Fluorescence of Nitric OxidePart 3.-Determination of the Rate Constants for Predissociation,Collisional Quenching, and Spontaneous Radiation of NO C2n[(v = 0)BY A. B. CALLEAR AND I. W. M. SMITHDept. of Physical Chemistry, University of Cambridge, EnglandReceived 15th January, 1964The intensity of the resonance fluorescence from NO C2l7(v = 0) was compared to the enhancement of the y-band intensity caused by deactivation of C2n(v = 0) to AW+. It was concludedthat the C2n(u = 0) state dissociates at a rate which is 30 times faster than the rate of spontaneousradiation to X2n. This observation demonstrated that the predissociating state has d>O, and itwas suggested that the simplest interpretation is that the a4l7 state is the connecting link betweenC2n(v = 0) and normal atoms.The D2Z+ state interacts less strongly with the a4n state becauseof the change in A.The following scheme was proposed to account for the fate of NO Czn(v = 0) :NO X217+hv NO A 2 B + h.k, = 9.7 x 10-10 cm3 molecule-1 sec-1, kF = 22x lo7 sec-1, kl = 6.6 x 108 sec-1 and k2 = 5.2 x10-16 cm3 molecule-1 sec-1.The potential curves of nitric oxide are shown in fig. 1. The X2IT, A2Z+,B2H, C2II, and D2Z+ states are well defined from the ultra-violet spectrum.1-7 Ahomogeneous perturbation results from the crossing of the C2II and B2ll potentialcurves.5~ 6 The positions of the a4IT and the flat 2Z+ curves have been calculatedfrom the data of Vanderslice, Maisch and Mason.* The excitation energy andmolecular parameters of the a4II state were estimated by comparison with 0; a4II.The position of the NO a4II potential curve can probably be relied upon to within0.2 eV or 0-05 A.The Ogawa bands 9 define the relative positions of the b4Z-and a4II states. The potential curve of the flat 2Z+ state is rather uncertain.If nitric oxide is irradiated with light from a xenon arc, the A2Z+, C2n, at- ‘D2Z+ states are excited, giving rise to the 7-, 6- and &-bands respectively. Some ofthe absorption bands of NO are shown in fig. 2. It is demonstrated in this paperthat NO C2II(v = 0) is weakly predissociated in all its rotational states. By cal-culating the Franck-Condon factors, the spontaneous emission rate of the C2II(u = 0)-X2IT transition has been determined. Hence, by means of a quantitative deter-mination of the weakening of the fluorescence due to dissociation, the rate constantfor predissociation of C2II(u = 0) has been found.It is suggested that the dissoci-ation energy of nitric oxide is close to 6.493 eV. Thus the rate of preassociation of N(4S)and O(3P) into the C2II(u = 0) state may be determined by calculation of the equi-librium constant. The photolysis of nitric oxide and the recombination of N and0 atoms are important chemical reactions in the earth’s atmosphere.There is already good evidence that NO C2II(u = 1) can dissociate; the &bandswith v’>O have never been observed in emission. However, the rotational lines9A. B. CALLEAR AND I. W. M. SMITH 97of the b-bands have Doppler width at - 183"C,5~ 6 and any predissociation maytherefore be described as weak ; the first-order rate constant must be less than about1 x 1010 sec-1.Flory and Johnston irradiated nitric oxide with light from a varietyof sources and concluded that decomposition resulted largely from absorption oflight by the (1,O) b-band,lo though a small fraction of the decomposition may haveresulted from absorption at longer wavelengths.11I *o 2 . 0internuclear distance R (A)FIG. 1.-The potential diagram of nitric oxide.In a study of the combination of N(4S) and O(3P), Young and Sharpless 12observed the (0' = 0) &bands in emission and concluded that the C2n(u = 0) statewas populated via the a4II state by a preassociation mechanism. The a4n state waspostulated as the precursor to combination in preference to either the b42- stateor the flat 2X+ state, because the observed infra-red system C2n-A2Ct exhibitedP, Q and R branches at low pressure. By considering the selection rules it followsthat the consecutive processeswould have no Q branch, and the stepswould give neither P nor R branch. The detection of the P, Q and R structure of theC2lI-A2ZS- bands therefore demonstrates that the precursor to the C2rI(v = 0)state has A>O, provided that the preassociation mechanism is correct.Thus,according to fig. 1, the a4n state is the only possible precursor to the C2n(u = 0)state, though states at present unknown could conceivably be involved. Young andSharpless suggested that A2X+(u = 0) can be populated by radiation or by collisionprocesses with N2.Their interpretation of the population of the b4Z- and theB2II states does not appear to be straightforward; our work is only indirectly2Z++C2II(u = O)+A2Z++hvb4X:-4C2rI(v = O)+A2X++hv98 FLUORESCENCE OF NITRIC OXIDErelated to these problems, and to the cause of the cut-off of the /I-bands in emissionwith v’ < 6.Young and Sharpless’ experiments provide evidence that the NO C2IT(u = 0)state is predissociated. However, there could be alternatives to the preassociationmechanism, such