J . Phys. Chem. 1986, 90, 466-471
466
Surface-Catalyzed Formation of Electronically Excited Nitrogen Dloxlde and Oxygen Ang-Ling Chu, Robert R. Reeves, Rensselaer Polytechnic Institute, Department of Chemistry, Troy, New York 12180
and Judith A. Halstead* Williams College, Department of Chemistry, Williamstown, Massachusetts 01 267 (Received: July 22, 1985)
-
The surface-catalyzed recombination of NO and 0 atoms to produce excited NO, molecules has been studied. By using the reaction of N atoms with NO (N + NO N2 + 0),the 0 atoms were produced in the absence of molecular oxygen. The resultant orange glow attributed to the electronically excited NO, molecule was observed directly over a nickel metal catalyst in a flow system operated in the 20 mtorr range. Measurements of the attenuation of the intensity as a function of distance from the catalyst yielded an observed effective lifetime of approximately 79 ps, which is in the range of radiative lifetimes of NO2*. Using molecular oxygen partially dissociated to 0 atoms in a discharge and reacting with NO produced an orange glow directly over the nickel catalyst. The observed effective lifetime was 960 ~.lsand this much longer lifetime is attributed to a precursor of an excited oxygen molecule. It is concluded that more than one reaction mechanism is needed to account for emissions of the orange glow of NO2*. The Herzberg emission from the 02(A3Z,+) state was observed as a function of distance from the catalyst surface. With increasing 0-atom concentration the observed effective lifetime appears to increase. However, since the Oz(A3Z,+) state is known to be quenched strongly by 0 atoms the mechanism for emission must include another excited state as a precursor. When the reaction of NO with N atoms was used to produce 0 atoms at the equivalence point, i.e. with no excess NO and N atoms, no Herzberg emission was observed.
Introduction The formation of electronically excited gas-phase product molecules as a result of heterogeneous catalysis at a gas/solid interface was first reported by Reeves et al. in the early 6 0 ' ~ . ' - ~ Although it has been more than 20 years since these initial observations were made, the processes involved are still not well understood. This is in part a result of the relatively few experimental investigations in this area in the 60's and 70's. In recent years, however, there has been a renewed interest in surfacecatalyzed excitation (SCE) of this type."8 There is strong evidence that the heterogeneous production of an electronically excited gas-phase moelcule may play a role in the formation of the much-publicized orange shuttle glow.9-'2 Since SCE processes occur in a localized region around the catalyst, surface-catalyzed processes can provide a relatively high concentration of excited states. It has also been suggested that the heterogeneous production of electronically excited gas-phase molecules may be used in the development of chemical lasers.' In a recent paper,s kinetic investigations of the heterogeneous formation of 02*and NO2* were reported. Kinetics of the surface-catalyzed recombination of 0 atoms on a nickel catalyst were investigated by observation of the variation in the Herzberg emission with 0-atom concentration. The intensity of the Herzberg emission was found to vary from first order to second order with respect to the 0-atom concentration as the ratio of [ 0 ] / [ 0was 2 ] increased. The results were interpreted in terms of a second-order heterogeneous recombination of 0 atoms to form the 02(A3Z,+)state. The 02(A3Z,+) state then diffuses in the gas phase where it may undergo quenching by 0 atoms, quenching by O2 molecules or emission to the ground state. The dependence of the light emission on the temperature of the catalyst was also (1) 946. (2) 1648. (3) 636. (4)
Reeves, R. R.; Mannella, G. G.; Harteck, P. J . Chem. Phys. 1960,32, Harteck, P.; Reeves, R. R.; Mannella, G. G. Can. J . Chem. 1960, 38, Mannella, G. G.; Reeves, R. R.; Harteck, P. J . Chem. Phys. 1960,33,
Kenner, R. D.; Ogryzlo, E. A. In?. J . Chem. Kinet. 1980, 12, 501. ( 5 ) Brennen, W.; McIntyre, P. Chem. Phys. Lett. 1982, 90,457. (6) Kenner, R. D.; Ogryzlo, E. A. J . Chem. Phys. 1984, 80, 1. (7) Caubet, Ph.; Dearden, S. J.; Rorthe, G. Chem. Phys. Lett. 1984, 108,
217. (8) Halstead, J. A.; Triggs, N.; Chu, Ang-Ling; Reeves, R. R. In "Gas Phase Chemiluminescence and Chemi-Ionization", Fontijn, A,, Ed.; North Holland: Amsterdam, 1985. (9) Banks, P. M.; Williamson, P. R.; Raitt, W. J. Geophys. Res. Lett. 1983, 10, 118. (10) Slanger, T. A. Geophys. Res. Lett. 1983, 10, 130. (11) Torr, M. R. Geophys. Res. Lett. 1983, 10, 114. (12) Swenson, G. R.; Mende, S. B. Geophys. Res. Lett. 1985, 12, 97.
