Embed Size (px)
Citation preview
SCIENCES INC ANDOVER MA L G PIPER ET AL JAN SBPSI -076/TR-593 AFWL-TR-86-95 F29681-84-C-876
UNCLASSIFIED F/G 7/2 UE ENEE; C H E I FE E EIEiiEEEEEEEEElliEBBhEEEBhEEEEEE
EE~hEEEEEE~IElhlhlhhhElhhEEhEEEEEElhlhEEEhEEEEEEE~hlhI
Eu...- --
100uI - .1 :
1-fl2 1 6
• • • • •(( •12 51.4 * .,4..
AFWL-TR-86-95 AFWL-TR-86-95
III LE Opy
CONAN: CHEMISTRY OF NITROGEN-ASNASCENCE
L. G. Piper, et al
Physical Sciences Inc.I Research Park
Andover, MA 01810
MAR 29 1988
Jan uary 1 988
i'*- Final Report
Approved for public release; distribution unlimited.
AIR FORCE WEAPONS LABORATORYAir Force Systems CommandKirtland Air Force Base, NM 87117-6008
88 3 29 46,~ ~~~U W6- ...- " I'Ii
AFWL-TR-86-95
This final report was prepared by Physical Sciences, Inc, Andover,Massachusetts under Contract F29601-84-C-0076, Job Order 33261W14 with theAir Force Weapons Laboratory, Kirtland Air Force Base, New Mexico. CaptainBrian D. McFeeters (ARBI) was the Laboratory Project Officer-in-Charge.
When Government drawings, specifications, or other data are used for anypurpose other than in connection with a definitely Government-related procure-ment, the United States Government incurs no responsibility or any obligationwhatsoever. The fact that the Government may have formulated or in any waysupplied the said drawings, specifications, or other data, is not to beregarded by implication, or otherwise in any manner construed as licensingthe holder or any other person or corporation; or as conveying any rights orpermission to manufacture, use or sell any patented invention that may in anyway be related thereto.
This report has been authored by an employee of the United StatesGovernment. Accordingly, the United States Government retains a nonexclusive,royalty-free license to publish or otherwise reproduce the material containedherein, or allow others to do so, for United States Government purposes.
This report has been reviewed by the Public Affairs Office and isreleasable to the National Technical Information Service (NTIS). At NTIS,it will be available to the general public, including foreign nations.
If your address has changed, if you wish to be removed from our mailinglist, or if your organization no longer employs the addressee, please notifyAFWL/ARBI, Kirtland AFB, NM 87117-6008 to help us maintain a current mailinglist.
This technical report has been reviewed and is approved for publication.
BRIAN D. McFEETERSCaptain, USAFProject Officer
FOR THE COMMANDER
GERALD A. HASEN HAR0 ACKERMANNMajor, USAF Lt Col, USAFChief, Advanced Chemical Laser Branch Chief, Laser Science and Technology
Office
DO NOT RETURN COPIES OF THIS REPORT UNLESS CONTRACTUAL OBLIGATIONS OR NOTICEON A SPECIFIC DOCUMENT REQUIRES THAT IT BE RETURNED.
AF9L-TR-86-95ECt*;RIfl C.A4S5,;C.7ON OF 71 ;Ai~
REPORT DOCUMENTATION PAGEla. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS
UNCLASSIFIEDZa. SECURITY CLASSIFICATION AUTHORITY 3. DISTRIBUTION /AVAILABILITY OF REPORT
S D F D R CApproved for public release; distributim2b. DECLASSIFI TION /DOWNGRADING SCHEDULE unl imi ted.
4. PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)
PSI-076/TR-593 AFWL-TR-86-956a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
I (If applicable)
Physical Sciences, Inc Air Force Weapons Laboratory6c. ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (City, State, and ZIP Code)
P.O. Box 3100, Research ParkAndover, MA 01810 Kirtland Air Force Base, NM 87117-6008
8.. NAME OF FUNDING/ SPONSORING I 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION I(if applicable)
I F29601 -84-C-0076Sc. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS
PROGRAM IPROJECT I TASK WORK UNITELEMENT NO. NO. No. ACESSION NO
*62601F 3326 [N1W14
11. TITLE (Include Security Classfication)
CONAN: CHEMISTRY OF NITROGEN-A NASCENCE12. PERSONAL AUTHOR(S) Piper, L.G.; Davis, S.J.; Murphy, H.C.; Cunmings W.C.; Walkauskas, L.2.;
DeFaccio, M.A.; Cowles. L.M.: Rawlins, W.T.; Marinelli. W.J., and Green. B.D.13a. TYPE OF REPORT 13b. TIME COVERED 114. DATE OF REPORT (Year, Month, Day) IS.. PAGE COUNT
Final FROM4 fl4ToA 61 1988 January I 19216. SUPPLEMEN"ARY NOTATION
17 COSATI CODES 18 SUBJECT TERMS (Continue on reverse if necessary and ident f by block num)er) .
_,ELD GROUP SUB-GROUP Active nitrogen, Energy transfer, 7N2(A) NF(a)Nitrogen metastables, Quenching rates, \N2(B) NF(b)Energy pooling, Excitation rates,-'NF2 (over)
19 ABSTRACT (Continue on reverse if necessary ad ikenrify by block number) 7-Rate coefficients for production of excited nitrogen have been measured, and energyt-dransferreactions utilizing the excited nitrogen have been studied Three reactions were involvedin chemical production of excited nitrogen (N*) when H and)JF were mixed:---d
HNF - HF + NF (1L =, _- '()
H + NF (ala) .HF + N (2D) (2)
H (2D) + NF (ala) N2* + F (3)
Reaction (2) is slow, k2 =(3.1 ± 0.6) I0" cm molecule" sec". Reaction (3) is essen-tially gas kinetic, k, = (1.6 0.7) * 10"1cir molecule'-sec'1, and produces mostly_ .(Bl-). The N 2(A-zu +)which is observed is probably produced from radiative cascade (over).
,z0 -IS-OIBUT'ON , AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION3,.-jC-AS,3I'EUNLIMITED r0 SAME AS RPT ODTiC USERS UNCLASSIFIED
'2a ';-IE OF PESPONSIBLE ,NOIVIDUAL 22b "ELEPUONI (Include Areoa COde) 22c CFr'CE i MBOL
Brian 0. McFeeterS. Cantain. ,,, (505) 844-0661 ARBI00 FORM 1473, aM VAR 83 APR edition may be i$ed until eiiausted SEC,;iTY CLASSIFfCATION OF -IS PACE
All other e tioln% re obsolete UNCLASSIFIED
UNCLASSIFIED.
UNCLASSIFIEDT 2 C.Ia.P CLAePCATsa0 7040S PAGE
18. SUBJECT tERMS (Continued)
Nitrogen atomsHydrogen atoms,Iodine monofluoride' ___
19. ABSTRACT (Continued)
from N2(B). The major impediment to production of large number densities of N2(A,B) usingthis system is the rapid quenching of N2(A) by atomic hydrogen.
Energy transfer reactions from N2(A) to potential laser species revealed that NO QuenchesN2(A) with a rate coefficient of approximately 7 * 10-11 cm3molecule-'sec" I, efficientlyproducing NO(A 2w, v'=0,1). When IF is added to a flow of active nitrogen three differentband systems are excited, one of which is IF (B 0+). The other two IF systems remain tobe identified. The excitation rate of these systems, if N (B) is the exciting species, is anorder of magnitude faster than gas kinetic. Therefore, N2 B) is thought to be only a tracerof the actual species exciting IF; the most likely source is excited ground state nitrogen.The quenching of NI(B) by molecular nitrogen is quite rapid, with a rate of approximately2 * 10-Ilcm3molecule-'sec1. Argon does not quench N2(B) electronically, but strongly redis-tributes vibrational energy to lower vibrational levels.
Energy pooling of two N2(A) molecules proceeds with a rate coefficient of aporoximately3 * 1O cm'molecules-'sec-1, when both molecules are in v'=O, and about 50 percent fasterwhen one molecule is in v,=l. In both cases, the products are N2 (B), N2(C
3 r) and bands ofthe Herman infrared system.
Ii UNCLASSIFIED'A C OP -OM"S"VI"01
SUMMARY
This report summarizes work aimed toward :nderstanding chemical pathwaysfor producing excited nitrogen molecules, and energy-4 ransfer reactions of theexcited nitrogen molecules which might prove important in Potential laserdevices. The main thrust of the effort on chemical production of excited
nitrogen involved detailed studies of several :;pcrtant reactions 4.1 thefollowing reaction sequence:
H + NF2 + HF N F(a 1 l) (1)
H + N(aIA) . HF - N(20) (2)
N(2D) + NF(al') - N2 F (3)
All observations lend supporm to this mechanism as being the route to akling
N2* when H and NF2 are mixed together. Measurements show that the rate coef-ficient for reaction 2 is relatively slow, k2 - (3.1 = 0.6) x 10-13 cm'
molecule - I s - . In contrast, the rate coefficient for the third reaction isfessentially gas kinetic, k 3 - (1.6 = 0.7) x 10- 10 cm3 molecule "I s "-. The
quoted rate coefficient for this reaction is actually the rate cofficierit forproducing N2(3 3 7g), v' - 1-12) rather than the total reaction rate coeffl-cient. Observations of other products, e.g., N2 (al.'g), indicate that thechannel which produces N,(3) is far and away the largest one. Some N,(A3".-u '
)
from the reaction is observed, as demonstrated by observing Vegard-Kaplanemission in the ultraviolet, but these observations indicate that the ".(A)most likely is produced by radiative cascade from N2 (3). This ;oint rte.ainsto be clarified.
The species NF(bl_" ) and N(4S) were not significant precursors to :2(3)formation. The population of NF'o) was too small to affect any obser'ations,while N( 4 S) was formed subsequent to N2(3) formation.
Reaction 3 appears to be a rapid and efficient sourze of N2(3), and onemight indeed be able to generate large number densities of "12'A,B), if one :Inproduce large number densities of NF(a) and :(2J). The najor impedinent weforesee in using this scheme to produce large number fens.:ies of N;-'A,5) ;sthat atomic hydrogen appeared to quench 12 (A) rather efficiently. Thus onewould have to adjust conditions to minimize the existence of free hydrogen
a toms.
Observations on N2 energy transfer reactions focussed upon the transfer
of electronic energy between N2 (A32u*) and NO, upon the products of N2(A)
energy pooling, upon the excitation of IF by active nitrogen, and the charac-
terization of vibrational* excited N42(X,v) in active nitrogien. We demon-strated that N2 (A) excites NO(A) very efficiently (k 7 x I " 1 1 cm3
molecule-1 3-1). The efficiency of tne reaction, f-artnernore, does not
degrade as the pressure in the reactor increases. Spectroscopic observati.,nsof 'JO(A) emission indicate a significant variatio.n in the transition :toment o
the NCO(A 2 1---X2) trans-'tion with r-:entroid. :he effect is such as to;ncrease the relative ;.nnortance of transitions witn larger jv, implyin: :[a
laser gain for the red shifted transitions 0,2-0,5 will be somewhat largerthan anticipated. This increases t.'e probability of adequate gain on taesetransitions fczr a laser *ev-4ce. Finally, it has been determined that theradiative lifetime of Nn'A) is about 30 percent longer than previously!though t.
Studies on N-61) energy pooling determined the production rates cf vari-ous vibrational Ieel of .&C~) N3 3 ), and the Herman infrared systemas a function of vibrational level of the N2(A) molecules involved in thepooling process. R~esults indicate a total1 rate coefficient for the energ:ypooling process of about 4.0 x 10-10 cm2 molecule-1 s:1 when we use the NZBA)Einstein coefficients of Shemansky. if the hypothesis is correct that theseEinstein coefficients are abcut 30 percent too large, then the actual value ofthe energy pooling rate coefficient will be roughly 2.5 x 10-10 cm3
molecule 1I 3-. -learly, the issue of the !42 (A) lifetime m-ust be resolved toclear up this point.
Observations both in the Nl2(B) -formnation from reaction 3 as well as thatfrom energy pooling shows fairly dramatic changes in the vibrational leveldistribution of the X2ft3) a3 a function of argon pressure, even for pressuresbelow 3 torr. M4olecular nitrogen, on the other hand, appears to quench the
* electronic energy in the 11,(3) quite efficiently,k 2x10lcm
moleculae1 s-1, out does not change the vibrational level distributionstrongly. These issues of N2 (B) quenching and vibrational redistribution aravital to modeling potential laser operation. Clearly, if the end result ofN2(3) quenching is formation of N-3 v' - ;),I)--levels we could observeeither poorly or not at all--or N2'(A), then apparent quenching of the hiighervibrational levels of N-3. robably; will not affect efficiencies as thesystem is scalea to higher pressures of argon or nitrogen. Similarly energypooling losses will be reduced in that energy-pooling reBactions which poduce
;2 or N-IS) ultimate>y regenerate N-C(A). Thus the effectJie loss is onlyone molecu e pe r pooling event ratner than two. if, however, N2 (3) quenchingproduces 3ro-urn-state nitzrogen, then at higher pressures, effective .12* fornia-tion from reaction 3 would be reduced and the pooling losses would double.
A diagnostic was devel.oed for N-AX,';) in active nitrogen which is basedupon the Penning ionization reaction between metastable helium atoms and
* molecular nitrogen. This reaction produces N2 "*(B2 ~ stt in Franck-Condontranlsitions from the gro:und state. The resultant vibrational distrinbution of
the _l -u : ) emission, therefore, mirrors the vibrational distribu-tion ofthe N2(X,v). We have also demonstrated the production ofNi 2 )in Penning ionization of metastable helium with *N2 (X,v). This indicates thepresence of vibrational excitation in excess of 3.3 eV in the ground-statenitrogen.
:t has been shown that three different band systems of 17 are excitedwhen :F is added to a flow of active nitrogen. One or tnese systens is thewell-known 'F(23~-X ) system, while we have been unable to identifytne other two zysiams as ;fet. The exzitation rates of all three systdeus
etv
varies in pr4:pcrti4on to the number density of '42(3) in the reactor. Theeffective exci:ation racca coefficient implied by the data however, is an orderof magnitude greater than gas kinetic-. Thus the N.)(3) is only a tracer of theaactual species resconsibi--e for the excitation. We have, so far, bzeenunab-le to ident'fy the excitation source, but think that the most likely
- -.hypothesis is "2(X ,v).
v/v ±
r- VP 1 1 1 -1
1 111' R P I I
ACKVOWLZCGE4 ENT
Partial fundinig for this wor'k has also been received from the Air F'orceGeophysics L~boratorias under ':ontract No. F19628-62-C-0050 and from SODJ/13Tand Los Alamos National :.aoratories. This extra support allowed a moredetailed and comnplete stuidy.
v~HMO=i
CONTENTS
Paragraph 1 IN:RO.UCT ION ................... I2 H - NF2 REACTION SE.2UENCE STUDIES ........ 3
2.1 Introduction . . . . . . . . . . . . . . . . . 32.2 H - NF2 chemiluminescence studies ... ....... 42.2.1 Flow tube description ..... ............. 42.2.2 Visible and near infrared studies ... ....... 72.2.3 Ultraviolet emission studies ..... .......... 92.2.4 Phenomenological kinetic investigations .... 12
2.3 Kinetic rate coefficient determinations .... 132.3.1 Apparatus .. ........ .............2.3.2 Rate constant for H + NF(a) .. .......... 132.3.2.1 H atom calibration procedure ............ . 172.4 Production rate for N2(B) excitation ...... . 242.4.1. Background ...... .................. 242.4.2 Absolute photon emission rate measurements. 252.4.2.1 General procedure .... ............... 252.4.2.2 Air afterglow calibration procedure ...... 23
2.4.3 N(2 D) resonance lamp calibration ....... . 30
2.4.3.1 Background ...... .................. 302.4.3.2 Review of resonance fluorescence ano
absorption . .
2.4.3.3 Measurement technique for N2 (3) excitationrate. .................... .28
2.4.3.4 Viorational relaxation in N;2 (3) ............ 472.4.3.5 Otner observations ...... .............. 482.4.3.; Modeling results 513 N -; .GY-TRANSFER. STUDIES ............ 57
3.1 State-to-state energy transfer fromN 2 (A3Z
, v' - 0,1,2) to
N( A 2:+ , v' = 0,1,2) ..... ............ 53
3.1.1 Introduction . . . . . . . . . . . . . . . . .3.1.2 Experimental ...... ................. 593.1.3 The electronic transition moment for the
- ..) transition. . ............ 63.1.4 The kinetics of the N2 (A) plus NO reaction 663.1.4.1 The quenching of "2(AX5, + , v' = 0) by NO 703.1.4.2 The excitation of NO(A , v' - 0) oy
N2(A3. , -+' -0) ..................... 743.1.4.3 State-to-state excitation of NO(A, v' - 0,1,2)
by N2 (A, v' - 0,1,2) ... ............ 753.1.4.4 Discussion ...... .................. 813.1.- Summary. ......... ....................
CONTENTS (Continued)
Paragrapn 3.2 N2(-t u ) energy pooling ........... 8
3.2.1 Introduction . . . . . . . . . . . . . . . . .3.2.2 Experimental ...... ................. 903.2.3 N2 (C3-uE v' "0-4) formation in
2 (A v' = 0-2) energy pooling .....
