THEREACTION OF H
WITH
1639
HI
The Reaction of Hydrogen Atoms with Hydrogen Iodide’” by Ralf D. Penzhornlb and B. deB. Darwent The Martin Maloney Chemistry Laboratory, T h e Catholic University of America, Washington, D . C. 80017 (Received October 86,1967)
The photolysis of HI has been investigated in the gas phase, between 303 and 533”K, in the presence of vary-
ing concentrations of Izand GOz. The results show: (a) the efficiency of COzin deactivating,thephotochemical “hot” H is given, with reasonable accuracy, by the simple “billiard ball” calculation; (b) the ratio of k’s for the reactions of thermal H with Iz and HI is 4.95 exp(640/RT); (c) the ratio of the k’s for the reactions of “hot” H with Iz and HI is approximately the same as that of the preexponential factors for the thermal H
reactions.
Introduction The mechanism proposed by Bodenstein and LienewegZaand Bonhoeffer and FarkasZbfor the photolysis of H I has been widely accepted. Ogg and Williams3 showed that H atoms produced by the photolysis of H I reacted 3.8 h 0.4 times as rapidly with Izas with H I and that, in the absence of an inert gas (cyclohexane), the activation energies of the two reactions were identical. I n the presence of cyclohexane the ratio of the rate constants increased from 3.8 to 7.0 at 155“. Ogg and Williams suggested that the photochemically produced H atoms were hot and therefore did not require any activation energy to react with Izor H I and that the ratio 3.8 was that of the preexponential factors. The addition of cyclohexane served to deactivate the hot H atoms. Schwarz, et aLj4confirmed the presence of hot H and and the diffound the ratio of A factors to be 4 X ferences in E’s to be 4.5 kea1 mole-’ for the reactions of and HI. thermal H with 1% Sullivan’s data516on the thermal reaction of HZand IZ suggested AE := 0.5 kcal mole-’, in marked contrast to Schwarz, et aL4 However, Sullivan’s result was obtained in a rather indirect manner, and it seemed desirable to reinvestigate the reactions more directly.
Experimental Section Reagents. Pure anhydrous H I was prepared by dropping previously boiled 85% orthophosphoric acid (J. T. Baker) onto potassium iodide (Fisher Certified) under vacuum. The gas was passed through a trap of solid carbon dioxide, to remove moisture and iodine, and then frozen in liquid nitrogen. After outgassing, two more distillations were made from a -123” bath, the middle fraction being retained in each distillation. The HI, which appeared as a white solid when frozen, was then stored in a black painted storage bulb. The use of an alternate method of preparing hydrogen iodide, suggested by Martin and Willard,’ did not cause any detectable change in the experimental results.
Mass spectrometric analysis showed the H I to be 99.1% pure, the main impurity being HCI. Carbon dioxide (Matheson Coleman grade) was further purified and outgassed. At least two distillations, where only the middle fraction was retained, were made from a solid CO, bath. The purity of the gas was then tested by photolyzing H I alone and in the presence of large quantities of COZ. It was found that there was no detectable change in the rate of formation of hydrogen due to the presence of Con,even if the ratio of COz to H I was as high as 11O:l. Apparatus. The high-vacuum apparatus used in this investigation, which was of the conventional allPyrex glass type, was kept in the dark during each run to prevent decomposition of H I by light. The H I was measured in small volumes (23.7 or 69 em3) and then transferred to the reaction system. A Kern glass spiral manometer, sensitive to =kO.O1 torr, backed by a mercury manometer, was used to measure the pressure and to isolate the gas from mercury. A cathetometer was used for small pressures and the precision of the measurement was better than *0.02 torr. The pressure of COz was measured, with a Wallace and Tiernan (400 0.5 torr) gauge, in a calibrated volume, and the gas was then transferred to the reaction system. The reaction system had a total volume of 7000 50 cm3 and was provided with a unidirectional circulating pump. Greaseless Viton A rubber diaphragm stop-
*
*
(1) (a) Acknowledgment is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for partial support of this research (Grant No. PRF 1612-85). (b) Abstracted in part from a dissertation submitted in partial fulfillinent of the requirements for the degree of Doctor of Philosophy of the Catholic University of America. (a) (a) M.Bodenstein and F. Lieneweg, 2. Physik. Chem., 119, 123 (1926); (b) K. I?. Bonhoeffer and 1,. Farkas, ibid., 132, 235 (1928). (3) R. A. Ogg, Jr., and R. R. Williams, J . Chem. Phys., 13, 586 (1945).