0022-3654/86/2090-0466$01.50/0
measured at various 0-atom concentrations. Very little temperature dependence was observed in the 30-80 OC region. This suggests that the absorbed 0 atoms which subsequently participate in the formation of an electronically excited state must be very weakly bound to the nickel surface. Addition of nitric oxide to an 0-atom stream with a nickel catalyst reduces the Herzberg emission from 0, and results in the appearance of a bright orange emission attributed to NOz* formation. Like the well-known greenish white homogeneous airglow, the orange emission appears to be a continuum. The shift to the red of the heterogeneous emission relative to the homogeneous emission is interpreted in terms of energy loss to the solid catalyst during product formation. The dependence of the NO2* emission on the concentration of NO and 0 atoms was investigated by simultaneous observation of the orange glow above the catalyst and of the airglow upstream and downstream of the catalyst. The heterogeneous orange emission intensity was found to be linear with the homogeneous airglow intensity over a wide range of intensities as either 0-atom concentration or nitric oxide concentration was varied. Since the airglow emission is well-known to be proportional to the product of the NO and 0-atom conc e n t r a t i o n ~it~was ~ concluded that the orange emission was the result of heterogeneous production of gaseous NOz* which is first order in both 0-atom and NO concentration and second order overall giving a rate expression of Z(orange) = k [ N O ][O] Kenner and Ogryzlo6 earlier reported the observation of an orange emission attributed to the formation of electronically excited NOz following heterogeneous generation of 02(A3Z,+). The catalyst was a nickel mesh screen perpendicular to the direction of flow. Observations were made downstream of the catalyst. These workers also found the intensity of the orange emission to be first order with respect to the NO. The orange emission, however, was also found to be first order with respect to both 02(A3Z,+) and 02(X3Z,-). These results are interpreted in terms of initial heterogeneous generation of 02(A3Z,+) followed by homogeneous reaction of 02(A3Z,+) with ground-state 0, to form vibrationally excited ozone. The excited ozone subsequently reacts with nitric oxide to form NOz*. The observation of an orange emission rather than the well-characterized red emission16from NO + O3 NO, + 0, + hv (red)
-
(13) Wright, A. N.; Winkler, C. A. "Active Nitrogen", Academic Press: New York, 1965. (14) Kaufman, F. J . Chem. Phys. 1958, 28, 352. (15) Reeves, R. R.; Mannella, G. G.; Harteck, P. J. Chem. Phys. 1960, 32, 632. (16) Greaves, J. C.; Garvin, D. J. Chem. Phys. 1959, 30, 348.
0 1986 American Chemical Society
Electronically Excited NOz and O2 A C glow
discharge
The Journal of Physical Chemistry, Vol. 90, No. 3, 1986 467
condensed DCdischarge
Orange Glow Intensity with Nitrogen Atoms and Nitric Oxide - Catalyst Downstream of Mixing Region
z 6k
Gas Mixture: 2 5 % Ne in Ar Pressure: 2 2 millitorr Catoivst: Ni at 25’C.
