REACTIONS OF ACTIVE NITROGEN WITH NO
March, 1960
AND
NOz
319
THE REACTIONS OF ACTIVE NITROGEN WITH NITRIC OXIDE AND NITROGEN DIOXIDE' BY G. J. VERB EKE^ AND C . A. WINKLER Upper dtniosphere Chemistry Research Group, Physical Chmlistry Laboratory, McGill University, kfontreal, Canada Received February 18, 1969
Both reactions have been studied over a range of reactant flow rates. NzO has been found as a product of the NOz, but not of the NO, rertction. The NO2 reaction also yielded NO. The chemical activity in the active nitrogen, inferred from the amount of NO destroyed, haa been found to be approximately 1.5 times that indicated by the maximum yield of HCN in the active nitrogen-ethylene reaction, for a wide range of flow rates of active nitrogen. Moreover, the ratio of the activities measured in these two ways increased with total pressure. These observations might imply that NO reacts with more than one reactive species in active nitrogen and that measurement of the nitrogen atom concentration by the extent of the NO reaction might be subject to considerable error.
In previous studies in this Laboratory, it has been assumed frequently that, active nitrogen concentrations could be takeii to correspond to the essentially constant limiting yield of hydrogen cyanide produced in the reactions of active nitrogen with various hydrocarbons a t elevated temperat u r e ~ . ~On . ~ the other hand, the concentration of active nitrogen so estimated was much higher than that indicated by the amount of ammonia destroyed under comparable conditions, and it was suggested that ammonia reacted with one, and ethylene reacted with a second reactive species in the active nitrogen. Since earlier work by other investigators6-' had indicated i;hat the extent of the actiGe nitrogen-NO reaction might be used as a measure of the active nitrogen concentration, it was decided to re-investigate the reactions of active nitrogen with KO and NOz, in the hope of obtaining more definitive information about the composition of active nitrogen.. The reactions of active nitrogen with oxides of nitrogen have been the subject of a number of recent investigations.8-l1 The reactions of nitrogen dioxide are necessarily complex and will be discussed further when the results of the present study have been presented. It was, however, generally concluded that n:itric oxide reacts completely with atomic nitrogen by the fast reaction
Experimental
Nitrogen was removed from commercial NO (Matheson Co.) by evacuation while the NO was kept a t li uid nitrogen temperature, and then by repeated distillationqthree times) from a frozen pentane bath (-130'). Infrared analysis showed the product to contain no traces of higher oxides and less than 0.3% of NzO. NO2 was prepared by treating purified NO with excess oxygen. When the deep blue color of Nz03 was no longer visible in the condensed material, excess oxygen was pumped off. The NO2 was kept frozen when not in use to prevent photochermcal decomposition. When traces of Nz03 were found to be present, the oxidation procedure was repeated. Halocarbon stopcock grease was used in all parts of the system in contact with NOZ. Ethylene, ammonia and nitrogen were of the grades, and were purified by the methods, described in previous papers. Active nitrogen was formed in a condensed discharge through nitrogen in an apparatus that waa essentially similar to that used in earlier investigations of active nitrogen reactions in this Laboratory ( e . g . , ref. 1). For some experiments the spherical, 300-cc. reaction vessel generally used was replaced by a tube 140 cm. long, 18 mm. i.d., into the side of which were sealed 6 reactant inlet jets at intervals down the length of the tube. Furnaces were provided so that the spherical or tubular reaction vessels could be heated aa dcsired, and the discharge tube and reaction vessels were poisoned with metaphosphoric acid to minimize wall deactivation of the active nitrogen. The products of the ethylene reaction, and residual ammonia from