducing a number of thin glass tubes into the reaction vessel. There is an increase in the velocity of decomposition, and the reaction follows the same course as before. A similar increase in the velocity is obtained by cleaning the reaction vessel with a dilute solution of hydrofluoric acid. There must be an increase in the number of active centers of adsorption as the result of this procedure, but this does not explain the induction period. We could assume that iodine is a catalyst for the decomposition. If the reaction were mainly homogeneous it would be of the second order, as Pollissar found in carbon tetrachloride. There is undoubtedly a slow homogeneous reaction at 65"C., but the heterogeneous reaction is the more rapid, and we can only conclude that there is some reaction between the ethylene iodide and the iodine on the glass walls of the reaction vessel. Iodine is not so strongly adsorbed on glass as on calcium fluoride, according to de Boer (1). But these effects are relative, and iodine is found to be adsorbed more readily on freshly blown glass surfaces, or on glass that has been treated with hydrofluoric acid. We find that in the vacuum system from which mercury vapor has not been removed by liquid air, and into which iodine vapor is admitted, deposits of mercuric iodide are found mostly on the internal surfaces of the glass joints that have been heated by the hand blowpipe. This is a suggestive but by no means conclusive demonstration of the selective nature of iodine adsorption on glass. In all the experiments we observed that the reaction did not appear to go to completion; there seemed to be some undecomposed ethylene iodide remaining, as the pressure was not sufficiently great for complete decom-
THERMAL DECOMPOSITION
371
OF GASEOUS ETHYLENE IODIDE
position. Mooney and Ludlam (4) found an equilibrium constant a t 65°C. which indicated almost complete decomposition, and for this reason we felt justified in using the simple kinetic equations. We also came to the same conclusion in some preliminary experiments at 65"C., in which ethylene and iodine vapor were mixed; there was no perceptible pressure TABLE 3 T h e decomposition of ethylene iodide at 75°C. Experiment no. 14. Short induction period TIMl
PRESSURE
minutes
mm. Hap01
10 15 20 25 30 35 40 45 50 55 60 65 75 85 90 105 120 135 150 165 180 195 210 225 240 255 300 345 360 375
18.1 18.1 18.5 18.9 19.2 19.6 20.2 20.6 21.0 21.3 21.7 22.2 23.0 24.0 24.3 25.9 26.9 27.5 28.4 29.2 29.9 30.5 30.9 31.3 31.6 32.0 32.4 32.6 32.6 32.6
kl
80 80 73 75 84 83 83 80 80 82 82 84 83 87 82 78 74 74 72 69 66 63 60 58 51
x
105
194 197 182 188 214 215 217 211 214 223 229 245 243 272 267 265 271 275 278 279 274 270 264 264 238
kz X 103
68 70 72 76 72 67
kdI2)
171 209 215 210 197 187 175 169 167 159 157 151 153 151 135 131 128 125 121 116 112 106 106 93 82
x
106
344 252 222 211 177 171 154 144 137 122 113 108 99 89 81 75 70 66 63 59 56 53 51 44 38
change over a considerable period. We prefer, therefore, to ascribe the anomaly to the loss of iodine in the gas phase due t o adsorption, iodine being more strongly adsorbed than ethylene i0dide.l This also accounts 1 There is also the possibility of ethylene adsorption on glass. Investigators differ in the estimation of the magnitude of this effect.