i
/
w w 4-
>
I
NOlNOp
pressure gauge
Catalyst Ni at 3OoC
’ p:Omps
Input
Figure 1. Schematic of apparatus for surface-catalyzed excitation. is explained by the additional energy in the product NO2*. This additional energy in the product results from the formation by the reaction of NO with vibrationally excited ozone rather than ground-state ozone. Kenner and Ogryzlo6 give the rate law corresponding to this mechanism as
It was the purpose of the current work to determine if the orange emission observed upon the addition of N O to an 0-atom stream in the presence of a nickel catalyst is the result of a heterogeneous or a homogeneous process and whether or not this emission can be observed in the absence of ground state 02.In addition, the mechanism for the heterogeneous formation of 02(A3Z,+) has also been further investigated and is reported to be more complex than assumed earlier. Experimental Section A diagram of the experimental arrangement for studying the surface-catalyzed excitation is shown in Figure 1. For the investigation involving nitrogen atoms, a stream of N 2 was passed through a flowmeter and was then split into two streams. One stream passed through a second flowmeter and then directly into the reaction tube while the other passed through a third flowmeter and through a condensed dc discharge tube operating at 4-8 kV. The concentration of nitrogen atoms was varied by varying the Nzflow through the discharge tube while maintaining the total flow of nitrogen through the reaction tube constant and by varying the power to the discharge tube. For the generation of oxygen atoms, either O2 or an Oz/Ar mixture was passed through an ac glow discharge tube. The atoms, along with the carrier gas, flowed into a 3-in.-diameter 55-in.-long Pyrex reaction tube. The catalyst was a thin sheet of nickel resting on a water cooled U-tube located approximately 15 in. downstream of the gas inlet. An ironconstantan thermocouple was used to monitor the temperature of the catalyst. Pressures typically varied from 20 to 45 mtorr with flow rates of approximately 1 cm3/s at NTP. Oxygen and nitrogen atom concentrations were determined by standard gasphase titration technique^.^^-^^ Catalysts were initially polished and washed in dilute HC1. These samples were then periodically cleaned with hydrogen atoms and then allowed to accommodate to the N-atom or 0-atom stream to be used. All nickel samples placed in these streams are presumed to react, probably forming nitrides and oxides, respectively. It is then these accommodated surfaces that actually catalyze the reactions under investigation. It should be noted that, to date, effective catalytic action has not been achieved when a nickel oxide powder rather than nickel sheet is initially introduced into the system. It is not clear whether this is due to a physical or chemical difference between commercial nickel oxide powder and a surface which is the result of the reaction of 0 atoms on a nickel sheet. For measurement of the heterogeneous orange glow as a function of reactant concentration, a Corning 2-63 filter was placed in front of an RCA 1P28 photomultiplier tube. For measurements of intensity as a function of distance from the surface, Corning
0 30
3 5
40 45 5 0 % NITRIC OXIDE ADDED
5 5
60
Figure 2. Relative NO2* orange emission intensity as a function of the concentration of nitric oxide. 2-63 and 7-60 filters for the orange glow and the Herzberg emission, respectively, were placed in front of an EM1 6256B photomultiplier tube. For the heterogeneous emission in the oxygen system, spectra were taken to verify that the dominant emission was from the 02(A3Z,+) state. All bands observed in the region transmitted by the Corning 7-60 used in the kinetic studies were identified as part of the Herzberg (I) system. The photomultiplier was placed on a cathetometer mount and the height varied at 1-mm increments. The emission being observed was collimated into a beam 0.6 mm in height by a series of slits. For each measurement the background homogeneous glow was measured upstream and downstream of the catalyst. This verified that the reactant concentrations did not vary substantially across the catalyst surface and that the homogeneous emission was small by comparison to the heterogeneous emission except at the largest distances from the surface that measurements were taken. The average of the upstream and downstream background measurements was subtracted from the heterogeneous emission intensity. Results The heterogeneous orange NOz glow has been observed in the absence of molecular oxygen. Oxygen atoms were formed by the addition of excess nitric oxide to a nitrogen atom stream. 0 atoms are produced by the rapid reaction:” NO + N Nz + 0 +
The NOz* responsible for the orange emission above the surface is proposed to then be produced by the direct recombination of N O and an 0 atom on the surface of the catalyst. Either the 0 atom, the nitric oxide, or both may be absorbed on the nickel sheet. The relative intensity of this emission is given in Figure 2 as a function of NO added for two different nitrogen atom concentrations. This data is also consistent with the previous conclusion* that the orange emission is first order with respect to 0 atoms as well as [NO]. This can be verified by observing that, at a particular emission intensity, the product of [O] times [NO] is the same for the two initial N-atom concentrations. The concentration of 0 atoms is given by the initial concentration of N atoms, [N],. The concentration of nitric oxide is given by [NO] = [NO], - [N],. [NO], - [N],) is approximately constant for The product [N],( any emission intensity. (For example, from the data presented in Figure 2 at an intensity of 6.0, and 2.94% N atoms, the net nitric oxide [NO], - [N], would be 4.50 - 2.94 = 1.56%, or a product of [O] [NO] = 2.94 X 1.56 = 4.59; at the same intensity and 4.15% N atoms, the net [NO] is 1.10% and the product is 4.56, approximately the same.) In the reaction of NO with N atoms, the equivalence point for NO and N atoms can readily be observed by eye during the experiment and from the graph where the orange glow intensity originates from the abscissa in Figure 2. At this point, the predominant reactive species at the surface is oxygen atoms and therefore emission from the Herzberg system is expected. To date,
468
Chu et al.