as bimolecular energy transfer. Our observations on the fluor-escence of the &bands, make possible a comparison of the rate of direct excitationwith the rate of 6-band emission from the C2IT(v = 0) state. The NO C2II(u = 0)molecules are deactivated by collisions with N2 and we have shown that the rateof photolysis depends in a similar manner on the N2 pressure as does the intensityof the 6-band fluorescence. These observations, taken with those of Young andSharpless, offer strong proof that NO C2IT(u = 0) has a much larger probabilityof predissociation than of spontaneous radiation.A quantitative study of theeffect of Nz on the nitric oxide fluorescence, makes possible a determination of thefirst-order rate constant for predissociation of NO C2II(v = 0).EXPERIMENTALThe apparatus and general experimental method have been described.13 The &systemwas photographed on Ilford Q3 plates using exposures of 3+-h duration. The subsequentdevelopment and plate photometry have been described.14 The calibration curve used toconvert the densities of the bands on the plates to relative intensities is shown in fig.3.///Y5 0 I00 150intensity (arbitrary units)FIG. 3.-Density-intensity curve for Ilford Q3 plates at about 2250A with the NO ybands aslight source ; time of exposure, 33 h.0, (0,O) band; a, (1,l) band; 0, (3,4) band; 8, minimum between (0,O) and (1,l) bands.The intensity of the y-bands was varied with a number of quartz plates acting as a step wedge.The variation of intensity was measured on a Unicam ultra-violet spectrophotometer.The complete curve was obtained by fitting together the separate curves obtained fordifferent regions of plate density.For measurements of the quenching of the &bands, the total pressure was kept con-stant by addition of an excess of either argon or helium.Pressure broadening the absorp-tion lines increased the intensity of the fluorescence, and also kept the rate of excitationconstant. Plate photometry of the bands was rendered difficult by the partial overlap ofthe &bands with bands from the y-system due to the use of wide slit openings. However,the (0,3) &band was suffciently isolated and strong enough for quantitative measurement,and the difficulty of overlap diminishes at high nitric oxide pressures when the (1,O) y-banFIG. 2.-Absorption spectrum of nitric oxide. NO pressure 30.0 mm ; patA. B . CALLEAR AND I . W. M. SMITH 99is considerably self-reversed. In later experiments, by improving the optical arrangementand using a new lamp, the (0,2) &band was also recorded with sufficient plate density foraccurate photometry.CALCULATION OF SOME FRANCK-CONDON FACTORS AND THE RATE OF RADI-ATION OF NO c2n(U = O) TO x2nThe mean radiative lifetime of a state can be calculated if the electronic f-value andthe Franck-Condon factors of all transitions are known.13 The f-value for the NO (0,O)&band has been measured by Bethke 15 but the Franck-Condon factors have not been re-ported for either the C2II-A2Z+ or the C2II- X2II transitions.However, Bates 16 hascompiled data from which Franck-Condon factors can be calculated for transitions betweenlow-lying vibrational levels.TABLE 1 .-FRANCK-CONDON FACTORSTc2n - A2XfV” 0 1 2V’0 -995 -001 -0001 -001 -986 a002c2n- X2IIV” 0 1 2 3 : 4V’0 ‘1 63(.153) *256(*255) .250(.238) *194(.166) : *O51(‘097)t Note added in proof.Ory (J Chem. Physics, 1964, 40, 562) has recently made an accurate calculation of the Franck-Condon factors for the 6 bands, and these are shown in brackets in table 1.The values of We, re and x,, were taken from Herzberg, 17 and Lagerqvist and Miescher. 6The Franck-Condon factors for the NO C2II(v = 0 and 1)-A2Cf are given in table 1.The equilibrium internuclear distances in the two states differ by only 0.011 A and theAv = 0 transitions are therefore much the strongest, with the Av = f l transitions weak,and the rest close to zero. The Franck-Condon factors for the (v’ = 0) 6-bands are alsoincluded in table 1.The strongest are clearly those with v”<3, for which Bates’ tablesgive accurate Franck-Condon factors. Therefore, it is possible to calculate the span-taneous emission rate for the C2n(v = 0)- X2lT transition using the f-value for the (0,O)8-band. The overlap with the weak (7,O) P-band has been allowed for by comparisonwith the f-value for the resolved /?-bands also given by Bethke. The $value used in thecalculation was fo,o(h) = 2.4 x 10-3. The rate of spontaneous emsision from C2II(u = 0) -X2lT was calculated to be 2 . 