the ammonia reaction, were trapped in liquid nitrogen, while liquid nitrogen under reduced pressure ( -210") was found necessary to trap quantitatively the NO from the nitrogen oxide reactions. A small coil of silver wire was placed ahead of the trap to remove oxygen atoms and prevent formation of small amounts of ozone that N NO + Ny 0 tended to cause explosions if trapped. To obtain maximum precision in the experiments, the the extent of which could be used to measure the power supply to the discharge tube was voltage-controlled, concentration of nitrogen atoms. This "titration" and the flash-rate in the discharge tube was controlled a t actually has been used to study t'he kinetics of either 12 or 20 flashes per second, as indicated by an oscilloscope or by a 929 phototube recording on a Brush oscillorecombinat'ion of nitrogen atoms. 1 2 - 1 4 graph. (1) With financial assistance from Imperial Oil Ltd., Sarnia. and the The extent of decomposition of NO was obtained by analyNational Research Counril. Ottawa. sis of residual reactant. No attempt waa made to distin(2) Holder of a National Research Cooncil bursary, 1954-1955, guish NO from NO2 in the products, since excess NO was studentship, 1958--lElX,and a Provlnce of Qiiebec Scholarship. 1950partially oxidized to NO1 by reactions with atomic and 1957. molecular oxygen. An excess of oxygen was added to the (3) (a) P. Gartaganis and C . A . Winkler. Con. J . Chem., 84, 1457 trapped products of reaction, after which the mixture was (1956); ( b ) B. Dunford, H. G . V. Evans and C. A . Winkler, rbid.. 34, allowed to warm to room temperature to ensure complete 1074 (1956): (c) D. Wiles and C. A . Wlnkler, i b i d . , 81, 1298 11957). oxidation of unreacted NO. The NO2 was then frozen out, (4) N. V . Klassen. M. Onyszchuk. J. McCabe and C . A. Winkler, excess oxygen pumped off, and the amount of NO, finally i b i d . , 86, 1217 (1958'1. measured by pressure measurements in a known volume at (5) R. J. Strutt, E'roc. R o y . Soc. ( L o n d o n ) , A86, 56 (1912). constant ternperature.l6 It might be noted perhaps that (6) E. J. B. Wil1e:r and E . K . Rideal. J . Chem. Soc., 1804 (1926). generally accepted methods of analysis based on oxidation of (7) M. L. Spealman and W . IT. Rodebush, J . Am. Chem. Soc., I?, NOz to nitric acid were not found to yield satisfactorily re1474 (1935). producible resulk.
+
+
(8) P. Harteck and 5. Dondes, J . Chem. P h y a . 43, 953 (1954). (9) P. Harteck an83 S. Dondes. ibid., 34, 619 (1956). (10) P. Harteck and S. Dondes, ahid., 47, 546 (1957). (11) G . B. Kistiakowsky and G . G . Volpi, ibid., 47, 1149 (1957). (12) J. T. Herron, J . L. Franklln, P. Bradt and V. H . Dibeler, i b i d . , 39, 230 (1958).
(13) J. T. Herron, J. L. Franklin, P. Bradt and V. H. Dibeler, ibid., SO, 879 (1959). (14) P. Harteck, R. R. Reevea and G. Manells. ibid., 39,608 (1958). (15) F . H . Verhoek and F. Daniels. J . Am. Cham. SOC.,I S , 1250 (1931).
G. J. VERBEKEAND C. A. WINKLER
320
The products of the NOz reaction were analyzed by adding a known pressure p l of oxygen to oxidize all the NO, and the total pressure p , , was measured. Unreacted oxygen was then pumped ob while the other gases were frozen with liquid nitrogen, after which they were allowed to warm to room temperature and their pressure, pa, determined. The partial pressure of NO then waa given by 2(p, p3 - p2) PNO Finally, the aacs were surrounded with a Dry Ice-acetone ~ off to permit the pressure p , of residbath and N z pumped ual NO, to be obtained when i t WM again restored to room temperature. The partial pressure of NzO waa given, of course, by pa - p4, while p4 gave the sum of NO and unreacted NO?. Hydrogen cyanide and ryanogen in the ethylene reaction were analyzed by titration methods, as in previous investigstions, and ammonia waa estimated by absorption in standard acid, followed by back titration with standard alkali. Experiments with known amounts of the various gases showed that the analytical methods gave results within +2%.
