I. REACTIONS AND PHOTOCHEMISTRY OF ATOMS AND MOLECULES Introduction to Chemical Aeronomy BY MARCEL NICOLET Centre National de Recherches de l'Espace, 3 Avenue Circulaire, Bruxelles 18 Received 12th March, 1964 Thirty-three years ago, Chapman 1 demonstrated that the dissociation of molecular oxygen is important above 100 km, and, therefore, that the photochemistry of atmospheric oxygen 2 is an important problem in aeronomy. However, the treatment of oxygen dissociation at high altitude must be examined 3 by studying the departure from photochemical equilibrium conditions. In fact, any detailed investigation requires a knowledge of aeronomic conditions in the various atmospheric regions. Theoretical studies are simplified by dividing the atmosphere into two parts : the homosphere, in which the composition of the principal constituents (N2 li 78 %, O2 N 21 % and Ar N 1 %) remains constant and the heterosphere, in which the dissocia- tion of oxygen and diffusion affect the air composition.If the temperature distribu- tion is introduced as the second aeronomic parameter, the homosphere is subdivided in three regions : the lowest region is the troposphere where the temperature decreases with height up to the tropopause. The tropopause has a temperature of about 220°K in the polar regions at a height of some 8 km and about 190°K at the equator for an altitude of about 17 km. The stratosphere is essentially that region where the tem- perature increases or at least does not decrease, with altitude. It extends from the tropopause to an altitude of about 50 km where the temperature reaches a peak of about 273°K.The third region belonging to the homosphere is the mesosphere situated between the stratopause (50+5 km) and the mesopause (85+5 km) where the temperature reaches a minimum as low as 160- 170°K. Above the mesopause, there is an increase of the temperature which has its largest gradient (up to 20" km-1) near 150 km. In this region, the thermosphere, the problems of the chemosphere change in that ionization must be considered along with dissociation. The vertical distribution of atomic oxygen makes it possible to understand the different roles of the three regions of the homosphere and the heterosphere. In the thermosphere atomic oxygen is more abundant than molecular oxygen above about 120 km.In the mesosphere, atomic oxygen is more abundant than ozone for an atmosphere illuminated by the sun. The day and night-time conditions are different. In the stratosphere, atomic oxygen is always less abundant than ozone, the presence of which depends on photochemical reactions. Finally, ozone is a pzrmanent element of the air mass in the troposphere and is subject to variations associated with advective and dynamical transport. After considering the principal constituents, it is necessary for chemical aeronomy to introduce the minor constituents. An inert gas such as helium (ratio 5-24 x 10-6 per volume) which has no chemical importance is, however, a tracer for atmospheric diffusion processes since it is not a minor constituent in the upper thermosphere.Several of the minor constituents observed at ground level (table 1) can play a role in 78 INTRODUCTION TO CHEMICAL AERONOMY the chemosphere. No systematic study of their abundance has been made at high altitudes and their behaviour is known from infra-red spectroscopic identifications or from chemical analysis at low levels. molecule TABLE MO MOLECULAR CONTENT OF MINOR CONSTITUENTS ratio by volume 3~ 10-3 10-5 to 10-2 10-7 to 10-8 2 . 5 ~ 10-7 5~ 10-7 5 x 10-8 to 2 x 10-7 1 . 5 ~ 10-6 5 x 10-10 to 2 x 10-8 remarks Mixed in the troposphere, small variation. Variable ; dissociation in mesosphere. Variable ; peak in stratosphere. Mixed in troposphere ; dissociation in stratosphere. Mixed in troposphere ; dissociation in stratosphere. Mixed in troposphere ; dissociation in thermosphere. Variable ; industrial.Variable ; industrial ; chemical origin in meso- sphere and thermosphere. The photochemistry of atmospheric water vapour was studied in considerable detail by Bates and Nicolet 4 after Meinel 5 discovered that the vibrational rotational bands of the hydroxyl radicle OH appear in the airglow with a total energy6 of about 3 ergs cm-2 sec-1. Such an emission arouses