3.2.4 HIR formation from N2 (A) energy pooling . . .. 99
3.2.5 Formation of N2 (B3g, v' - 1-12) in
energy pooling .... ............... 107
3.2.6 Discussion of energy-pooling results3.3 Active nitrogen excitation of IF ....... . 1253.3.1 Introduction ...... ................. 1253.3.2 Development of N2 (v) source and diagnostic . 126
O 3.3.2.1 introduction ...... ................. 1263.3.2.2 Experimental ...... ................. 1263.3.2.3 Theory behind the Penning-ionization
measurements ..... ................ 1293.3.2.4 Results of tie Penning-ionization
measurements ..... ................ 1333.3.2.5 Observations of nitrogen first-positive
emission . . ................ 1423.3.3 The excitation of IF by active nitrogen . . . 1473.3.3.1 Spectroscopy of active N, plus IF ....... 1473.3.3.2 IF* rates in active N2 . ...... 153
..F....NC.. . .................... 153
9
FIGURES
7-igure Pg
1 Block diagram of the CONAN flow tube .......... 52 Detailed view of observation chamber ... .......... 53 Chemiluminescence in the 500 to 900 rm region produced
by the reaction of H2 with discharged NF3 . . . . . .. . 74 Dependence of NF(a-X), NF(b+X), and N2 (A) emissions as
a function of H2 flow ...... .............. . 85 Mi (c~a) and (A-X) systems produced in the H2+
discharged NF3 reaction ..... ............... . 106 NO(A-X) and N2fA+X) system produced by reacting H2 +
discharged .TF3 ....... .................... ... 107 Emission spectrum of the N2 a
1!g X1 g system .... 11
8 Temporal profiles of NF(a), N2 (A), and N2 (3) ...... ... 12
9 Block diagram of CONAN flow tube ... ............ ... 1410 Relative concentration of NF(a) as a function of
F 2 Flow for three different NF3 flows:
a) CNF 3] = 6 -imoles-z- 1; b) [NF 3 3 = 2.7 -gmoles-s-1;
c) [NF 3J 1.25 -moles-s - . . . . . . . . . .. . .. . . . . . 1511 Scans of NF(aX) emission showing removal by hydrogen
atoms: a) H2 discharge off; b) H2 discharge on . . .. 1612 Typical data for H + NOCl titration ... .......... . 1913 Plot of !HNO (proportional to H) as a function of
microwave power ...................... 20
4 .No as a function of [ 2]added for constant 'N]0 andconstant microwave discharge power .. .......... ... 21
15 :jN0 versus [NOadded for constant [H] .. ......... ... 2216 Plot of Zn(NF(a)] as a function of added [.H ....... 217 Plot showing 0 + NO titration procedure ........ 2918 Details of the resonance fluorescence/absorption lamps
used for N(2 D) concentration measurement ........ . 31
19 Energy level structure of N ..... .............. . 32203 Plot of N(20) resonance fluorescence signal 1F
versus Zn(i/!-A,. ......... ................... 3821 Chemiluminescence spectrum (thin line) showing
NF(b X), N2 (3+A), NF(aX), and HFt(3+0) systems . . 3922 Plot showing linear dependence of (Ni(B)] with
respect to the product [NF(a)]x1N( 0)] ... ........ 4123 Composite plot of [N2 (3)] as a function of the product
[NF(a)] x [N(2D)] for all data collected ........ . 4524 Populations of N2 (B;v') as a function of v' for
two Ar presaures ....... ................... ... 4625 Dependence of measured excitation rate coefficient, k, as a
function of v' for two Ar pressures ... ......... 48
i :( i
FIGURES (Continued)
Figure Paae
26 Plot of £'NF(a)J LN2(3), and :N 2. ja functionsof CH21 added ...................... 49
27 Pl~ot of L'NF(a)], (Nn(Bfl, (N(20)1 and relativeCN2Ca)] as functions of CH2] added. ........... 50
28 Comparison of predicted and measured populationprofiles for NF(.a), N(2 0), and N2(3) .. ........
29 Temporal profiles of N2(B) and N(4S) produced in
the H +NF 2 reaction sequence .............. 5630 Cheoiluminescence spectrum (dark line) and spectral
fit (light line) produced from the H +NF2reaction sequence .................... 56
31 Flow tube apparatus configured for N2 (A) decaykinetics measurements .................. 59
32 Vegard-Kaplan emission in flow reactor 9 :as downstreamfrom the discharge. ................... 61
33 Variation in electronic transition moment withr-centroid for th OA~ -
2 ) transition. ...... 6534 Experimental (light line) and synthetic (heavy line)
spectrum for the NO y-bands assuming a constantelectronic transition moment. ............... 9
35 Experimental (light line) and synthetic (heavy line)spectrum of the NO -f-bands using the electronictransition-r-oment function determined in this 4ork 69
36 Decay in the natural log of the N2(A, v' - 0) numberdensity as a function of NO number density atthree different reagent mixing distances. ........ 73
37 N2(A, v' - 3) decay constants in NO -as a function ofreaction tine . ..................... 73
38 Variation in the peak intensity of the NO(A-X, 0,1)'iand as a fuiction of added NO number density . ... 76
39 Ratio in the rate coefficients for excitingNO(A, v' - 1) to NO(A, v' = 0) by N2(A, V' - 3)as a function of argon pressure ............. 76
40 Spectrum of NO(A-X) and N2(A-X) in the absenceof CF4 . . . . . . . . . . . . . . . . . . . . . . . . . . 77
41 Spectrum of NO(A-X) and N2(A-X) in presence ofCF4 which relaxes most of the M2(A) vibrational
4Penergy .. ........................ 7742 Excitation of NO(A, v' - 0) as a function of the
ra3tio of '42(A, v' - 1) to NnCA, v' = 0) ......... 7943 Exci.tation of NJO(A, v1 - 1) as a function of the
ratio of :4-(A, v' - 1) to N-)(A, v' - 3) ............ 7944 Excitation of NJO(A, v' - 0 ) as a function of tne
ratio of N-'(A, v' = Z) to N2(A, v' '0 ... .. .. . ...
xi.
9t 105 MU2,11 'l'
A FIGURES (Continued)
Fiure Page
45 Variation in the excitation of NO(A, v1 = ai a
function of the ratio of N2 (A, v' = 2) to'42(A , v = 0) ....... .................... . 80
46 Electronic transition moments for N2 (33
- A3, ) . . . 8347 Variation of electronic transition moment for
N2(A3.u+. Xlg ..+ ) ...... ..................... 8548 Cross-sectional view of flow tube illustrating the
geometry germane to the radial number-densitygradient problem ...... ................... .. 92
49 Variation in the correction factor for energy
pooling measurements with the ratio h/r ....... .. 9350 Observed (heavy line) and computed best fit
(light line) to the spectral region between250 and 400 nm ........ ...................... 95
51 Observed (heavy line) and computed best fit(light line) to the spectral region between220 and 400 nm ............... ..... 95
52 Variation in the number density Of NL2CC, v' =0)as a function of the square of the numberdensity of N2 (A, v1 - 0) ..... ............... ... 96
53 Variation in the number density of N2 (C, v' = 1)
as a function of the square of the numberdensity of N1(A, v, = 0) ..... .............. ... 96
54 Variation in the number density of N2 (C, v' = 2)as a function of the square of the numberdensity of ";2 (A, v' = 3) ..... ............... ... 97
a 5Variation in the number density of N 2(C, v1 - 3)as a function of the square of the numberdensity of N2 (A, v' = 3) .... ............... .... 97
_56 Variation in the number density of N2 (C, v' = 4)as a function of the square of the numberdensity of N2 (A, v' = 0) ..... ............... ... 98
57 Variation in the ratio of the number densityof N2 (C, v' = 0) to the square of the numberdensity of N2 (A, v' = 0) as a function of the
ratio of the number densities of vibrationallyexcited to unexcited N2 (A) .... .............. ... 100
' 53 Variation in the ratio of the number densityof N2-(C, v' - 1) to the square of the numberdensity of N,(A, v' - 0) as a function of the
ratio of the number densities of vibrationallyexcited to unexcited N2 (A) ........ .............. 00
i Xiii
FIGURES (Continued)
Figure Pag_
59 Variation in the ratio of the number density of
'N2 (C, v' - 2) to the square of the number density of42 (A, vi - 0) as a function of the ratio of the numberdensities of vibrationally excited to unexcited42 (A) ......... ........................ . 101
60 Variation in the ratio of the number density of N'2 (C,
vI - 3) to the square of the number density of N2 (A,
v' - 0) as a function of the ratio of the numberdensities of vibrationally excited to unexcitedN2 (A) ......... ........................ . 101
61 Variation in the ratio of the number density of N2(C,v1-4) to the square of the number density of "2 (A,
v'=0) as a function of the ratio of the numberdensities of vibrationaly excited to unexcited
4M r N2 (A) .......... ......................... ... 102
62 Spectrum of tne Herman infrared v'=3 system excited in
the energy pooling of N2 (A, v'=O). Ptotal - 7.5tor , 7L - 0.20 . . . . . . . . . . . . . . . ... 104Storr, 0.0..............................0
63 Spectrum of the Herman infrared v'=2,3 systems excited
in the energy pooling of N2 (A, v'=0,1). Ptotal
7.5 torr, XN2 = 0.20 ....... ................ .. 104
64 Variation in the number density of N2(HIR, v'=3) as a
function of the square of the number density ofv =0) . . . . . . . . . . . . . . . . . . . . . . . . . 105
65 Variation in the ratio of the N2 (HIR, v'-2) numberdensity to the product of the number densities of
,.-(A) v1=0 and v'=1 as a function of the ratio of
the number densities of N2 (A) v'-1 to v'-0 ....... ... 10666 Spectrum of the nitrogen Herman infrared v8 = 3 and
first-positive systems excited in the energy pooling
* of N2 (A, v' = 0) for a nitrogen partial pressure of0.46 torr ........ ....................... .... 108
67 Spectrum of the nitrogen Herman infrared v' = 3 and
first-positive systems excited in the energy poolingof N2 (A, v' - 0) for a nitrogen partial pressure of0.027 torr .......... ..................... 10.
* 6a Variation in the ratio of the square of the numberdensity of N2 (A, vI - 3) to that for N2(3, v' - 2)
as a function of the molecular nitrogen number
density .......... ...................... .. 10969 Variation in the ratio of the square of the number
density of N2 (A, v' - 0) to that for N2 (B, v ' -4
as a function of the molecular nitrogen number
ensity ......... ........................ .. 109
x:
%,
FIGURES (Continued)
Figure * Page
70 Variation in the ratio of the square of the number
density of N2 (A, v' = 0) to that for N2 (B, v' 9)as a function of the molecular nitrogen number
density ............. ........................ 11071 Variation in the ratio of the square of the number
density of N2 (A, v1 - 0) to that for N2 (3, v' = 10)
as a function of the nolecular nitrogen numberdensity ............ ........................ 110
72 Variation in N2(B.v) excited from energy pooing of, N2(A, v'=O) with neon pressure .... ............ 11
73 Variation in N2(B,A) spectrum produced in N2 (A) energypooling with changes in nitrogen and argon partial
pressures. ....................... 11374 Spectrum of the nitrogen Herman infrared v' = 2,3 and
first-positive systems excited in the energy poolingl.- of N2 (A,v' - 0,1) for a nitrogen partial pressure of
38 mtorr and a methane partial pressure of
1.4 mtorr ........... ....................... 11475 Spectrum of the nitrogen Herman infrared v' = 2,3 and
first-positive systems excited in the energy poolingof N2 (A,v' - 0,1) for a nitrogen partial pressure of38 mtorr and a methane partial pressure of
4.9 mtorr ... .............................. 11476 Variation in the ratio of the number density of N2 (3,
vI = 2) to the square of the number density of N2 (A,
vI - 0) as a function of the ratio of N2 (A) vI = 1 tov'= 0 ......... ........................ ... 115
77 Variation in the ratio of the number density of N2(B,
v6 - 9) to the square of the number density of N2 (A,v' = 0) as a function of the ratio of N2 (A) vI - I tov' - 0 ............. ...................... 115
78 Variation in the ratio of the number density of N2 (3,v7 - 10) to the square of the number density of N 2(A,
vI = 0) as a function of the ratio of N2 (A) vI - 1 tov'= 0 ............ ........................ 116
79 Spectra of -IR v' - 3 and N2 first-positive systems when
the N2 (A) is formed by direct discharge of the N2with the argon and when it is formed in the
oconventional manner by adding the N2 downstream fromthe discharge ........ .................... . 117
80 Variation in N2 (3,A) spectr-m produced in N2 (A) energy
pooling with changes in nitrogen and argon partial.-resasures ........ .................... ... 124
81 Flow reactor for studies on the vibrational energy" con:ant of active ni:rogen and of IF excitation by
acti'e nitrogen ....... ................... 12i
Xv
FIGURES (Continued)
Figure *Page
82 Ratio of the Popoulations Of N12 +(3, v') to N2 -(B, v' 0cretedin e*Penning-ionization of L42(X, v") as a
function of the vibrational temperature of theground-state nitrogen .................. 131
83 Spectrum of the jv--2 sequence Of the N2 + (B27u+_X 2 E +)
system excited in the Penning-ionization ofnitrogen by He*(2 3S). ................. 134
84 Spectrum of the jv--2 sequence of the N2+( 2 u EX2 g)
* system excited in the Penning-ionization ofactive nitrogen by He *(2 3S)...................134
85 Spectra of the jv - -2 sequence of the N2+'(B2 - +_X2Zg+)
system excited in the Penning-ionization of activenitrogen by :.e*(2 3S) in the presence and absence ofSF6 . . . . . . . . . . . . . . . . . . . . . . ... . . . . 135
86 Comparison between experimental and calculatedvibrational distributions of N2+(B) created in thePenning- ionization of active nitrogen by metastablehelium atoms for a nitrogen mole fraction of 0.011
4(p - 1.5 torr, transit time from discharge - 11 ms). 13887 Comparison between experimental and 'calculated
vibrational distributions Of '42+(B) created in thePenniig-ionization of active nitrogen by =etastablehelium atoms for a nitrogen mole fraction of 0.346(P - 1.5 torr, transit time from discharge - 11 ins). 138
88 Effective vibrational temperature from thenonequiliz-riim model versus nitrogen mole fractionthrough the discharge .................. 139
89 Spectra from active nitrogen between 184 and 208 rim inthe absence and presence of metastable helium atoms. 140
90 First-posizivte spectrum, N2(B3 *-g 3,37. from activenitrogen in helium (p - 1.5 torr, XN2 - C011) . . . . 143
91 First-positive spectram resulting from N-atomrecombination in helium.....................143
92 Vibrational distribution of N-)(3".,) in activenitrogen for various attenuations by a Ni screendownstream from the discharge, but upstream from theobservation region. ................... 145
93 Variation in. N.)(3) number density with number densityof vibrationally-excited, grouna-electroni;.c-statenitro-4en .. ....................... 146
94 Variation in !12(3) number density with the signal froma photcnetar centered at 580 am....................146
xvi
FIGURES (Continued)
Spectrum -if active nitrogen between 450 and 780 rim Inthe absence and presence of IF. .............. 48
96 Spectrum between 480 and 680 rim from active nitrogenplus IF with tie Nn(B) spectral features subtractedout. .......................... 143
9e plot for :F system 1. ................. 14998 ue plot for IF system 2 .................. 14*999 Fitting of the !F(3 7 +--X 1 +) bands between 450 and
500 rim to the spectrum excited by adding IF toactive nitrogen ..................... 151
100 Fitting of the 17-(B370 -XI bad+ewe 0 n600 rim to the spectrum excited by adding ZF toactive nitrogen ..................... 151
101 Fittng of th IF3ho-X' bands between 450 and650 rim to the spectrum excited by adding IF toactive nitrogen ................ ........ 152
102 Vi'Crational distribution of IF(3 3'10;)*excited byactive nitrogen. .................... 152
103 Residual soectrum of active nitrogen plus I? afterspectral features beloni ig to the nitrogenfirst-positive an~d IF(Byg .V 4"'lI) systems have beensubtracted out. ..................... 154
I12.4 Spectruia of active nitrogen plus IF wita the nitrogenfirst-positive features subtracted out taken under=onditions which enhance IF system 1 rel.ative toother IF spectral features. ............... 154
10Z Spectrum of active nitrogen plus IF with the nitrogenfirst-positi.ve features subtracted out taken underzonditions which enhance the :F(S-X) system relativeto other IF spectral features .............. 155
106 Var-.ation In the intensity of the IF(B, 5,0) band as afunction of added IF number density for variousa umber densi.ties of N2(3) in the flow reactor .. . . 15;
107 Excitation rates of IF(B) versus N2(3) number density. 156108 intensities of the IF series 2 band at 636 rim as a
function of added IF number density ........... 157109 Intensities of IF series 1 and series 2 emaissions as a
function Of N2(3) number density for constant IFnumber density. ..................... 157
xvi i
TABLES
Table
1 Compilation of Conditions for leasurement ofH + NF(a) *Products Rate Coefficient.... .. .. ..... 3
2 Resonance Lines for Metastable Nitrogen Atoms ....... 373 Sample Data Showing 1XL2(B)3, CNF(a)] and
EN(2D)] as a Function of [H21 Added ........... 414 Mleasured k3 for v' - 1-12 at Various Bath
Gas Pressures. ..................... 435 Summary of Conditions for Measurements of N2(3)
Excitation Rate Coefficient ............... 456 Rate Package Used in "lodeling Study ............ 527 Einstein Coefficients for NO(A2E+ - X2 r) . ......... 67
State-to-State Ralative *Excitation-Rate Coefficients . 813 Revised Optical Gain Predictions for NO(A-X)
Transitions ....................... 68Rate Coefficients for N2(C nu) Formation from
%~N 2(AZu+) Energy Pooling .. .............. 10211 Rate Coefficients for Herman Infrared Formation
from N2(A) Energy Pooling ........ ........ 10612 Rate Coefficients for N2(B3n g) Formation from
N2(A3 ,u+) Energy Pool.ing .. .............. 116
13 Rate Coefficients for N42(3) Quenching by N2. . . . . . . . 11914 Franck-Condon Factors of N2+(B u, v) --~2(Xg, ' 131
Einstein Coefficients for N2+(B Zu - 1361S pectroscopic Parameters of IF and Similar Moleculez 15-1
Xvi ;,1
1. INTRODUCTION
The lowest-lying metastable sta:e of N2 , 3 is a well-known energy
reservoir (-6 eV) which has potential as an energy-storage candidate for use
in short-wavelength laser systems. As such, knowing the means to generate
this species efficiently and understanding its energy-transfer kinetics in the
lasina medium are essential requirements for laser-device design. This pro-
gram was designed to quantify the fundamental kinetics and mechanisms of
several key aspects of N2 (A) generation and utilization. These investigations
provide a fundamental data base which is a necessary first step in developing
*an N2 (A) transfer laser.