+
Results Nitric Oxide Reaction. Products of Reaction and Reaction Flame.-The amount of NO destroyed increaaed linearly with flow rate until it reached a rather critical value. This behavior is characteristic of a rapid reaction in which the reactant present in smaller amount is completely consumed. The plateau of NO destruction a t higher flow rates was found to be essentially independent of temperature up to 400°, the maximum temperature studied. No N20 was produced at any flow rate in the range studied. Ozone was the only product condensed a t flow rates below the critical value, where a blue reaction flame (p- and ybands of NO) was observed.16 Om the other hand, both NO and NO2, but no ozone, were found a t flow rates above the critical, under which conditions the blue glow contracted to a layer of few millimeters thick just below the reactant inlet and was largely (though not completely) replaced by a yellowish-green glow (NO2continuuin). Effect of Active Nitrogen Concen tration.-The amount of NO destroyed by active nitrogen was compared with the amount of HCN produced in the reaction of active nitrogen with ethylene when the concentration of active nitrogen was changed by alteration of the capacitance in the electrical circuit over a 16-fold range. Experiments were made with excess of NO or C2H4 and with the reaction vessel heated to 300 and to 400°, over which interval it was found that the extent of reaction with both reactants was independent of temperature. Analysis of products showed that cyanogen production from ethylene did not exceed 1% of the hydrogen cyanide production and was therefore negllected in drawing the comparison. The results for a range of capacitance values are shown in Fig. 1A. The ordinates represent maximal values for NO destroyed or HCN produced, read from plots of NO destroyed or HCN produced against the corresponding reactant flow rates. Within the error of the experiments, the ratio of NO destroyed to HCN produced appeared to be constant, as indicated by the data Capacitance (rfd.): Ratio
HCN produced
0.25 0.50 1.O
2.0
4.0
1.52 1.71 1.49 1.11 1.51
(16) F. Kaufman and J. Kelso, J . Chsm. Phys., 47, 1209 (1957).
Yol. G 1
Comparison of the Reaction of Active Nitrogen with NO, NH, and C2H4.-Figure 1B shows the amounts of NO and NHI destroyed, and of HCN formed from C2H4,when excesses of these different reactants were introduced into the tubular reaction vessel at different positions. The molecular nitrogen flow was 170 micromoles/sec., a t a pressure of 1.90 mm., with a discharge rate of 20 flashes per second. No difference was observed in the amount of NO destroyed a t the top reactant inlet over the temperature range 100 to 300”. The reaction tube was not heated during any of the experiments with NH3, but was heated to 300’ or higher for all the experiments with C2H4. The ratio of NO destroyed to HCN formed from ethyleiie appeared to increase slightly, though the increase was of the same order as the experimental error, as decay of active nitrogen progressed down the tube. Ilistanre k d o w first inlet (cm.): 0 20 40 60 80 120 Ratio NO destroyed: ______ 1 35 1 3 9 1.36 1.37 1 40 1.50 HCX produced
An increase was observed in the ratio of NO destroyed to HCN produced, when the total pressure was increased,17as shown by Ne- -pressure (mm.): . . Ratio HCN produced
1
1.4
2
1.6
4
8
1.9
1
2.3
6
2.4
Nitrogen Dioxide Reaction.-In Fig. 2 (L4,B and C) are shown the results of experiments with NOz a t the top reactant inlet, in an unheated reaction tube and with the tube heated to 300”. The blue glow indicative of oxygen atoms again was observed, and increased in intensity as the flow rate of NO, was increased up to a critical flow rate where a sharp transition to the yellow-green glow occurred. Nitrous oxide was the only condensable product a t flow rates where only the blue glow could be seen. The green glow was gradually reduced in extent and finally disappeared, as the flow rate of NOz was further increased. The ratio of NO2 flow a t which the green flow disappeared to the flow at which it first appeared was found to be 1.8 f 0.1 for three different values of energy input to the discharge tube. For flow rates of NO2 in excess of that where the green glow appeared, both NO and NO2, as well as N20, were recovered in the products. Rate Constants for the Nitric Oxide and Nitrogen Dioxide Reactions.-Rate constants for t,he reactions of NO and NO2 with active nitrogen were determined by a gas “titration” for active nitrogen analogous to that proposed for oxygen atoms by Kaufman.18 The analysis was based on the observation, established experimentally, that the green glow appeared simultaneously with the first analytically detectable excess of either reactant in the effluent gases from the reaction vessel. Apparently the process responsible for the green (17) A microwave generator waa used t o produce t h e active nitrogen, owing t o difficulties in maintaining a condensed discharge at t h e higher preaauren. (18) F. A. Kaufman, J . Chsm. P h y s . , 28, 352 (1958).