2 ~ 107 sec-1.The C2II state also radiates to the A%+ state and Young and Sharpless suggest thatthe rate is comparable to the rate at which the &bands are emitted. If this were so, thetransition matrix element for the CTI(u = O)-A2Z+(v = 0) transition would have to beabout 400 times that for CW(v = O)-XZll in order to compensate for the (frequency)3and degeneracy terms in the expression for the transition probability.This is possiblesince the f-value would be below unity ; for a given $value, the rate of radiation dependson (frequency)2. Our rate constants are determined relative to the &emission and are notdependent on the unknown rate of the C”(v = 0)-A2Z+ transition.RESULTSQUENCHING OF THE NO@’ = 0) BANDSIf nitric oxide is irradiated with light from a xenon arc, an ammonia filter canexclude excitation to states higher than A2Z+(v = O).139 14 The stationary stat100 FLUORESCENCE OF NITRIC OXIDEconcentration of molecules in this level, when excited by light filtered throughammonia, can be compared to that when the fluorescence is excited by unfilteredlight, in order to determine the extent to which this state is populated via higherenergy levels.Thus it was shown that population of NO A2C+(u = 0) from higherenergy levels by collision processes with argon or helium is extremely slow. Inthe presence of nitrogen, however, a considerable enhancement of the intensity ofthe u‘ = 0 y-bands is observed when the fluorescence is excited by unfiltered light,and this is attributed to the occurrence of two energy transfer processes :(i) vibrational energy exchange between NO A%+(u = 1) and N2 XlC+(u = 0),(ii) quenching of NO C2rI(v = 0).The rate of these processes has been measured and, at low nitrogen pressures, quench-ing of NO C217(v = 0), and not vibrational relaxation is clearly the dominantprocess.Since neither argon nor helium quench the 6-bands, the intensity of the systemin fluorescence can be increased, and kept constant, by adding a large excess ofargon or helium to the nitric oxide in order to pressure-broaden the lines.Quench-ing by nitrogen can then be observed by measuring the intensity of the (u’ = 0)&bands excited in mixtures of nitric oxide, nitrogen and inert gas. The nitric oxideand total pressures are kept constant, but the partial pressure of nitrogen is changed.Quenching of the (0,2) and (0,3) 6-bands by low partial pressures of nitrogen isclearly shown in fig. 4. Five series of experiments have been carried out at differentnitric oxide and total pressures.These results have been treated in the conven-tional manner by plotting Io/l against the partial pressure of nitrogen, where I0is the intensity of the (0,3) or (0,2) 8-band without N2 added, and I is the intensityin the presence of added N2, at the same nitric oxide and total pressure. The re-sultant Stern-Volmer diagram is shown in fig. 5. Within experimental error, fornitrogen quenching half-pressure at a total pressure of 700mm was constant, thenitric oxide pressures between 2 mm and 12 mm. From fig. 5, it appears that thenitrogen quenching half-pressure is slightly dependent on the argon pressure presentin the system. The argon quenching half-pressure for the &bands is certainly greaterthan 380 mm.The substitution of helium for argon as inert gas in these mixturesdid not affect the measured nitrogen quenching half-pressure. These results demon-strate that helium is as ineffective as argon at quenching the &bands. By comparingthe intensity of the 6-bands excited in the same excess of argon and helium, it wasshown that any difference in the rate of excitation as a result of different pressurebroadening by helium, on the one hand, and argon and nitrogen, on the other, wasvery slight.In order to investigate the result of quenching NO C2n(u = 0) by N2 in moredetail, the intensities of the various y-bands were measured with, and without, theaddition of a small partial pressure of N2. Any increase due to quenching of NO C2IIis most apparent at low nitric oxide pressures when self-quenching of the y-bandsis slight.Therefore, the fluorescence was excited in 0.6 mm of NO with (a) 700 mmargon, and (b) 45 mm N2+655 mm argon. To bring the plate densities of they-bands into the most convenient and sensitive range, exposure times were reducedto l s h . Fig. 6 shows clearly the enhancement of the (u’ = 0) y-bands and thesmaller increase in the (u’ = 1) progression on adding nitrogen. With a smallnitrogen partial pressure, this effect cannot be solely due to vibrational relaxationof NOA2C+. To treat these results quantitatively, the ratio of intensities inmixtures (a) and (b), (Ib/Ia), was predicted for each y-progression, assuming thatpopulation of the excited levels occurred only by direct light absorption and bA .B. CALLEAR AND I . W. M. SMITH 101vibrational exchange with N2 at the rates previously measured. This was done bysetting up the necessary stationary state equations for population and depopulationof the various levels and substituting in these equations the rate coefficients previously3 -2 -l o l lII I I20 4 0 6 0 00pressure of N2 (mm Hg)FIG. 5.-The ratio of the intensity of the (v’ = 0) &bands without (lo), and with (I), added nitrogen,as a function of its partial pressure, for different nitric oxide and total pressures.8 2.0 mm NO0 12.0 mm NO9 5.0 mm NO (0,3) band, (He+NO+N2) 700 mm0 5.0 mm NO(0,3) band, (Ar+NO+Nz) 700 mm0 5.0 mm NO (0,3) band (h+NO+N2) 300 mm(0,2) band}reported,l3, 14 and the relative intensities from exposures (a) and (b).For example,the concentration of NO A2Cf(u = 3) with nitrogen present, is given byr3(n31N2 = 1 + k,[NO] + kQ[M] + k,[N,]’and without nitrogen bySo for the (v‘ = 3) progression= 0.81. 