(4) H. A. Sohwarz, R. R. Williams, Jr., and W. H. Hamill, J . A m . Chem. SOC.,74, 6007 (1952). (5) J. H. Sullivan, J . Chem. Phys., 30, 1293 (1959). (6) J. H. Sullivan, ibid., 36, 1925 (1962). (7) R. M. Martin and J. E. Willard, ibid., 40, 2995 (1964). Volume 7 8 , Number 5
M a y 1968
RALFD. PENZHORN AND B. DEB.DARWENT
1640 cocks were used in the gas-measuring and reaction sections. The photolysis took place in a 180-cm3reaction cell provided with a 0.36 mm thick Vycor window. The light source was a Hanovia Type SH mediumpressure quartz mercury vapor lamp which was operated with a Sola constant-voltage transformer. The most important lines were at 2400, 2482, 2537, 2652, and 2504 A. Procedure. Iodine was prepared by photolyzing H I (3-6 torr) for various times depending upon the amount of iodine desired. Prior to each experiment H I was condensed from the storage bulb into a trap, evacuated, and distilled into the gas-measuring section from a -123" bath. The pressure was then measured and the gas condensed into the reaction system, expanded, photolyzed, and again condensed at liquid nitrogen temperature. The hydrogen was transferred to a Toepler pump and measured in a gas buret. The assumption was made that Hz and 1 2 were produced in equal amounts. The H I remaining from the photolysis was distilled into the measuring section from a - 123" bath, the desired pressure was measured, and the H I was recondensed into the reaction system. A clean separation between iodine and H I could be accomplished by this technique because of the large differences in vapor pressures. The effectiveness of ,this technique was demonstrated as follows. (1) The H I recovered from the preliminary photolysis was condensed as a white powder. (2) A 0.5-hr. photolysis of the residual iodine did not produce any measurable amount of uncondensables, indicating essentially complete removal of the HI. (3) The intensity was checked frequently, by photolyzing a standard sample of HI, and the rates were corrected as necessary. The mixture of Iz,HI, and COPwas expanded, mixed with the circulating pump for a t least 0.5 hr, and photolyzed. The amount of hydrogen produced was usually about 50 and never less than 30 hmol. The temperature was varied in a region (303-533°K) where there was no detectable thermal decomposition of hydrogen iodide.
0.000 0.001
0.059 0.075 0.100
0.119
Figure 1. The effect of inert diluent on the quantum yield at various L/HI ratios at room temperature.
shown in Figure 1. For constant 12/H1, 4 decreases with increasing COz and becomes constant and independent of C02 a t the higher concentrations of COz. The effect of C02 on 6 was shown to be directly related to the concentrations of H I at constant Iz/HI; thus, by halving [HI] and keeping I2/HI constant, 4 became independent of COZ at approximately half the value required formerly. Accordingly, conditions can be and were attained such that is independent of the concentration of CO,. We assume that, under those conditions, essentially all of the hot H were deactivated and the reaction of thermal H could be investigated. The quantum yield was studied as a function of Iz/HI and temperature in the presence of a large excess of COz, and the results are presented in Table I.
+
Discussion There is ample evidence that hot H atoms are produced in the photolysis of HI. Those atoms may possess translational energies of 41 kcal mole-', if I (zPps,,) are produced, or 19 kcal mole-' if I (zPpl/J are formed. The reactions
+ hv+H* + I H* + H I +Hz + I H* + 1% HI' + I HI
----t
Results The rate of photolysis of HI, as measured by the rate of formation of Hz, was shown to be independent of time, up to a conversion of 0.28%, and to be unaffected by the addition of COz even when the ratio CO2/HI was as high as 11O:l. Thus the photolysis was unaffected by the small amounts of 1 2 produced in the photolysis and the COZwas adequately free of reactive impurities. I n addition, the rates of photolysis of H I were independent of temperature, between 303 and 533"K, when the concentration of H I was kept constant. The quantum yield of Hz (+) was assumed to be unity in the photolysis of HI. The effects of the addition of Iz and COZ on + are The Journal
l2/HI
of
Physical Chemistry
(0) (1) (2)
certainly occur. Whenever the energy of H* is greater than 35.5 kcal mole-' ( D I ~ )the , HI' formed in I (2) will have enough energy to decompose to H and may do so on every vibration
+
HI' -+H'
+I
(2')
The H ' atoms may have translational energies of between zero and 5.5 kcal mole-' and so may also react as hot atoms. Accordingly, in the absence of CO,, we should consider at least three types of hot atoms, possessing 41,19, or 0-5.5 kcal mole-', and the actual situation may well be very much more complex. We shall not attempt further analysis of this problem.