I~EACTIOKS OF AlcwvENITI~OGEN W ~ T HNO A N D XUz
March, 1900
+
32 1
+
glow (NO 0 + NO2* -+ NOz hv) is slower than the reactions of active nitrogen with either NO or NO2, and does not occur in the presence of active nitrogen. Hence, essentially all the active nitrogen reaction occurs in the region where the blue glow is visible. A measure of reaction time should therefore be possible from the length of the (contracted) blue region and the linear velocity of the gas in the reaction tube. The length of the blue region (genrmlly a few millimeters) was easily measured in a darkened room. The linear velocity was calculated from the cross-section of the reaction tube and the volume of flow, without attempting a correction for a possible velocity gradient between the center of the tube and the wall. For the calculations, the reaction was assumed to be 90% complete in the reaction time, hence the values obtained for the rate constants must be regarded as lower limits. The integrated rate expression, assuming a bimolecular rate-controlling step with complete lateral mixing, takes the form k =
1
t(h'i - R.)
in
1
2
3
20
4
CAPACITANCE p F d
60
40
100
80
I20
D I S T A N C E BELOW F I R S T I N L E T
CM
Fig. 1.-(A) effcct of active iiitrogen concciitrstioii 011 the rate of destruction of KO and on the rate of formittion of HCN from C2H4; ( U ) aniouiits of NO aiid ?;H) destroycd, and HCN produced from C2H4,after difrcrciit tiines of decay of active nitrogen. 40r -
-
-
1
A
'
30.
--I
i
, "
RiLV I
NiRf
where k = second-order rate constant t = reaction time Ni and Ri = initirtl concn. of active nitrogen and reactant
,Vf and lif = concn. of active nitrogen and reactant after time t
1
From thermocouple measurements with comparable flow conditions in the spherical reaction vessel, the temperature was estimated to be approximately 500°K. Collision numbers were calculated from 10 20 30 40 50 kinetic theory in the usual way, assuming collision NO2 F L O W R A T E - M O L E / S E C X I O 6 diameters to be: N atom, 2.98 A.; NO, 3.4 A.; Fig. 2.-Ratrs of fornii~tioiiof products i n thc rcactioii of NO2, 5 A. The (data used in thc calculations and with active nitrogen: A, 0 NO2 (urihe:ittd), NO2 the values of the second-order rate constants were NU2 (300'); A NO + NO2 (unheated), A NO + SO2 (300"). .4rtivc Nr flow rate iiiolcs/sec'.
RPnm:t,ant
Row rate, iiioles/scc.
Length, of blue zone. inm.
ltcaction time, see.
A. Nitric oxide 18.7 X 20 X: lo-" 4 2 . 5 X lo-' 18.7 X 40 X 3 1 . 9 X lo-' Collision no. = 2.0 x 1014 cc./Inole fiec.
+
B. 0 NzO (unheated), 0 N20(300'); A N 2 0 produccd nitrogen oxides lost (unheated), A NtO producd nitrogeii lost (300'). C. 0 NO (unheated), 0 NO (300'); cc./molc BCC. oxides A On (unheated) A Os (300'). kl.