1 +O*GjO*91+ 700/14001 + 0.6/0.91+ 700/14OO + 8.3 x 45/760- 102 FLUORESCENCE OF NITRIC OXIDE13 is the rate of formation of NO A~Z+(U = 3), and n3 is the concentration.rate coefficients are for the processesThekfk ,kQNO C(U = ())-+NO A(u = O)+hvNO A(u = O)+NO X+NO X+NO XNO A(zj = O)+Ar-+NO X+Ark3NO A(u = 3)+N2+N0 A(v = 2)+N2The observed and predicted ratios differ significantly for the U' = 1 and v' = 0and have all been divided by the spontaneous emission rate from the A2Z+ state.bands, and this is due to quenching of NO C2ll by nitrogenkc, NO C+NZ+NO A(u = l)+N2NO C+N2-+N0 A(v = O)+N2.It is essential to determine the relative rates of these processes in order to determinethe rate of excitation to NO C2II.It is first necessary to reconsider the stationarystate equations and equate the differences in table 2 between the observed andpredicted ratios, to the terms previously neglected. This givesobs. &/Iapredicted I,[&,TABLE 2difference -0.10 - 0.03 0.32In case (a) (nO)Ar/(dAr = 1.5% of total NOA2C+ pro-duced by quenching ofand- 22 1 - NO C2II by 45 mm of N20.7878Similar measurements have been made on the (u' = 0) and (u' = 1) y-bands ex-cited in mixtures of 6.0 mm of NO with (a) 700 mm of argon, and (6) 700 mm ofnitrogen.These results are compared to the intensity ratios predicted for thesemixtures assuming that populatiun of these levels occurs only by direct excitationand by vibrational relaxation of NO The results can then be treated in themanner which has been described above. It is thus shown that the ratio of therates at which molecules are produced in A~Z+(U = 0) and in A2X+(u = l), as A . B. CALLBAR AND 1. W. M. SMITH 103result of quenching NO CHI by 700 mm of nitrogen, is equal to the ratio of therates in the presence of only 45 mm of nitrogen.TABLE 3difference 0.47 0.75In case (a) (no>A?/(nl>A, = 2.524% of total NOA2V produced byquenching of NO C2lI by 700 mmof N276EVIDENCE FOR PREDISSOCIATION OF NO c2n(v = 0)The enhancement of the (u' = 1) and (u' = 0) y-bands as a result of quenchingof NO C2II by N2, is considerably larger than might be expected by comparing theintensity of the unquenched &bands with the intensity of the y-bands in fig.4. At700 mm, k, [N2][NO c]/(nl)N2 = 2.1, i.e., over twice as many molecules are quenchedfrom NO C2II to A2C+(u = 0) per unit time as radiate in the same time fromA2X,+(u = 1). However, fig. 4 clearly shows the (u' = 0) &bands to be very muchweaker in the absence of nitrogen than the (u' = 1) y-bands. This cannot beattributed to low Franck-Condon factors for the observed &bands; for the (0,3)6-band the Franck-Condon factor is 0.166.A second apparent anomaly in regard to the &band intensities is that at low nitricoxide pressures and in the absence of foreign gases, the &-bands are considerablystronger than the &bands.This is clearly illustrated in fig. 7. Under these con-ditions the reverse is to be expected; the two (0,O) bands have roughly comparable$values, the &band, in fact, having slightly the larger value. Moreover, the 6-band is at larger wavelength and is, therefore, in a region of greater quantum input.Finally, because of the large A-doubling in the C2II state reported by Herzberget aZ.,59 6 the (0,O) &band contains more separate rotational lines than the (0,O)&-band. Therefore, despite the comparable $values, the (0,O) &-transition resultsin a greater fraction of the incident light being absorbed near the front wall; at adepth of 2cm inside the reaction vessel, the exciting radiation suffers a greaterreduction in intensity than that exciting the (0,O) &band.Allowing for the Franck-Condon factors, calculations based on measurements from the plate shown in fig. 7show that the overall rate of radiation of the (u' = 0) &-bands is about 3 times thatof the (u' = 0) &bands. Gaydon 2 has reported that the 6 and &-bands excited in atransformer discharge show similar relative intensities.Final proof that NO C2II(u = 0) dissociates was obtained by studying the extentof photodecomposition, with and without added N2. Table 4 shows the results ofsome of these experiments in terms of the pressure of gas produced in the reactionvessel, which was condensed at- 196°C and was not removed by pumping.TABLE 4.