3327
RATESOF REACTION OF ATOMIC OXYGEN
Rates of Reaction of Atomic Oxygen. 11. Some C, to C, Alkanes' by John T. Herron and Robert E. Huie National Bureau of Standards, Washington, D . C. 20284
(Recehed March 21, 1969)
Rate constants have been measured for the reactions of atomic oxygen (08P)with eight straight and branched chain alkanes from 250 to 600"K, and for six other branched chain alkanes at 307°K. The reactions are interpreted on the basis of a hydrogen atom abstraction mechanism, and Arrhenius parameters are derived for abstraction of primary, secondary, and tertiary hydrogen atoms. The activation energies are found to be in the order primary > secondary > tertiary. For abstraction at secondary C-H bonds, the activation energy is observed to decrease with increasing molecular complexity of the alkane. A method for estimating the rate constants of other atomic oxygen-alkane reactions is given.
Introduction In part I of this series2 we reported rate data for the reactions of atomic oxygen with some chloro- and bromoalkanes from 336 to 622°K. These reactions were interpreted in terms of a hydrogen atom abstraction mechanism. It was suggested that the observed curvature in the Arrhenius plots for some of these reactions was due to the presence of different kinds of C-H bonds. However, the degree of curvature was not sufficiently great to lead to any firm conclusions. In the case of the ordinary alkanes, the curvature is more apparent, and it is possible to interpret the Arrhenius plots in terms of abstraction at different C-H bond sites. Experimented Section The experimental method has been fully described,2 and only a brief outline is given here. The apparatus consists of a flow system coupled to a mass spectrometer. Atoms are generated by passing a mixture of several per cent oxygen in an argon carrier through a 2450-MHz electrodeless discharge. Alternately, atomic oxygen was generated by passing pure nitrogen through the discharge and titrating with nitric oxide to yield atomic oxygen via the very fast reaction N NO --t Nz 0. The partial pressures of the reactants were monitored by means of the mass spectrometer. The atomic oxygen measurements were based on the m/e 16 peak a t a nominal ionizing energy of 20eV. This was put on an absolute basis by titration with nitrogen dioxide or ethylene. The ethylene titration was not used previously by us, but is necessary in the present case because of the breakdown in the nitrogen dioxide titration method a t temperatures below about 273"K, as noted elsewhere3 and confirmed here. The use of ethylene as a titrating agent is based on the observation that with an excess of ethylene, the ratio Of atomic Oxygen consumed to consumed' A [O]/A [CZHd],has a value of 2.04,6(with an estimated
+
+
*
uncertainty of 0.2) independent of temperature over a wide range.5 We have repeated the measurement of A [O]/A [C2H4] a t about 300°K and confirmed this value using the NOz titration to measure the oxygen atom concentration. At lower temperatures we find no evidence for a change in this ratio. Rate constants are derived using the expression
k
= In
{ [A]o/[AIt)/[Olo.&
where [A10 and [AI t are the concentrations of reactants at times zero and 2, respectively, and [010.5is the oxygen atom concentration at the midpoint of the reactor (ie., at 0.5t). The basic assumptions involved in the rate measurements are: reactant is lost only through reaction with atomic oxygen, reactant is not reformed, and the oxygen atom concentration is unaltered by addition of reactant. The latter condition is met by keeping the oxygen atom concentration much larger than that of the second reactant. The atom concentration is determined by titrating the oxygen atoms along the length of the reactor and from a linear plot of concentration vs. reaction time reading off the value [0]0.6. To avoid problems associated with "background" in the mass spectrometer we have used ethane-l-dl for the ethane experiments and n-butane-l-dl for the hightemperature n-butane experiments. For some of the lowtemperature experiments the reactor was lined with a snug-fitting sheet of polytetrafluoroethylene. This is so indicated in the data tabulations. Results The data are given in Tables I-VI11 and Figures 1-8. (1) Contribution of the National Bureau of Standards, not subject to copyright. (2) J. T. Herron and R. E. Huie, J . Phys. Chem., 73,1326 (1969). (3) F. 9.Klein and J. T. Herron, J . Chem. Phgs., 41, 1285 (1964). (4) J. T. Herron and R. D. Penzhorn, J . Phys. Chem., 73,191 (1969). (5) A. A. Westenberg and N. deHaas, Twelfth Symposium (International) on Combustion, The Combustion Institute, 1969, t o appear.