+
atoms, and a good deal of evidence supports this view. In particular, when hydrogen cyanide formation is plotted against flow rate of ethylcnc, a plateau value for hydrogen cyanide produetioil is 13. Nitrogen dioxide reached which is independent of ethylene flow 18 >: 10- 2 1.2 X lo-' 5 . 3 X 1012 rate a t high flow rates, and independent of tcmpera14.0 X ture in the range of about 250 to 450". Moreover, 14.0 X 10-6 28 >I: 1 . 5 1 . 0 X lo-' 4 . 5 X lW9 Collision no. = 2.0 X lo1' cc./molc sec. this maximal hydrogel1 cyanide formation is found Total flow rate = 2 0 0 x 10-6 molc/scc. to be essentially the same as that from such diverse Total pressure = 2.2 mm. reactants as propane,'9 isobutane,*O ~yclopentane,~ methyl- and dimethylacetylene,21 and methyl Temperature = 500'K. chloridez.z2under comparable conditions. Linear velocity = 16 m./sec. The lower concentration of active nitrogen Discussion indicated by the yield of hydrogen cyanide from Nitric Oxide ]Reaction.-The most striking ob- ethylene, compared with that inferred from the deservation of t,he present study was that nitric composition of nitric oxide, is not likely to be due oxide was destroyed by active nitrogen in sig- to loss of nitrogen atoms by recombination catanificantly larger amounts than hydrogen cyanide lyzed by ethylene or one of the reaction productsz3 was formed from ethylene for conditions of maxi(19) M. Onyaachuk. L. Breitman and C . A. Winkler. Can. J . Chsm.. mal hydrogen cyanide production. I n previous 89, 351 (1954). (20) R. A. Back and C . A . Winkler, i k d . , 89, 718 (1954). studies from this Laboratory, the maximal produc(21) A. Schavo and C. A . Winkler, $bid., in p r w . tion of hydrogen cyanide from ethylene (together (22) 8. Sobering and C . A. Winkler, i h d . , S6, 1223 (1958). with a small amiount of cyanogen) has been taken (23) W. Forst, H. G. V. Eva- and C. A. Winkler, THISJOURNAL, to be a measure of the concentration of nitrogen 6 1 , 3 2 0 (1067). 4 . 0 X 10" 1 . 2 X 1Ols
G. J. V E ~ S E AND ~ E C. A. WINKLER
322
since the plateau from which the active nitrogen concentration is inferred in the ethylene reaction is not dependent on either flow rate of reactant or temperature. If it is accepted that the nitrogen atom flow rate in the present study is given by the hydrogen cyanide plateau value (11 f 0.2 ICmole/sec. for a condenser capacitance of 2 fd.) the diBerence between this value and the corresponding plateau value for decomposition of nitric oxide (16 i 0.2 pmole/sec.) is presumably due to deslmction of nitric oxide by some reaction other than dlirect attack by nitrogen atoms. The extent of the NO reaction might therefore be a quite unsatkfactoiy measure of the nitrogen atom concentration in active nitrogen. There is little doubt that the main reaction of active nitrogen with NO is N +NO+N2+0+75kcal." (1) The spin forbidden reaction N('5)
+ NO(%) +N&('Z) + 114.5 kcal.
(2)
apparently did not occur Since no NzOwas fouiid as a product at any flow rate of NO, yet it is known that NzO is stable to active nitrogen at temperatures UP to %ooo,l1,u and was observed to react to less than 5yoin some experiments at 300" in the present study. A possible chain reaction for the decomposition of NO by nitrogen atoms is N + NO +N-NO (3) NO + N-NONz + Or + N (4) N-NO +Nt
+0
(5)
However, the complex, N-NO, should be comparable with a m y excited N& molecule, and should not have significant lifetime. However, if this lifetime were sufficient to allow further reaction of the complexes with NO, some of them would presumably be stabilized to M 2 0 by coWon with nitrogen molecules. A scheme involving excited oxygen atoms seems even less attractive since there is only enough energy available from reaction 1to produce the low lying ID state. A difference between the maxjmal amount of HCN produced from Ca and the amount of NO decomposed is explained readily if it is Bssumed that, in addition to atomic nitrogen, with which both GIt and NO react, there is a second species in active nitrogen that is capable of decomposing NO but is not capable of reacting with GH, to produce HCN. If, as suggested previously," this second species were an excited molecule, destruction of NO might occur by IVO + N 2 * +N2 + N + 0 (9) Decompogition of NO by electrollieaiiy excited Nt mms plausible, especially since the A%.+ state lies parficularly close to the disrsocistion energy of NO (6.49 e.v.) which would favor collisions of the second kind. The A state is the end product of the radia4ive recombination of nitrogen atoms (24) Vduea for the h a t s of formation, med to d&te the h a t s of r d o n . -t&en from circpkr 500. N a W Bureau of 8tandude. W M W . 19a. (26) E . G. V. Evms. G . 11. Freeman and C. A. W a a r . Cam. J . C k r . U.1217 (19.56).