-hESSURE OF THE CONDENSABLE PRODUCTS OF PHOTODECOMPOSITION (ITU'Tl)expt .20 rnm Not300 mm Ar 20 mrn N0+300 mm N2blank 0.041 0.452 0.503 0.500.010.190.220.21 04 FLUORESCENCE OF NITRIC OXIDEDecomposition in the presence of N2 occurs via NO C2II(u = 1) and NOD2C+(v = 0 and 1); the latter appear to undergo collisionally-induced predissoci-ation. In Ar, the NO C2II(v = 0) also predissociate. More NO C2rI(u = 0) isproduced than NO D2Z+(v = 0), because of the smaller f-values of the individuallines of the (0,O) &band and also because the (0,O) &band is at longer wavelengthsthan the (0,O) &-band. Details of this work and a full account of the study of thephotolysis will be described elsewhere.18RATE OF PREDISSOCIATION OF NO c2n(U = o), RELATIVE TO SPON-TANEOUS RADIATION TO x2nThe results show clearly that in argon or helium a large proportion of the NOmolecules which are excited to CXI(v = 0) do not revert to the X2II state and emitthe &bands, but are lost by a second process which is practically independent ofinert gas pressure.To compare the number of molecules radiating from NO C2II(v = 0) with those produced in NO A2C+(v = 0) as a result of quenching of theC2rI state, the intensity of the (8' = 0) &bands were measured relative to the y-bands in the presence of argon. This ratio was compared to the relative increasein intensity produced in the (u' = 0) bands by quenching of NO C2II in the samepressure of nitrogen.The stationary state equations for population and depopulation of the AzZ+(u = 0) level in nitrogen, and in argon arerO+kf[No](nl)Ar("o)Ar = 1 + k,[NO] + kQ[Ar]'where kl and kl are the rate coefficients forkiNO A(u = I)+N2-+N0 A(v = O)+N2,NO A(u =l)+NO X+NO A(u = O)+NO X.klIf [MI = "21 = [Ar],andThe relative concentrations no)^,, (no)&, ( n l ) ~ ~ and ( n l ) h were determined inmixtures of 6-0 mm of NO with (a) 700 mm of argon and (b) 700 mm of nitrogen,by measuring band intensities, and are shown in table 5.With the rate coefficientsreported previously,13~ 14 it is then possible to calculate k,[NO c)~,[N~l/(nl)~,.18.1-k,,[NO C] ~,[Nz]/(n 1)N2 = 1-35 - 0.07 - 0.24,kc~[No clN~[N21/(nl)N2 =8*4A .B. CALLEAR AND I . W. M. SMITH 105To compare the rate of quenching of NO C2ll to A~Z+(U = 0) with the rate of radi-ation from NO C2ll(u = o), the relative intensities of the (0,3) 6-band and the(1,l) y-band were measured in 700 mm of argon. Correcting by the respectiveTABLE ~.-RELATNE STATIONARY STATE CONCENTRATIONS(no) 100 52.5(n1) 35.1 21.06.0 mm N0+700 mm Nz 6-0 mm N0+700-mm ArFranck-Condon factors, kf[NO C ] / ( n l ) ~ , was thus determined, where kf is the spon-taneous emission rate of the 6-bands divided by the spontaneous emission rate ofthe y-bands :kf[NO C ] j ( r ~ , ) ~ , = 0.63.- rate of production of NO A2X+ju = 0) by quenching of NO C2Hrate of radiation of NO C211 to NO X211= 8-4/0.63 x 35*1/21.0 = 22.3.In order to convert this ratio into the percentage of molecules excited to C2Il(v = 0)which revert to the ground state and emit the b-bands, it is now necessary to con-sider quenching to A~X+(V = 1).It is shown below that this also occurs as a resultof quenching of NO C2ll(u = 0) and, as the rate of quenching to v = 1 is knownrelative to that to v = 0 (table 3), thentotal rate of production of NO A2C+ by quenching of NO C211rate of radiation of NO C21T(v = 0) to NO X21122.3 x 2476~ 2 2 . 3 +- = 29.3.Finally, the fraction of molecules remaining unquenched in the presence of 700 mmof N2 must be allowed for. Sorate of excitation to NO C211(v = 0)rate of radiation of NO C211(u = 0) to NO X21129.3 x 41= 29.3+ = 31.0.700The non-appearance of the (v' > 0) &bands in emission despite no apparent broaden-ing of the absorption lines has led Herzberg et aZ.59 6 to postulate a weak predis-sociation in these levels.The above evidence shows that only 3 % of the mole-cules excited to the C2II(v = 0) state radiate ; this level is also weakly predissociated.DISCUSSIONMECHANISM OF PREDISSOCIATIONYoung and Sharpless proposed that the a4ll state provides a connecting linkbetween NO C2II(u = 0) and the separated ground state atoms; their reasoning issummarized in the introduction. Our experimental results also show that theintermediate state has A>O. If the C2II state predissociated via a C state, onlyone-half of the A components would interact, since NO C2II approximates veryclosely to Hund's case (b), and the rate of radiationless to radiative depopulationof the C2II state would have a maximum value of 0-5.It is unlikely that the flat2Z+ curve crosses the C2rZ and R2Z+ curves near the minima of these curves a106 FLUORESCENCE OF NITRIC OXIDEshown by Young and Sharpless, since this would cause an observable broadeningof the rotational structure of the 6 and &-bands. Moreover, the &-bands have beenseen in emission with u’ up to 3.The observation that only some 3 % of the molecules in the C2ll(u = 0) stateradiate, carries with it the implication that C2ll(u = 0) interacts with a con-tinuum. The continuum is probably of a state which correlates with ground-stateatoms, though there are possible alternatives which are discussed below.It seemshighly unlikely that the reduction of the intensity of the 6-fluorescence to the degreeobserved could be due to a perturbation by a state that is itself only weakly pre-dissociated. Herzberg et al., following Sutcliffe and Walsh, have suggested that theperturbation of C2l-I by the B2l-I state may account for the break-off of the &bandsin emission; the B2ll state is supposed to dissociate via the 2C+ state, where a fullinteraction is possible because the former belongs to Hund’s case (a). However,rapid transitions could occur only