interest in the photochemistry of hydro- gen+oxygen compounds 7, not only of water vapour but also of methane and perhydroxyl radicles. The attack of methane by atomic oxygen was studied by Bates and Witherspoon 8 and there is little doubt that the concentration of methane in the mesosphere depends on collision processes involving atomic oxygen. The enteric fermentation corresponds to a production 9 of CH4 of at least 1010 molecules cm-2 sec-1.The chemical oxidation of methane in the stratosphere leads to a production of hydrogen, the accumulation of which is limited by the possibility of the upward transport into the mesosphere. Unlike oxygen, nitrogen is extremely difficult to dissociate and it is so stable that it remains in the molecular form up to very great altitude. This low degree of dis- sociation was suggested 10 in studies of nitric oxide which was considered to be an important ionic constituent of the terrestrial ionosphere. The photochemistry of tropospheric nitrous oxide has been investigated by Bates and Witherspoon 8 who indicated that this molecule is not a member of the main photochemical family of nitrogen oxides which were studied by Bates 11 and Nicolet 12, 13.Free nitrogen atoms in an oxygen atmosphere make possible a large number of reactions which are now studied in the laboratory. A difficulty in giving a systematic account of the chemical aeronomy has been the grevious lack of reliable basic data. Our knowledge concerning the experimental rate coefficients has increased rapidly, however, in recent years and systematic accounts can be found in several review papers presented at the symposium an aeronomy held in Berkeley in August 1963 : Three-body reactions by Barth ;14 reactions involving nitrogen and oxygen by Schiff 15 and aeronomic reactions involving hydrogen by Kaufman.16 Much progress must still be made in the elucidation of chemical re- actions for a complete application to aeronomy. It cannot be overemphasized that laboratory investigations under controlled conditions are of fundamental importance for a useful interpretation of space observations.SOLAR RADIATION AND ITS ABSORPTION A knowledge of the radiation available for dissociation in the atmosphere is required before conclusions can be reached regarding the relative importance of aeronomic processes. The principal gases in the thermosphere, molecular nitrogen,M. NICOLET 9 atomic oxygen and molecular oxygen, limit the penetration of solar radiation into the heterosphere at wavelengths iz < 796 A, A < 910 A and iz < 1025 A, respectively. Since absorption cross-sections are not less than 10-18 cm2 between 1000 A and 100 A, the solar radiation in this spectral range is absorbed above 100 km and ionizes N2, 0 and 0 2 .The total number of solar photons available at the top of the earth’s atmosphere is not greater than 2 x 1011 photons cm-2 sec-1, and corresponds to the number of ionizing processes in the E and P ionospheric layers. Molecules such as NOa, HzO, 0 3 , NzO, CH4 and OH can be neglected in the study of the ionization of the atmosphere since their ionization potential is greater than that of molecular oxygen. No radiation will be available to ionize these molecules below 100 km where they are subject to dissociation processes. At 1750 A (see fig. 1) where the Schumann-Runge continuum of molecular oxygen begins, the total number of photons available at the top of the earth’s atmosphere is 17 about 2 x 1012 photons cm-2 sec-1 and consequently this value also represents the total number of oxygen molecules which is dissociated in a vertical column of the atmosphere for iz < 1750 A.wavelength (A) FIG. 1. At 2420 A where the Herzberg continuum of 0 2 begins, the total number of solar photons is about 2.7 x 1014 photons cm-2 sec-1 and such a value corresponds to the maximum number of oxygen molecules which could be dissociated per cm2 sec-1 in a vertical column of the earth’s atmosphere. Thus the number of photodissociation processes of molecular oxygen D(O2) in the earth’s atmosphere is The maximum value which can be reached depends on ozone absorption which needs to be taken into account in estimating the dissociation rate in the Herzberg continuum.Ozone shows an important absorption which begins near 3500 A, extends below 3000 r f with rapidly increasing cross-section to a maximum at about 2550 A, and is still important in the Herzberg continuum, iz <2400 A. Due to the presence of O3 molecules, it will be shown that the rate of dissociation of 0 2 is strongly affected below the stratopause. Under these conditions, the dissociation of molecular oxygen must be studied before determining the behaviour of other constituents. 