VOl. 64
and is also one of the states into which two 's atoms can recombine directly. The radioactive half-life of the A state is now recognized to be of the order of 0.1 sec., which is approximately equal to the time required for the active nitrogen to travel the length of the long reaction tube used in the present experiments, and considerably longer than the 0.02 sec. required for it to reach the spherical reaction vessel from the discharge tube. Rigorous treatment is not possible without knowing the initial extent of dissociation in the discharge tube. However, approximate calculations show that the rate constant for the homogeneous recombination of N atoms to form excited molecules86is of the right order to permit a concentration of molecules in the A state sufficient to account for the excess of NO destroyed over HCN formed from GH, under conditionsof complete reaction. Ground state vibratiody excited molecules (v 5 27) might also be effective in causing dis Socistion of NO.% If the half-life of a vibrational level of ground state nitrogen is of the order of 10' c~llisions,~ the half-life for the experimental conditions used would have been about 0.1 sec., Le., about the same as the probable half-life of the A state. It does not seem possible, from the present study, to come to any conclusion about the relative importance of electronically and vibrat i o d y excited molecules in promoting the dis socistion of NO. Nitrogen Dioxide Reaction.-The reaction of active nitrogen with NO2 was found to yield considerable NzO. The relative amount of this product was larger than that found by Kistiakowaky and Volpi" in their mass spectrometric investigation." These ditrerences are almost certainly due to the difficulty, mentioned by these authors, in obtaining reliable mass spectrometer data in the presence of NOz. There seems little doubt that NnO was formed from NO2 by the reaction N + NOs .--) NtO + 0 + 42 k d . (10) Since NtO is not sign3cantly attacked by active nitrogen, the amount of NSO formed may safely be taken as a measure of the extent to which this reaction occurs. Since the NZO yield was never more than ti fraction of the NOz that was not recovered as either NO or NOZ,some N G must have been converted to non+ondensable products. This would suggest an additional reaction such 8s N
+ NO,+Nz + 02 + 122 k d .
(11)
which, however, might have occurred by a more complex mechanism, e.g. NOt
+ 0 +NO + Or + 47 kcal.
(12)
followed by reaction 1. The data for the production of atomic oxygen in the system reveal that reactions 10 and 11 are not adequate to explain all the experimental d t s (26) (a) J. Berkowits. W. A. Chupka rad G. B. gisti.torsly. J . C k n . PhUJ., lT, 1417 (1957); (b) R. K d l y and C. A. W a r n . Cor. J . C k n . , ST, 62 (1959). (27) 8. I. Luk& and J. E.Young, J . C h . PhuJ..m, 1149 (1957). (28) The rate of formation wan amooth function of NOI flow rate with neithar maxima nor minimr.
March, 1960
INTRAMOLECULAR PROCESSES IN UNIMOLEWR REA~ONS
+
with NG*. The green afterglow due to NO 0 + NO, became visible when the flow rate of N G exceeded 14 pmoles/sec., and disappeared rather suddenly when the flow rate was increased to 25 pmoles/sec., at which point it may be assumedll that all the oxygen atoms were consumed by reaction 12. Sia: no NO or NOI was recovered at flow rates of NOI smaller than 14 pmoles/sec., it may be concluded that the minimum flow rate of atomic oxygen was 11 pmoles/sec. Obviously, this cannot be explained by only reactions 10 and 11, since the maxjmum production of atomic oxygen from these reactions cannot exceed the flow rate of NzO, which was never greater than 4.9 pmoles/sec. The most plausible reaction to produce additional atomic oxygen is probably NO, + N +NI + 2 0 + 4 ked. (13) If this reaction were responsible for all the destruction of N% by active nitrogen (14.0 moles/ sec.), apart from that converted to N20 (4.9 pmoles/sec.), an atomic oxygen flow of at least 18 pmoles/sec. would result. Since this is considerably in excess of 11 pmoles/sec. indicated by “titration” to the disappearance of the green glow, it would be nece,ssaryfor reaction 13 to occur to a limited extent of about 3 pmoles/sec. destruction of NO*. The recovery of NO NOS was linear in NOi flow rate, with a, slope of unity, and extrapolated to 14.0 pmoles/a?c. of NOSfor zero recovery of the oxides. This suggests that, up to a flow rate of 14.0 pmoles/sec., all the NOS reacts with active nitrogen while, at higher flow rates, some of the excess is converted to NO, presumably by reaction of oxygen atoms formed in the active nitrogen reaction, i.e., reaction 12.