Several laboratory investigations have identified methods for forzing
N2 (A) chemically. Two methods, in particular (Refs. 1 through 5), may Ce
capable of producing N2 (A) in sufficient quantities to drive a transfer laser.
The first method consists of a set of reactions initiated by the reaction of
atomic hydrogen with NF2 , and appears to involve the metastables NF(aI ) and
N(2D) as key intermediate species (Refs. I through 3). An intriguing second
method is the bimolecular disproportionation of the highly energetic azide
radical, N3 (Refs. 4,5). The former method has a potentially high efficiency
per unit mass, but requires the handling of the hazardous species N2 F4 ktne
thermal source of NF2 ), and may present serious kinetic complications upon
scale-up from laboratory experiments. The latter method, which uses ionic
metal azides such as NaN 3 as thermal sources of N3 radicals, is also poten-
tially efficient per unit mass.
Additional aspects which govern the efficiency of potential chemical
laser devices are the efficiency of transferring energy from N2 (A) into the
lasing medium and the efficiency with which energy is drained from the N2 (A),
*£ both by other species in the medium as well as by itself. Several candidate
energy-transfer schemes have been identified whose kinetics and efficiencies
need to be quantified in the laboratory. Under previous sponsorship
(Refs. 6,7), PSI studied the energy transfer from N2 (A) to iF in detail. %ea-
surecients shown here demonstrited a ripid, efficient transfer which results in
fluorescence fr-cin t-e 3 -,,r XZ transition of 7Fcetween 500 and 600 n-m.
CAdditicaal measurements -_sing dischar~ed nitrogen showed that other excited
forms of nitrogen, perhaps vibrationall.y excited N42, also contribute to :F
excitation. it is conceivable that the '42(A) generation schemes will also
produce such excited N-) spec_,es.) 'Durinq the course of this dcrlk, another
excitation process was studied which appears to have potential in a laser
device: the excitation of N0(A n-) bY "i2(A). In addition the kinetics of
N2(A) energy pooling was studied in great detail. The efficiency of the
energy pooling reactions govern the maximu.m X-) excited-state number densities
available in the laser medium. Finally, the preliminary measurements on the
excitation of ::' ty active nitr.ogen were extended.
All experimental measurements were carried out on discharge-flow reac-
tors, one of which .,,z developed specially for this program. Introducing
selected chemical reagients and excited species into the main reactor isolated
individual elementary reactions for unambiguous kinetic measurements. A vari-
ety of well-characterized ultraviolet, visible, and infraired spectroscopic
techniques identified the species present, determined their -lumber densities,.
and allowed their keinetic behavior to be quantified. All diagnostics were
calibrated using well-established, but ntot always well understood, methods.
Section 2 tetails the k-inetics oZ the H 4- N,, 4, + NF(a L), and
(2)+NF(al-) reaections. tIxtens.".e measurements on metastable nitrogen
energy-transfer kinetics comprise Seto .This section includes studies on
the N2 1(A) + NO reaction, the N,1 A) energy-pooling reactions, and the active
* nitrogen plus I.- reaction in addition to the development and implementation of
a diagnostic for N;2 (X,v) in active nitrogen. Section 4 summarizes findings.
This report -does not have a separate experimental section. Rather, the
apparatus and techniques germane to each set of experimental measurements is
described along with the :n.eisurements.
2. H NF2 REACTION SEQUENCE STUDIES
-. 1 :NTROCUCTIQN
: his Task was concerned with excited molecular nitrogen production. frz-m
the H + NFT reaction sequence. This reaction has been the subLject of several
previous studies and its history in the literature is intriguing. The first
report was by Clyne and White in 1970 (Ref. 8). They observed that the reac-
tion of H -NF2 produced NF(ak A), NF(b 1Z), and N2( 3 g9), and proposed that
recombination of N(4S) atoms was the source Of N2(B)-
in 1973 Herbelin and Cohen (Ref. 9) performed a similar chemiluminescence
study and suggested the following mechanism
H + NF2 - NF(a) + HF 1i)
H + NF(a) +N( 2D) + HF (2)
N('D) + NF(a) .N 2(3) +F (3)
Although Wey could not prove this model they presented indirect evidtence
for its validity and argued that spin and angular momentum conservation wonli
be =aior constraints in reaction product channel availability. in. particular
thiey emphasized that the reaction of H NFa would produce :;1,20D) exclusively:
if these correlation rules rigorously held. In spite of these early coserva-
tions the reaction mech~anism for this potentially important source of N-b' has
remained unclear.
In the early 1980's Clyne and coworkers embarked on a series of detailed
experiments designed to clarify this interesting reaction sequence
(Refs. 1,2,3). Usin4 sensitive diagnostic techniques, they observed that
N(20) was indeed the primary product of reaction (2) and concluded that
Hertelin and Conen's proposed mechanism was probably correct. In addizion
tn.ey made :prelinL nary estimates for the rate coefficients of reactions -,2)
3
now
k, = 2.5 x I0- 13 cm3 -molecu.le - I -- I
k 3 = 7 x 10 - 1 1 cm3-molecule - l S-1
The purpose of this Task was to measure rate coefficients for c-act:o.ns Z
and 3 and to study the efficiency of N2 * production from the sequence. =y
deter-mining these rate coefficients systematically we had hoped to be able to
provide further evidence concerning the validity of Herbelin and Cohen's
model.
The remainder of this section is presented in several su Usecons. 3rst
the initial studies of the chemiluminescence produced by the H - reaction
sequence are studied. Next the rate coefficient determination results for
reactions 2 and 3 are presented. Some modeling results are 31so desz-rloed.
Finally, we attempt to relate these results to possible development of t:as
scheme as a chemical source of N2*.
2.2 H * +IF 2 CHEMILUMINESCENCE STUDIES
2.2.1 Flow Tube Description--The new fast flow reactor ,zA ;as -o . -
pleted and tested specifically for these experiments. The flow tuce is con-
structed from 5 cm id Pyrex, and the observation region Is a 5 cm iz! stalnle-s
steel c.amber coated with Teflon (Dupont Poly TFE 4652-201). Prior to :ef!orn
coating the entire observation chamber was painted with a flat black prl=er.The black pricer/Teflon combination serves two functions. The primer reduces
scttered light which facilitates all spectroscopic observations. The :eflon
also reduces wall reactions and recombinations. The entire CCNA:- :low tuce Is
shown in Fig. 1, and a detailed section of the observation chamber i-
presented in Fig. 2.
!s indicated in Fig. 1, the flow tube is modular and configuration
changes are conveniently made. All gas flows are monitored by electron.: -ass
flo-meters. The flowmeters were calibrated by measuring the chan.ge In o)res-
3ure versus tine of gas flows into a standard volume (6.5 i or 12.2 ;las s).
.he pressure chanle was meisared using a "alidyne pressura transzucer ta:t hId
been caliorated a.:ainst a mercury or oil. manometer.
4
~%~i.vv -i,. -- *
L- C3IPUTER!
RESONANCE
Ar/H2 .MMNOHOAO
L 0 0 TO SLOWER
u-WAVE H2DISHARE NHIGH PRESSURE
tVOLTAGE PORT
Ar/NF3 P~HOTOMULTIPLIER
DISCRIMINATOR. vuVMON OCH ROMATOR
* PHOTONCOUNTER
STRIP CHART
RECORDER
Figure 1. Block diagrim of the CONAN flow t~e
To
rI ~ I TO
Fiar 2. StetnLled S~~~ oeiteel Cel!e
SI
.fo Iner
The pressure -n the flow tube is measured 1;sinz a Bara:rzn C to 1,0 torr*
capacitance nano-eter. Typical flow tube pressures ranged frcm 0.13 to
2.5 torr. Flow velocities could be varied from 500 to 5300 Im/s. The flow
velocity was de-ermined using the
v k ~tot
where
f E gas flow rate in number of molecules/s
Ntot total number density in the flow tube
A flow tube area
v bulk gas velocity.I
The term f is ieasured with the mass flowmeter and Ntot is calculated as the
sum of tie component densities found from the capacitance manometer pressure
reading.
If the total pressure in the flow tube is PT(torr) at temperature B(K),
then the number density of reactant i is given by:
N i ' p" o
where fi - flow rate of reactant i, Zfi total flow rate of all reactants in
flow tute, No is Nvogadro's number, and . is the gas constant.
Four of the eight viewing ports in the observation chamber are indicated
in Fig. 2. In the present configuration, three ports are used for optical
detection. Ultraviolet, visible, and near infrared chemiluminescence are
monitored using a 0.3 m McPherson monochromator and a Hammamatsu GaAs ?17. A
second port is used to monitor NO emission used in calibrating H atom number
* densities (described later), and a third port is used to observe resonance
fluorescence for N atom detection.
1t torr = 133.3 Pascals.
06
ID04'53 UM "o m
ii
2.2.2 Visidle and Near Infrared Studies--To study the chemical -roduc-
tion of excited states from the reaction of H + NF2 , we utilized the reaction
of hydrogen wi-th the products of a microwave discharge of NF3 dilute in "r.
~.%Initial run.s showed that the emitting species produced by the NF3 dlscharge
consisted almost exclusively of NF(b-X) and NF(a+X) emissions at 528 and
843 nm, respectively. Weak N2 (B+A) first-positive emission was also observed,
but since the radiative lifetime for N2 (B) is orders of magnitude shorter than
those for NF(a) and NF(b), the N2 (3) is a minor product.
Upon the addition of hydrogen to the effluents of the NF3 discharge, the
NF(a and b) emissions became approximately an order of magnitude more intense.
:n addition N-(B-A, emission became more intense. The NF and N2 emission
intensities also varied as a function of H2 flow. A spectrum covering the
range 320 to 2O nm is shown in Fig. 3.
~4F .-p-
7,7
5~ ~ ~ il 7 .. j;? ' 0 7501) 3(100 "i5J' ,. ,
WAVELENGTM "
74pure 3. 1hemi.'iminescence in the 500 to 900 nm region produceJty the reaction of -Hn with dischariea N,4F. Totalp.ressure ias 3. 16 tort.
The observation of enhanced NF(a) emission when H2 was added to the .is-
charge effluents is evidence that in addition to NF(a,b) both TF2 and F are
products of the NF3/Ar microwave discharge. Rapid reaction of F with added H2
produces atomic hydrogen which can react via the abstraction reaction:
H + NF2 +NF(a) + HF
These observations are also in accord with those reported by Herbelin and
coworkers (Ref. 9). The dependence of the Nr(a+X) emission upon the H, flow
rate is shown in Fig. 4. Note that as H2 is added, the LNF~a)] increases,
%. goes through a peak, then appears to reach an asymptotic limit at high H2
--lows. The fact that the NF(a) concentration appears to be essentially inde-
pendent of [H2 ] at high H2 flows is utilized in the measurement of the quench-
ing rate of NF(a) by H atom as described later.
4,k A)
. 4 (b) NOTE:
-NFia) -N(A);MAX - 4 x i09/cn
3
- I! t
0/ \ \
/I, ,
,
Fiure 4. Dependence of NF(a+X), NF(b-X), and N,(A) emissions as afunction of Hq flow.
8
2.2.3 Ultraviolet Emissicn Studies--The spectral region 220 to 20C rnm
was scanned to study any production of N2 (A) from:
H + N F - NF(a) + NF
NF(a) + H -* N( 2D) + HF
-..
NF(a) + N(29) - N2 (A) + F (4)
it is important to note that Clyne et al. (Refs. 8,9,10) proposed that
t-e final step was:
N(2D) + NF(a) N2 (B,a) - F
DIirect A state production was presumably discarded because of orbital angular
momentum correlations. However, spin correlation which is typically more
rigorously obeyed in these types of reactions would allow N2 si .glet and
triplet formation. Since orbital angular momentum correlations are often
v_;ia-.d especially for nonlinear collision complexes, it seems reasonaole to
expect t:iat direct ;2 'A) production should be possible from reaction 4.
Typical scans are presented in Figs. 5 and 6. Figure 5 shows intense NH
emi-csion originating from two band systems Mi(c' - a 1 ) and NiH(A 3 l -).
Clyne Ref. 1,2,24 observed the identical systems in the H + NF2 system f'r-
ther supporting the contention that under our experimental conditions, the
H - NF-) system was effectively imitated by using discharged NF3 . Upon removal
of H2, tfhe NH spectrum vanished. The most likely production source of N- * is
ty the reaction:
N( 2 D in 2 N :H(X) + H (5)
followed ty:
-2 (A) + NH(c,A) + N2(X) (6)
Th.Ls had been previously argued as indirect evidence for the presence of NI(A)
in the flow cf rrodu:ts resultina from the reaction of N + ;FN-.
-'r-r. xi- .
20) - H2 • NH-x| 4
,%
L
10O0 3 ;134) 340() 35030 J600
,. -. Figure 5. NH (c-a) and (.X-X) systems produced in the Hi-t
,.4% discharged NFi reaction.
.X
-- l ] SGIA J tv4
II"
* ~~OSSISVLE H (A) UTCNaOT
''Igre 6. NOIA-.'(: and ";2(A-X) system produced by reacting H2 + diachar~ed NFI.
* '0'
* NcA
Curing the course of tne initial experiments on the CONAN fiow :-e -. e
chemical production of N2 A) was observed by detecting the Vegard Kap-an (V-K2
system near 250 nm. A typical spectrum is shown in Fig. 6. The strong 170 k
bands are probablv due to slight contamination by NO produced in tne N-'Ar
discharge. The distinct red degraded N2 (A+X) system is clearly evident and
vanishes upon extinction of the H2 flow. Thus, it is not a discharge =roduct.
To our knowledge, this is the first direct observation of N2 (A) chemical 7:ro-
duction and graphically de=onstrates that at least some branching of the reac-
tion energy into this highly metastable state. Some measurements were made of
the number densities of N2 (A) produced, and these results and the source of
the N2 (A is discussei later in Subsections 2.4.3.5 and 2.4.3.6.
The N2(a-X) L3H system near 150 am as shown in Fig. 7 was also observed.
This emission also vanished when H, was removed. The production of N2(a pro-
vides a graphic demonstration of the extreme exothermicity of the H - NF2
reaction sequence since N2 (a) lies 8.5 eV above N2 (X).
0-2
N2(a x) SYSTEM
0-1
0-32-0
1-1
, 043_
4-0'
t'
S(nrn)
Eal.s:n sz;ectrlim of n N al- + av* Si i*;e.
2.2.4 Phenomenoloaical :inetic investigations--Z_ was believec that to
interoret the olanned kinetic rate measurements, it would be advantageous to
examine some of the dependencies of the various excited species upon reactantconcentrat ions. Relevant to N2 (A) production are the data presented in Fi;. 4
which show relative [N2 (A)] as a function of H2 flow at a constant NF-/Ar
flow. The Nn(A) concentration rises rapidly then turns over at
"[21 - 2 x i0 1 8 s- 1.
The data in Fig. 4 were obtained at a distance of 36 cm from the H2
injector. This represents a reaction time of -7 ms. To obtain a temporal
profile of -he '2 (A) produced in the reaction, the fixed 2 injector "'as
replaced with a movable injector. By holding all flows constant, the N2(A)
concentration profile was obtained by moving the H2 injector in the flow
direction. The results of this scan are shown in Fig. 8. Also shown are
CN32 6.7 X i03/m [N NT3 17 OF ~..::-20 - 2.3 X *Io icv j [N2 (8)3 N:2S F .
1Ar:o . 4.1 X iAi-/cm3 A [N2 (A)] UNITS OF (C,: -8
30 ' ,
28 -
25-aS24
-%2 -)
20
Figure 3. Temporal profiles of NF(a), N-,(A), and 5J-(3).
rA4
21
etQ0 J.i
~~'w/w.M (,,is)) ~V
nearly the same t:4e suggesting that they are kinetically related. Since tne
radiative lifetime of NF(a) is 6s, its decay is not due to radiative -:ss.
Rather, observations support the hypothesis that at least some NF(a) reacts
eventually tz form N-2(A). The rapid decay of N2 (A) is also nonradiative and
4s -rcbably due to a combination of wall loss and reactive loss. Some "-)(A)
may be consumed in pumping NH(X) to the A and c states. Time constraints have
*precluded a detailed study of these processes, and these data are preliminary
obserations. Systematic studies of the production of NF(a), N2 (3) and :;N-A)
are presented in Subsection 2.4.
2.3 KINETIC FATS CCEFFIZ:21'NT DETERMINA-IONS
2.3.1 Apparatus--Several changes were incorporated into the apparatus
* for the kinetic studies. The output of the photon counter was connected to a
COMPA2 microcomputer to collect chemiluminescence spectra on floppy disk. The
data were later ana!yzed on PSI's DEC MICROVAX computer system using the spec-
.tral fitting routine. Resonance fluorescence and absorption lamps were incor-
porated into the flow tube to measure 1(2D) atom concentrations. Finally, the
0.3 m monochromatcr was moved to a different viewing port which permitted
monitoring of N(20), %F(a), NF(b), N2 ' N2 (2) and N2 (a) at the same spatial
pcsition ia the low tube.