2 x lOlz< D(o2)<2.7 x 1014 cm-2 sec-1. (1)10 INTRODUCTION TO CHEMICAL AERONOMY The photodissociation process, 02+hv(;l5 1750A)+O(3P)+0(1D), (2) J,,(AI 1750 A) = 5 x sec” (3) leads to a photodissociation rate coefficient JoZ of at zero optical depth. The continuum of the Schumann-Runge system has an absorption cross-section varying from about 2 x 10-19 cm2 at the threshold with a peak not less than 10-17 cm2 between 1500 A and 1400 A.The penetration of solar radiation into the atmosphere is limited where the total content of 0 2 molecules is between 1019 cm-2 and 1017 cm-2. For example, a vertical column of 2 x 1018 0 2 molecules cm-2 (about 100 km) leads to photodissociation rate, Jo2(z- 100 km) = 2 x sec-l. (4) Thus the life-time of an oxygen molecule in the sun’s radiation field is very long at heights where photochemical equilibrium conditions should be applied if transport process are ignored. The penetration of solar radiation to greater depths than 1019 molecules cm-2 occurs only in various “ windows ” between 1225 A and 1 100 A.The most important “ window ” is situated at 1216 A of the important solar radiation, Lyman-a, for which the unit optical depth corresponds to a cross-section of the order of 10-20cm2. Since the number of photons available in Lyman-a is between 2 and 4 x 1011 photons cm-2 sec-1, the dissociation rate coefficient is ~ , , ( ~ y - a ) = 2 to 4 x 1 0 - ~ sec-l. ( 5 ) For an overhead sun, the dissociation rate coefficient at 75 km becomes Since the concentrations of molecular oxygen at 75 km 402) N 2 x 1014 cm-3, the minimum dissociation rate is still 2 x 105 cm-3 sec-1. A difficulty occurs, however, in the determination of oxygen dissociation in the mesosphere. This difficulty comes from the impossibility of obtaining a sufficiently accurate value for aeronomic purposes.The dissociation rate in the Herzberg continuum, particularly near 2000 A where the Schumann-Runge bands occur, is not sufficiently precise;18 we cannot enter into details here and an approximate distribution has been deduced (fig. 2). A plot of Jo, against height between 50 km and 100 km referring to an overhead sun indicates that .lo2 decreases by about a factor of 100 in this 50 km height interval. The effect of the Schumann-Runge continuum is apparent in the thermosphere and the decrease of Jo, in the mesosphere is related to the absorption in the Herzberg continuum. The photolysis of ozone is due to the absorption in the ultra-violet and the visible. It is customary to take absorption cross-sections as dissociation cross-sections and to adopt average numerical values for aeronomic purposes.It should be pointed out, however, that errors of about 10 % still seem to occur in recent data.19~ 20921 In applying average values, the total rate coefficient .loz at zero optical depth is about J , ~ = sec-l, (7) J,,(visible) = 3.5 x sec-l. (8) while the visible part of the spectrum leads to onlyM. NICOLET 11 It can be assumed that ozone and molecular oxygen are the two molecules which absorb solar radiation between 3000A and lOOOA, i.e., that the other constituents constitute a negligible role. Ide dissociation rate coefficient (sec-1) FIG. 2. I I I I I wavelength (A) FIG. 3. Photodissociation of water vapour, which begins around 2400 A, is related to the absorption of molecular oxygen. The absorption cross-sections is less than 10-20 cm2 at 1900 A (see fig.3) and increases rapidly in the spectral range of the12 INTRODUCTION TO CHEMICAL AERONOMY Schumann-Runge bands system. It reaches about 10-18 cm2 near 1800 A. Thus the photo-dissociation of H20 in the mesosphere is related to the complicated structure of the absorption spectrum of 0 2 . At the present time, only a rough estimate can be obtained. The total dissociation rate coefficient at zeroth optical depth JH,O is about which becomes about 