The Journal of Physical Chemistry, Vol. 90, No. 3, 1986 O2 and 3.0% 0-otoms, balance Argon
78% Ar, bolonce 0,
2 4 % N and 4.2% 0-atoms, bolancg Argon (No
4)
With 12% 0-atoms
-
-, .I5 0
I
t
\
Pressure = 18millitorr Temperature = 30'C
.8
IO 0
U
-
0,
1
,20
-
Nickel Catalyst Pressure = 15millitorr
0
1
2
I
1
I
I
4 6 8 IO HEIGHT ABOVE SURFACE (CM)
Figure 3. Relative NO2*orange emission as a function of distance above the catalyst. such emission has not been observed and is only observed in the p r w n c e of discharged molecular oxygen. Ground-state molecular oxygen was added to the flow system at the equivalence point and the region above the catalyst was again observed to be dark. This may be due to the absence of some other species important in the heterogeneous production of the 02(A32,+). Alternately, this may be due to differences in surface conditioning with or without molecular oxygen presence. Further investigation on this phenomenon is needed. This observation supports the suggestion that the mechanism for the formation of NO2* in the absence of molecular oxygen does not involve the 02(A3Z,+) state as an intermediate. During the experiments described above, it was observed that the orange glow resulting from the addition of N O to an N-atom stream appears to extend a smaller distance than the orange glow described previously,8 which was observed in an 0-atom stream from an ac discharge through molecular oxygen. The intensity of the orange NO2* emission has been observed as a function of the distance above the surface of the catalyst both in the presence and absence of molecular oxygen. The results are given in Figure 3. The emission observed in the presence of molecular oxygen does clearly extend farther from the surface than the emission observed in the absence af molecular oxygen. If we assume the NO2* is created on the surface, the steady-state concentration of NO2*in the gas phase is controlled by diffusion into the gas phase and by the rates of quenching and radiative emission. Solution of the rate equation yields18 In ([NO,*]
/ [NO2*],)
Temperature = 45°C
I
12
=-x/(DT)~/~
where [NO2*], is the concentration of NO2* at the surface of the catalyst, x is the distance above the surface, D is the diffusion coefficient for NO2* in the gas mixture, and T is the effective lifetime for NO2* considering quenching as well as the natural radiative lifetime. Since the emission intensity, I , is proportional to [NO2*],In ( I / & ) is expected to be linear with distance above the surface as verified by Figure 3. Assuming a mutual diffusion coefficient39of approximately 0.1 1 cm2/s at 1 atm yields effective lifetimes for NO2* of 79 and 960 pus respectively in the absence and presence of molecular oxygen. The intensity of the Herzberg emission from the 02(A3z,+) state resulting from the recombination of 0 atoms on a nickel catalyst was also measured as a function of distance from the surface of the catalyst. Figure 4 shows that the effective lifetime is larger (8.4 milliseconds) when the major component is argon than when the major component of the gas mixture is molecular oxygen (0.93 ms). Measurements of the emission intensity as a function of distance from the surface were also made for three (17) Clyne, M.A. A.; Thrush, B. A. Proc. R. SOC.London, Ser. A 1961, 261, 259.
(18) Weinreb, M. P.; Mannella, G. G. J . Chem. Phys. 1969, 50, 3129.
.25 -
4"
~
0
0.3
0.6
1.2
0.9
1.5
1.8
HEIGHT ABOVE SURFACE (CM)
Figure 4. Relative 01*Herzberg emission as a function of distance above the catalyst.
I
\a\
Nickel Catalyst Temperature = 35OC Pure Oxygen
I
I
0
.4
I
I
1.2
I
I
1.6 2.0 HEIGHT ABOVE SURFACE (CMI
.8
I
2.4
Figure 5. Relative 02*Herzberg emission as a function of distance above the catalyst.