Volume '78, Number 10 October 1969
3328
JOHN T. HERRON AND ROBERT E. HUIE
Table I: Summary of Rate Measurements for the Reaction 0 Velocity, om 880-1
219 329 252 303 252 276 229 275 284 275 295 290 291 283 307 299 293
528 631 594 620 679 693 654 797 79 1 787 1530 1720 2060 2180 3500 3530 3530
-
Temp, OK
255 276 276 276 307 307 307 357 357 357 386 44 1 513 532 595 603 607 a
Total pressure N m-ea (1 Torr 133.32 N m-1)
+ n-CdHlo
----Reactant
10-"(0)."
--c
OH
+ C4H8
Concentration, mol cm-s--IO-IS(A)o 1O-'S(A)
12.1 10.1 19.1 22.4 21.6 58.9 34.9 13.7 12.1 77.1 15.3 16.4 38.7 36.0 34.8 36.6 37.0
432 583 935 886 969 560 627 525 886 190 93.0 47.0 107.0 628 681 405 678
t
413 492 891 825 874 468 535 464 574 080 78.5 32.5 26.0 114 132 81.9 121
10-10 k, cma mol-] sec-1
0.414 0.635 1.18 1.25 2.21 1.44 1.59 4.44 6.31 4.83 10.1 24.2 48.2 63.6 124 112 100
(E) (E) (E) (E) (NE) (NE) (NE) (NE) (NE) (NE)
N m-2 = newtons per square meter.
~
Table 11: Summary of Rate Measurements for the Reaction 0
Temp,
OK
255 255 258 276 276 276 276 307 307 307 321 336 338 357 386 437 437 44 1 508 508 593 597
Total pressure, N m-2 (1 Torr = 133.32 N m-3)
Velocity; om seo-1
219 228 216 263 272 227 310 248 223 254 237 323 289 250 295 288 283 29 1 291 289 289 281
528 513 533 589 584 684 685 646 633 727 683 786 1350 742 1530 2520 2530 1720 2880 2870 3420 3540
R -eacatnt 10 -"(o)
&"
31.8 30.8 37.0 14.7 19.0 28.1 16.6 21.1 13.7 21.5 7.73 30.5 14.9 9.19 15.2 9.74 10.0 14.7 9.17 7.15 3.97 4.90
I n the tables, the symbols N, E, and T have the following meanings: N, oxygen atoms produced by means of NO --+ Nz 0 reaction; E, oxygen atoms the N calibrated by titration with ethylene; T, reactor lined with polytetrafluoroethylene. For most of the compounds we have examined, the Arrhenius plots are distinctly curved, although the degree of curvature is quite variable. At a given temperature the rate of reaction can be related to the
+
+
The Journal of Physical Chemistry
+ n-CsH12
+
OH
+
concentration, mol om-S---10-'8(A)a 10-'*(A)t
66.3 82.0 78.2 56.1 56.8 55.3 40.9 43.9 79.4 64.6 52.6 30.1 60.0 40.3 17.0 20.0 6.78 11.2 15.1 8.53 8.84 9.17
57.0 73.3 63.4 50.8 50.4 48.3 36.6 35.1 69.9 51.6 46.1 20.2 50.4 33.3 12.2 15.4 4.97 6.42 10.3 5.32 5.94 5.79
10-10 k,
cma mol-' sec-1
1.53 (E) 1.22 (E) 1.32 (E) 2.38 (E) 2.43 (E) 2.06 (NET) 2.86 (ET) 4.72 (E) 4.02 (E) 4.57 (ET) 6.90 (E) 11.3 9.46 11.1 (E) 21.1 43.0 37.8 43.2 72.3 116 217 197
numbers and kinds of C-H bonds in each compound. These general observations strongly favor a hydrogen atom abstraction mechanism for these reactions. If the abstraction mechanism is correct, then it should be possible to resolve the data into contributions due to reaction a t primary, secondary, and tertiary C-H bonds and derive the relevant Arrhenius parameters. The starting point for this procedure is the choice of a rate expression for abstraction a t any primary C-H
3329