10-6 sec-1 near the mesopause level after absorption in Schumann-Runge bands. It is still about 10-6 sec-1 in the middle mesosphere due to the effect of solar Lyman-a. With such low values for the dissociation rate coefficients it is difficult to consider photochemical equilibrium conditions for water vapour in the mesosphere.The life time of water vapour in the solar radiation field is sufficiently long to lead to departures from photoequilibrium conditions. Fig. 3 shows also the absorption cross-section of C02. Its photodissociation in the mesosphere is related to the effect of Lyman-a which leads to the following value of Jco,, Methane also has an absorption spectrum in the region of the Schumann-Runge continuum of 0 2 and its direct photodissociation in the mesosphere depends on the penetration of Lyman-a. At zero optical depth, it is Even if collisions involving atomic oxygen are effective as a loss process for methane in the mesosphere, photodissociation by Lyman-a at 75 km is sufficient to reduce its initial concentration to 50 % in about 3 days of 12 h.J~~~ = 10- sec- I , (9) Jco,(Lyman-a) = 3 x sec-l. (10) J,,,(Lyman-a) = 5 x sec-l. (11) PURE OXYGEN ATMOSPHERE The photochemistry of an oxygen atmosphere has been studied by a number of investigators (see, for example, ref. (4), (22), (23)) since Chapman 1 gave the complete set of equations. We follow here the analysis made by Bates and Nicolet.4 The dissociation of oxygen obtained by photodissociation O2 + hv(A < 2420 A)+O + 0; coefficient J02, (12) (1 3) (14) (1 5) (16) is followed by the three-body recombination, and by The bimolecular process occurs, therefore, and of course the photolysis of ozone is considered, The equations governing the rate of change of the concentrations n(Oz), n(0) and n(03) are dn(O2))ldt + n(02)Jz + k~n(M)n(Oz)n(O) = kln(M)n2(0) + 2k3n(O3)n(O) + n(03)J3, dn(O)/dt +2kln(M)n2(0) + k2n(M)n(O~)n(O) + k3n(03)n(O) = 2n(02)J2 +n(03)J3, O+O+M-+02+M+118 kcal; coefficient kl, 0 + 0 2 + M-+O3 + M + 24 kcal ; coefficient k2.0 + 0 3 -+ 0 2 + 0 2 + 94 kcal ; coefficient k3, O3 + hv-+O, +O; coefficient Jo3. (17) (1 8) (19) dn(03))ldt + n(03)J3 + k3n(O)n(O3) = k~n(M)n(O2)n(O>.M. NICOLET 13 The conditions for the simultaneous variation of n(0) and n(O3) can be con- veniently written At sufficiently high altitudes, i.e., in the thermosphere, (20) becomes, n(03) <n(O), For the day equilibrium of ozone in the mesosphere, In the stratosphere, the day-time conditions become, n(O) <n(03), indicating that the equilibrium conditions for ozone, is reached depending on t being the time measured from an initial time to.The time t increases with lower heights since the dissociation rate coefficient J 2 decreases more rapidly than 4 0 2 ) increases. Departure from photochemical equilibrium conditions takes several days below 40 km. Any variation in the ozone content of the stratosphere modifies the value of J2 and affects the vertical distribution of ozone. If numerical values of the various parameters are considered, it is evident that (1) the ratio J2/kl is important in the thermosphere ; (2) the ratio k~J2/k3J3 is important in the stratosphere. In the mesosphere all parameters are involved, since the day- time equilibrium conditions are and dn(O)/dt + dn(03)dt + 2kln(M)n2(0) +2k3n(O3)n(O) = 2n(02)J2. (20) dn(O)/dt +2kln(M)n2(00) = 2n(02)J2.(21) (22) dn(03)ldt N 2n(02)J2, (23) dn(O)/dt + 2kln(M)n2(03) + 2k3n(O)n(O) = 2n(02)J2. nW3) = (k2lk3)n(M)n2(02)(J21J3), nt(O3) = nt,(O3) + 2n(Oz)J24 (25) n(03)/n(O) = [k2n(M)n(O2) - k3n(03)1/J3, (26) in which the terms k3 n(X) can be neglected. If there is no theoretical obstacle in discussing the ozone atomic oxygen problem, numerical results differ by a large factor. The difficulties mentioned concerning the photodissociation rate coefficients are not important compared with the inaccuracies in the chemical rate coefficients. The rate coefficient of the three-body reaction (13) of oxygen atoms 14 can be taken as, for aeronomic purposes, kl = (3 k 1) x 10-33 cm6 sec-1, (284 corresponding to the values assumed by Bates-Nicolet 4 k3 = 5 x 10-34Ti cm6 sec-1.The values of k2 and k3 are not yet certain. Using the laboratory values of Eucken and Patat,24 Bates and Nicolet 4 adopted and However, recent