2.3.2 .t:e Constant for H * (a)--The rate equation for NF(a) in the
flow tube is given by:
dL'NF(a2 ;kl I-d "NF(a), (7)•dt 11 - ".n - wj
where ;uenching of [NF(a)] by ground state NF and undissociated NF3 was
neglected. Experimentally we keep the initial NF3 concentration low
< i10 3 /cc). :n addition the NF3 flow rate is constant for each run. Even if
some quenching were present, it should be a constant term independent of
cnanges in H. 7f we neglect wall collisions, the integrated rate equation
becomes:
-em'.v.i[ ~a ] e k H ( z v 3
wnere z -'- the d Jstance fro H in-ection to t-ae ocservat.._on iQrt, and s - ;e
bulk zas veloci: in the flow tube. :f the concentration of [H: is much
greater than that of NF(a), -lots of in(NF(a) chemiluminescence intensit" as
a functiin of [Hiz/v) yield k, the desired rate coefficiert. A diagram of
the flow tube used for these studies is shown in Figure 9.
Although the form of Eq. (6) is simple, there were several demandin3
experimental requirements that had to be met. The first was to ascertain
that a detectable flow of NF(a) could be established. The initial studies
demonstrated that this could be obtained. We next had to investigate whether
H 2 was a quencher of NF(a). The data presented in Fig. 10 indicate that con-
ditions can be established where NF(a) is essentially independent of i2 lcw.
VHRESONANCE
LAMPPM
0. o
-i- --.. TO BLOWER
4, DISCHARGEHIGH PRESSURE
V PORT
Ar/NF3 _ PHOTOMULTIPLIER
DISCRIMINATOR VUVMONOChROMATOR
PHOTONCOUNT ER[
- STRIP CHART]RECORDER
F u r -2 21ock di :ram- of C.-... flow tute.
01 .14
25
20
Z~10
'aab
C 015 10 15 20 25 30
[H12] umoles-s - 1
Figure 10. Relative concentration of NF(a) as a function of H, Flowfor three different NVI flows: a) [NF 3 ] - 6 ' moles-s';
t) T, F3 = 2.7 moles-s']; c) [NF 3] - 1.25 Lmoles-s - l
The activation for adding a large excess of H2 with respect to NF3 was to
react away any free F and NFI produced in the discharge. In addition, by
injecting a large excess of Hi in the formation region of NF(a), su*bsequent
addition of smaller amounts of undissociated H2 further downstream would have
minimal effect on the NF(a) concentration. All removal then could ce attrib-
uted to H atoms.
The H atom quenching experiments are thus ran as follows. A small flow
of :F3 (typically 0.2 umoles s-1) in a flow of Ar (2 mmoles s-1) are dis-
charged in an tvenson cavity using -6W of microwave power. Approximately
0.2 mmoles s-1 of H2 was added 20 cm downstream from the NF3/Ar inoection
re-lion. A mixture of H'/Ar was discharged in a second cavity (50W) and
iniectei 15 *:m downstream from the first H2 injector. For clarity this 4,
fiow ls referred to as the secondary H2. The Ar flow carrying the secondar-
H2 -was <ept constan: typcailly 0.4 mole s-1 ). To assure that NF(a; decay
et 15
due to : atoms could indeed be ocserved, some preliminary runs wer- maze.
Figure 11 shows two scans of :;F(a-X) emission under ident-za" cond.::ns
* except tha: in Fig. 11b the secondary H2 discharge is on. :.e diminitio. of
NF(a) due to H atoms is clear. The :NF(a)3 was monitored as a func:on.
adde'd schar~el H. . These experiments were performed 'y hci;v n; f!:--;s
constant except for the secondary H2 . No detectable cnange In [NF(a;, was
observed for any secondary H2 flow without the discharge. Lire changes (-. an
orter of magnitude) could be observed when the secondary ':2 d-schar;e was
initiated. Although we had not yet calibrated the detect.on systa. to .aeasure
absolate H atom concentrations, it was felt that it was prudent to aonitor
relative [NF(a,' as a function of secondary H2 - At relatively low secondary
12 flows (< 4 moles 5s1), Zn (NF(a) chemiluminescence intensity) variei line-
arly with the secondary! H2 flow. At these low flows one wou-d expect that the0dissociation fraction of A2 is essentially constant. Thus, the A ato-a concen-
tration would 'e a constant fraction of the secondary H2. At higher lows of
secondary dary H2 the plots displayed a curved behavior indicative of a diain-
ishing dissociation efficiency. Only a few such runs were made to become con-
-,vinced4 that NF(a) decays could be detected in the presence of H ato;:s. Tne
next step was t. calibrate the detection system, outlined as folows.
/ IV (a- X) NF a- I
8675 '750 8675 3750
Fi;ure 11. Scans of *#F(a:<) emission showing removl by hydro4en at:u;as:a) H: discharge off, b) H0 discnarge on. Condi:ions as
follows: pressure = 1.24 tort; secondary fiow=5a0 uoles-s - .
16
%%A
T!! he t~r:~:. c-f NO was -sei wi=t. H to meas.:re * atom concen.r-. .
.The reco minaticn reaction of H -r NO produces HNO emission. The intenity .s
Proportional to :Hj [O0 and is ressure independent (Refs. 10,11). The -rc-
V. : portionality constant is dletermined by t.he following titration.
32..-1 H atom :alibratio, orocedure--Having estaolished t-- we
could indeed detect removal of NF(a) by H atoms, the next step was to develop
a calibration standard for H atom number density. These calibration exper;.-
ments are described as follows.
The titration of H + NCC_ first described by Clyne and Stedman (Ref. 10)
was crosen. The reaction sequence i-
H + NOC1 - HC1 i.NO (9)
.d. * H + NO + M * H:;O" * :4 10).,*°.
followed by radiative emission from HNO*. When [H) > [NOC1I, unreacted H co'-
bines with NO to produce the HNO* + H40 + hv red afterglow (Ref. 11). W'-hen
[NOCI] > [Hj the glow is extinguished. To use this titration one adds a
metered flow of NOC1 to an unknown flow of H while monitorinz the HNC* e=is-
sr .:his chemiluminescence has four characteristic features at A27.2,
692.5, 762.5, and 796.5 nm (Ref. 11). We used a 9 am bandpass filter centered
at 69a.5 nm to monitor the 692.2 nm peak with a Hamamatsu R955 PMT.
Earlier work by Clyne and Thrush (Ref. 11) had shown that the HNO, emis-
sion intensity resulting from the three body recombination H - NO + M
HNO* - ;4 is descrided by
:,-O= K [HINO]
and is independent of total pressure over the range of I to 2.5 torr. We note'"
. t :-aZat is orooortional to fH! for a given (NO]. To utilize this titrat;ion
as a calibration standard, the proportionality constant K must be deter-nlnd.
-Snce ,/: 0] is proportional to [H] ".r is lso tr'=e that i[/NOCl]added ;3
17
proportiona. to :-i. Note that [H] - :Hl 0 - [NOCI] where [H', is the atom--
hydrogen concentration in the absence of NOCI.
:herefore
.vI K' [HCM] - CNOClY'[CNOCJ]
A plot of 1/[NOC1] versus £NOCIJ will be linear with a slope of K anh an
abscissa intercept of [HL 0 . with K determined by this method, the H atom
concentration can readily be measured by adding a known concentration of NO to
Pthe unknown H flow.
Although straightforward in principle the NOCI calibration proved to be
*tedious with several subtle problems. The initial calibratizon produced nun-
linear behavior and irreproducible results when plots of !/[NCCI] versus
[NOCI] were constricted. The probable cause of this problem was traced to a
wall reaction. A detailed study of Clyne and Stedman's i0 paper revealed
* that they had observed a similar behavior on bare pyrex walls. The postulated
reacticn sequence including wall effects is
H NOC1 - HCI + NO
H - NO + %I - .,iO* - :4
H + HCI H2 + Ci (11)
Cl + Cl W4lC12 (12)
,H + C - HCI 113)
UU The catalytic recoobination of Zl on the wall forms C1 2 which then con-
sume H via the ripia C1 2 - reaction. By treating their flue' ' abe ;al:s with
,chocnorl . acii wnch innibits 1wall recombination of Cl, Clyne and 3ted:ian's
* reau" s cecame mun --ore re')roduci-le.
1 , 1
% ! .o
A sect,.ion -e .1w e as uncoated .yrex. This design per. -
visual examination cf chemiluminescence in the flow tube. Sunsequent-y t"-
reaction cell was coatad with Teflon in an effort to solve this Proble:. The
calizra:i.n plots of :/N:0CI versus :N0C1] did become much =ore reproucIL.e
and the linearity improved. Although the NCC1 titration technique proved- to
be more ..ficult to ise than anticipated, the titrations performed in tzne
Teflon coated flow 7-be gave satisfactory results.
A typical plot of I/..NCC as a function of added [NOCIJ using the full7
coated reactor is shown in F g. 12. An average of five calibration r-,ns at
variou..s oonditicnz jieldeo a irzportionality constant K = 0.46 x 10' Ai'(mtorr"
with a standard deviation of 15 percent. With a known K one can add a 'known
concentration of NO to the flow and determine [H] from I.{NO/K[NC] =-].
1.2
K = 0.43 x 10- 7 A/(mTorr) 2
* Z 0.90 0.31.
F ,AFN<0-
0. -
C.0
0 1 2 3D,,,, I0 '7rr ,
Fiatre 12. :701ca1 fata for H , NOCI titration. Shown is a plot of.5'0i!C/[N C :added versus [NClJadded. The slope of the olot
Is used to :eter.ine the proportionality constant
KSNo
-. ?H at-ore or~o~ctln at a -ziven microwave power and at a .:orstant .r.
rate was not constant with respect to time. The data in Fg. 13 a.h cal._
demonstrate this beha'vior. These data were obtained by keepin; the -,/-Nr and
NC flow rates constant while varying the microwave power. ..e nu--hers
attacnea to the data points indicate the order in which the data were
recorded. Note the reduction in the intensity of the HNC* emission for a
constant microwave power (points 4 and 5) and the apparent hysteresla-like
behavior when the power was reduced from 90 to 40 W. The cause of tnis
problem was heating of the microwave cavity. Even though a flow of co:wressec
air was foroec through the cavity, it provided insufficient cooling for
unizfcrm H at om prouction. The difficulty was resolved by flowin: the cooling
air through an ice bath prior to introduction into the microwave cavity. :his
allowe, ccnstant H. atom prodaction for at least a minute at any Mizrcwave
power. Since 1t required only a few seconds to measure both the NF~aX) and
HNO- emissicn intensities, this was a satisfactory remedy. A further
improvement mihnt be realized using a different design such as that of Zing
et al. (Ref. I1 who used a fluid as a coolant.
I.
1.9
C.7
0 .6
0.4
C.2
4..
0,. o 1o 2u 36 40 50 60 7 :C
MICROWmAVE POWER (WATTS)
~~~ ~~to -:) as a~r~z- )Z ~ aw,ote tnat the -u n efficiency for -H at.n trou':t-n is~~~~stror.g"!, Jepentlen-. _poni tne microwave pc,,..re : o er. 4m e
re:_:r to tha ten-ral orler in whichi ,aa -s o a wea.
et
1.8-
1.6-
1.4-
,o 1.2-
Z 1.0-
0 0.8-
0.6
0.4
0.2
0.0'
0 10 20 30 40 50 60 70 80 90 100 110 120 130
P2 (sccM)
Figure 14. i,,n as a function of [ for constant [NO] and
constant microwave discharge power.
in practice the microwave power was held fixed and [Hj variations were
obtained by changing the H2 flow rate. A typical plot of I.:4kO) versus H-
flow rate at constant (NO] is shown in Fig. 14. Note that as 'H2 3 is
increased, the dissociation efficiency is reduced. This observation coupled
wit' t-e dependence of H atom production on the temperature of the imicrowave
cavity clearly indicate that (H) must be measured for each H2 flow. One
cannot rely upon comparable conditions to give the same [H].
Next the response of IHNO at constant [H] was checked while changing the
NO flow. Figure 15 clearly shows that 'HNO is indeed proportional to added
NO. With these checks and the previously described calibration completed we
were in a position to measure the desired H + NF(a) + Products rate constant.
it was noted also that typical [H) was >1014 atoms cm- 3 while (NFl (whic
reoresents the maximum possible [NF(a)]) was -10 1 3 molecules cm" 3 . (Later
experiments showed that the maximum "NF(a)] was no greater than
2 :< )12 _clecules :,- 3 . Thus, the pseudo first order condition wi: H in
ex.za3.5 wa3 sat. f--aI.
21
I.. U
W- 1W 9 W .W- M~r W ~ rrr ~ W r'Y 1- '. . n .r1i
I p:J I I , I :
2
bI
0j
0 2 4 6 8 10 12 14 16 18 20 22 24
NO (sccm)
Figure 15. IoNn versus [NO],H , for constant [H).
Typical scans of the NF(aX) emission with and without H atoms was pre-
viously presented in Fig. 11. in practice the peak NF(a+X) signal was moni-
tored at 374.2 nm but pericdically scanned the entire band to check for any
possitle spectra. interferences from "12 (3+A) and HF'v). At tne high H- flows
used in these studies no interferences were detected.
A sample plot of !n'NF(a)j as a function of [Hi is shown in Fig. 16.
These data yielded a rate coefficient of 2.9 x 10- 13 cm3 molecule-1 s- . The
rate coefficient was measured under several different sets of conditions.
Table 1 summarizes the results.
The average of all r-uns is
k (3.1 :t 0.6) x 10- 13 cm3 molecule " 1 s-1
The error represents the best estimate of the systematic uncertainties in the
mea surements. These arise from the H atom calibration which is the dominant
sour:e of error. This valle 4s in excellent agreement with the valle of
22
d, . ' I ' I i i I i I I I
CH (,A.rr
7.u re .. of-[F ja ucino de H.Ttlpesr
;,-,
for'= ta' u s4g rcrirgs a .9tr
Rat Coefcet
- v I z c 4 0 i R Cofi ce
-i 2 3 4 5 6 7 8 9 10 11 12"--.". H] (rntorr-)
'-'" : re 1o. P)*: of Zn['NF(.a)] as a function of added [HI. Total pressure
N 1 1 ffor tgas rn using Ar carrier gas was 1.79 torr
i~tatrt of orditions for w!easureent of cm + NF(a) r
Rate Coefficient.
i Effective
I Flow Reaction?rsu e ..... y Time Rate Coefficient
I;[ ,or (c - ) (33)(cm 3 molecule - a -1)
* .47 2046 21.9 2.31 x1 - 1
"".79 I 1641 26.6 2.65 x1 - 3
I 3 913131 14.1 3.3 x 1 -1
1. : 1604 42.9 3.38 xI - 1
.... u::fer 33s irn 31t 3SeS 'was Ar. Distance from H aton
ir.-ec:zr to observation port was 45 cm.
23
.N
.5x IG- 1 3m mo.lecule-1 _-
2. x estimated by Clyne at al. (Ref. 2,3). 7t is
apparent that the first step in the proposed excitation scheme of N- i s
relatively slow.
Throughcut these studies it was observed that [NF(a)] was essen-ially
independent of Hj] added. No decrease in [NF(a)] could be seen even though
sensitivity was mcre than adequate to see a: 10 percent degradation. From
these observations, an upper limit for H2 quenching of NF(a) was estimated to
be 1 x i3-14 cm- nolecule 1 s- This can be compared to an earlier measure-
ment of D2 quenching of NF(a) by Kwok et al. (Ref. 13) who reported a value of
k < 7 x 10- 14 cm3 molecule- I s-1
2.4 =RODUCTION PAZE FOR N2(B) EXCITATION
2.4.1 2ackground--Determination of the production rate of excited N-
from the H + NiF2 reaction scheme is of critical importance to the possible
utilization and development of this system into a chemical laser pump source.
Since N2 (B) is relatively easy to monitor and is the likely source of '.-IA)
produced in this scheme via N2 (3BA) emission, we chose to test the zechanism
previously postulated in the Refs. 3,9,10. The salient model is described by
two stels
NF(a) + H - N(2D) + jF
NF(a) 1 !D) N2 (3) + F
The initial approach was to add N(2 D) atoms produced by a microwave discharge
to a flow of microwave discharged N-c3 dilute in Ar (which produces NF(a)).
5 The difficulty of strong N2 (3A) emission from three body recombination of
- )4) atoms (produced from our discharge source of N(2D) atoms) was encoun-
teed. This wouli have been the preferred approach since interferences caused
by H ato:m chemls:r; would have been eiiiinated. The strong recocibination Nn
first positive emission, however, made data from this approach diffizult to
S24
'h i-% % ,%- ,, %%%% ,, %%
An alternate a.proach involved the entire H + Nr2 reaction sequence. Al:hough
nore complex than the first method attempted, it does offer the advantage cf
permitting a detailed study of the actual chemical production of N2* in an envi-
or.ment similar (but not identical) to what one would expect in a laser fev;.e.
The model described by Eqs. (2) and (3) leads to predictions of the '12(3)
production that can be tested experimentally. If reactions (2) and (3) are an
accurate description, then the rate of change of [N2 (3)] with time is given by
diN,(3)]
d = k3[4F(a)JfN(2D)] - (B) I-N A + K (14)d4t 3~ 2 B +A 4'()
where AsA is the radiative rate for N2 first positive emission and ". is the
first crder juenching rate of N2 (3) by species in the flow.