RATESOF REACTION OF ATOMIC OXYGEN Table I11 : Summary of Rate Measurements for the Reaction 0
Temp,
Total pressure, N m-1 (1 Torr 133.32 N m-n)
-
OK
329 252 303 276 385 22 1 229 289 275 371 284 275 295 288 283 290 291 289 289 281
276 276 276 307 307 307 307 338 357 357 357 357 386 437 437 441 508 508 593 597
Velocity, am sec-1
631 594 620 693 719 659 654 1350 797 826 791 787 1530 2520 2530 1720 2880 2870 3420 3540
Total pressure, m-2 (1 Torr = 133.32 N m-*)
Temp,
247 255 255 258 276 276 276 276 276 307 307 307 307 321 335 336 338 357 386 437 437 441 508 508 593 597
OK
224 219 228 216 263 2 72 227 221 310 248 223 215 254 237 259 323 289 250 295 288 283 29 1 291 289 289 281
Velooity , cm sec-1
583 528 522 533 589 584 684 639 685 646 633 702 727 683 1490 786 1350 742 1530 2520 2530 1720 2880 2870 3420 3540
OH
-c
+ CsHii
Reactant concentration, mol cm-8lo-"(o)&V 10-"(A)o 10-'"(A)t
7--
36.6 36.5 44.8 28.0 11.9 39.7 34.1 152 92.4 110 25.9 92.0 26.3 89.7 68.0 27.5 6.97 59.9 22.2 29.5
128 89.9 112 66.5 40.7 78.7 95.9 16.1 15.4 33.6 60.5 65.4 16.3 10.5 8.37 2.25 9.29 7.31 4.10 4.93
Table IV: Summary of Rate Measurements for the Reaction 0 N
+ neo-CsHn
+ ?&6Hl4
+
OH
17.8 30.6 27.9 33.6 13.3 18.2 23.6 20.2 16.2 19.8 13.2 7.61 21.5 7.63 39.6 31.7 15.1 7.98 15.2 9.81 9.78 14.7 9.10 6.99 4.07 4.90
bond. For this quantity we have used the results for neopentane. The neopentane data, although having more scatter than we would wish (Figure l), can be fitted to a straight line, leading to the Arrhenius parameters givein in Table IX. The rate expression
57.2 47.5 57.5 36.3 48.6 40.3 42.9 44.2 49.0 36.8 34.3 49.6 22.4 38.4 31.5 11.7 26.2 27.1 13.3 14.1 8.44 8.03 11.8 7.18 3.57 7.11
k, sec-1
33.7 34.3 42.1 26.7 11.6 36.8 31.5 148 89.3 104 22.4 77.6 24.1 84.9 64.4 23.8 6.45 55.3 19.8 26.2
0.250 (E) 0.255 (E) 0.218 (E) 0.378 (NE) 0.294 (NE) 0.342 (NE) 0.322 (NE) 1.48 1 . 0 9 (NE) 0.805 (NE) 1.14 (NE) 1.25 (NE) 5.30 8.29 7.73 9.03 15.5 19.5 59.7 52.6
10-q~)t
10-10 k, cm3 mol-] see-1
+ CBHia
mol om ---lo-"(o)av Reactant concentration, 10 -"(A) o
10-10
om33 mol-1
-3-
52.0 37.9 47.4 24.5 42.7 34.2 35.9 37.7 41.9 26.7 28.2 44.2 15.7 32.5 20.8 6.07 20.3 21.2 8.05 9.31 5.69 3.52 6.80 4.46 1.98 3.53
1.95 2.36 2.37 2.47 3.86 3.53 3.16 3.17 4.15 6.87 6.19 6.55 7.50 9.17 14.6 17.5 14.0 28.0 31.8 65.7 64.5 60.7 107 118 317 299
(ET) (E) (E)
(E) (E) (E) (NET) (ET) (ET)
(E) (E) (ET) (ET)
(E)
(E)
for a single primary C-H bond is then given as k , = 0.49 X 1OI2 exp(-2920/T) cm3mol-1 sec-*. We now consider the data for the Cq t o C8 normal alkanes which contain only primary and secondary C-H bonds. For each measured rate constant we have Volume 73,Number 10 October 1969
3330
JOHN T. HERRON AND ROBERT E. HUIE
Table V : Summary of Rate Measurements for the Reaction 0 Temp, OK
Total pressure, N m-2 (1 Torr = 133.32 N m-2)
Velocity, cm sec-1
247 249 249 254 255 255 276 276 276 276 276 307 307 307 307 32 1 335 336 338 357 386 386 437 437 44 1 508 508 593 597