investigations such as Benson and Axworthy,2s Zaslowsky 26 and Kaufman 27 lead to with an undetermined activation energy. k2 = 5 x lO-36Pcm6 sec-1, k3 = 1.5 x lO-llT* exp (- 3000/T) cm3 sec-1. kz = (5 k2.5) x 10-34 cm6 sec-1 (29) (30) (31)14 INTRODUCTION TO CHEMICAL AERONOMY As regards the value of k3, an exact evaluation is difficult.15 The uncertainty is illustrated by the values deduced from laboratory measurements, namely,15 and k3a = 5 x 10-11 exp (- 3000/T) cm3 sec-1, k3b = 7 x 10-12 exp (- 1600/T) cm3 sec-1. If H20 is an impurity in the measurement, k3b could be the result of the bimolecular reaction between ozone and atomic hydrogen and k3c;c should be the exact rate coeffici- ent.It is clear that only approximate numerical solutions can be derived as the uncertainties in the coefficients are too great. The results obtained with (28a), (31) and (32a) or (32b) are given in fig. 4 and exhibit the same general features found by concentration (CM-3) FIG. 4. earlier workers. But the absolute values are essentially different. An activation energy of 6 kcal for the bimolecular process O3+0 with a relatively small ratio kl/k2 5 6 leads to large concentrations of ozone and atomic oxygen in the mesosphere. Also the limit set for night-time conditions by or by (33) (34) leads to an ozone concentration in the mesosphere much larger at night than during the day for k3b.Consequently, there is a need for careful experimental work on ozone reactions. It must be stressed that extremely precise data are required for the analysis of aeronomic conditions in the mesosphere, in which it is possible to study photochemical and chemical processes without additional effects such as advective and dynamical transport and without a practical inff uence of solar activity. Without a perfect knowledge of the ozone+atomic oxygen behaviour in a pure oxygen mesosphere it becomes difficult to study departures from photochemical equilibrium conditions in the stratosphere and thermosphere. Finally, the introduction of other minor constituents necessary in the study of the terrestrial atmosphere cannot be made if the idealized atmosphere is not properly defined.M.NICOLET 15 ORIGIN OF A HYDROGEN OXYGEN ATMOSPHERE A hydrogen oxygen atmosphere is very complicated. Photoaction on water vapour in the mesosphere and oxidation of methane in the stratosphere are im- portant processes leading to the production of hydrogen atoms.7~ 8 Bates and Nicolet 4 made the first attempt to estimate the various aeronomic processes 13 years ago and are now again considering (with the new experimental data16) the very complicated situation resulting from chemical actions and atmospheric mixing effects. It has been shown 28 that there is a continuous escape of atomic hydrogen atoms at exospheric levels corresponding to a diffusion flow FD(H) at the 100 km level of F~(H)loo = 2.5 x 107 cm-2 sec-1. (35) This must correspond to a total loss of about lO7H20 molecules cm-2sec-1 or 6 x 106 CH4 molecules cm-2 sec-1.Under mixing conditions, the diffusion flow 29 of methane with concentration n(CH4) = 1.5 x 10-6 n(M) is FD(CH~) = 7 x 106 cm-2 sec-1, (36) which must be compared with a production rate 9 of CH4 of at least 1010 molecules cm-2 sec-1. Thus, the escape flow is always supported by the diffusion flow of CH4 and is a small fraction of its total production. Atomic oxygen attacks methane through CH4 + 0 + CH2 + H20 + 30 kcal (37) with an activation energy of 7-8 kcal.30~ 31 Adopting a(CH4, 0) = 2 x 10-11 exp (-4OOO/T) = 4 x 10-12 exp (- 3600/T) cm3 sec-1 (38) for the rate coefficient of (37), it apears that the life-time of CHq in the mesosphere is relatively short.8 As the re-formation of CH4 is a very slow process, its concentration in the mesosphere must be very small.There is no diffusion or atmospheric mixing process able to maintain an adequate vertical flow of methane. It is almost certain that its fractional abundance begins to fall off well below the stratopause. With a rate coefficient of the order of 10-19 cm3 sec-1 for (37), adapted to tropopause condi- tions corresponding to about 5 x 1012 CH4 molecules cm-3, it can be seen that a 10 km layer with about 107 oxygen atoms cm-3 will lead to a production of about 107 H20 molecules cm-2 sec-1. Judging