THEREACTION OF H
WITH
HI
1641
Table I5 Expt no.
Q
PHI, mm
PI2
Pccz,
mm
mm
Iz/HI
RHa
11.0 14.4 21.7 25.4 37.5 40.8 199.0
68c 70b 7 Id 65b 72a 59b 1l a
1.131 1.483 2.018 2.165 3.126 2,916 18.65
0.147 0.177 0.187 0.166 0.188 0.136 0.15
150 151.5 196.5 150 196.5 150 150
303°K 0.13 0.119 0.093 0.077 0.060 0.046 0.008
70e 71c 72c 59a
1.478 2.021 3.118 2.921
0.18 0.186 0.192 0.133
196.5 196.3 196.5 150
415'K 0.123 0.093 0.062 0.0455
13 -5 19.5 33.0 35.9
68a 70f 71a
1.133 1.473 2,028
0.146 0.182 0.182
150 196.5 196.5
478°K 0.129 0.124 0.090
68d 70d 70c 64a 71b 63a 72b
1.133 1.481 1.481 1.497 2.03 2,167 3.12
0.146 0.178 0.178 0,165 0.185 0.165 0.191
151.5 196.4 151.5 150 196.5 150 196.5
533'K 0.129 0.12 0.12 0.11 0.091 0.076 0.061
RHZ
31.5 39.8 50.0 53.0 69.0 65.0 224.0
l/d(Hz)
kdkr
2.88 2.76 2.30 2.08 1.84 1.60 1.12
14.46 14.80 13.99 14.11 13.97 12.95 15.0
31 40 55 53
2.3 2.05 1.67 1.48
10.56 11.25 10,88 10.55
10.2 12.5 18.8
22.7 28 36
2.23 2.23 1.92
9.61 9.77 10.25
10.2 12.5 12.5 13.0 19.0 19.4 28.0
22.6 25 25 26 32.5 33.5 44
2.22 2.00 2.00 2.00 1.71 1.725 1.57
9.46 8.33 8.33 -9.09 7.81 9.55 9.3
The pressures were those observed a t ambient temperature, and thus have not' been corrected for the temperature of the reaction
cell.
I n the presence of COZ, the hot H may be deactivated H*+COz+H+COz
(3)
and me do not attempt to differentiate between the various types of hot H. I n reaction 3 the H* may require several collisions with CO, to be effectively deactivated and we define k3 to allow for that possibility. Now the fraction of the photochemical H which will be kam), deactivated is approximately k3m/(klx kzy where x, y, and m are concentrations of HI, 1 2 , and COz, respectively. Obviously, m can always be made large enough, relative to x and y, for that fraction to be close to unity. The results of these experiments show clearly that essentially all of the H react as thermal species when na is, typically, 100 torr for x = 3 and y = 0.2 torr. Reference to Figure 1 shows that COZ had attained 50% of its effectiveness at about 10 torr when x = 0.78 and y = 0.078 torr and at about 20 torr when x and y are 1.6 and 0.16 torr, respectively. Let us assume that 50% of the hot H were deactivated at the 50y0 effective pressures. For a rough order of magnitude calculation we assume kz = 5kl and the results then suggest kl % 10 k3. The assumption kz = kl leads to kl % 5k3. Accordingly, if reaction 2 occurs on every collision, as seems likely, COZrequires between 5 and 50 collisions to deactivate
+
+
Figure 2. Effect of temperature on the quantum yield for various L/HI ratios: 0, 303°K; 0 , 415'K; A, 478'K; V, 533°K. (Data are in Table I.)