224 241 237 200 217 228 207 260 241 221 310 215 223 257 215 237 259 323 289 243 267 295 288 283 291 291 2 89 289 281
583 600 590 512 527 522 565 581 652 639 685 618 633 731 702 683 1490 786 1350 743 2450 1530 2520 2530 1720 2880 2870 3420 3540
--Reactant lo-yo),,
Temp,
OK
247 249 249 255 255 255 276 276 276 276 276 276 307 307 307 307 321 335 336
338 357 386 437 437 44 1 508 508 593 597
Velocity, cm sec-1
224 241 237 219 217 228 263 272 227 241 221 310 248 223 215 254 237 259 323 289 250 295 288 283 291 291 289 289 281
583 600 590 528 527 522 589 584 684 652 639 685 646 633 702 727 683 1490 786 1350 742 1530 2520 2530 1720 2880 2870 3420 3540
The Journal of Physical Chemistry
17.1 26.3 22.0 26.5 20.4 27.2 12.8 17.4 25.3 19.6 16.8 15.8 19.9 11.8 7.35 11.4 7.43 39.4 31.2 14.9 8.52 15.2 9.52 9.67 14.8 9.01 6.99 3.94 4.93
28.2 34.5 29.6 24.9 16.6 27.1 34.9 36.0 13.7 30.2 14.4 34.0 41.2 25.0 18.5 43.6 26.0 18.3 29.6 42.1 5.97 13.1 13.2 10.6 11.0 20.7 8.09 13.5 17.2
-
+ n-C~H16
---Reactant lo-"(o)&v
4
OH
concentration, mol cm-310-13(~)0 lO-'a(A)
20.5 27.2 21.8 45.2 21.5 29.7 16.6 15.2 25.8 19.1 18.8 11.3 14.7 10.9 11.0 5.43 38.9 30.9 14.7 5.42 49.8 15.5 9.78 9.52 14.8 9.17 7.15 3.83 4.85
Table VI: Summary of Rate Measurements for the Reaction 0 Total pressure, N m-2 (1 Torr = 133.32 N m-*)
+ CeHla (2,3-Dimethylbutane)
OH
19.2 20.3 21.3 7.66 9.54 14.3 21.7 24.7 7.13 20.2 9.22 21.3 25.1 17.0 13.7 35.9 13.1 5.76 19.1 30.2 1.86 6.09 8.54 6.06 4.11 11.8 4.14 7.86 10.3
10-10 k , cm3 mol-' sea-1
6.80 (E) 7.14 (ET) 5.48 (ET) 9.00 (E) 8.40 (E) 7.23 (E) 11.2 (E) 9.49 (E) 10.2 (E) 8.47 (ET) 10.0 (ET) 18.1 (E) 15.0 (E) 1 6 . 1 (ET) 12.0 (ET) 16.0 (E) 24.7 32.1 25.1 28.0 (E) 47.0 35.8 71.9 88.4 73.5 109 165 295 230
+ C?Hu
concentration, mol o m - 8 - F 10-1"A)t lO-l3(A)o
70.5 16.3 34.1 33.5 39.2 22.6 28.1 25.5 25.6 25.6 53.2 16.5 10.0 35.4 34.6 25.3 18.1 18.7 9.44 26.6 14.6 7.87 10.3 7.39 4.06 9.10 5.75 4.31 4.85
t
+ CBHla
62.4 12.9 28.9 22.4 30.8 14.9 23.6 19.7 19.9 20.4 43.8 12.7 5.95 27.9 29.1 19.5 14.3 10.6 4.11 19.8 10.1 3.82 6.56 4.42 1.57 4.52 2.85 2.00 2.06
10-10 k , om3 mol-' sec-1
2.52 3.23 2.71 5.25 3.85 5.01 5.26 5.80 4.30 4.75 4.60 6.47 11.2 8.97 10.4 10.3 12.7 20.3 22.5 16.6 21.6 44.8 79.6 83.3 69.6 136 181 400 363
(ET)
(ET) (ET) (E) (E) (E) (E) (E) (NET) (ET) (ET) (ET) (E) (E) (ET) (ET) (E)
(E)
3331
RATESOF REACTION OF ATOMICOXYGEN Table VII: Summary of Rate Measurements for the Reaction 0
Temp,
OK
247 254 255 255 255 276 276 276 276 276 276 276 276 276 307 307 307 307 307 32 1 336 338 357 386 437 437 441 508 508 593 597
Total pressure, N m-2 (1 Torr 133.32 N m-*)
224 200 219 217 228 207 263 260 272 238 227 241 221 310 215 223 257 215 254 257 323 289 250 295 288 283 291 291 289 289 281
Velocity, crn sec-1
283 512 528 527 522 565 589 581 584 682 684 652 639 685 618 633 731 702 727 680 786 1350 742 1530 2520 2530 1720 2880 2870 3420 3540
.-L
OH
+ C&
_--Reactant concentration, mol ern----10-"(0)."