from this value it seems probable that, in the stratosphere, the fractional concentration of CH4 is affected by its transformation into H20, and consequently the formation of H20 depends on the methane exchange between the troposphere and the stratosphere.The tropospheric mixing time is short enough to lead to a uniform vertical distribution of CH4,32 and its injection rate into the stratosphere should be known with precision in order to determine the actual production of stratospheric H20. With uniform transport due to diffusion, its fractional abundance would be about 3 x 10-6. However, since the life-times T(CH~) at 30 km and 40 km are of the order and this implies that the exchange between troposphere and stratosphere is controlled by " turbulent " processes rather by diffusion. Consequently, an abundance than of H2O greater than 3 x 10-6 can result from methane oxidation in the stratosphere.16 INTRODUCTION TO CHEMICAL AERONOMY REACTIONS OF ATOMIC HYDROGEN The products of dissociation of H20 in the mesosphere give rise to a complicated series of chemical processes.More than 30 processes are involved 4 and we retain here the more important processes that a detailed study of the situation suggests. The principal reactions are listed below. Those involving hydrogen atoms are : H+ 0 2 + M-+HO2 + M + 46 kcal, a1 = 3.3 x 10-33 exp (SOOlT) cm6 sec-1 (41) (42) with a rate coefficient according to Clyne and Trush 33 of showing a negative temperature coefficient, and for which the rate coefficient a2 16 is very large, H + 03+OH + 0 2 + 77 kcal, a2 = 1-5 x 10-12T* cm3 sec-1. (43) This reaction was introduced by Bates and Nicolet 4 in 1950 to explain the observed airglow emission of the hydroxyl radicle OH up to the vibrational quantum number 9 (75.2 kcal) but not up to v" = 10 (81 kcal).The same reaction, leading to HO2, is less important and instead of the three-body association between OH and 0, the bimolecular process is noted as also having a high rate coefficient with practically no activation energy :I6 Since 02(X3Xg),=4 = 17.4 kcal only the first 3 vibrational levels of 0 2 are involved. The reaction of HO2 with 0 is also a fast process,l6 HOz + O-OH + 0 2 + 55 kcal, OH+O-+H+02+16.6 kcal (45) a5 = 3 x lO-12T* cm3 sec-1. (46) (47) with a rate coefficient a7 which may reach a7 = 1 -5 x 1 O - l W cm3 sec-1. Assuming as a first approximation that only the preceding reactions are involved, the following ratios corresponding to chemical equilibrium conditions are obtained : and Since atomic oxygen is present in the mesosphere with concentrations not less than lOlocm-3 during the day, (49) and (50) are representative of day-time conditions. At the stratopause level aln(M)n(O2)-5 sec-1 is the most important term in eqn.(49) and (50), and, therefore, n(HOz)>n(H) and n(CH)>n(H). Atomic hydro- gen becomes important only in the mesosphere (and above). The rapid decrease of the term aln(M)n(O2) with height compared with that of bimolecular processes leads to the approximation, somewhere above the stratopause, This indicates, since n(O3) < n(0) in the middle mesosphere and thermosphere, that the hydroxyl radicle OH must have an upward sharp decline. Under the same conditions, i.e., a2n(03) > aln(M)n(O2), we have n(0H) > n(HO2), which indicates that hydrogen is in atomic form to such a degree as to play an important role in meso- spheric processes.n(OH)/n(H) = (aln(M)n(O2) + a2n(03))@n(O), (49) n(H02)/n(H) = aln(M)n(02)/a7n(O). (50) n(OH)/n(H) = n(03)/240)* (51)M. NICOLET 17 If atomic hydrogen is sufficiently abundant in the mesosphere, eqn. (20) must be dn(0)Jdt + dn(O&dt + 2kln(M)n2(0) +2k3n(O3)n(O) + 2a2n(03)n(H) = 2n(02)J2. Again, in order to derive any numerical value, it is necessary to know first the exact numerical expressions in a pure oxygen atmosphere. In any case, comparing the numerical values of k3 and a2, it is clear that for atomic hydrogen concentrations in the mesosphere greater than 107 cm-3 Since n(H) is of the order of 107 cm-3 at 100 km 28, 34 and increases downwards in the lower thermosphere, it is certain that the ozone + oxygen equilibrium in the mesosphere is affected by atomic hydrogen which acts as a catalyst.As a example, let us assume a concentration of n(H) = 3 x 108 cm-3 at 80 