372
T. IREDALE AND L. W. 0. MARTIN
in part for the prolongation of the induction period, which has the appearance of a steady state, rather than of a slow chemical reaction. The iodine adsorption evidently reaches saturation at low pressures. The rate of reaction is therefore independent of the iodine pressure, and does not appear to be of the second order. But we would expect the apparent incompleteness of the reaction to be more pronounced the greater the activity of the surface, and this is not the case. TABLE 4 T h e decomposition of ethylene iodide at Y5"C. Ex riment no. 15. No induction period TIME
PRESSURE
minutes
mm. H&'O,
5 10 15 20 25 30 35 40 45 55 70 85 100 115 130 145 160 195 220 250 310 355 385
17.9 18.9 19.7 20.2 20.9 21.6 22.2 22.8 23.6 24.4 26.2 27.5 28.8 29.7 30.5 31.1 31.5 32.1 32.7 33.0 33.0 33.0 33.0
ka X 103
ki X 105
125 144 129 132 133 130 129 133 122 122 115 110 104
298 353 319 332 346 346 349 370 350 374 375 385 379 376 366 348 322 305
k2 x 105
I
5
84 87 87 86 88 85
We cannot be certain, therefore, if the glass surface itself without iodine is also catalytic to some extent. The introduction of freshly made glass tubing increases the velocity so much that no observation of an induction period is possible. There is not sufficient time for thermal equilibrium to be established and readings on the manometer are unsatisfactory. There remains the question as to whether the anomalous final pressure is due to some polymerization process. We do not believe this to be the case because of certain independent experiments we carried out. Ethylene iodide was sealed up in a glass vessel and the equilibrium pressure at 35°C.
373
THERMAL DECOMPOSITION O F GASEOUS ETHYLEKE IODIDE
found to be 22 mm. of mercury. The vessel was then cooled to 16°C. (pressure 5 mm.) and afterwards heated to a much higher temperature, 65°C. It was then allowed to drop to room temperature again, and after some time heated to 35°C. (final pressure 22 mm.). If any irreversible polymerization had taken place during the decomposition at the higher temperature, the initial and final pressures at 35°C.would not have been the same. The addition of ethylene seems to have a slight accelerating effect on the reaction, It is not easy to explain this, but as a chain mechanism is not TABLE 5 T h e decomposition of ethylene iodide at 65°C. Experiment no. 19. Ethylene (29.2 mm.) added at beginning of reaction. induction period TIME
minutes
5 10 15 20 25 30 35 51 65 80 95 125 140 155 170 185 230 245 275 305 335
PRESSURE
-mm.
k o x 103
k, x 105
(180) 140 (180) (170) 106 130 126 121 116 114 109 106
292 230 302 290 187 237 239 239 261 259 253 258 246 232 226
Short
k2 X 106
H I P ~ ~
26.5 27.2 27.2 28.1 28.6 29.9 30.6 31.0 33.7 35.4 36.9 40.0 41.5 42.4 43.6 44.0 45.8 46.2 46.8 47.0 47.0
'
1
36 35 37 37 36 34
69
34
27
altogether improbable for this reaction the ethylene could have some influence in delaying the breaking of the chains at the walls. For such a chain mechanism we could write: I (glass)
+
+
C2H412 = 1-1 (glass) CZHJ CzHiI -k CzH412 CzH4 1 2 f CzHrI 1 CzH4I = CzH4 CzH4Iz 1 CzH4I Iz C2H4 1 = CzHiI I (glass) CzH4I = CZH4 1-1 (glass) I (glass) I = 1-1 (glass)
++
+ + + + + I + I* = +
I8
(1) (2) (3) (4) (5) (6) (7) (8)
374
T. IREDALE AND L. W. 0. MARTIN
It seems doubtful if reaction 1 is exothermic, and reactions 2, 4, and 5 are certainly endothermic, and require more energy of activation. Reactions 6, 7, and 8 are the chain-breaking reactions. But the energy of the C-I bond is not known with any certainty, and it would be unwise to generalize about any such chain mechanisms. The great increase in reaction velocity brought about by the introduction of a larger glass surface is only consistent with the idea that the chains begin at the walls to a greater extent than they end there. TABLE 6 The decomposition of ethylene iodide at 66°C. Experiment 21. Enhanced activity of surface. Ethylene 17 mm. No induction period TIME
PRESSURE
minutes
mm. HaPG
5 10 13 15 17 20 23 25 30 35 40 45 50 55 60 75 108 120 165
14.2 23.2 24.1 24.7 25.5 26.3 27.1 27.9 29.3 30.6 32.1 33.3 34.1 34.8 35.6 37.2 38.5 38.9 38.9
ko x 103
121
x
106
330
685
i
718
292
kz X 10;
1
110
112 114 119 125 110 107
From experiments at 65°C. and 75°C. it should be possible to calculate the energy of activation. But we have never found a satisfactory reproducibility for any one set of values of the velocity constant at either temperature. The catalytic surface does not necessarily remain constant from one experiment to another, and only gross comparisons can be made. But we estimate that the activation energy, calculated2 from the zero-
* This calculation is only possible if the density of the manometric liquid does not vary appreciably, which, in this case, is only approximately correct. We can also calculate the energy of activation from the monomolecular constants. I n any case, it is not of a high order.