16.2 40.2 25.9 17.4 22.7 16.0 11.8 15.8 17.7 21.6 24.0 19.4 18.9 16.2 10.5 10.7 10.0 6.88 11.4 7.08 29.8 14.9 9.05 14.7 9.34 9.37 16.6 8.47 6.45 3.62 4.61
subtracted the contribution due to reaction a t the six primary C-H bonds. These corrected rate constants were then fitted to a straight line which is represented in Figures 2 to 6 by the solid line. The derived Arrhenius parameters are given in Table IX and correspond to abstraction a t secondary C-H bonds. I n a similar manner, by considering the data for 2,3-dimethylbutane (Figure 7), we can arrive a t the Arrhenius parameters characteristic of abstraction at tertiary C-H bonds. These are given in Table IX. The broken lines in Figures 2 to 7 represent total rate constants, which are derived by summing the contributions due to the number of primary C-H bonds to those for the secondary or tertiary bonds. The rate constant for a primary C-H bond is that derived from neopentane and given above, while that for a secondary or tertiary is taken from Table IX. For example, the total rate constant for the n-pentane reaction is equal to 6[0.49 X 1013 exp(-2920/T)] 6[1.34 X 1013exp(-2320/T)] crn3 mol-' sec-I. I n our treatment of the data we have excluded all data taken at temperatures less than 307°K. At these lower temperatures the data show a marked deviation from the predicted line. The most likely explanation
+
+n-cdh
10-'a(A)a
43.4 25.2 29.5 29.1 69.1 33.9 47.3 29.9 14.5 78.1 49.9 29.9 40.5 43.4 59.6 44.8 28.9 49.9 24.5 26.8 17.1 18.0 20.7 14.3 12.5 9.55 4.33 14.7 11.3 7.97 8.56
10-13(4
34.8 7.66 13.7 13.5 42.6 19.8 38.5 21.1 9.6 69.9 35.2 22.4 30.8 34.8 42.3 31.7 21.8 42.0 18.8 19.9 6.71 11.9 11.7 6.78 7.49 5.71 1.03 6.88 5.04 4.50 4.32
10-lQ k, cm3 mol-' sec-1
4.80 10.2 9.37 6.29 6.70 12.6 6.52 8.52 8.96 2.13 5.74 5.05 5.77 8.12 13.6 13.9 12.9 10.9 10.7 17.1 27.3 23.0 31.1 50.2 88.0 88.2 92.2 159 216 328 326
(ET) (E) (E)
(E) (E) (E) (E)
(E) (E) (NET) (NET) (ET)
(ET) (ET)
(E) (E) (ET)
(ET) (ET) (E)
(E)
for this behavior, aside from "tunneling," is the occurrence of reactions on the walls of the reactor. There is some qualitative evidence to support this argument. A limited number of experiments were carried out using a reactor of internal diameter of 1.5 cm rather than of 2.0 cm which was that used in obtaining the data reported in Tables I-VIII. The measured rate constants were found to coinicide with those measured using the larger reactor at temperatures greater than about 300°K but were somewhat higher below 300°K. We also note that the rate constants measured using a polytetrafluoroethylene-lined reactor were somewhat lower than those measured in an ordinary glass reactor. These effects are most noticeable for n-octane. I n the case of n-octane we also observe a great increase in the scatter of the measured rate constants a t the lowest temperatures used. Table I X summarizes the results of our work along with results for CH4 and CzHs based on the data of Westenberg and deHaasS6 The uncertainties attached to our measurements are the standard errors of the (6) A. A. Westenberg and N. deHaas, J. Chem. Phys., 46, 490 (1967).
Volume 73, Number 10 October 1969
JOHNT. HERRON AND ROBERT E. HUIE
3332
~
~~
~
Table VIII: Summary of Rate Measurements for the Reactions of Atomic Oxygen with Ethane and Some Miscellaneous Alkanes Total pressure
N m-* (1 Torr = Temp, OK
133.32 N m-2)
Velocity. om see-1
-Reactant 10 -11(0)*v
concentration, mol c m - b - - 10-"(A)o lO-La(A)t
0 + CgHbD + OH 336 336 429 508 595
331 323 317 291 307
762 786 984 2880 3500
307 307 307
241 280 193
664 697 605
307 307 307
241 2 80 198
307 307
250 213
+ CzHiD