O
The ,2(3) emits radiatively essentially instantaneously with reszect to
the spatial volume in which it is formed in the flow tube. Thus, N2 (3) is in
steady state with N'(a) and N(2D). Hence, from Eq. (14)
1k3 [F(aJEN(2 D)] = [N2(S))(AB+A + K (15)
Note that a plot of [N2(3) versus t-he product of [NF(a)] x [N( 2D)] shoulj be
linear with a slope of k3/(AB-A + X2). To extract an excitation rate from
such an analysis one must saiuitaneously measure absolute concentrations of." N'(a), "N.(3), and N(2 D). Careful and systematic calibrations are required an-
are described as follows.
The [N2 (B)J and [NF(a)] were deter:.ined by measuring the intensities of
the '12 first-positive and 'F(a X) systems. The relative response function of
the monochromator and detection system was determined using spectral standarj
deiteri-m and juartz-halogen lamps. The next section describes how thie
relative response curve was placed on an absolute basis.
a 2.4.2 .tsolute photon emission rate measurements
2.4.2.1 Gener3l procedur-e--The observed signal is relates to the
r.... vrli;ume emission rat through
a,.
04
__ % % %%
-- T
obs= true 4-T 'k
.where j/4- is th e effective solid angle subtended by the detection system.,
ais the quantum efficiency of the photomultiplier at the wavelenzth of inter-
eat, T. is the transmission of the optical system (e.g., mirror reflectivi.ies
and grating efficiency) and V is the observed volume of luminous gas in the
reactor. The wavelength dependence of the product 2/ 4 7 ,%TkV is given by the
relative monochromator response function R . Absolute values of that roduct
are obtained in a calibration experiment using the O/NO air afterglow at or.e
or several specific wavelengths. Absolute values at wavelengths other than
those chosen for calibration experiments are obtained by scaling with a,.
As stated earlier, the relative spectral response of the monochr.mea:zr
was calibrated between 200 and 900 nm using standard quartz-halogen and :-2
continuum lamps.* Additional confirmation of the calibration between 5d0 anc
800 nm is obtained by scanning the air afterglow spectrum and comparing
observed relative signal levels with the relative intensities given by Fontijn
et al. (Ref. 14). The absolute spectral response of the detection system is
measured at 530 rim using the O/NO air afterglow as described in the following
paragraph.
When atomic oxygen and nitric oxide are mixed, a continuum emission
extending from 3 5 am to beyond 3000 nm is observed (Refs. 14-Z2). The
intensity of this emission is directly proportional to the )rc,1uct of the
number densities of atomic oxygen and nitric oxide, and independent of pres-
sura -,f batn gas, at least at pressures above about 0.2 torr. Thus, the
volume-emission rate of the air afterglow is given by
Itrue - kx[Oi[NO1 IA (17)
a wnere k is the air afterglow rate coefficient in units of cm3 molecule "I --
nm-I and '\ is tne m:onochr:mator bandwidth. Literature values for tha-- rate
coeffi.cient scan a range of more than a factor of twdo (Refs. 14-20) but recent
*"Optrtn..: Lacoratrles Inc., "ziler 3prinz, Maryland.
V1 1rI 'er
---- |,
studies (Ref. 19) Indicate that the original work of Fontijn et al. (.:,ef. 14)
is probably correct at wavelengths shorter than - 800 nm. We use a value of
1.25 x 10- 19 cm3 molecule - l a-' nm-1 at - = 580 rim. Combining Eqs. (16) and
(17) ;-ives the observed air afterglow intensity:
O/No1)I k 0O110 - T 7; k6"1 "
Air afterzlow calibration experiments give a calibration factor,
< = k \-- -\V (19)
the determination of which will be described in the next paragraph.
Absolute number densities of emitters are obtained by dividing absolute
volume emission rates by known transition probabilities. The air afterglow
calibration factor, <., and the moderately well established value of the air
afterglow rate constant, k\, are used to convert observed emission intensities
to volume emission rates:
*I osk X cRX. i !~obsX\A c~X
, c c
true
where 'c represents t*e wavelength of the calibration experiments and \obs i
the wave7ength of the transition of interest. The lobs must be the total
integrated band intensity. :n this work the measured emission intensities
were measured in terms of teak intensities. A calibration factor consisting
of the ratio of the integrated band intensity to the product of the spectrome-
ter bandwidth and the peak intensity was determined for cases in which t"otal
intens~ties were needed. :n calibration experiments, the band areas were
integrated numerically using spectral scans which were greatly expanded along
the wavelength axis.
27
A ser;.es :- =alibration experiments taken over a eri4d :pf -4me eIstab-
fished the callbration factor, <560, to ±15 percent. An additional ncer-
tainty of :25 rer:ent exists in the absolute value of the air afterglow rate
coefficient ksz 0 . Further uncertainties in the determination of the absolute
photon-emission rate for N2 (3) come in through the relative monochromator
resoonse function. There were also some errors introduced -n determining
absolute NF(a-x) emission rates because the a-X emission band intensities were
needed. in calibration experiments, the band areas were integrateUo numeri-
cally using spectral scans which were greatly expanded along the wavelength
axis.
2.4.2.2 Air afterglow calibration procedure--in air afterglow
%calibration experiments, the intensity at the calibration wavelength is mea-0
sured when known quantities of both atomic oxygen and nitric oxide are added
to the reactor. Known quantities of atomic oxygen are prepared by titration
of nitrogen atoms with excess nitric oxide:
S+ 'Q + N2 +
(21)
(k2 1 = 3.5 x 10- 11 cm3 molecule "1 s-1) -?Refs. 23-25)
in the absence of added nitric oxide, N atom recombination produces chemiluni-
nescence from tn nitrogen first-positive bands, the intensity of hnich is
:proportional tc the square of the N atom number density. The kinetic 7;ro-
ceases are described as follows.
N %I + N2(3,g) (22)
N2(3 3 .g) +N 2 (A3u +) + hv (first-positive bands) (23)
Upon addition of NO the first-positive emission intensity 4ecreases until the
quantit: of NC) added balances the amount of N atoms initialiy in the flow. At
this point, the end point of the NO titration, all N initia:ly in the reactor
nas been quantitatively converted to 0, and no emission is observed 4n tAe
,
reactor. As NO is further added to the reactor, the air afterglow emission
begins to be observed, and the intensity of the emission varies linearly with
the amount of NO added. Such a titration plot is shown in Fig. 17. -he
equation describing the change in the air afterglow intensity as a fanztion of
added NO for NO additions beyond the titration end point is
IO/NO = <[O](NO1 = <CN2]o(NOJo - Io) (24)
where < is the constant of proportionality relating the air afterglow inten-
sity to the product O][NO], [NJ o is the number density of N atoms initially
in the reactor crior to NO addition, and [NO] refers to the NO number ensity
which would be octained in the absence of reaction (21). The factor < then
is determined to be the ratio of the square of the slope to the intercept of
the line describing the change 4n air afterglow intensity with [NO].
50 2.0
40
S30
01.0
@ 2
0 10.00 1 2 3 4 5 6 7
[NO] -ntrr
?:~ure 17. Plot showin z - NO titration procedure.
ii
Two major comolications can lead to problems in using the technicue in
Eq. (24) to calibrate the apparatus for absolute photon-emission rates or for
[0] measurements. 7f the initial N.1-atcm number density is greater than about
1014 atoms cm-3, some 02 spectral features from 0-atom recombination will
V begin to be observable above the air afterglow continuum, and can lead to
incorrect air-afterglow intensity determinations (Refs. 26,27). A more seri-
ous proclem, however, liesin the slow removal of 0-atoms in a three-body
recombination with NO. The important reactions are:
0 + NO +- M - NO2 -M
2 5 o3 m oeue for M = Ar) (:.ef. 23)
and
0 + N02 + NO + C2
(26)
(k26 = 9. x 10-12 cm3 molecule-1 s-1) (Ref. 28)
Reaction (26) is fast, and essentially acts to maintain a constant '40 number
density and to double the effective rate at which 0 is removed in reaction
(25). This effect tecomes a problem at higher pressures I(> 1.5 torn), longer
mixi.ng times (; 30 mns) and large NO concentrations (> 1014 molecules cu -,;.
It is just these adverse conditions which give the best signal-to-ncise
in the calibrati' n experiments. The calibrations for the experiments
described in this report were corrected when necessary for the effects of
removal of 0 in reactions (25) and (26).
2.4.3 N(2 ,,) resonance .amo calibration
2.4.3.1 Background--Az mentioned earlier resonance fluoreso:ence
was used to detec:t :;(2 0) at:os. The lamp is of a special design and is shown
in 71g. 18. Two crucial featares are: (a) the gas i4nlet to tne lamp is near
the '.am- output and the :as z~ is out the bazk; and (b) the :zicrowa:e cv
30
".S _.. .?ITAILES
"NT 0
-PLASA MICROWAVECAVITY
FLUORE SCENCE
P LAWP
'GA
4,..
rLAA
*~v Fig-ure 16. Details of tne resonance flu~orescence/abscrrion lamos~used for N(VD) concentration measurement.
"is a ite.d so t> : t.he o~scharce plasma extends to the lamo output "inow
* This minimizes the de~ree of lamp self-reversal.
%_For the present stuies the lamp was run at 1.5 mtorr of He "iha rc
'N of "N2 introduced through a Oranville Phillips calibrated Leak valve. .This
'Nproceciure ensures reprocuciole lamp operation. The microwave power the
*lamp was keot a:. ::e moderate value of 20 W. These operating points were
chosen to utilize the results of previous studies (Ref. 29) that reported ga
~temperatures of -6,2') K( in the lamp under these conditions. As is dweli known
4TO
in resonance fluorescence or absorption measurements, the temperatures of the
.• amp and the absorbing medium both influence the observed signal strenuth.
4.' The calibration orocedure for the N(2D) resonance fluorescence lamcp in-
volved relating the : (2:) flilorescence intensity IF to N(2:) concentrat~rns
* e~s'ured~ by resonance azscrmption. Two identical Lamps were u/sed. The ahsurr-
_ tizrn .amp was ::tachedi to tr, e flow tube Docosite tne V'JV mcnocnromatzr wh;ie
PUMP
-v - - 'NMIROWAVE~.%~N
CAVITY' ' "
A :n~te fluoraescance lamp was normal to both th-e monochiromator and the ~srtolamp. The calibration is complicated by the ener;y level Stracture of 'I atons
as shown in Fig. 19. Since the 2s2 2p3(2p0) and 2s22p3 (2De ) levels are oth
radiatively connected to tlhe 2z2202js(2,) level, fluorescence at 149 .3 n= can
contain zontributi ons from both 2po and 2 D0 . The degree oZ contaminatlcn, -f
course, depends upon the ratio of the resonance lamp intensity at 174.3 nm to
that at 149.3 rn. in these exoeriments a dielectric coated filter '"gl sub-,vstrate& was used at the lamp output. The filter rejected 174.3 nm while -ass-
ing 149.3 nm. With the filter in place the lamp output was measured at both
wavelengths and it was found that the ratio of the intensity at 174.3 n= to
that at ;49.3 nm was approximately 1 percent.
.. ENERGY/eY N
(2s22p22s) (2s22p2 lp) (2:Z)12- Ar(4s)
4i2... . - 3 P 4 D,', .';P 2
I 1345.6 nm11- '
"" 871 nm - P
-4p// ,'C7 0 -7A
p 174.3 nm /1 i 120.0 nm /.. /j4 !9. l/, 113.5 nm
V ,/, .,// //
•2s 22p3 4 ,
2S43L D
2-
°,2s22 3
,.,Ort 1 . -nervy -evel str'ictlre of N. ReLev.in trinsiti ,-.s. are Znlte.
a z te
VWV%
As an add :i_.nal cneck selective q;enching studies wera :erfore_4
Nt 23) and N(vP?. :hese exzeriments -4ere done Cy adding CO2 to a flow of k _j
and N(2P) The N(29) and N1 Pp) atoms were pro uced in a micrwave dJiscnarze
of N dilute in Ar. The relative concentration of N(2D) was easureo usin
resonance fluorescence and N(20) was monitored using (2o : 4S) emission at
347 nm. Since the quenching rate of N20 Dy CO2 is greater than that for N:P,
the N D is selectively quenched. If part of tie observed 149 am fluorescence
originates from the resonanze lamp exc:tlig N( 2 D) atoms, then plots of in
:F ( 14 9 nm)) versus [CO 2 and Zn:1(347 nm)' versus !CC 2 J should becoue carl-
lel when the N2_ i.s totally quenched. However, it was observed that tne
149 nm fluorescence signal was still :-ecaving faster than tne 347 nm si:nal
even at the highest flow rates of C2. it was concluded from the study that
t .e upper limit of N(2?) contribution to the N(2 ,) fluorescence signal was
l Iess thnan 3 cer-ent.
Following these systematic checks, the resonance lamp calibrations w;are
performed. These measurements were accomplished by constructing a curve o:gcrowth for N(29). A flow of N2 dilute in Ar was discharged in a icirowave
cavity to produce an unknown concentration of N(2 D). Resonance absorption on
the 149.3 nm line was used to measure the concentration of N(2 0). Since reso-
:, nance fluorescence and absorption display some subtleties in atomic- "4 .hese
phenomena will oe descrized as follows.
2.4.3.2 .eview of resonance fluorescence and absorpticr--:he gern-
Veral technijues of resonance assor'ution and fluorescence nave been descr~bed
in detail in Ref. 30. Only the salient features are outlined as follows.
For an isolated line Beer's law is usually written
K) Tran( e 1 1) )e 4 (27)- Trans 0 0
wh ere
-3
,0,
*%,
lo incient intensity at frequency '
:Trans transmitted intensity after transversing the absorbing medium
a distance
N = nuzcer densi'.7 of absorbers
= absorption cross-section at frequency ,
k(v) = absorption coefficient at frequency u
For small absorptions (N: (,)Z 4 0.1)
I .Trans ) I0 (,)( 1 - Nc( lZ) Z)
or
Zn(I0/ n) = n- = Nc('lZ = k(.)Z (29)
. where
0 TransA (30)
Thus, for azsorption at a single frequency the measured fractional
a'sor:tLcn can be used to determine N. In practice, however, one cannot mea-
sure the fractional absorption at an isolated frequency. Rather, the frac-
tional azsorption coverinz a frequency range that typically encompasseZ t.he
entire absorpt,:n line is measured. This integrated fractional absortin is
the -.uantity that is relevant to the present discuss-on.
In this zase the fractional absorption A is related to the absortign
coefficient through
- I.(fl'1) - e~ d,
A rrans
l I(,j; dv
When the line Is predominantly Dooler broadened the atsorption c)e::i-
b, cient can be r r,: < . .mat . yI0 ;I
k k~, 3 k2)
whera the atsorption coefficient at line center, and
2 ( ; - ' ) )
In ~ ~ ~ ~ ~ ~ - is.(3 ' ~) (.2?l the Doppler wiitni atIn Eq (3)(2Z8x1 0 11frequency '.0.
:t can a-so '-e shown that in general the absorption coeffiolient inte-
grataed over all frequencies is :-iven by
k d,) - Nf (34)
Up where -4 is the oscillator strength of the transition, m is the electro n zaIs
e is the electron charge, and c is the speed of light. if the approxi-ation
k.) = 'koe 2. in. 34) is used, we find that
'-4. 2 J -
:n manv. :-itIat1ionsm -in=zlud;ng the present one, the source of resonanze
Kraiic. lamp) a--4 the a'd-sorb-ers (N21) in flow tube) are at different
te~ner~tres.~s Z usuzally treated using the empirical expressior,
(36
for t-e intensity Of the lamp as a function of frequency.
where sonl2 breadthAisrztonline breadth
35
0~%
"hus, we arrive at an excression for the integrated absorption for tne aze of
the lamo and absorber a: different temperatures both being Doooler bro a dened.
/ 2
- oe 2
i3 - Trans - (A "0 0 -( 2/a )
The actual emission line shape may differ from the Gaussian for= of
Eq. (37) due to self-absorotion (which preferentially affects the center of
the ine) or moncenergetic excitation processes (e.g., energy transfer frcm
metastable Ar). '4athematical treatments of these effects have been given by
Kaufman and coworkers (Refs. 31,32) and by Clyne and coworkers (Refs. 33-35).
O The sel--abscr-tion model of Rawlins and Kaufman 3 1 gives an excellent -th to
0(i) azsorpticn data over a wide range of lamp optical thickness. However,
the i.-icl'sion in some applications of self-absorption (Ref. 36) can be vitally
important. it will not be considered here since it is not a factor in the
experiments. Monoener~etic effects are often collisiona!I moderated in
steaiv-state line sources, in any case, since only the inte-cral over the line
is observed, l '.37' is often an adequate description.
--ua tion -3-) can be integrated in a straightforward manner by -Ie method
-*.'. or 5al:ss:3an .uabrature. However, for the purpose of the present ':ssion,
.. thne :ziwlng analytic expression for A is useful.
( )1/2.'4 n-IA = 2: (-1n) (36)
n=1 2 1/2'4..n~ n'(1 + na
-- IVe exp-essions hold only in the case of a single line. However, the 149
1W and 174 rm transitions consist of multiplets which are difficult to resolve
fully. 7or the case of unresolved multiplets (line separation << instrument
resol'.in), an expression similar to Eq. (37) holds:
36
-' ". % ", % % , , , ' 'h % '% ". " " " ".".'% ' .",.,%,%,% % % % . ,-, . .
2
- ' " O-e i
C e d'd
eeh C rerent the relative intensities of the components of the
*,ne h scectros;c'Ji2 croperties of the N(20) and P(22) resonance
4 l-.'':ies ar :i'en i T3~A.- 2.