different 0-atom concentrations at constant pressure and temperature. Figure 5 represents only the relative changes in the intensity as a function of height above the surface and not the relative intensities. The integrated relative intensities over the entire emission area above the catalyst are increased with increasing 0-atom concentration in a manner consistent with the previous report8 that the Herzberg emission is second order with respect to 0-atom concentration at high 02/0ratios. The effective lifetime is observed to increase nearly linearly with the oxygen atom concentration from 404 ps at 8.2% 0 atoms to 756 hs at 13.8% 0 atoms in O2 at 20 mtorr. Both spectra of the Herzberg Emission from 02*and the orange glow from NO2* have been measured. The spectrum of the Herzberg Emission is shown in Figure 6 and most bands have been assigned. The spectrum of the orange glow with molecular oxygen present is shown as curve B in Figure 7. The correction of the orange glow from the background airglow (curve A in Figure 7) was accomplished as described by Kenner and Ogryzlo.6 The spectrum of orange glow obtained by Kenner and Ogryzlo6 (curve C in Figure 7) was also included for comparison. Discussion The observation of an orange emission resulting from the heterogeneous recombination of 0 atoms and nitric oxide in the
Electronically Excited NO2 and O2
The Journal of Physical Chemistry, Vol. 90,No. 3, 1986 469
Nickel Catalyst Pressure = 20 millitorr Temperature = 3 5 O C 10% 0-atoms in O2 IV””)
1 (0.7) I 11.6)
I (2.4)
1
1
11.8)
11.7)
I
1
(2.5)
I
(2.6)
(2.7)
I
1
1
(0.8)
(0.9)
10.10)
1 11.10)
1
(1.9)
(3-60 (5.4)
I
I
I
300 nm
350nm
I
400nm
450nm
I
500nm
Figure 6. Spectrum of electronically excited oxygen over nickel catalyst
Wavelength (nm 1 Figure 7. Spectra of electronically excited nitrogen dioxide: (A) greenish-white emission of the air afterglow; (B) orange emission on nickel surface (PO,= 20 mtorr, PN0 = 1.5 mtorr); (C) orange emission obtained by Kenner and Ogryzlo.6
absence of molecular oxygen indicates that there are at least two different mechanisms which can result in the formation of this glow when nitric oxide is added to an 0-atom stream with a nickel catalyst. In the absence of molecular oxygen the following mechanism previously proposed by Halstead et a1.* is expected to dominate: NO
+0
surface
- +
NO2* (surface)
(1)
NO,*(gas)
(2)
N02*(gas) N02(gas) N02* NO2 hv
(3 )
N02*(surface)
M
(4)
Step 1 is the direct heterogeneous recombination to form electronically excited nitrogen dioxide. Either nitric oxide, atomic
oxygen, or both may be initially absorbed on the surface. Step 2 is diffusion of NO2* into the gas phase and step 3 represents quenching by all species present. With this mechanism, the effective lifetime observed by measurements of the intensity as a function of distance from the surface is expected to be the effective lifetime of NO2* calculated from the natural radiative lifetime and the appropriate quenching coefficients. The choice of a value for the radiative lifetime of the heterogeneously formed NO2*is by no means straightforward. Quantum calculation^^^ indicate that NOz has a bent 2A, ground state and the first two excited states are a linear 2Bl state and a bent 2Bz state. The ,B1state may be considered to be the most likely product of the recombination of N O and an 0 atom since it correlates with the ground-state reactants. It has been suggested, however, that the zBI state may cross over rapidly to the 2B2state even under collision-free conditions.20 It is also not clear whether the ’B, state, the 2B2, or both are created in the much more thoroughly studied case of the gas-phase recombination of N O and an 0 atom or airglow.21 There is a large body of work dealing with the radiative lifetime of electronically excited nitrogen dioxide in the visible r e g i ~ n ; ~ however, ’ - ~ ~ a clear understanding of the processes involved has still not emerged. There is some indications that the lifetime or lifetimes in this region may be wavelength dependex~t.~~JO At some wavelengths nonexponential behavior and (19) Gangi, R. A,; Burnelle, L. J. Chem. Phys. 1971, 55, 843, 851. (20) Paech, F.; Schmiedl, R.; Demtroder, W. J . Chem. Phys. 1975, 63, 4369. (21) Newberger, D.; Duncan, A. B. F. J . Chem. Phys. 1954, 22, 1683. (22) Douglas, A. E. J. Chem. Phys. 1966, 45, 1007. (23) Schwartz, S. E.; Johnston, H. S. J . Chem. Phys. 1969, 51, 1286. (24) Sakura, K.; Capelle, G. J . Chem. Phys. 1970, 53, 3764. (25) Sackett, P. B.; Yardley, Y. T. J. Chem. Phys. 1972, 57, 152. (26) Paech, F.; Schmiedl, R.;Demtroder, W. J . Chem. Phys. 1975, 63, 4369. (27) Hakala, D. F.; Reeves, R. R. Chem. Phys. Lett. 1976, 38,510. (28) Donnelly, V. M.; Kaufman, F . J. Chem. Phys. 1977, 66,4100. (29) Donnelly, V. M.; Kaufman, F. J . Chem. Phys. 1978, 69, 1456. (30) Donnelly, V. M.; Keil, D. G.; Kaufman, F. J . Chem. Phys. 1979,71, 659. (31) Keil, D. G.; Donnelly, V. M.; Kaufrnan, F. J. Chem. Phys. 1980, 73, 1514. (32) Kaufman, F. In “Chemiluminescenceand Bioluminescence”,Cormier, M. J., Hercules, D. M., Lee, J., Ed.; Plenum Press: New York, 1973.