91.4 38.8 42.9 9.70 37.6
0
172 86.7 36.5 99.1 17.6
11.8 15.2 14.8
917 1214 1245
250 213
307 307
215 257
664 697 605
+
10.5 15.8 14.8
+
C&6 (2,4-Dimethylpentane) 661 7.27 659 11.2
+
661 659
5.78 9.9
618 731
0 + C& 276 385
693 719
11.9 10.2
7.75 8.53 7.75
+
OH C7H16 12.7 21.4 17.7
+
OH 56.1 44.7
11.1 17.6 14.0
6.71 5.96 6.69
47.9 34.8
9.10 11.5
98.7 32.9
6.05 8.08
7.01 12.2
3.32 (E) 2.62 (ET)
9.43 30.9
0.964 (E) 0.708 (E)
c7&3
+ C8Hl7
110.0 38.4
0 + CsH18 (2,3,4-Trimethylpentane) OH
307 307
0.164 0.216 1.41 5.05 19.9
756 886 948
0 + CsH18 (2,2,4-Trimethylpentane)4OH 307 307
167 84.2 33.4 96.3 13.2
+ C3H12(Isopentane) + OH + CbHll
0 + C7H16(2,2-Dimethylpentane)
0
10-10 k, cma mol-1 see-1
+
cS&7
17.5 22.1
(2,2,3,3-Tetramethylbutane)4OH 60.3 10.7 40.7 33.0
+ C8H17
Table IX: Arrhenius Parameters for Abstraction Reactions of Atomic Oxygen a t Primary, Secondary, and Tertiary C-H Bonds and Total Rate Constants a t 298 and lOOOOKa Reactant
Log A
----Total
10-18 A (per C-H bond)
13.34 13.60 13.77 f 0.17
0.55 0.68 0.49
13.798 i. 0.089 13.905 3~ 0.090 14.031 f 0.083 14.077 f 0.079 13.967 f 0.053
1.57 1.34 1.34 1.20 0.77
E/R
Primary 4460 3300 2920 f 140 Secondary 2570 2320 2250 2190 2030
f 80 80 ct 70 f 70 f 50
rate constantc---lO-l*k (1000OK)
10-lok (298OK)
0.0007 0.056 0.33 1.3 3.5 5.6 7.7 10.1
0.25 1.4 3.2 6.4 9.6 13 15
13
Tertiary 13.494 f 0.095
1.57
1650 f 80
12
9.1
Uncertainties for this work are standard errors of the reported a This work except as noted. Units of A and k are cm3 mol-' sec-1. See ref 6. The Arrhenius parameters were recomputed by us using all the experimental values based on a least-squares treatment. data. They differ somewhat from those given in the original paper, k(C&) = 1.7 f 0.2 X 1013exp(-8700/RT) cmS mol-' sec-1 and See text. lc(CzH6) = 1.8 & 0.2 x 1018 exp(-6100/RT) cm8 mol-lsec-l. The Journal of Physical Chemistry
3333
RATESOF REACTION OF ATOMIC OXYGEN
n -PENTANE THIS WORK
1010
Figure 1. Arrhenius plot for the reaction of atomic oxygen with neopentane.
n-BUTANE o ELIAS AND SCHIFF THIS WORK A AVRAMENK0,etd ’--AZATYAN MARSH AND HEICKLEN X WRIGHT
1
I
1.0
I
I I 2 .o 1 0 3 1 ~K-I ~
I
I
3.0
I
4.0
Figure 3. Arrhenius plots for the reaction of atomic oxygen with n-pentane. The solid line is the rate constant for abstraction at secondary C-H bonds; the dashed line is the total rate constant.
I
1
I
I
I
i
1
I
-
n -HEXANE THIS WORK
-
+
-
-
-
-
1.0
1.0
2.0
3.0
4.0
2.0
3.0
4.0
IOVT,K-1
1 0 3 1 ~K-I~
Figure 2. Arrhenius plots for the reaction of atomic oxygen with 12-butane. The solid line is the rate constant for abstraction at secondary C-H bonds; the dashed line is the total rate constant.
Figure 4. Arrhenius plots for the reaction of atomic oxygen with n-hexane. The solid line is the rate constant for abstraction a t secondary C-H bonds; the dashed line is the total rate constant. Volume 73,Number 10 October 1969
JQHNT. HERRON AND RQBERT E. HUIE
3334
Figure 5 , Arrhenius plots for the reaction of atomic oxygen with n-heptane. The solid line is the rate constant for abstraction at secondary C-H bonds; the dashed line is the total rate constant.
Figure 7. Arrhenius plots for the reaction of atomic oxygen with 23-dimethylbutane. The solid line is the rate constant for abstraction a t tertiary C--H bonds; the dashed line is the total rate constant.
I
1014
I
I
I
I
I
t
I 013
Figure 6. Arrhenius plots for the reaction of atomic oxygen with n-octane. The solid line is the rate constant for abstraction a t secondary C-H bonds; the dashed line is the total rate constant. The Journal of Physical Chemistry
Figure 8. Arrhenius plot for the reaction of atomic oxygen with ethane.