A welnrte av-ra.e 3f the cscillator strengths for the three unre .
.":- ne- near %- -m combined with Eq. (35) gives
- <i x i - NZ (43)
2. Resonance Lines for Metastable Nitrogen Atoms
Wavelength Jper Lower Relative C
I State State Cy Intensit7
5 6 9 3 ,. -.763/ 2 5/2 62' 3 /2 3 /2 4 ,.3
"-"- I I.9. *, - 3/2 32 2
", .2725 I,/ 4 0 '3.02 " • 3/2•W.'"" 4.2725 ' o - - -3
,....74.5245 P1/2 p1/3 2 2.]4""I 7.555:P/2 j2?e3/2 41 3.3011
. Ler :Ce, 3. . anz anage, B.D., "Radiative Lifetimes of UV Multiplets
-- n rr C rn ana Nitrogen," zhys. Rev. 141, 67 (1966).
37
% ,.t..,%% % .N l ,
0 'P
q
Finally, since . = 5.0 cm
N" = 9.14 x 10 1 1 <'O> (41)
A determination of <k">, thus, allows ',s to calculate N via -q. (38. A
Gaussian-quadrature numerical method was used to calculate k0 Z from the
observed fractional absorption using Eq. (37). A typical calibration plot of
I; versus Zn(1/1-A for N(2z) is shown in Fig. 20. Using this calitration
curve allowed "s to determine 7:N(2D)] from a measurement of -'-e fluorescence
intensiJty iF .
S -
4 -4
£4
.C2 C..14 0.18 0.2I
Figure 20. Plot of N(20) resonance fluorescence sicnal Irvers:s -)n(I/1-A).
2.4.3.3 Measurement Techniuqae for N2(3) Excitation Rate--The
actual 4ala coilection for measurement of the excitation rate coefficient, k 3,
conszeis of scans of tie cnemilun-inescence due to N2(=.A) and ;FaX). inr(i j, t 1.3or, and : t - )were -nei-
ate.n ~t..e ,n 2: was cet:ermined eacn tiene N2 3- nd sia dosiio -- sre~ *~~eci~nof all three species were done at one sl)a5t1 posi~tion -as
'p r?'It NNAMM&KSl
noted earlier. A -:-n cons-sted of the measurements just descriced eaic a, :aconsan .2 flow. This was repeated for several H2 flows with all other cor-
ditions held constant. Most runs were done using a fixed H2 injector, 'ut one
set utilized a sliding H2 injector. The bath gas pressure '0.7-3.2 :orr),
bath gas species (Ar and we), and flow velocity (1.1 x 103 cm/s to
5.5 x 103 cm/s) were varied in completing a series of runs. The spectral
scans were recorded on a CCMPkA microcomputer and stored on floppy disk for
later analysis using the spectral fitting code. A sample scan and correspond-
ing fit are shown in Fig. 21. Ht(3,0),
NF(a X), ANDHFT,3,2- ANON2(B A)2 1' %h (3 *A>;,
711
0.40 N
-NF (b-X)
0 0,0
540 ei- 640 690 740 790 840 890WAVELENGTH (nm)
Fi urr 21. Chemiluminescence spectrin (thin line) showing NF(b+X:,NBI3-A:, - ;F(a-X), and HF (3+O) systeus. The thick etrace is a computer fit to the data derived from thespectral fitting code.
Figare 21 points out the prominent emission features originating from
the excited staces NIF(a), NF(b), N2 (B), and HFt. During the scans, due to
9 *Ie '.;F- -X) (0, , band, a iarge off-scale feature was observed at 523 .
This system is not shown in Fig. 21 since it was used in the computer fits.
To record, on scale, a spectral feature several times more intense than all
other features would decrease the sensitivity for detection of tne weaker fea-
tures. -ather the (3,1) band at 560 nm was used to mon4tor [NF(b)]. This
39
0
approacn was used so that a complete spectral scan could be record-u 4 a
single sensitivity setting for an entire scan. The intensty of the N-Ftb-X)
(3,1) band was related to that of the (0,0) band by making careful scans over
the range 525-570 nm. it was found that the (0,1) band was 33 = 2 times
weaker than the (0,0) band.
NIt is important to note that the NF(b.X) (0,0) band appeared to be much
stronger than the NF(a+X) (3,0) band. This is due almost entirely to the
ratio of the Einstein A coefficients for the two states (Ref. 37):
43 s 241% ' AN ( /AF(a) - = 24
oF(b) NF(a) 0.178 s
..-
In actuality it was found that [NF(a)j/(NF(b)] was greater than 100 for these
conditions.
The spectral fitting routine utilized the absolute spectral response
" calibration data, consequently it calculated absolute number densities for the
species N2 (3), NF(a) and NFkb). Since v1 = 0 in N2 (B) could not be detcted
the results are for N2 (B; v=1-12). The N(2 D) concentrations were determined
separately. A sample of the results for a run is presented in Table 3.
To obtain an N2 (3) excitation rate, plots of [N2 (l;v)J versus (NF(a)] x
EN( 2D), were constructed. A typical plot is shown in Fig. 22. Note that the
(NF(a)]LN(2D)i product was varied by nearly two orders of zagnitude. R.ecall-
ing Eq. (15) it is noted that the slope of this plot should equal
k3/(A=-,A - K,). To accurately determine k 3 one must determine the production
rates into the individual vibrational levels of N2(3). This stems frou A5+A
being dependent upon v (Ref. 38). The total rate coefficient, k 3 , is then
calculated from
.W%
9%,%
:BLE 3. Sa~z2.e :ata Show-J19 £N()[NF(a)j anid £'N( 2Dn)7 as a Funct.in
of W Added..
LA2 4a) LN(2 0)) 7Nk)) 4F(a)lx[N( D)j N(3J:N-)
.37x10 1 3 71x1 1 3.4 0' 2.20 2.3x10 2 2 6.,3x.0-4
7.95x3~ 3 7.13x10 1 1 324x0 1 0 2.22x10 7 161 2
3..C 1013 7.23xO 11 1.84x'.010 .10 .x0 2 661 4
4.0x114 7.3x1 1 1 7.7x0 .8xIO0 5.7x102 2 6 0x'10;4
2.4a3xI0' 2 1.110 1 1 J2.92x103 4.4x10 5 4.1x10 2 0 .xY
NCTE: All entries are in molecule crn-3 .Te t ga*as raitntotal oressurs was 3.06 torr. Tebt a a r n
7.00.-
7.00
- 5.30
S4.00
3.00
S1.00
A u
u o 20 .iU
N (a ] N 2)l (t"1 ' -6LN~, N
2)~(0 molecules'-cm
p~u r . 22. Plloz showini- linear ez-endence -of ";1(3)) with respect tothe product ENF(a)Jx[N(1D):. Run conditions: press-ure =
* 3. : t,:rr ;r o-ath caa.
W~ ~*~i~~~. -~ ?~y ~ ~ *~ I j4 *
12 vk3* ~ k (42)
- Using plots similar to that shown in Fig. 22, the total excitation rate
coefficient was computed from Eq. (42). Note that we could not =onitor vt =3
so the k3 is only for levels v1 = 1 to 12.
The detailed '-3 (vi) are presented in Table 4. Also shown are the suns,
'K3 , The total excitation rate coefficients for the six runs are tabulated ii
Table 5.
In Fig. 23 a plot is presented of [12(B)l versus [.NF(a)j x 1.V(2D)j a!!
the data collected. The spread in the data is typical of absolute photometric
measurements and considering that these runs were made under differing condi-
tions and over a :Deriod of three weeks, the reproducibility is satisfactory.
- Although the data in Fig. 24 do not indicate a discernable dependence of
r upon bath gas pressure or species, the data for the individual rulns,
summarized in Table 5, perhaps show a slight quenching of L2(3) by Ar and He.
-Nt the n;.ghest pressure (-3 torr) the excitation rate coefficients (Propcr-
tl:rnal to iN()3 re -15 percent reduced from the values at the lowest o)res-
sures >-0.7 torr). In addition Ar may be a slig htly more efficient :quenchez
-C.of N -3) .
otshown in Table 5 are populations fr ZJ().As staited earlier
[NF(bJ1 was typically two orders of magnitude smaller than [NF(a). In light
of k3 being nearly gas kinetic, NF(b) + N(2D) reactions can be discarted as a
significant source of N2(3) excitation.
Iis to be noted that the spectral region of 974 am contains contr~bu-
tin rom adthe (,)band of the "42(3-A) system- The iF(3,)
overoneemission was in some data observed. Altnough the snectral itn
ruiecan treat this situation, problems can arise if tie ';-(3-.;) or 4
I ........2
X3LE 4. Measured for v' = 1-12 a: Various Bath Sas ?=ressuris.Con='.onsara i.ndicated.
Excitation RatePressure Coefficient, kv i
*~v 3 ~sr -1 -1Buffer -as torr) 'cm-molecule _
Ar 0.7 1 1.273 x1012 2.335 x 10- 11
3 1.447 x 10- 11
4 1.912 x I - 115 1.612 x 10 - 1 1
6 1.628 x 10 - 11
7 1.546 x 10-11
8 S 1.4"2 x 10- 'I
9 1.140 x 10-1110 8.473 x 10- 12
11 5.882 x 10-1212 2.644 x 10 - 1 2
Sum=1.600 x 10- 10
Ar 1.36 1 1.220 x 10-112 2.231 x 10- 11
3 1.528 x 10- 11
4 1.804 x 10 - 1 1
5 1.186 x I0-1 1
6 1.226 x 10- 1 1
7 1.262 x 10- 113 1.353 x 10"1 1
9 9.102 x 10 - 1 2
10 5.534 x 10- 12
11 3.533 x 10 - 1 2
12 7.047 x 10-13
Sum=1.379 x 10-10
Ar 3.06 1 1.116 x 10 - 112 2.313 x 10-11I
3 1.545 x 10-11
4 1.726 x 10-11
5 9.785 x 0- 12
6 1.128 x 10 - 11
7 1.363 x 10-118 1.534 x 10-119 8.865 x 13-12
10 4.086 x 1 - 1 )1 1 2 .4 4 8 x - ,2
12 .771 x 0- 13
Sum=1.422 x 1,)-
43
',. ,.
., 2S A3LE 4. (Concluded'
IExcitation Ra:e
Pressure Coefficient, k.1. Buffer Gas (tor) V ' CM3 _molecul - Il - I "1
He 0.79 1 2.215 x I2 2.948 x 10-13 1.482 x 10- 11
4 2.208 x 10 - 1 1
5 2.583 x .0-116 2.914 x 10-11
7 2.163 x 1-3 1 1
8 1.560 x I9 1.196 x 10-11
10 1.035 x 10-1111 8.394 x 10-1212 5.071 x 10- 12
Sum=2.165 x 10- 10
2.18 1 1.762 x 1 -11
2 2.165 x 1 - 1 1
3 1.058 x 10-114 1.599 x 10- 1i5 2.047 x 10 - 1 1
6 2.536 x 10-11
* 7 1.959 x 10-11
8 1.399 x 10-11•I9 1.0315 x 10 1 1
10 8.426 x 10-1211 6.742 x 1;3-1212 3.944 x 10- 12
Sum=1.745 x 10-1 0
SI
He 3.30 1 2.047 x 10- 11
2 1.958 x 10 "113 9.322 x 10- 12
4 1.576 x 10- 1 1
5 1.559 x 1'0" 1 1
6 1.984 x 1-11
7 1.642 x 10 - 1 1
.r 3 1.215 x 10- 1 19 7.626 x 10- 12
10 5.666 x 10-1211 4.152 x 6-0 12
%ll12 2.3 3(66 x 12
Sum=1 469 x 10- 1 0
44
if or W I
N ,- ar.-~- ,w No
TABLE 5. Summary of Conditions for "leisurements of N-3)
Exciteaion Rate Coefficient.
Time from
"Distance ( 2 In-eetor to ExcitationH2 inTector to Observation Rate
Flow Observation Port (cm-
3uffer Pressure Velocity (V) Port (D/V) molecule - I
<. Gas (torr) (cm/s) (cM) (10 - 3 s) s-
Ar.36 285 30 10.0 1.36xIC -1 0
0.710 3047 30 9.8 1.C0xv0-
Ar 3.06 1155 2 17:3 1 42x'.--I 5460 0 2.1. 12%tHe 5480 5.5 1 Exi '-
-e 3.30 j 1333 20 15.0 1.49x!0 - I
!e 2.18 1961 Variable 2.0-10 1.74x13 - 3
( Sliding
%injector) .31< j
O__ ,___________________________
. ,' , I I I I
. . 0
A
0.I c0.1 1.0 1.5 2.0 2.5 ..' (2D)1xr'NF(a)! (1,022 moecules2 cm-6)
Figare 23. Composite plot of :N2(3)] as a function of the product
NF(a)]x[N(o)] for al data collectei; A: Ar buffer gas
(0.70 torr), 0: Ar buffer gas (1.37 torr), ,:I Ar buffer
3as (3.06 torr), 0: we buffer gas (0.79 torr), Q: -e
b ffer -as (3.30 t-rr)) <: -e buffer gas (2.18 torr
45
ZI
.000,
A 0 0r .S. .r'
a A 0- 43
A 0
= A A
0
%- A
I A
Figure 24. Populations of N-.(3;v') as a function of v' for two Ar oreSsures.
overtone emissions are stronger than the NF(a X). in that case the :'F(3)] is
underrepresented; the N 2 (3) and HFt are too heavily weighted. It is a problem
that we are attemp ting to alleviate. For the present this was cirzumvente "
working at conditions where ';Fta+X) was at least as strong as either the
HF(3,O) or N2 (3-A) in the region of 'IF(a.X) near 574 nm. The problem was
especially prominent when the aliding injector was used. Data recorded froa 0
to 2.5 ms downstream of the H2 injector were heavily contaminated with HF(3,0)
emission. Consequently only those NF(a) data for t > 2.5 ms were used. :t is
significant, however, that when clean NF(a X) spectra in the region of
2.5 * 10 ms were utilized the data gave an excitation rate for N2 (B) in good
agreement with all fixed H2 injector data. (These movable injector data are
included in Fig. 14 and in Taoles 4 and 5.)
The initial result is .uoted for the excitation rate coefficient, '.3, for
'I(3-A) _rmation in vibrational levels v' = 1+12 as
- .3 (1. _- '.7 x 10 - 1 ' ) cm3-.oi -' -
40
•~~~~~ I%. W V',, ',, . - .. ,. ,I , ,,I I%- ",
where i-.e error s "the best estimate of the systematic uncertainties -tnerent
in the neasure-enat. These include the calibrations for the a(2D)],:NFa):,iand .'2 3): diagnosz-zs. *:' '.C/O calibration procedure introduces a
29 percent error: 4 sercaen' for the 0 + NO rate coefficient and 15 ter-ernt
recrcauciziliz in che :easurements of the NO/O titration). The estimated
error in the N(2-,, calibration is -15 percent. The reported value of the
NF(a) lifetime probably carries an additional 25 percent error. Comoining
this wi:h the results presen:ed in Table 5 gives -42 percent. The apparently
l3re error is cue to the errors quoted in the literature for results that e
have used to reduce data.
2.4.3.4 Vibrational relaxation in Nn(B)--Since these studies were
oerformed at several pressures fromn 0.7 to 3.0 torr some vibrational redistri-
buticn was observed in the NK3) manifold. Examples of this are shown in24 wnere t-e popuain N)
' Fig . opu)ati-s in N2(B;v) are plotted as a function of "
for two Ar pressures. At 3 torr the effects of vibrational relaxation ara
seen, especially for v' > 8. (The bottlenecking in v' = 8 has also been
observei in the N2kA) ener;y pooling work.) A much flatter v' distribution 4-3
-.bserved at 0.7 crr.
:t is clear t.-at the citribution at 3.0 torr is not nascent. While we
cannot state with ertainty tact the data at 0.7 torr represent a nascent
.istriu:ion it is certainly much closer to the initial distrioution produced
by the N-7.'a) Clearly reaction Clr one must exercise caution when -nter-
pretin7 populatirns in N,(3) from chemiluminescence intensity measurements.
't is of interest to examine the individual rate coefficients (kv,) for
';2 (3;v) production as a function of v'. Figure 25 presents plots of kvI
versus -.' at two Ar both -as pressures (3.06 and 0.80 torr). Also seen arethe effects of viorational relaxation in the vibrational manifold of .Z(3).
The differences between the respective kv , at the two different pressures ara
apparentl' due t vibrational relaxation prior to N2 (3+A) emission.
47
- I I
fucto o ' o tw Ar5 prsue. t-7 0
Fi-ure 25. Denendence of measured excitation rate coefficient, kj, as afujnction of v1 "or two Arpressures.
2.4.3.5 Other observations--The data presented in Taole 2 snow an
interesting relationship between the excited species prouced and te am:,unt
of H! a ided. :n Fig. 26 we plot NF(a)], N2 (3)1, and ['k :)' all as a -
tion of 7i " ad ei using the data from :able 3. The rNF(a) seems t. reach a
Sstaassy va-.:e as previous!y ooserved. 7-n contrast j ('2); first incre =sas
reas.na-li consistent with the previously measured quenchin3 of ,,2 o
I H2 (R~ef. 39). The P 2 C)l follows the same pattern.
Since 12 (3) is in steady state with 4(20) it is expected that at constant
[NF(a) .the ratio of N 2 (3) / IN(2 D)J will be constant. In the last column of
Table 3 it is seen that for the first five entries <NF(a )> = (7.1 - 0.2) x
1 0 1' c m 3 a n d '' 2 ,)' /'( 2 D) ] > ( 7 . 2 t 1 .4 ) x 1 0 - . S i m i l a r b e h a v i o r w a s
observed for all rns. As expected, the value of [N2 (3)1/' !;2D)1 decen4ei
5 , v n .> ; ,a , e . g. , f o r o n e s e t o f r a n s [ N F ( a ) ] = ( 1 . 3 1. ) x 1 C 0 a n d
2 < ;= (1.3 i 0.2) x 10 - 3 . However, for any set of runs wit- cor.-
stant wF , , :::- / ;2 W as zonstant to within -2') er ent.