between a restricted number of levels and sincethe predissociation of B2l-I by the 2C+ state is very weak, this explanation mustbe rejected.If the C2ll(u = 0) state is to interact with another state with A>O and whichcorrelates with ground-state atoms, the only possibilities are the a4II and the 6Hstates.The latter is continuously repulsive,8 and since the a4l-I state is reasonablywell defined, we conclude that the simplest and most reasonable explanation for theweak fluorescence of the (u’ = 0) 6-bands is to postulate a predissociation of theC2H(u = 0) state by the continuum of the a4l-I state. The rate of the predissociationis obtained by multiplying the spontaneous emission rate of the 6-bands by the ratioof the rate of predissociation to the rate of radiation.Hence, forNO C~II(U = 0) -[NO a4n] +N(4S) + O(3P)the first-order rate constant = 30 x 2.2 x 107 sec-1 = 6.6 x 108 sec-1.A detailed account of the quenching of the &-bands will be described elsewhere.It is important, however, to discuss briefly the fate of molecules in the D2Z+ state,because solution of this problem provides information concerning the characterof the state which predissociates NO C2ll(u = 0).In the past, the appearance ofthe (u’ = 1) &-bands in emission whilst the (u’ = 1) &bands have not been seen,has been particularly puzzling. Apparently the &-bands are quenched by added gasesbecause of collision-induced predissociation. Whereas NO CzII(v = 0) dissociatesat low pressures and can be deactivated by collision, NO D2C+(u = 0) radiates atlow pressures and is dissociated by collision. The &-bands are strongly quenchedby helium, argon and nitrogen, without population of radiative states lower in energythan D2Z+(u = 0).It is possible to develop a partial explanation for these ob-servations by means of the information given on fig. 1, an explanation which isconsistent with the assignment of an intermediate state with A>O. If two inter-acting states have AA = 1, a perturbation is said to be heterogeneous and thedegree of interaction increases with J, the rotational quantum number. Thus, ifa predissociation is both spin-forbidden and heterogeneous, it should have a lowerprobability in low rotational states than a predissociation that is homogeneous andspin-forbidden. Therefore the postulation of NO a4ll as the intermediate for dis-sociation of the C2II state is consistent with a comparatively weak interaction ofthe intermediate with the D2C+ state.Unfortunately the lack of informationabout this type of process makes it impossible to assess accurately the importanceof the change in A. The effect of the collision may be to increase the probabilityof the DzX+ - a4II radiationless transition. Alternatively collisions could inducA . B . CALLEAR A N D I . W. M. SMITH 107the transition D2E+-b4Z-, by breaking the AK = 0 rule; the quenching of theiodine fluorescence provides a similar example. The b4E- state would have to bestrongly predissociated by a4II in order to present a continuum to the D2X+ state.Thus, although the mechanism is not uniquely determined by the potential curvesof fig. 1, the difference in behaviour of the C2IT and D2Z+ states, at low pressure,is interpreted.Fig.8 illustrates the overlap of the continuum of the a4n state with C2II(u = 0)and CXI(u = 1). The position of the a4n state was calculated from the data ofVanderslice et al.8 and accounts without adjustment for the predissociation of the&bands. The overlap with C2ll(u = 0) is small, as it must be, compared to theoverlap with CXI(u = 1). The interaction of the a4n continuum with the BXIFIG. &-Overlap of vibrational eigenfunctions for the C217(u = 0 and l)-a4n transition. Theeigenfunction for the a417 continuum is approximate.(u = 7) state may also account for the break off of the /I-bands in emission. Theflat Z+ state may interact with the B2II state at u = 7, as suggested by Herzberget al.(though the lines are not broadened) and thus there may be two modes ofpreassociation of normal atoms into the B2II(u = 7) state. The apparent absenceof emission from B~II(v = 7) in Young and Sharpless’ work may be due to the smallemission rate of the P-bands compared with the &bands, coupled with the fact thatthe strongest (v’ = 7) P-bands either overlap with the much stronger &bands orare at wavelengths greater than 6000 A.Gaydon 19 favours another mechanism for predissociation of C2II(u = 0) andD2Z+, viz., that the C2II and D2X+ curves cross a 4A state, thus forbidding predis-sociation of the D2C+ state. The 4A state would dissociate via the a4IT state.Gaydon suggests that the change of A in the 2Z+--4II transition is insufficientrestriction to explain the difference in behaviour of the D2X+ and C2II states.A4A state cannot correlate with ground-state atoms and therefore it would have tobe strongly predissociated by the a4n state, in order to present a continuum to theC2ll(u = 0) state108 FLUORESCENCE OF NITRIC OXIDEDISSOCIATION ENERGY OF NITRIC OXIDEBy considering the percentage of NO C2II(u = 0) molecules which radiate,and the energy of the rotational levels in this state, an upper limit can be placedon the dissociation energy of nitric oxide. Certainly less than 10 % of these mole-cules radiate and this would correspond to the fraction of molecules in levels belowJ = 2+ in the higher spin-component. This level corresponds to an energy of6.495 eV.In fact, the ineffectiveness of both argon and helium in quenching NOC2II(u= 0) shows that predissociation is not limited by molecules in the lowerrotational levels possessing insufficient energy to dissociate. If this were so, addedgases would cause rotational transitions fast enough to be very effective quenchingagents. For this reason and because the &bands do not exhibit irregular intensityin emission, it is concluded that the dissociation energy lies below the energy of allthe rotational levels of CzII(u = 0). The lowest rotational level of C2II(v = 0)corresponds to an energy of 6.493 eV.The most accurate value available for D(N0) is that calculated from the spectro-scopically determined heats of dissociation of N2 and 0 2 , and the heat of formationof nitric oxide.20 This last quantity has not been directly determined for manyyears and the most accurate value is that determined indirectly from the heat of thereactionwhich leads to AHiNO(g) = 21.45, 0.07 kcal/mole,NOb) + CO(d = 3N2(g) + C02(g),D(0,) = 5.114, +Om002 eV21 and D(N,) = 9.758 fr0.0025 eV.D(N0) = *D(N,)+*D(O,)-AH; NO@)= 6.507 fr 0.006 eV.In the light of the results described in this section, it is concluded that this value isslightly high.Our own observations show that D(N0) is very close to 6.493 eV.QUENCHING OF THE NO &BANDSThe quenching half-pressures, obtained from the Stern-Volmer plots shown infig. 5, are given in table 6. It has been described how strong self-quenching ofthe y-bands causes the observed quenching half-pressure of an added gas to varyconsiderably with nitric oxide pressure.13 For the &bands, however, the nitrogenquenching half-pressure is independent of NO pressure from 2mm to 12mm.TABLE 6.-QUENCHING HALF-PRESSURES (m) FOR THE &BANDSN2 (+ - 700 mm argon)N2 (+ - 300 mm argon)NO > 15Ar >38041 mm (h8)30 mm (f5)Even allowing for the possibility that experimental scatter masks an actual variationof 16 mm in the nitrogen quenching half-pressure, the nitric oxide self-quenchinghalf-pressure must be at least 15 mm.Similarly, the insensitivity of the nitrogenquenching half-pressure to the argon pressure shows that the argon quenching half-pressure is at least 380 mm.The determination of the collisional efficiency of quenching depends on the life-time of the C2IT(u = 0) state.The mean radiative lifetime of this state is 4.6 x 10-8sec, ignoring radiation to AG+, but because of predissociation the mean lifetimeA. B. CALLEAR AND I . W. M. SMITH 109even in the absence of quenching, will be considerably shorter. The mean lifetimewith respect to radiation and predissociation can be estimated by multiplying theradiative lifetime by the fraction of molecules excited to this state which radiate.This results in a value of 1.5 x 10-9 sec. Assuming the collision diameter of NO C2IIto be equal to that of NOA2Z+, this lifetime was used to calculate the quenchingefficiencies per collision which are shown in table 7. The probability of NO C2II(v = 0) being quenched per collision with N2 is approximately unity.The shorterlifetime of NO C2II(u = 0) compared to NO A2Z+ is one cause of the differencein the extent of self-quenching for the two states.TABLE 7.4UENCHlNG PROBABILITIES PER COLLISION FOR THE &BANDSN Nr P* (-) Pnitric oxide 7 . 1 4 ~ 109 10.7 > 15 < 4.7nitrogen 7 . 3 6 ~ 109 11.1 22 3.1argon 6-55 x 109 9.8 > 380 < 0.21 atm pressure/sec.1 atm pressure, per mean lifetime at zero pressure ;argon pressure ;N, the number of collisions made by NO C2II(u = 0) with the molecules of a gas atN,, the number of collisions made by NO C2rI(u = 0) with the molecules of a gas atp+, the quenching half-pressure (mm); the N2 half-pressure has been feduced to zeroP, the probability of quenching NO C2II(u = 0) per collision = 760/iVrp+MECHANISM OF QUENCHING OF NO c2n(V = 0) BY N2The relative vibrational yields in A2C+ were shown to be independent of thepressure of nitrogen, thus demonstrating that deactivation occurs exclusively fromNO CzII(u = 0) and not from NO C2II(v = 1). Because of its shorter lifetime,we should expect the latter state to have a larger N2 quenching half-pressure thanthe former.Thus the independence of the ratio of the intensities in the two y-progressions, shows that NO C2II(u = 1) is too short-lived to be deactivated underthe conditions described.The most obvious quenching reaction isNO C211(v = O)+N, XIZZ-+NO A2Zf(v = 0 or 1)+N, XIZ,f.The experimentally determined yields in the various vibrational levels of NO AG+are slightly different from the calculated Franck-Condon factors given in table 1.It might therefore be concluded that the N-0 bond distance in the transition com-plex is little