I
1
RATESO F REACTION
OF
3335
ATOMICOXYGEN
reported values based on a least-squares treatment. The data will possess a higher degree of internal consistency than absolute accuracy, since unknown sources of error and bias will probably cancel out when making internal comparisons. As we have noted p r e v i ~ u s l y the , ~ errors associated with the measurement of pressure, flow, temperature, and reactor geometry are small in comparison with the reported uncertainties. The principal source of uncertainty, aside from spurious effects such as those arising as a result of heterogeneous reactions, is in the determination of In [AolAt]. When [&/At] becomes very small (e.g., secondary > tertiary; second, for a particular bond type, the activation energy decreases as the molecule become more complex. The order of activation energies primary > secondary > tertiary is the same as the order of the corresponding bond energies and is that normally observed for the abstraction of hydrogen atoms by other atomic and free-radical rea~tants.’~J’However, the selectivity of attack a t different bond sites is quite variable, ranging from fluorine atoms which are only slightly selective to bromine atoms which are strongly selective. Atomic oxygen is intermediate in selectivity, having approximate primary :secondary :tertiary relative rates a t 307°K of 1.0:21:185. (7) V. V. Azatyan, A. B. Nalbandyan, and Ts’ui MengYuan, Dokl. Akad. N a u k , SSSR, 147,361 (1962). (8) D. Saunders and J. Heicklen, J . Phye. Chem., 70, 1950 (1966). (9) L. I. Avramenko, R. V. Kolesnikova, and N. L. Kuznetsova, Izv. Akad. h’auk, S S S R , Otd. Khim. iVauk, 620 (1963). (10) L.I. Avramenko, R. V. Kolesnikova, and G. I. Savinova, ibid., 976 (1963). (11) L. I. Avramenko, R. V. Kolesnikova, and G. I. Savinova, Izv. Akad. Nauk, SSSR, Ser. Khim., 253 (1967). (12) L.Eliaa and H. I. Schiff, Can. J . Chem., 38,1657 (1960). (13) V.V. Azatyan, Arm. Khim. Zh.,20, 577 (1907). (14) G.Marsh and J. Heicklen, J.Phys. Chem.,71, 250 (1967). (15) F.J. Wright, Tenth Symposium (International) on Combustion, The Combustion Institute, 1965,p 387. (16) J. M.Tedder, Quart. Rev., 14,336 (1960). (17) A. F. Trotman-Dickenson, Advan. Free Radical Chem., 1, 1 (1965). Volume 79,Number 10 October 1969
3336
JOHN T. HERRON AND ROBERT E. HUIE
The second conclusion, that the activation energies vary with molecular complexity, has been shown by Knox and Nelson'* to also hold for the abstraction of hydrogen atoms by chlorine atoms. Their data, given in Table X for comparison, indicate that the general
D(C-HI,kJ mol-' 350
I
450
400
I
I
Table X : Data of Knox and Nelson on the Reactions RH C14R HC1"
+
Reactant
+
10-18 A (per C-H bond)
Primary 0.6 f 0.1 1 . 5 rrt 0 . 2 1.7 f 0 . 2 1.4 f 0.2 1 . 3 f 0.2 1.4 f 0.2 Secondary 3.6 =k 0 . 5 2.2 f 0.3 0 . 9 i 0.1 3 . 2 f 0.4 Tertiary 1.7 =k 0 . 3
E/R~
1938 f 30 523 =k 38 503 f 38 398 f 40 413 4 40 463 f 40
10 f 45
See ref 18. Uncertainties are those given in the original paper, suitably converted as noted below. b Original data given in terms of E. To convert we have used R = 1.987 cal OK-' mol-'.
rate expression used by us for reaction a t a primary bond based on the results for neopentane will probably be slightly in error when applied to the other alkanes. Fortunately, the degree of correction for a primary is small, so that an error in this quantity is not very significant. The general trends in the activation energies of Table IX are better illustrated by fitting the data to an Evans-Polanyi expression of the form E = a[D(C-H) - Cl], where E is the activation energy, D(C-H) is the strength of the bond broken, and cy and C1 are empirical constants." I n Figure 9 we plot E vs. D(C-H) for the alkanes, cycloalkanes, and acetaldehyde. Considering only the data on the noncyclic alkanes (Le., that from Table IX) we derive values of the Evans-Polanyi coefficients Table XI : Values 01 and C1 for the Evans-Polanyi Expression E = or[D(C-H) - C11"
I NFa Br CHs 0
a
0.97 1.1 0.86 0.49 0.36
01
69.0 76.1 82.5 74.3 82.0
01 and C1 for other a Data on atomic oxygen from this work. reactants from J. A. Kerr, Chem. Rev.,66,465 (1966). ~
The Journal of Physical Chemistry
L
t
Bi
W
342 =k 38 136 f 40 2084 i 40 4152 f 43
a
Radical
-
I
BO
/I 85
I
I
I
I
90
95
I00
I05
D (C-H),kcol mol'!