THERMAL DECOMPOSITION O F GASEOUS ETHYLENE IODIDE
375
TABLE 7 T h e decomposition of ethylene iodide at 65°C. Experiment no. 22. No ethylene TIME
PRESSURE
minutes
mm. &POI
3
7 12 17 23 27 35 37 57 72 87 102 117 132 177
ko
13.4 18.1 18.9 19.5 20.3 20.8 21.6 22.3 25.0 26.9 29.0 30.5 31.5 32.0 33.2
TABLE 8 T h e decomposition of ethylene iodide ut 65°C. Experiment no. 25. Glass capillary tubing in reaction vessel TIME
PRESSURE
minutes
mm. HiPo4
5 7 10 12 17 22 27 32 37 42 48 59
8.5 13.9 21.1 24.8 31.5 34.8 38.3 41.9 44.9 47.4 48.2 48.5
x
108
133 I50 138 144 135 139 140 138 136 136 130
ki X 106
333 312 358 363 363
TABLE 9 T h e decomposition of ethylene iodide at 65°C. Experiment no. 17. Free iodine at t h e beginning of reaction TIME
minutes
2 7 12 17 22 27 32 37 42 47 62 77 82
Velocity constants indeterminate.
PRESSUREl
mm. Hap04
11.8 39.2 42.2 46.2 49.9 52.8 55.7 58.3 60.0 60.8 64.2 63.8 63.8
Velocity constants indeterminate
order constants, is not greater than 12,000 calories. The heat of the endothermic reaction C2HJz =
C2H4
+ Iz
T H E JOURNAL OF PHYSICAL CHEMISTRY, VOL. XXXVlTI, NO.
3
376
T. IREDALE AND L. W. 0. MARTIN
is 20,000 calories, calculated from the thermal data, assuming the heat of sublimation of ethylene iodide to be 15,000 calories (4), and the heat of combustion of ethylene, 344,000 calories (2). We hope to explore more deeply the main features of this reaction, particularly as it is affected by the shape of the reaction vessel and the material of the catalytic surface. We do not wish to express any definite opinion as to the mechanism of the reaction, as it is obvious that there are a number of factors involved. SUMMARY
The decomposition of gaseous ethylene iodide a t 65°C. and 75°C. in a glass vessel seems to be heterogeneous and to a great extent autocatalytic; iodine adsorbed on the glass is a probable catalyst. There is a short induction period which disappears when traces of iodine are present at the beginning of the reaction, and when the catalytic surface is greatly increased. The reaction is of zero order after the induction period, and follows higher orders as the pressure diminishes. A chain mechanism is suggested for the reaction, but this has not yet been completely verified. REFERENCES
(1) DE BOER:Z.physik. Chem. 13B,134 (1931). (2) MIXTER:Am.J. Sci. [4] 12,347 (1901). (3) MOONEY:Trans. Chem. SOC.1931,2597. (4) MOONEY AND LUDLAM: Proc. Roy. SOC.Edinburgh 49, 160 (1929). (5) POLLISSAR: J. Am.Chem. SOC.62,956(1930). (6) SCHUMACHER A N D WIIO: Z. physik. Chem. 11B, 45 (1930).