km, namely, a normal value if atomic hydrogen is in mixing between 80 km and 100 km. The ozone concentration which is about 109 cm-3 in a pure oxygen atmosphere (see fig. 4) decreases to 108 cm-3 and the atomic oxygen concentration decreases from 7 x 1011 cm-3 to 1011 cm-3. It is clear, therefore, that the mesospheric behaviour of a hydrogen + oxygen atmosphere is completely different from a pure oxygen atmo- sphere. modified by adding another term, (52) k3n(O) < azn(H). (53) REACTIONS I N A HYDROGEN + OXYGEN ATMOSPHERE Numerous secondary processes, involving the destruction of atomic hydrogen, hydroxyl and perhydroxyl radicles occur in the mesosphere.Among the data re- quired for a complete discussion are the rate coefficients of the various reactions of all hydrogen + oxygen compounds.16 Recent publications on the subject were also used as main sources 35-38 since a number of investigators have studied these re- actions under various aspects. The source of hydrogen atoms (free or combined) is the photodissociation of H20 (and H202). The final loss processes which are directly related to hydrogen atoms are not important compared with collision processes where OH and HO;? are involved. This is due to the fact that at the stratopause and in the lower meso- sphere n(H)<n(OH)<n(H02) and also that the reactions are not very rapid.The main final reactions are as follows : between two hydroxyl radicles, OH + OH+H20 + 0 + 17 kcal, which is discussed by Kaufman 16 and has a reaction rate (54) between hydroxyl and perhydroxyl radicles, for which a high rate coefficient is suggested 16 and is taken as OH + H02-+H20 + 0 2 + 72 kcal, between two perhydroxyl radicles HO2 + HOpH202 + 0 2 + 42 kcal, (58) with a rate coefficient of the same order as a16, i.e.,18 INTRODUCTION TO CHEMICAL AERONOMY Among the loss processes of atomic hydrogen which could be added to the reactions just described, we may consider the three-body reactions : with rate coefficient a21 of The exothermic bimolecular process with rate coefficient, H+H+M+Hz+M+ 103.2 kcal, (60) a21 = 3 x 10-32 cm6 sec-1.(61) H+OH+Hz+O+ 1.9 kcal (62) (63) a22 = 2 x lO-W’”* exp (- 3400/T) cm3 sec-1, can be compared with the endothermic process, with rate coefficient Hz+O-)OH+H- 1.9 kcal a24 = 5 x lO-13P exp (-aOO/T) cm3 sec-1. (65) Such a process is important in the thermosphere since it leads to the final dissociation of molecular hydrogen. Finally, the reaction with perhydroxyl radicles H+H02+H2+02+57 kcal (66) (67) with a rate coefficient a23 = 5 x 10-1W exp (- lOOO/T) cm3 sec-1 should be a normal production process of molecular hydrogen. The expression for complete equilibrium is written as follows : n(H20)JH,0+ n(H202)JH,0,+a2,n(0)n(H2) = a16n2(0H)+a17n(0H)n(H02)+ a2,n2(H02) + azln(M)n2(H) + a,,n(H)n(OH) + a2,zt(H)n(H02). (68) which can be applied in the mesosphere for daytime conditions.The vertical distribution of the production function depends on the values of n(H2O) and J - , o for which exact values are not available. Because of absorption due to molecular oxygen, JH,O is a very sensitive function of the optical depth of molecular oxygen. Various estimates can be made, but it is not possible to discuss here all aspects which must be considered for a complete discussion. The omission of reactions involving atomic hydrogen at sufficiently low altitudes (above the stratopause) where n(H) must decrease rapidly according to (49) leads to a simple way of considering the result of night-time conditions. The differential equation is simply, see (68), - [dn(OH)/dt + d(HO~)/dt] = al&(OH) + ~2~n2(HO2) + al7n(0H)n(H02).(69) Since (55), (57) and (58) show that a16 = a27I+a17, the relevant solution to (69) is, t being the time measured from to which is sunset, The nocturnal decay is important in the lower mesosphere since for a16 = 2-5 x 10-12 cm3 sec-1 and t = 4 x 104 sec the initial concentration is reduced to less 107 cm-3. Thus, even if the aeronomic problem considered in this section is idealized, it does indicate that the whole mesosphere is a transition region in which the freeM. NICOLET 19 hydrogen atoms are formed and diffuse upwards. Water vapour diffuses from the stratosphere into the dissociation region with an equivalent current which can be furnished by