Figure 9. Activation energy as a function of the strength of the C-H bond broken. Activation energies for the alkanes from Table IX, for the cycloalkanes from W. K . Stuckey and J. Heicklen, J. Chem. Phys., 46,4843 (1967), and for acetaldehyde from R. D. Cadle and J. W. Powers, J. Phys. Chem, 71, 1702 (1967). Bond dissociation energies from J. A. Kerr, Chem. Rev.,66, 465 (1966).
a and CI given in Table XI. Values of these coefficients for other radicals are given for comparison. The atomic oxygen reaction cannot be as well fitted by an Evans-Polanyi expression as can the other reactant species listed in Table XI. This is not only a consequence of the generally lower precision of flowtype experiments as compared with static-type experiments on which the other data of Table X I are based, but also to the fact that the range of activation energies found for abstraction of secondary hydrogens from the normal alkanes precludes an exact fit, assuming it is valid to use a common value for the secondary C-H bond energy. A further consequence of the inconstancy of activation energy within bond type is that it is not possible to derive a set of exact rate parameters for individual primary, secondary, or tertiary bonds which could be used for estimating rate constants for other alkanes. Fortunately, the spread of values of rate coefficients is not so great that we cannot at least set up some approximate rate expressions. (18) J. H, Knox and R. L. Nelson, Trans. Faraday Soc., 55, 937 (1959).
3337
RATESOF REACTION OF ATOMIC OXYGEN On the basis of Table IX we suggest the following rate expressions (per C-H bond) for reaction a t primary ( p ) , secondary (s), and tertiary (t) bonds
Table XII: Comparison of Predicted and Measured Rate Constants for the Reactions of Atomic Oxygen with Some Alkanes"
k , = 0 . 5 X 1013 exp(-2900/T)cm3 mol-l sec-I
k (pre-
k, = 1 , 3 X
dicted) k (measured)
1013
exp( -2250/T)cm3 mol-' sec-l
k t = 1 - 6 X l O l 3 exp( - 1650/T)cm9 mol-' sec-' Using these rate expressions, we have computed the rate constants given in Table XI1 for some miscellaneous alkanes and compared them with measured values taken from various sources. I n general, we find the predicted values to be within a factor of 2 or 3 of the measured values. The most striking exception is methane for which the ratio of predicted to measured values is 150. This is simply a consequence of the usual "uncharacteristic" behavior of the first member of a series and the manner in which the value of IC, was chosen. The other notable exception is isobutane, for which the difference between predicted and measured values is so great that it casts considerable doubt on the accuracy of the measured value. A more reasonable value would be about 3 X 1O'O cm3mol-' sec-I a t about 300°K.19 This method of estimating rate constants can be used a t present only for the alkanes. I n the case of the chloro- and bromoalkanes, although it is known that a halogen substituent has a strong inductive effect,2 there is no detailed knowledge as to the relative importance of different reaction sites; therefore, it is not possible as yet to derive general rate parameters.
Temp,
10-1%. cm8 mol-' sec-1 Predioted Measured
Reactant
OK
Methane Ethane Propane n-Butane Isobutane Isopentane ZJ2-Dimethylpentane Z14-Dimethylpentane 2,2,4-Trimethylpentane 2,3,4-Trimethylpentane 3-Methylheptane 2,2,3,3-Tetramet hylbutane n-Octane
307 307 307 307 293 307 307 307
0.16 0.24 2.0 3.7 6.0 9.4 3.9 17.
1.6 0.36d 8.0 6.5 10
150 2.8 2.0 2.3 17 1.2 0.60 1.7
307
9.9
7.1e
1.4
a
0 .O O l l b 0.08jb 1.oc
307 300
23. 13.
30 6.5'
0.77 2.0
307 307
0.70 10.
0.84 12
0.83 0.83
This work except as noted (experimental data from Tables
I, VII, and VIII). b Computed from rate coefficients given in D. Saunders and J. Heicklen, J . Phys. Chem., 70, Table IX. 1959 (1966). d Reference 15. The data of Marsh and Heicklen14lead to a value of 4.6 X 1Olo cm3 mol-' sec-1, in fair agreeH. W. Ford and N. Endow, ment with out measurements. J . Chem. Phys., 27,1277 (1957).
'
(19) The principal peaks in the mass spectrum of isobutane correspond t o major background peaks, so that we were unable to make a direct check on this rate constant.
Volume 7 3 , Number 10 October 1969