different from its value in NO A2Z+ or NO C2n; there is very littlechange in internuclear distance during the quenching process.It is possible, however, that quenching of NO C2II by N2 is not a single step,but consists of two energy transfer processes,followed byNO C211(v = O)+N2 XIZ,f+NO X211+N2 A3Cy+,N, A 3 C i + N 0 X211(v = O)+NO A2C,'(u = 0 or 1)+N2 X'C,,f.Young and Sharpless observed that addition of N2, in their experiments, gave riseto the two progressions of the y-bands and in addition the (v' = 0) P-progressionwas excited.Setser and Thrush 22 have recently described chemiluminescenceresulting from addition of NO to a mixture of atomic oxygen and cyanogen.Agai110 FLUORESCENCE OF NITRIC OXIDEthe two progressions of the y-bands were observed, together with a weak (u' = 0)p-emission. It is possible that NOAzX+ is excited by the same mechanism in allthree types of experiment, and the only possible process would be energy transferfrom Nz A-7Z;.However, we have searched in vain for the (u' = 0) P-progression in the fluor-escence experiments. Simultaneous population of A2C+(u = 0) and B2II(u = 0) bycollision of N2 A3X: with NO P I T , would provide a very unusual type of process,considering the large difference between the internuclear distances of B2ll and A2C+.Until a full interpretation of these observations has been developed, it is not possibleto draw any conclusions from the fluorescence expzriments.Quenching of C2lJ(u = 0) will be studied in the presence of other added gases, and this may providenew evidence about the mechanism. If the second mechanism is correct, it maybe possible to discover a substance which will intercept and deslctivate N2A3Clfbefore it collides with NO X2II. Our rate constants are neither dependent on theintermediate state of predissociation nor on the mechanism of quenching, providedthat one NO A2X+ is produced for each NO C2II(u = 0) quenched.CONCLUSIONSThe fate of NO C2II(u = 0) can be described by the schemekz kc(Nz)N(4S)+ O ( 3 P ) ~ [ N 0 a411];.N0 C211(v = 0)-[N, A3Z,f]-+N0 A2Z+NO X211 NO A%+kt / k F \GThe ratio of kl to k2 is the equilibrium constant which is 1.26 x 1024 molecule cm-3,if AE," is taken to be zero (this introduces a possible error of +lo %).Thus k2is determined.At 293°K:k, = 9.7 x 10-10 cm3 molecule-1 sec-1;kl = 6.6 x 108 sec-1;kF = 2.2 x 107 sec-1;kZ = 5.2 x 10-16 cm3 molecule-1 sec-1.kk is not yet known.Young and Sharpless measured the rate of fluorescence of NO C2II(u = 0) toNO PIX, as a function of the N and 0 atom concentrations. Their second-orderrate constant is 1.1 x 10-17 cn13 molecule-1 sec-1. This is equivalent to our kFkz/(kl + kF> = 1.7 x 10-17 cm3 molecule-1 sec-1. There is good qualitative and quanti-tative agreement in the interpretation of quite different experiments. However,kl%kp and consequently this overall rate is very insensitive to kl, but is largelydetermined by the equilibrium constant and kF, the rate at which the &bands areemitted.The photolysis of nitric oxide and iodine provide the simplest known examplesof photochemical decomposition of diatomic molecules resulting from absorptionof light by bands that are not broadened by predissociation. An understandingof these systems may facilitate similar photochemical studies on triatomic or evenmore complex species.The authors would like to acknowledge a brief but valuable discussion withProf. Longuet-Higgins, on the effect of changes in A on the rate of predissociation.They would like to express their gratitude to Imperial Chemical Industries for theloan of a spectrograph. I. W. M. S . is indebted to D.S.I.R. for a studentship, duringthe tenure of which part of this research was carried outA. B. CALLEAR AND I. W. M. SMITH 1111 Jenkins, Barton and Mulliken, Physic. Rev., 1927, 30, 150.2 Gaydon, Proc. Physic. SOC., 1944, 56,95, 160.3 Sutcliffe and Walsh, Proc. Physic. Soc. A, 1953, 66, 209.4 Brook and Kaplan, Physic. Rev., 1954, 96, 1540.5 Henberg, Lagerqvist and Miescher, Can. J. Physics, 1956, 39, 622.6 Lagerqvist and Miescher, Helu. Phys. Acta, 1958, 31, 221.7 Barrow and Miescher, Proc. Physic. SOC. A, 1957,70,219.8 Vanderslice, Mason and Maisch, J. Chem. Physics, 1959, 31, 738.9 Ogawa, Science of Light (Tokyo), 1954,3, 39.10 Flory and Johnston, J. Amer. Chem. SOC., 1935,57,2641.11 Flory and Johnston, J. Chem. Physics, 1946, 14, 212.12 Young and Sharpless, Disc. Faraday SOC., 1962,33,228.13 Callear and Smith, Trans. Farday SOC., 1963,59, 1720.14 Callear and Smith, Trans. Farday SOC., 1963, 59, 1735.15 Bethke, J. Chem. Physics, 1959, 31, 662.16Bates, Mon. Notes Roy. Astron. SOC., 1952, 112, 614.17 Herzberg, Molecular Spectra and Molecular Structure: I, Spectra of Diatomic Moleciiles18 Callear and Smith, to be published.19 Gaydon, private communication.20 Evans, private communication, quoting results from a Ph.D. Dissert. of Frisch.21 Brix and Henberg, J. Chem. Physics, 1954, 21, 2240 ; Can. J. Physics, 1954, 32, 110.22 Setser and Thrush, Nature, 1963,200,864.@. Van Nostrand Co., New York, 1950)