methane in the troposphere. If gentle mixing winds occur and carry up H2O at a greater rate, then there will be regions of abnormal specific humidity which will persist for a certain period of time.But the final process must be a down- ward transport of water vapour. The reactions (54), (56) and (58) obviously leads to H20 since hydrogen peroxide is destroyed (see, e.g., ref. (35, 39, 40)) by the following chemical processes : with a rate coefficient of H + H20pOH + H20 + 69 kcal, (71) a29 = 5 x 10-12Ti exp (- 3000/T) cm3 sec-1; (72) OH + H202-+H02 + H20 + 30 kcal, (73) a30 = 1.5 x 10-1W cm3 sec-1, (74) O+H202-+02+H20+86 kcal (75) (76) with a rate coefficient which may be and with a rate coefficient a31 = 1-5 x 10-13Ta exp (-2OOO/T) cm3 sec-1, The exact role of hydrogen peroxide near the stratopause level must be re- determined from the solution of the differential equation dn(H202)/dt+ n(H202)[J~,o~+ a29n(H)+ a3on(OH)+ a3dO)l = 2a27n2(H02)* (77) An approximate value for JH~o, is 10-4 sec-1, and almost equilibrium day-time conditions can be imposed.Molecular hydrogen should exist in the mesosphere since its dissociation pro- bability is small in this region. In addition to processes (62) and (67) leading to the formation of H2, and to (65) leading to the loss of H2, we may consider also with a rate coefficient a19 of The general equation being OH+H2-)HzO+H+15 kcal (784 a19 = 0.5 x lO-llT+ exp (- 3000/T), &(Hz)/dt + n(H2)[a24n(O) + aign(OH)] = n(H)[a22n(OH) + a23n(H02)1 +n(H20)J~,-o. (79) This leads only to equilibrium conditions when a2#(0) is sufficiently large, i.e., in the thermosphere where the temperature is high.1 Chapman, Phil. Mag., 1930,10, 369. 2 Chapman, Report Progr. Physics, 1943, 9,92. 3 Nicolet and Mange, J. Geophys. Res., 1954, 59, 15. 4 Bates and Nicolet, J. Geophys. Res., 1950, 55, 301. 5 Meinel, Astropliys. J., 1950, 111, 207 and 433. 6 Chamberlain and Smith, J. Geophys. Res., 1959, 64, 611. 7 Bates and Nicolet, Compt. Rend., 1950, 230, 1943 ; Pub. Astronom. SOC. Pacific, 1950, 62, 106. 8 Bates and Witherspoon, Month. Notices Roy. Astro. Soc., 1952, 112, 101. 9 Hutchinson, chap. 8, in Kuiper (ed.), The Solar System, vol. II, The Earth as a Planet (The University of Chicago Press, Chicago, 1954.)20 INTRODUCTION TO CHEMICAL AERONOMY 10 Nicolet, Inst. Mktior. Belg., Mkmoires, 1945, 19, 124. 11 Bates, Ann. Giophys., 1952, 8, 194. 12 Nicolet, J. Atm. Terr. Physics, 1955, 7, 152. 13 Nicolet, Aeronomic Chemical Reactions in Physics and Medicine of the Atmosphere and Space (John Wiley and Sons, New York, 1960). 14 Barth, Ann. Gkophys., 1964,20, 182. 15Schiff, Ann. Giophys., 1964,20, 115. 16 Kaufman, Ann. Giophys., 1964,20,106. 17 Detwiler, Garret, Purcell and Tousey, Ann. Gkophys., 1961, 17,9. 18 Ditchburn and Young, J. Atm. Terr. Physics, 1962,24, 127. 19 Vigroux, Ann. Physique, 1953, 8, 709. 20 Inn and Tanaka, Y., J. Opt. SOC. Amer., 1953,43, 870. 21 Hearn, Proc. Physic. SOC., 1961, 78, 932. 22 Diitsch, in Chemical Reactions in the Lower and Upper Atmosphere, (Interscience Publishers, 23 Paetzold in Chemical Reactions in the Lower and Upper Atmosphere (Interscience Publishers, 24 Eucken and Patat, 2. physik. Chem., 33B, 1936, 459. 25Benson and Axworthy, J. Chem. Physics, 1957, 26, 1718. 26 Zaslowsky, Urbach, Leighton, Wnuk and Wojtowicz, J. Amer. Chem. SOC., 1960, 82, 2682. 27 Kaufman, this Discussion. 28 Kockarts and Nicolet, Ann. Gkophys., 1962, 18, 269. 29 Nicolet, Me'm. SOC. Roy. Sci. LiGge, 1962, 7, 1960. 30 Steacie in Atomic and Free Radical Reactions, (Reinhold Publishing Corporation, New York, 31 Wong and Potter, J. Chem. Physics, 1963, 39, 221 1. 32 Goldberg and Miiller, Astrophys. J., 1953, 118, 397. 33 Clyne and Thrush, Proc. Roy. SOC., 275A, 1963, 559. 34 Bates and Patterson, Planet. Sp. Sci., 1961, 5, 257. 35 Foner and Hudson in Adu. Chem. Series no. 36 (American Chemical Society, Washington D.C., 36 Knox, Ann. Reports Progr. Chem., 1962, 59, 18. 37 Minkoff and Tipper, Chemistry of Combustion Reactions (Butterworths, London, 1963). 38 9th Symp. Combustion (Academic Press, New York, 1963). 39 Baldwin and Mayor, Trans. Faraday SOC., 1960,56,103. 40 Foner and Hudson, J. Chem. Physics, 1962,36,2681. New York, 1961), p. 167. New York, 1961), p. 181. 1954), p. 601. 1962), p. 36.