48
,,5 , ,,,. ,. ,,. , . % - % % % % - .. 59.5 % - ,
I rN- [,F(a)] LUNITS OF IC I I
'A UNITS OF 107 -3..
[ [N( 2 D) (UNITS OF 1010 cn-3
<-33
3
w.,,;. * -
C 1 2 3 4
CH2 . (1014 cm- 3 )
ADOED
.26. Plot of () and N(2 D)l as f-inctions cfadded. a) Total pressure = 3.06 torr (Ar); effectivereaction time = 17.3 ms.
-I
_n the third quarterly report it was pointed out that the N2 (A) proiuc-
tion snowed a strong dependence upon the amount of H2 added to the micro e
discnarzed .a3 he current results demonstrate a similar iependenze cor
- '3) and : a(2D) as indicated in Fig. 17. (Also included in F.;. 27 are data
0 for :2a) octained by monitoring the N2 ,a-X) system at 149.5 nm.) No:e that
2 were quenching ;2(3) directly a quenching rate coefficient of
_10- 9 cm3 molecule-1 s-1 would be required to explain its observed decay.
Rather, as st.ate-i above )2 quenching of N(2 D), the probable precursor to
42* 3), 2.s a much more likely explanation.
The data cresented in Fig. 27 cobined with the previous observition on
asA) support the reaction between ... a) and ,(29) as being the key excit-itin
s B,) r .
-~ 43
-V.-
_ _-_ _ _ _ _ _ _ _ _ _ _ _ _ -1O0 - CN(2D)] (UNI1TS OF 1iCo :-,I
A - rN2u8o1 (UNITS OF cr-
,,,%ELATDE JNTS)
G50. 75 1.0 1..5
CH1015 cn 3 )
7ure 27. =lot of N(a)], (N,(3) -j 2 D) and relat ive s-,2a 3as functions of (H.,] added. Total pressure = 3.3 "orr;
reaction t [ie = 15 ms.
Dur "'ng the . rse of -.his prograN other observations have been made tnat
are relevant to tne potentiil of the i NF 2 reaction sequence as a source o.
N- (A,. There is not an estimate for 'N2 (a)]. We observed N-,Ca;v'=6) and
N 2(3:v'=12) whcn are bot at the predissociation limit of 10 eV. The only
two reievant reactions energetic enough to produce the excitation are
"(2 ) - NFa) N-2 - F = 10.05 eV
* and
N:) - b) -, - F I= 10.97 eV
As pointed out above the [NF(b)] is probably too small to contribu:e sig-
nificantly. Thus, the conclusion is that all Nn* may be produced by reaction
o. wi, e reaction 3 is exothermic enough to produce excitation up to the pre-
dissociation andi, end since production of N' is indeedi observed all the way
to pre iissociatin tn.ee-a zy be an inaerent inefficitency in th-e system with
1A res.,ect t: ;artti r.in i . N-ecic N2 " states.
P'':;1 , 9, , ,. -, ,, -.y v , "vv . ', . ' ' *,' '. ,. ., .
-=" Z-vi- L3 Z :zSer':az:on made wnile performing some early surv-y -wr-. r . .. e .,-(A) was produced using the H2 Ar/NF 3 ;2ichr
:'e _-zr, nd -h4 ':n;-<.*X ",e-zard4-Kaolan bands were =onitored . -n the h, Ce of
r z ..-: . - a mcrowave d ischarge was initiated in the H- flow.
* :-.e ";- w s e;r3:ed a -actor of 2. Although only a qualitative obser-
t accear -nit . quenching of N2 (A) or a precursor coul i be a
str_*_S :r-cle--. -ne ;uenchina reaction to form MH + H is slightlv en4o-
* - _. e .er 3ne quantum of vibrational energy in N2 (A) would make this
i •:C~ -- '-r . Per.nats ten opening of a reactive channel may exp'aln Why
:IZ e :7 3 D 3s0 f a
2.4.3.6 odeilnres2ts--A major goal of Task 2 was to cetermine
- va- <e': r3te zoefficientz that could be incorporated into the cot,,rehen-
. -- , .FlA. * :Ys code has been recently tailored to describe
.?xc4t .at_ of excited N-. To aid in the data interoretation some-f the :i "F2 system has been performed. Tnile not in-ence
* - i * :znpre.e3:,ve tady, the modeling results described in the following
lfll in daca interpretation.
-:;ne:.z , o_- s -sed that integrated a system of first order -:-
.- :w" -",n :-:en an initial set of conditions which consisted o- nmner
i__ re-:evant sacenies. The model also assumed that all species
reactir.s considered and the corresponding rate package
------------- -------- . -he comprehensive rate package of Ref. 4C -,as used
-i'nce t. ".ntercretaticn was a major goal of the modeling investizati on,
-_ r -t rtz 4ere conzent:3ted on comparisons of model predictions wita expert-
* merta :eta. r.e -ost relevant comparisons were temporal profiles of the
t ... -entees:. :he ma-ority of the experimental work monitored N-(),
-, , 'aj and .. ., the set of species that is now consiler.
,aroratory, Kirtland AF3, New Mexi.co.
_5
2.%LE 6. Rate Package Used in ".odeiing 3uao,
., Reference
(iI -r H - HF(1) H RC = 7.30E-12 cm3 ccIecule
-' z- =f. -2
(2) F + -2 HF(2) - H RC = 2.407-11 cm3 ccecJ.e- _-1 43
(3) F - :iF(3) + H RC - 1.20E-i1 -m3 oiecue - F.el. 42
(4) . N HF(0) + JFa) RC - 8.OCE-12 cm3 mccie- . a 0
(5) H + NF2 HF(1) + NF(a) RC = 3.OOE-12 cm3 ,oIecm0 e-u 1. Ref. 43
(6) H - " HF(2) + NF(a) RC = 7.70E-13 cm3 moleule - Ref. 40
(7 3 + NF, HF(3) -+ NF(a) RC = 1.20E-13 cm3 m oecue - 1 -e ef. 40
(8) - .-F - HF(O) + NF(b) RC = 2.60E-13 cm3 olecuie - I s- ef. 40
(9) .i +NF HF(2) + NF(X) AC = 9.10T-13 cr3 Zo I eC 1
cu ,-f. 4
(13) H NF(a) , HF(0) R (2 D) _.C = 3.10E-13 cm 3 =oIe u--I -aia -41r
+(11) NF~n, H + NF(a) RC = 5.OOE-12 c 3 mo ecu le - -,-i R f .
12) H'( 2 ) +~ NFa) NF(b) + HF(0) RC = 8.30E-12 cm3 molecule - I s- 1 E=. 4
(13) 3) 'F-)a NF(b) + HF(I) -,C = 7.50E-11 c-m3 molecue - I 3- 1 -f 4
(14) HF 4) - (a NF(b) + HF(2) RC = 3.30E-12 cm3 o-i.e cuI s . 40
1(5 ) N -;F) N2 (3) + F .C = 1.90E-I0 cm3 moleule - - -hs 4or.. --- '' - 4
(16) H" -) HF(I) HF(0) + HF(2) RC 1.70E-11 cm L ciacue - , a f. 4
(17) HE(1) + HF(2) - HF(O) + HF(3) RC = 2.OOE-11 cm 3 moLe ule-1 z-- Ref. 4'
1-3) i?) HF3) HF(O) - HF(4) RC = 2.20E-11 cm 3 moecule- - ! Ref.
HF4) HF(0) HF(3) + HF(O) RC = 4.30E-11 cm3 molecule- 1 Ref 40(19)
(20) HF'[ ) - HF(3) + HF(2) + HF(0) RC = 2.00Z-11 cm 3 moLecuIe 1 s- e 4'
(21 H(2) HF(3; F, F(1) + HF(0) RC = 7.20E-12 cm3 ma ecuie - Rf. 4Q
(221 I +', 'F() HF(3) - "F(O) RC = 1.20E-12 cm 3 , .91eCUle- I a f ,
H(23) :4F(4) 2 HF(3) + H2 RC = 8.50E-13 cm3 aoiecule-I s- I R-ef. 4
) :F(5) 7 'Fk2) - 2 RC = 3.90E-13 c 3 (0 ecuI! - -- 3
(25) ?iF2) "2 HF(1) R H2 C = 1.30E-13 c33 molecule - i e. 43
026) -1 HF(C) - H2 RC = 2.30E-14 cm 3 ,-oIecue1 --1 Re . 40
(27) H:( - HF(3) F2 PC = 1.40E-14 cm moecul- 3-f 43
2 -. -:F! HF(I) -R N 2 PC = 9.70E-14 cm3 molecule -' a- - f. 40
3(9 -1 -1 40.29) F - HF(2) + N-) RC = 2.50E-13 cm 3 MOLecule- I s . 3
HF (3 ) N F2 .C = c.0 --- cm: .. e. S- -3 A e.. --
3' ' - .C = 4.50EI31 s- . 40
(-32) F F 2 1 , N F3 + M RC = 1.)OE-30 c- 6 . 32 R 4I% -
33a N. , -. ;FtX) - NF2 ,C = 2.70E-16 cm3 fo ecu e- . a:. 40
N2(A) h C = I .50E -,5 s - ef. 36
(35) N-!A) + WAL NZ(X) + .C = 2.18E+02 s- I CAlcullate-4
(36) , - WALL N(4 S) RC = 2.34E-02 s- 1
(37) N(iD) - H2 N(4S) + H 2 RC = 2.30E-12 cm3 molecuie-1 =-I Reaf. 39
4(3i) 2VO, + :4' .1N(4 S) + NE'3 RC = 3.OOE-13 cm3 molecu:e'1 3- T.Ii s workt*
(39) ( 4 3) -;, 2 N F(X) + NIF(X) RC = 3.OOE-12 cm 3 moiec'le" I s1 Ref. 1
(40 :; A) ',A N2 (3) + Ni2 (X) RC = 7.70E-11 =3 ecu'e -. ,I s 4ork
(41)1 .% N - .NCC) N M2(X) RC = 14.6E-1I cm3 .olecule- This 4ork
(42 :F'a) - NF (X) RC = 2O3 s'
rAme t:r
* .re2.:iria/ ::e43izre:aen t.
*5 Z
% 52N
:he model was rin by Specifying an initial set of condi:tons f.)r .1
species concentrations. The rate equations were then integrated and : e
species concentratirns were predicted from 0 to 10 ms at I ms intervals.
'.This range was identical to that of the experimental rans when the sliding
in:ector was used.i
initial conditions in the model were chosen to be identical to those ofthe actual expertzents. The [Arlo and [H2]0 were measured directi'- from the
mass flowmeter readings. The largest uncertainty in initial condition deter-
mination occurred for [:NF 2]3 and IF]0; since we had a diagnostic for neither
of these their .zoncentrations were estimated. The upper limit of the "-
was [.F-30 wnile :F]' is required to be less than 3LNF 32. Secause molecular
hydrogen was introduced into the flow tabe, atomic hydrogen is produced by the
reaction of F - HI.
As pointed out previously, when H2 is added to the procucts of the NF3
discnarge, [NF(a)] rises then reaches a plateau which is independent of
further H2 increases. This titration end point is interpreted to occur when
all F has been consumed, i.e., [H2] 0 - EF]O . Examination of the data in this
manner indicates that approximately four F atoms must be produced by the dis-
charge for every five NF3 molecules that are introduced into the cavity.
The estimate for [NF2J0 is less certain. Since this model best fits the
data when LNF 2 >iF0 1 1/2, it appears that the microwave discharge croduces
about a 50 perzent yield of NF2 from the NF3 . Note that the discharge is
always run at a relatively low power, -10 W. When the discharge was increased
to 30 W the NF(a) and N2(3) concentrations were drastically reduced indicating
that at high microwave fluxes NF2 is dissociated. In addition as the micro-
wave power increased, N(4S) (detected at 120 nm by resonance fluorescence)
also was enhanced.
In Fig. 23 a comparison of predicted and measured profiles for is shown
:?( , N( 2) , and N-(B) us-n~g the rates determined for reactions 2 and 3.
3 i:n that the estimated uncertainties for these two rate coefficients are --. 0
anl percant, res ectirely the a3reemant is quite good.
0.%
1016
ID N2 (~' *
1O' ~ G rN 2 0) EXPERIMENTAL DATA
10 4 SOLID LINES - ,IOC;-L PREDICTIONS
iC NE (a)
~ N(20)
i107
0 2 4 6 8 1
.- ura Cc=a-.4 or. Of Iredicted and mieasured po .ulat:nprof iles f' ,r NF(a) , N(4D), and N:5(3). Total--avi as :.ressure was 2.2 torr (rie).
=4
"s J
It is important to note that to obtain the predicted profiles an
removal rate for ',F*a) has to ze included. The actual source of this removal ,
not presently understcod, but wall reactions seem to be an unlikely explanation
- (Ref. 3). This removal of NF(a) is an open question and the cossibility of a
bimolecular reaction must be considered. More work on this problem is indicite-.
As indicated, the modeling study was completed to aid in data interoreta-
tion. The model supports two important features of the proposed H - NF2
mechanism. First it is consistent with a two-step productIon of 'i2 (3) via
reactions 2 and 3 with k2 << k3 . Secondly, it supports the hypothesis :nat
,j( 2 0) is the primary produc-, of reaction 2.
For completeness we also experimentally monitored by resonance fluores-'.
cence at 120 nm :(4S) production from reaction 2. Although not an abso'utely
calibrated diagnostic the relative N(4S) could be easily observed. in
Figure 29 we show relative concentration profiles for N(4S) and 42 k) (which
as shown previously closely follows [N2 (D)]). It is clear that N(4 S) ,s not
primary reaction croduct and is only formed at relatively late reaction. times.
Its role in N2 * production can therefore be only minimal.
:n an effort to shed some further insight into the mechanism for N-' pro-
duction from the H + NF2 sequence we made some computerized data acquisition
runs recordinz both '; (A X) and 42(B-A) chemiluminescence. A sample s7-ectral
_t Of, o he :12(3-;0 spectrum has been previously presented in Figure 21. in
Figure 30 we present a similar spectral fit for the N2 (A-X) reaion, :n addi-
tion to the N2)A+X) emission strong NO(y) bands are also observed; these were
almost c-ertainly from N12 'A) to NO(A) E-E transfer. The [N2 (A)] number densi-
ties we measured ranged from 2 x 109 cm- 3 to 8 x 109 cm- 3 for a variety of
conditions. The highest concentrations were obtained at relatively high pres-
sures (-1 Torr) wnere wall losses would be minimized. Typically EN2 (A)]
exceeded [N 2 (B)i by about two orders of magnitude. This is in qualitative
agreement witn our modeling predictions which has N-(B A) emission as the only
scur:e cf ;.-(A). While not wisning to extrapolate the results of these stul-
• j~i des bevond their range of alinil:, tey do support the supposition that t.e
,kinantrou...;n of " 2 ,. is y N2(3-A) radiati.-n.
55
2H9 e *( ):3, A) ,., J *12 . ±.., .moes-s--
$ U N(4S)
INTENSITY i(AR31TRARY 5 1UNITS)
-4w
3
2
0 5 10 15
-LREACTION TIME (1O-3s)
Figure 29. Temporal profiles of N,(3) and N( 4 S) oroduce- in the H -
.reaction sequence.
NO(A- X) v'=O
NO (A X) '-r- ,4"5 6 7
1 . 20 - L =!...N 2 (A-,X) v'=O
-i6 7
z0. 4C0•
S"0.40
249 259 269 279 289 299 309 3i?
WAVELENGTH (nm)
F;.gure 3k. Chemiliminescence spectrum (dar, line) 3nd spectral fit
(light line) produced from the H NF recion -e,:ene.Prowinent featires are indicated.
Or :85
3. N, E ERGY- NSFE UDIES
This section discusses work on understanding some energy-transfer rezc-
tions of electronically excited nitrogen. After Burrows (Ref. 41) demon-
strated opticall-i .pumped lasing of the NO y-bands, NO(A2 -- -- X2-), t.e i:ea
of using 12 (A) to pump the y-tands chemicaily in an energy-transfer react-r.
followed. The studies on the N2 (A) plus 'O energy-transfer reaction have two
different aspects. First, determination of variation in the electronic tran-
sition moment with r-centroid for the NO(A 22 -- X27) transition had to be
made. :t was found that proper interpretation of the enerzy-transfer reslts
required this information. Secondly, both the total quenching of N2 (A) by ';3
and the excitation of NO(A2 *) and NO(3'2. by N2 (A3Eu+) were studied. ne
excitatlion-rate coefficients have been zade state-to-state so that we can
soecifv the efficiency for exciting each of the NO(A) vibrational levels 3yeach of the different N2CA) vibrational levels present in our reactor. ?ara-
grach 3.1.2 details the experimental facility used in all three sets of mea-
surements under this task. Additional experimental details appear where
needed in Para. raohs 3.2.2 and 3.3.2.2.
A major concern in developing a chemical laser based upo . energy transfer
frcm a storage state of one molecule or atom into a lasing state of another
uolecule or atom is now large a number density of tie excited reservoir mole-
culas can be generated. The reservoir molecules can be destroyed either in
quenchin4 reactions with other species in the laser medium, or in self-
quencnin3, or energy-pooling reactions. :t has been found that energy pooling
of ")(A) molecules oroduces nitrogen in the C3 u and B3. states as well as
generating emission from the Herman infrared systeu, the upper state of 4hich
has vet to be characterized. Paragraph 3.2 describes state-to-state kinetic
measure!aents of the rate coefficients for producing various vibrational levels
of each of the products from the various vibrational levels of the pooling
N-(A) molecules.