+
-
323
It seems likely that the reaction 2MO
+ 0,
2N0,
(14)
occurred in competition with the reaction of NO with N G to form N,OI, which was evident in the trap from its blue color. If, as indicated above, the amount of NO, destroyed by active nitrogen may be taken as 14.0 pmoles/sec., it is necessary to account for the discrepsncy between this value and the 18.7 pmoles/sec. of NO destroyed under identical conditions. It could be explained if the reaction N+N4+2NO
(15)
consumed 2.35 jmolea/sec. of NOr, to f o m 4.7 pmoles/sec. of NO, which also was destroyed by active nitrogen. The total destruction of (NO NO2), in pmoles/sec., would then be (14-2.35) by reactions other than reaction 15, plus 7.05 by reaction 15 followed by reactions of NO with atomic nitrogen. However, this would require the rather remarkable coincidence that reaction 15 occur to just the extent necessarg to make the total amount of NOs destroyed correspond almost exactly with the amount (13.9 pmoles/sec.) of HCN produced from GEL under optimal conditions. It might be preferable to Bsrmme that part of the NOz suffers destructive collisionswith excited molecules present in the active nitrogen
+
+ Nz*+
xz + KO + 0
(16)
A study of the reaction of No2 with active nitrogen over a range of operating premms might help to assess the reality of reaction 16, although the rather obvious complexity of the reactions that occur in the N k t i v e nitrogen system is not particularly encouraging to a detailed study of it.
INTRAMlOLECULAR PROCESSES IN UNIMOLECULAR REACTIONS BY DAVID J. WWN Deporfment of C h i s t r y , University of R ~ 4 Ad w t
Rochcsier,Rochder, New York 18.1860
A general meFod is developed foy hand!ing intra- and intermolecular epe-rgy tnrnefer and chemical decomposition in unithat weak i n ~ l e c u l a rcoupling b e molecular mactaona. The methad IS apphed to two simple modela and it tween the d t o m representhy the reactant molecule leads to a tramatmn regan between the hgh and low pressurn Mta of the rate coilstantof a ununolecuiar reaction thst ia b d e r than that predicted by Kamel theory.
It is well established that unimolecular reactions M e place via a, mechanism involving collisional activation and deactivation of the reactant molecule, redistribution of internal energy within the reactant molecule, and reaction of activated reactant molecules. Most previous theories of unimolecular Iwlctions have assumed that the redistribution of internal energy in, and the reaction of, an activated molecule are characterized by a single half-life or relaxation time dependent only on the total vibrational energy of the molecule. We shall here canaider a model in which redistribution of internal energy may take place either by (1) This work wan oupported by the National Wenca Foundation under Grant W.
processes leading to a single, energydependent relaxation time, or by singlepuantumtransfer processes leading to several energydependent relaxation times, or by any type of process intermediate between these two extremes. It will be assumed that the rate of formation by intramolecular energy transfer of one state Aril of the reactant molecule from another state AI,) having the same internal energy is equal to a((m);(i])Arm), where (i) is a set of quantum numbers specifying Aril and a((m);(i))is proportional to the transition probabfity for the indicated transition. The nature of the collision mechsnism and the microscopic decomposition rate will be left general for the time beiig. This theoqy will include (in a
