Additional Flowing Afterglow Measurements of Negative Ion Reactions of D-region Interest

Some additional negative ion reactions of possible D-region importance have been measured in the ESSA flowing afterglow system at 300°K. It has been found that NO,does not react with atomic oxygen or atomic nitrogen with a rate constant as large as lo-” cm*/sec. The electron affinity of NO, is found to be greater than that of NOZ by more than 0.9 eV. It is pointed out that conversion of 02to Opfollowed by subsequent reactions of the O,may play an important role in D-region negative ion chemistry. Rate constants for the reaction of 04with NO, CO* and 0 are reported. The reaction of O,with COz produces CO,-. Rate constants for the reactions of CO,with NO and 0 are reported. INTRODUCTION A number of rate constants for reactions between negative ions and neutrals of atmospheric importance have previously been reported from this laboratory (Fehsenfeld et al., 1967; Fehsenfeld and Ferguson, 1968). The available laboratory data on negative ion processes have recently been reviewed (Phelps, 1969; Ferguson, 1969). Some additional measurements have now been made and are reported here. The flowing afterglow technique for ion-molecule reaction rate constant measurements used in this laboratory has been described in the earlier papers and has recently been described in considerable detail (Ferguson et al., 1969). NEGATIVE ION REACTIONS Negative ion chemistry is initiated in the D-region by the three-body attachment of electrons to 0, to form O,-. In our first paper (Fehsenfeld et al., 1967), we described a sequence of reactions which led predominantly to NO,. In a subsequent paper (Fehsenfeld and Ferguson, 1968) we reported that the reaction NO, + 0, --+ NO, + 0, (1) has a rate constant 1.8 x lo-l1 cm3/sec at 300°K and suggested that this could lead to NO, being a dominant negative ion in the D-region. The question as to whether the NO,produced in (1) can undergo further reaction is therefore an important one. The first possibility investigated was reaction of NO,with atomic oxygen : NO, + 0 --+ NO, + 0, (W -+ NO, + O,WI -tNO,+O,+e @cl --f NO + O,-. (24 No reaction of NO,with 0 was observed and it was possible to set an upper limit of lo-l1 cm3/sec to the sum of all possible channels, i.e. k, < lo-l1 cm3/sec. The failure to observe an ion-neutral reaction with a rate constant larger than lo-l1 cm3/sec often (although not always) implies that the reaction is endothermic. This does not seem likely to be the case for the most energetically favored channel (2a) however. Reaction (2a) is exothermic 1759 1760 F. C. FEHSENFELD, E. E. FERGUSON and D. K. BOHME unless the electron affinity of NO, exceeds that of NOB by more than 2.9 eV. The electron affinities of neither NO, nor NO, are known. However, the electron affinity of NO, exceeds that of OH (l-78 eV) since the reaction OH+ NO, + NO,+ OH (3) is known to proceed with a large (lows cm3/sec) rate constant (Ferguson et al., 1969). This would require that EA(N0,) > 4.7 eV in order for (2a) to be endothermic and this is a larger electron affinity than any molecule fo far is known to have. Calculated values of the electron affinities are EA(N0,) = 1.6 eV and EA(N0,) = 3.9 eV (Pritchard, 1953), however these values (from lattice energy calculations) cannot be relied upon. The other branches of reaction (2) are probably endothermic. Reaction (2b) requiring only that EA(N0,) > 3.3 eV, (2~) that EA(N4) > 2.9 eV and (2d) that EA(N0,) > 2.7 eV to be endothermic. The (2d) estimate assumes EA(0,) = 1.9 eV, which is also rather uncertain (Wood and D’Orazio, 1965). For most important purposes reaction (2a) would be of little significance, since conversion of NO,to NO,would be followed by the reconversion reaction (1). Reactions (2b), (2~) and (2d) however would all enhance electron detachment in the D-region and thus play an important role in determining the electron density if they did occur. Since nitric oxide appears to be fairly abundant in the D-region (Pearce, 1969), we have examined the reaction NO,+ NO + NO,+ NO,. (4) No reaction was observed, and an upper limit to k4 was determined to be lo-l2 cm3/sec. Indeed reaction (4) appears to be endothermic as the reverse reaction NO,+ NO, -+ NO,+ NO (5) was observed to occur, somewhat indirectly in the flowing afterglow system. The ratio of NO,-/NO,was found to decrease with NO, concentration in the afterglow. A rate constant, k, r 4 x IO-l2 cm3/sec was deduced from this effect. With increased NO2 concentrations the dimer NO,. NO, or NO,. NO was also observed but its concentration was small compared to that of NO,-. The occurrence of reaction (5) gives some useful data on the NOa electron affinity. For (5) to be exothermic, or AE5 = D(NO,-O)D(NO-O)+Ez4(NO,)-EEA(NO,)> 0 (6) EA(NO,)> EA(NO,)+ O-9 eV. Since we know EA(NO,) > 1.8 eV, it follows that EA(N0,) > 2.7 eV. This in turn implies that only wavelengths shorter than 4600 d could be effective in photodetachment of NO,in the D-region. Since it unlikely that EA(N0,) is just barely greater than that of OH, and that reaction (5) is just barely exothermic, it is probable that EA(N0,) is significantly greater than 2.7 eV and that ultraviolet light will be required for photodetachment. It was also determined that NO,did not react rapidly with N atoms, i.e. the reaction NO,+ N -+ products (7) had a rate constant k, < lo-l1 cm3/sec at 300°K. This reaction, to produce NO,+ NO, is more exothermic by 1.4 eV than reaction (2a). MEASUREMENTS OF NEGATIVE ION REACTIONS OF D-REGION INTEREST 1761 The implications of the present NO,studies for D-region aeronomy bolster our earlier conclusion that NO,should be an important D-region negative ion, and probably a ‘terminal’ one in a chemical sense. That is, NO,may only be lost significantly by ion-ion recombination and/or photodetachment, rather than undergoing further chemical reactions. This is not a firm conclusion however, as we have set only a modest upper limit, k, < lo-l1 cm3/sec and if reactions (2b), (2c), or (2d) should occur with rate constants only as large as lo-l4 cm3/sec, they would have time constants of only ~10~ set if the calculated 0 atom concentrations N lOlo cmv3 in the D-region are correct (Hesstvedt, 1968). It is very difficult to measure small rate constants involving negative ions and atomic oxygen so that to establish that k, < lo-l4 cm3 set experimentally would be very difficult. In this regard it would be most valuable to know EA(NO,). If, as we suspect, EA(N03) > 3.3 eV, then the reactions (2b), (2~) and (2d), which could lead to electron regeneration in the D-region can be dismissed as being endothermic. CLUSTER ION REACTIONS A new development in D-region negative ion chemistry, which may substantially alter our previous considerations of this subject, is the recent recognition of the importance of three body clustering reactions. For example, we have recently shown qualitatively (Fehsenfeld and Ferguson, 1969) that the positive ion chemistry of the D-region is significantly affected by the clustering of O,+ with 0, to form 0, +. The O,+ ion then reacts in a series of binary reactions with H,O to produce the major positive ion below 80 km, H,O,+. In a similar manner, the clustering of O,with 0, to form 0, may be quite significant in the D-region. The cluster ion O,has a bond energy 0.59 eV (Conway and Nesbitt, 1968), significantly more than that of 0, + which is 0.42 eV. The rate constant for