470
The Journal of Physical Chemistry, Vol. 90, No. 3, 1986
multiple lifetimes are observed even when narrow-band excitation is used.26,29*30 A number of studies indicate a lifetime in the 60-70-1s range.32 More recent work indicates lifetimes which increase smoothly from 88 ps at an excitation wavelength of 473 nm to 120 ps at an excitation wavelength of 562 nm.29 When interpreting the effective lifetime reported here for the case of NO2* formed by the direct heterogeneous recombination of NO and an 0 atom, electronic quenching coefficients as well as the radiative lifetime should be considered. Schwartz and Johnson23 state that the electronic quenching coefficient for self-quenching must be much smaller than 2 X lo-'' cm3 s-'. Keyser, Levine, and K a u f m a r ~reported ~~ an upper limit for the cm3 s-'. Since the electronic quenching coefficient of 4 X coefficient for the quenching of NO2*by N2 can be expected to be smaller than that for quenching by NO2,electronic quenching can be assumed to be nearly negligible at 20 mtorr. The observed effective lifetime of 79 ps can then be interpreted as an average radiative lifetime for the species created at the surface in a direct heterogeneous recombination process occurring in the absence of molecular oxygen. While it is not entirely clear which previous radiative lifetime this should be compared to, this value seems reasonable for emission from NO2* in the visible region. The observation of an orange emission rather than the well-known whitish green emission from the homogeneous air is interpreted in terms of an energy loss to the surface during product formation. Another reaction for which the emission from a heterogeneously formed electronically excited state is observed to be shifted to the red relative to the analogous homogeneously formed product is the case of N-atom recombination to form N2*.1-3,8 Unlike the case of the homogeneous airglow, the visually observed color from the NO2*heterogeneous recombination reaction is found to be pressure independent in the 10-400-mtorr range. The value of 960 ps observed for the effective radiative lifetime in the presence of molecular oxygen is clearly inconsistent with the direct heterogeneous formation of NO2*at the surface of the catalyst as the sole source of NO2*. An effective lifetime this long suggests the gas-phase formation of NO2* from an electronically excited precursor. This precursor, which is formed on the surface of the catalyst, must have a radiative lifetime longer than that of NO2*. The observations in the presence of molecular oxygen supports the formation of NO2*by the following mechanism previously suggested by Kenner and Ogryzlo:6 0 2 ( ~ % , + )+ O2
-
03* +0
Chu et al. 02(A3Z,+) state is formed exclusively in a primary heterogeneous step as proposed previously: and the quenching coefficients of Kenner and Ogryzlo4are used for the quenching of the 02(A32,+) by 0, 02,and argon then the effective radiative lifetimes expected can be calculated. For 12% 0 atoms and 88% O2 at 15 mtorr the expected effective lifetime is approximately 1.5 ms. For 8% 0, 14% 02,and 78% Ar at 15 mtorr the expected effective lifetime is 2.7 ms. The observed effective lifetimes for these conditions are 0.93 and 8.4 ms, respectively. The ratio of these two values as well as the magnitude of the latter suggest that the observed effective lifetimes are not dominated by quenching processes as expected from the mechanism proposed previously by Halstead et ale8 This mechanism suggested the direct second-order heterogeneous recombination of 0 atoms to form the 02(A3Z,+) on the surface. The 02(A32,+) state then diffuses into the gas phase where it may undergo quenching by 0 atoms, 02,or Ar. Figure 5 indicated that the observed effective radiative lifetime increases linearly with 0-atom concentration. This is also in direct contrast with the mechanism proposed previouslys which predicts a decrease in the effective lifetime with increasing 0-atom concentration due to quenching. An alternative mechanism must be proposed. While this mechanism cannot be uniquely identified from the current data it must involve two pathways by which 02(A3Z,+) can be formed. One of these involves an, as yet unidentified, electronically excited state as a precursor. The following series of steps is an example of one possible mechanism which fits these criteria: 0