H*. This agrees reasonably well with a simple billiard ball calculation to deactivate either 19- or 41-kcal mole-' H atoms by collision with COz. Volume 72, Number 6 M a y 1968
R. H. Cox AND A. A. BOTHNER-BY
1642
(4’)
found kj’/k4’ = 3.16 exp(800/RT) with an uncertainty of =k 1000 cal mole-1 in Est - E4t. The ratio of preexponential factors &/A4 was found to be 4.95. It is interesting to note that both the simple collision theory and the transition-state theory lead to values of about unity for that ratio. Schwarz, et al., found, by transition-state theory, that the ratio Howof preexponential factors (A5/A4) was 4 X ever they omitted the moment of inertia of the transition state in their calculation. The inclusion of that moment of inertia loads to a value of about unity for A possible explanation for the high value found by Schwarz, et al., for k5/k4 is that the presence of mercury in their reaction system, which reacts with Iz to form HgI, may have affected their results. The ratio of the rate constants for the hot reactions ( k 2 / k l ) was found to be 4.2 =k 0.2 by Schwarz, et aLJ4 and 3.8 =k 0.3 by Williams and Ogg.3 Since the hot atoms require no energy of activation in reactions 1 and 2, it seems that the preexponential factors are essentially the same for both thermal and hot atoms.
(5’)
(8) M. C. Flowers and 5. W. Benson, J . Chem. Phys., 38,882 (1962 )
I n the presence of an adequately large excess of C 0 2 we need consider only reactions of the thermal H
+ H I H2 + I H + +H I + I
H
---+
12
(4)
(5)
which lead to the relationship
The values obtained for l e ~ / J c 4are given in Table I. AIthough the effect of temperature on k5/k4is small, it is nevertheless significant (Figure 2) and leads to k5/k4 = 4.95 exp(640/RT). Since E6 is generally regarded as being zero, E4 = 640 cal mole-’ is in general agreement with Sullivan’s5Jvalue of 480 f 350 cal mole-l but not with that of Schwarz, Williams, and Hamil14of 4500 cal mole-’. Flowers and Benson,8who competed the reactions
+ H I +CH4 + I CH3 + 1 2 CHJ + I CH3
---+
Proton Nuclear Magnetic Resonance of Di- and Trisubstituted Pyrazines and Their Cations by R. H. Cox and A. A. Bothner-By Mellon Institute, Carnegie-Mellon Universitu, Pittsburgh, Pennsylvania
15.919
(Received October 27, 1967)
The proton nmr spectra of eight substituted pyrazines in neutral and acidic solvents have been analyzed. Long-range couplings over five and seven bonds were found. Protonation effects on coupling constants are discussed. Results obtained from the analysis of compounds bearing a hydroxy substituent suggest they exist in the keto form.
Introduction The factors affecting the nmr parameters of a variety of aromatic compounds are rapidly being elucidated. Considerable effort has been devoted to substituted benzenes’-5 and pyridines,6-8 focusing on the nature of substituent effects. I n an earlier paper9 we have shown that substituent effects in monosubstituted pyrazines parallel, roughly, the effects in monosubstituted benzenes and pyridines. Also, the magnitude of the vicinal coupling constant (JS6)and the chemical shifts were shown to reflect tautomeric structures in monosubstituted pyrazines. The Journal of Physical Chemistry
By determining the effects of protonation on nmr parameters often, information regarding the site of (1) S. Castellano and C. Sun, J . Am. Chem. Soc., 88, 4741 (1966). (2) 9. Castellano, R. Kostelnik, and C. Sun, Tetrahedron Letters, in
press. (3) S. Castellano, private communication. (4) J. M. Read, Jr., R. E. Mayo, and J. H. Goldstein, J . M o l . Spectry., 21, 235 (1966). (5) J. M. Read, Jr., and J. H. Goldstein, ibid., 23, 179 (1967). (6) J. B. Merry and J. H. Goldstein, J . Am. Chem. Soc., 88, 5560 (1966). (7) S. Castellano, C. Sun, and R. Kostelnik, J . Chem. Phys., 46, 327 (1967), and references therein.