0
+0
+0
surface
surface
02*(surface)
(9)
02(A3Zu+)(surface)
-
+ - +
O2(A3ZU+)(surface) Oz(A3Z,+)(gas) 02(A3Z,+)(gas) + 0 Oz(A32,+)(gas)
O2
02(A3Zu+)(gas)
(14)
02(gas) + 0
(15)
202(gas)
(16)
O2
hv
(17)
Reactions 9-1 1 represent the formation of the 02(A3Z,+) via a gas-phase reaction with the heterogeneously formed unidentified precursor Oz*. Reactions 13 and 14 represent the direct heter03* NO NO2* O2 (6) ogeneous formation of the 02(A3Z,+) state followed by its subNO2* M NO2 M (7) sequent diffusion into the gas phase. For this mechanism to be consistent with an increase in the observed effective lifetime with NO2* NO2 + ~ Z J (8) increasing 0-atom concentration, the total rate of quenching of the 02(A32,+) state by all species present must be greater than The 02(A38,+) state is assumed to be produced by the het~ total rate of quenching of the 02*state. erogeneous recombination of 0 atoms as reported p r e v i ~ u s l y . ~ ~ , ~the The ozone formed in step 5 is vibrationally excited. The observed Conclusions effective lifetime for NO2*formed via this mechanism will differ Measurements of the attenuation of emission intensity as a from that for NO2* formed via the mechanism discussed above. function of distance from the catalyst for the surface-catalyzed In this case the observed effective lifetime will be determined formation of NOz* indicate that there are two different mechaprimarily by the effective lifetime of the 02(A3Z,+) state under nisms which can result in the formation of this glow when nitric the conditions of the reaction. oxide is added to an 0-atom stream with a nickel catalyst. In The intensity of the Herzberg emission from the 02(A38,+) the absence of molecular oxygen, mechanism I, the direct hetstate resulting from the recombination of 0 atoms in an 0-atom erogeneous recombination of nitric oxide and an 0-atom on the stream was also measured as a function of distance from the surface as previously suggested by Halstead et a1.,8 appears to surface under various conditions. The observed effective lifetime dominate. One mechanism, mechanism 11,which would meet this when the major component is argon is expected to be larger than criteria is that previously described by Kenner and Ogryzlo.6 the observed effective lifetime under similar conditions when the Unlike mechanism I, mechanism I1 involves two gas-phase steps. major component is molecular oxygen. This expectation is An electronically excited state of oxygen is formed on the surface qualitatively verified by the data presented in Figure 4. If the and subsequently diffuses into the gas phase where it reacts with ground-state molecular oxygen to form vibrationally excited ozone. (33) Keyser, L.F.;Levine, S . Z.; Kaufman, F. J. Chem. Phys. 1971, 54, The vibrationally excited ozone then reacts with nitric oxide to 355. form NO2*. (34) Fortijn, A,; Meyer, C. B.; Schiff, H. I. J . Chem. Phys. 1964, 40, 64. The laboratory results presented here which indicate that NO2* (35) Mannella, G. G.; Harteck, P.J . Chem. Phys. 1961, 34, 2177. may be formed via a direct heterogeneous reaction not involving (36) Harteck, P.; Reeves, R. R. Discuss. Faraday SOC.1964, 37, 82.
+
+ -
+ +
(5)
J . Phys. Chem. 1986, 90,411-471 any gas-phase steps support the recent suggestionsl1J2that the source of the space shuttle glow may be NO2*. This emission was first discovered in photographs taken during the third and fourth space shuttle missions (STS-3 and STS-4) while the vehicle was on the night side of the earth.37*38An orangish glow initially thick and later revised to 20 cm estimated to be 5 to 10 cm37*38 thickI2 was observed adjacent to the windward side of the vehicle. This glow is the result of an interaction between the spacecraft and the surrounding atmosphere. At the very low pressures in the vicinity of 240 km where the shuttle glow has been observed, the emission of light must be due to an excited molecule formed (37) Banks, P. M.; Williamson, P. R.; Raitt, W. J. Geophys. Res. Lett. 1983. 10. 118.
(38) Mende, S.B.; Garriott, 0.K.; Banks, P. M. Geophys. Res. Lett. 1983, 10, 122.
(39) The diffusion coefficient of electronically excited NO2 may be different from that for ground-state NOz. The value of 0.1 1 is selected as a value typical of triatomics in air.
471
on the surface, escaping from the surface, and then emitting light. Any subsequent reaction in the gas phase can be neglected as a result of the long mean free path. The formation of 02(A3&+) above a nickel catalyst in an 0-atom stream formed by an ac glow discharge through molecular oxygen is proposed to also have two mechanisms contributing to the formation of the observed electronically excited state. The observed effective lifetime, as measured by the decrease in intensity as a function of distance from the catalyst surface, increases with increasing 0-atom concentration. Since the 02(A32,+) state is known to be quenched strongly by 0 atoms this suggests a mechanism involving two pathways for 02(A3Z,+) formation, one of which involves a longer-lived precursor. Acknowledgment. The authors thank Rensselaer Polytechnic Institute for support of this work. Registry NO.NO, 10102-43-9; 0,17778-80-2; N02, 10102-44-0; 0 2 , 7782-44-7; Ni, 7440-02-0.
Structure and Reactivity in Ionic Reactions Chau-Chung Han, James A. Dodd, and John I. Braurnan* Department of Chemistry, Stanford University, Stanford, California 94305 (Received: July 22, 1985)
Rates and mechanisms of selected bimolecular nucleophilic substitution and proton-transfer reactions involving anionic reactants in both the gas and solution phases are summarized and compared. The predictive power of various rate-equilibrium relationship, especially Marcus theory, is discussed. The nucleophilic reactivity (A&*) is found to correlate better with the nucleophile methyl cation affinity (MCA) than with basicity. The effect of solvent on Hammett p values observed for methyl transfer reactions is considered.
I. Introduction Ionic reactions play a central role in chemistry. For instance, many of the most powerful synthetic reactions utilize ionic reagents. The ability to make modest changes in reaction conditions, such as solvent, leaving group, or counterion, enables one to manipulate rates and chemical products with extraordinary control. Nevertheless, our true understanding remains limited; the delineation of conditions that optimize the yield of a desired product is a process that relies on experience-patterns found in previous studies-rather than on some theoretical footing. Useful predictive tools would find application to chemical reactions taking place in different physical environments. We have been pursuing experiments in the gas phase, where organic systems can display "intrinsic" reactivity in the absence of solvent molecules. The rate of a particular reaction is in general much different in solution than in the gas phase; indeed, the effect of including even one or two solvent molecules into the reaction system can be dramatic, especially for acid-base reactions.' Comparison of the solution- and gas-phase rates can often provide valuable information. In principle, a complete quantum mechanical description of a reaction-one that could describe the potential surface and the dynamics that take place on it-would allow prediction of all aspects of its chemistry. Unfortunately, an accurate potential surface for a complex polyatomic system is quite difficult to obtain, and even if one did possess an accurate surface, the best that could
be said for any subsequent dynamical analysis is that it would be far from routine. The alternative is to make use of the empirical and semiempirical theories that have historically furnished much of the predictive power in describing chemical reactions. Additivity methods that are used to calculate such quantities as heats of formation and molecular polarizability are well established and in widespread use. The predictions afforded by linear free energy relationships have provided many insights over the years. Indeed, many semiempirical rate-quilibrium equations have been applied with equal effect to reactions in the condensed and gas phases. One might conceivably criticize one or more of these theories for their apparent lack of sensitivity to gross environmental changes; nevertheless, their great predictive success-especially that of Marcus theory-makes for a strong rebuttal. There is no doubt that the effects of solvent are considerable, but these effects can be taken account of in uncovering basic similarities in solutionand gas-phase chemistry. The focus in this paper is on two of the simplest ionic reactions: proton transfer and nucleophilic substitution. We first discuss our approach to the study of gas-phase reactivity. We then retreat a bit to consider the history of rate-equilibrium relationships, as well as the more recent Marcus (and related) theories. The next section presents some examples of predictions that are made possible through the semiempirical theories, followed by a discussion of relevant empirical correlations. We conclude with a brief consideration of ionic reactions in solution.
(1) (a) Bohme, D. K. NATO Adu. Sci. Int. Ser., Ser. C 1984, 118, 111. (b) Moylan, C. R.; Brauman, J. I. Annu. Rev. Phys. Chem. 1983, 34, 187.
11. An Approach to the Study of Ion-Molecule Reactions The rate of a gas-phase ion-molecule reaction is determined by both the collision rate and the efficiency with which the collision complex is converted into products. Because of the characteristic long-range ion-induced and -permanent dipole interactions, a
(c) Reaction mechanisms may also change on proceeding from the gas phase to solution. For instance, see: Caldwell, G.; Rozeboom, M. D.; Kiplinger,
J. P.; Bartmess, J. E. J . Am. Chem. SOC.1984, 106, 809, and references therein.
0022-3654/86/2090-0471$01.50/0
0 1986 American Chemical Society