:t has been shown previously that active nitrogen excited IF(B3-0) effi-
ciently (.ef. 6). Although the precursor of the :F(B) emission unequivDcaily
c:i.J not be -3ntlfied, these 3ata i-.-ated th-4t N2(A 3 u), ',,3.
Wwr r 7
1) (a N)( and NC)orobably were not t:Ie :recuraar states.
Rather, our evidence pointed to N-2(X ,v.) as being the Jliely precursor state
with high vibrational levels of the ground-electronic state~ of nitrog en Y-bova
v" = ,I excitjing the - in a virtoa-oeetoiV-E, transfer. Para-
graphl 3.3 d-escribes zcmze add itional studies on the nature of active ntoe
and its subsequent transfer of energy to 1F. A diagnostic has been d!evel.zoe-d
4 for vibrationally excited, ground-state nitrogen, and have further character-
ized the energy transfer reactions between components in the active nitrog en
a nd :7'.
3.1 sr:-:-T: ENERGY TFANSFER FRO;M N-)(A -U, v1 ,,2 ''40(A2--, v' = 0,1,2)-
3.1.1 introducti-on-The excitation of the NO y-tarnds in t'.e
energyz.aasfr reaction between N-)(A) and NO is now well establi-shed(Rf.43,44,45,4'3). What is not well established is the fraction r o f to.;ta I
V N2 (A) quenching which results in NO(A) excitation. The published valaes of
the rate coefficient for excitation of NO(A) by N2)(A) (Refs. 43,44) are bcoth a
:actor of 2 greater than mcst of the measurements of the rate ccefficient for
the des!truction of N,2 IIA) by NO (:Refs. 42,43,45,46,47). The :.agnitude of this-
~. -discrecpancy deizan"is further investigation. In addition, the state-to-stati
partiti cr.n btween vlbratior. al levels of the N2 (A) pumping reagent and the
NO(A) is uncertain- (tallaa: and ',cod ' Raf. 45) claim a strono differen -ce n
ra:io of NO A) v' = 0 to v' = 1 excited by N2 (A-, v' 3 (3.8; ) c - are
to that excited by- :;2 (A, v' = :j (19 1 Some pr--li-minary resailts frou a
study* a number of years ago alimed at using NO y-bands as a monitor of systa~a
purity (Ref. 493)nict a much smaller difference (7:1 and 4:1,
respectively). It is also not clear if there is a strong difference i4q the
juench-Ling rate coefficients for the different N42 (A) vibrational levels.
Dreyer at al. (Ref. 47) found *'O quenched '- 2 (." 1) almost 70 percent faster
than %2 (A, v1 = 0) while Clark, and Set.ser (Ref. 43) and Young -and St. 3onn
(e.44) say both Nn2,A) levels are iuenched by NO witn equal. effici-enc'v-.
j -:)~nse-.'iently, -3 careful investigation w4as undert.Pken which Is rec,-ortei 'az~
3. ?n;s~i. c~r~;:1C., Rasear-. Par,, P.O. Box 30, An-cve r, '!A
)00 ISZk
3. ...2 apntal--T.e apparatus is a 2-in flow tube pumped by a
' eybold-Heraeus -oots blower/fore.ump combination capable of producin g !;.near
velocities up to 0,3 Cor s-1 at pressures of 1 torr. The flow-tube design% .
-cis modar (74=. 1), with separate source, reaction, and det'ction sections
% n'which -la =p toze:-r with C-ring ;oints. It has been described in its various
configurations a number of times (Refs. 49-54). The detection region is a
rectangular stainless-steel block bored out internally to a 2-in circular.
cross section and coated with Teflon& (Dupont Poly TFE 4852-201) to retard
surface recombination of atoms (Refs. 55-59). Use of a black primer crior to
the TeflonQ coating reduces scattered light inside the block dramaticall'.
The block has two sets of vlewing positions consisting of four circular or-s
each on the four faces of the block. These circular ports accommodate va :uum-
S., ultraviolet V V resonance iamos and visible monochromator interfaces, laser
deliver. side-arms and a spatially filtered photomultiplier/interference fit-
ter coMinration.
In these experimnents a suprasil lens collected light froa the center of
the flow tube and focussed it on the entrancL slit of a 0.5 m "4inute:aan moro-
chroator which :s outfitted with a 1200 groove mm- 1 grating blazed at 250 n7.
-ig -
F7]nPhot n P--"
i L, - Q= To
30 00m4 .6,., ---- ] Fn
INJEC7C A
Fz:re 31. t'lc" tube apoar3tlis configured for No-(A) decay ',netics
e : e:r
*59
* ~~ .'cno:c A -old tcmultiplier (:-WR/Da- Z -itecte" cnctcnsi
-wia of a P SR 1 1 C cnoton-co'unting rate meter. ~A 1zcratory corn':'::er
syst-a- 1: :itizaed the a n al outpult from the rate metert and Storedth :r-
tion on floppy c~ssfzr farthr proceszing. The computer systaeo co-_cr'_Se. an
:S',I P'2 :5: -' oZ t., wo -26C K disket tea dr i-ve s, a m:nc r ::e zC :-,, Lr , -a n:
a 160: cs iot-matrix printer with graphics capability. Data 7ranlatin
=anufazzaras :_ne '/0 sy:stem (D*T23O1A), which featu:res 16 channel-s of ;/D
inpts,':w _hanelsof /Aoutput, two 8-bit digital I/0 ports, softw'ara
prorim-z.le 33,single-ended or differential input, and dat:a aiii
rates as -fast az 14 'knz Laboratory Technologies Inc software -cackaz.e
Iabcrat::ry Notebo:ok, " interfaces the computer to the _-/A hoar:a
organizes .data In a form comcatible for analysis using the= Lotus 123 tcusIn*sS
screds~eetSoft,iara or for sendJing to the PRIME 40,0 compuiter .4- PSt's
pite r a tr fr anal/!sis on that machine. Much of the analy-sis e: e
around; last-s~uaras fitting of soectra. This code (Ref. 0,generates aa
functions consisting of a synthetic electronic spectru ocr _) unt o.1at
-n each vibrationqal level of each electronic state appearing in tne s:;ectra!
retnof intar- st. A least-squares routine then finds the popl'iatiorz5 of
aacn :r'ncband which when multip lied by the appropriate 1basis rcto
;jivas a cozcste spectrum most nearly matching the exeietlspec:zum.
Stan-ard uartz-;halogen and :22 lamps were used to calirr3ta tne sc:
sv,,ste_- :).-o rela -ive resconse as a fuinction of aen::.Exl r aer.
_twa-, :cosrved iind calculated intensities of a numoer of ofn~ : n
N~A:.'- X I __) systam tetween 20and 400r confi:e th~ rea it;
1Tie ratnbetween metastable Ar( 3 P0, 2 ) and molecular ntoncoaa
t~ie oataS'_ahle nitrogen molecules, N2 A 3,u) (Refs. 61,62). This trinsfer
exct3 - -53 wnicOh 1uic*-Iy cascades radiative.,. to thie cetastable .2:_
'5 th1 t e V 1 state. A hollow-cathode dizharge source crerati-
24: *',,z - >o-A pr:;.uces the : :on ,.ezastablez. :,Ie A:Ior a:1-4 r>~
7:)Ir fi2_ 1yf'- :ng them thr-o.i.:h trap filled witn 5 A mole:ular siev%, :.-* '
72 K ~ r :~~ nt-1vn it is iot niecessarv to: : t~e
hee*1cmn_ - e'.33t'Iein
C_-se r ations :of strono Ve -:a r d-K ~a oan , N, (A3~:. U_ X )eaiss-o cn
strea= -f the Ar,/' mixiriz -one conf irms the orc~duction of th.e nitro-ea --eta-
stables (Fig. 32). Codisctarging the N- withn the Ar increases the *;-,A) yield
by a facto:r of aocu-t .3 (Ref . A34, --,t it has been found that this n)rzce-d ra
ilso Z ocs.omne atnmic Nn~, vibDratiorallly excited Nn~ and rMetaitabh'e
Nka -' Ref. r,5;. 'Jnequivorca. meaoorazents or.nA - ect-rste-
*fore, demand that the N- Ice added downstreai of the dischar~e.
~\/ 3 2 0 91-m0
v-a 'N,(A 3 - X
V46 7 8 9 10
V" 3 4 57 8 90 ii:2
7j 8I 1 02II
O A 2 ,7+- X 2- C
V" 4
2p 300 3 50
* ~Fi.ire 22. VeaiYa~nemison in flow reactor 9 -is iownstriiao fr~-:-the :Iiscnar 'a.
Nitric ox_ te enters the flow tube tnrouga a 1-in-dia loop -niector Satei
onth end of a 1/4-in-dia ture which slid_-es alon,4 the bott.'n of the flo:w t.ub e
3ri4 nrallel to its axis. --his allows a viariety of reaction dista:ic- s fzor
a~'rt ~ntzstudies. A'iaing CH, CF3-{, or CF.; t) thle 3.55s streso:: t-.r-~u
a f~e: hc~-za~ein~cto, ust Jowistream from -.4here tile N-1 ent-_4 '?_
f~;Vr- ~~tu re.-ixe -A.'izraionl t:~n ithutsi3nif icant tr'i
I.-
mlA
L% % N
A* M A C
4t% ~ ~'.ss-ffizw -e'erS -or rz tamaters =oni tor g~as fiow rates. All floweers
we re c-allbratenJ '-v measuri,-. rates of: increase 0: pressure w ti-,.ie into
or Zflas'ks, us~rg apcrzpriatte dIf f r n tia!. oressurca transc Ucers (al:n
D? -1 w-. 4 zh t - -se Ives ca' een caliz-ratze3 wit'n sUi On oi :r Mer'-- -
me t-'zz. :y,. C flIIo-W r a tes -cor A.- , N a. nd e th ro h thea -:: r a r -
50CO, 130-530 an. 33 mo! s-, while the NO f low ra te ranzes -cetween 3 orI
and 3- and 0.01 _:C 3-1 ror decay or excitation rate measur mnts, rescec-
tive ly. :oDtal pressuires, as measured by a BaratronZ capacitance nianonlet-2r,
rance fronu 3.3 to 13 torr, and flow velocities vary frooa 500 to E5200 cm S- 1 .
The number of2nrt a af:raactat i' is given by
2tot 0
where the f4 represenit reagent flow rates, ptot is the total flow tube- pres-
sr ntrN Avogadr-o's nuber, R the gas constant '6.236 x 10-2 torr r
~uol~K 1 ),and T the absolute temperature.
N.tri: oxizl I's purfied by slOwly7 flowing it ait aoseic pressure
t: r ooah an -AscariteL trap and,, then through a cold fin,-ger su..rounied by :a
m e thanolz r,- * sIU S' h ath 0'- 10C~ . he ,;o 3s then Jute h-%r andth
mn*x,::ris store-d in -jrexD flasks. Mixtures of 5-6'. NO sufficed for d4ecay
rite eauent z the2 e: caon--rate deter-aination1s re iu;.red NO mole
- Vfractl~ns < 1)2~.
3.1.3 7he -lrroi ranztstion '-izaent for the NO(A 2 , - X
-ransitior.--7he issue of whetner or not there is a significant variatio)n in
the electron-4c transition moc-ent with r-centroid for the NO(A + - X) tran-
* sition h-as been thie suo;ect o f a number of papers over the last several
'eaes .e. 57-73). Wh-,le a number -of )acers have shown e~ec fast
nifi -ant trinsiti-r-oment crent i'jrjati :n (Refs. - 7-%L sev~eral :r:o,:os have
din~eiths ~tston~R: -7 . References 72 and ,; re-ji-a. %;ost 3z
thIe rllavant 14itera';r-!. '2e a eqne r al consensus in the scient.Lf2.cco ut
se : -D -e tha t no s . ni icanit trifls~t moment v a r-3t:r. occ:urs for t:,,e
"'2
.7 rIk * 1 4
*-and s Rece n t-, Whi 1e S tui Vi4n g the eiectronic: energy trinsoer rze-w"
-4H _ NC, :nconsistencies were round in NC(A excitLatio:n ratres
~IKJ easu:red uisin different zaisori..inating from v' = 0 If a constant trir'si-
tinoznt was =::<d ::;e excitation of the NO -- ands by N, I.A
a ner~ - vtran r. s f '1 (X o rovides a sourze of vt-band emisslon ihz s
free from ot'ner signorocant overlipoing tand systems in th'-e spectralrei.
was found~ that the coservei branchin2- ratios for transitions fron a non
%ibratiornal level in the ucper state cannot be explained b-y variatiorns on "e
A ra~c~~n~Onfact~r* Cbst-rv'atiors show variation 4in the electronic rs-
ti4on :mozi-nt or -acre than 42A over the r-centroi.d range 1.229 . hscc-
Ata-tizon is suoortei in t."e fz!Itowing paragrachs.
Co-rreact transition orzocabilities for the NO A-X svystem b-ear dirzzty i:_oa
*atmosconeric science throuch such processes as the measurement :)f NO co:.oon
densities in the cLesochere ~e.79) or the interpretation or emission.s in a
stronz aurora. n- addition, proper NO A-.( transition probabilities are -e.:e
to calc:ulate thea -ga3n for various transitions in the optic:Ally -U=Ped NC A3
laser. However, -,roper NOC(A-X) transition Probabilities affect a .4iierrne
of studies, ever. eccause theay are easy to excite, the NO -- ands are_ orten
AX. u~_sed to establisn tne rela tive spectra eoneo crohoctr
I t vIe 7 3, 7 3 rJsong incorrect bran~hing ratio)s f:r ztne A-I ran.-
sit.-on w-411, of course, reasult in an incorrect response fu nctio.n, anuo
-~ there'd- invalidate. al! other meazurienents -apcnnouon t:nc soc ctril
rescns teo~ncfrom t'e -.- ban.. branching ratio Lea ,re.uen ts
A nijef ae-nue u'esec on, NO~ A-X(taniin has -ecoac.3 an
increcasin:. importin: to(,l- for o:r7b ing t-e vitrational Jistri)utir- of
;romn elezronirstate ;NO -ouce, in checicz3l or photol,,tiz rati
('Refs. 321. ncorrect values for tae NC,: A-X tZrAnS;.tIor rabl
inva-14iates3 the r-esu. *s of. tnese rieassre-rents, anai the W .. o hvaioct.
inteoreatins f th,,e re3.olts -ecome ~ rt f.
3tt.i''
%~e n . :: ~
.4
Franck-Zcnlcn fa3ctor, the cube of the transition frequency, anc tne s.::r c
the el3.ztroni'z transition -moment. Thus
" A onst x N -v Rl.l (r 2.,,vil vi v V '' v V"
::~trr~s::u-cenles coulatior. in the urcer state is th-en c-eter7:-necd
from thea ratio of the integrated band intensity to the product :)f the
?ra~:-or~on a':tzr and tha cube of the transition frequency:
AV I (44)
:he rati-oS of Csetrans ition-momenttess oloculations to each- other shiibe
constan iss'~ transition 7-omnent varies -with r-centroi~d. -l rl-,v
variatio,.n in thea transition mioment -with r-centroid results fr-om the razio of
A*tne varicas mocoeat:lss populations to one reference 'oula-ir. scalinz the
relative transi:aon moments to experimentally determined Lfi.eor
cszcillator-strangth- data is then a relatively simple process. The srectral!
resionse finction of the monochro-mator,'detector systea was acc~irataly l-
bratead an-- verif--ed by fitting the IN2 (A-X) Vega ri-Kaplan tz-rsit;On fea:tores
w.~ncozurs in :.ne same sr-ectra. region.
:t --as czr-eot.-at transitior-moentless ooDwilations of t--.e ss
e~iaaun;frD:a v3and I using pectral fitting routine for each 2:sae.; e
r co o *~ = -. he fitting routine corractec. for s-ecztral overia-,
zetve;:te trar-.Siti ons for 1/l = 3 and 1. Tefticcao:ee~ fet~Th i nlue a-peer f-c.,-
*terocperatures for the two vibrational levels. Each -of these popula-
tion~s was rati'oad to the ones determined for the --- = -3 saquanze. -- t.s
jinc s In the midldle of tnie wavelength range for the band systeo -and,
4 ~thlos, shouldj -3--iize any systeaatic errors in the relative monochromat.or
reasrcorse f-isctl--r. The F'ranck-Condon factors of both Nicholl's (Ref. 62) and
A Lritton et aL. kRef. z2 oave similar resuilts. The final results incor:)orate-
e .- .e .- . - - -
%
% % %
il'"prCS
:~re:~5z: t r:; e a~2 rns t o r. mome nt re 1a z-v e zo it3~~
z: ::z -3g in r-cenltr iI 7me variatilon -z~.-
-rznth =-3 se.le nce , but r i s es mu ch mcre s r - l
-. - z~ >e-2~S. me -4amond-s in 7ig. 33 snow our raanalysis 04
7 o~ :fa.3oremerts M ,c~ee et al. (Ref. 7-7). They excited NO (A,
a ae ar in-1 rec;crta-4 relative band emission intensities. er
* wi dt:. zurs. 74-nally, the transition moment variation was ccon--
- -~var-,aori was cie y Brzozowski et al. (R.ef. 71 ) who obsefrv-2
-I -e t M'.g5 ui's n N
J T~~.+ 0V.1
I .'0 MCGEE
B RZQZOWSKI
'1.2
'5,7
7-ETRlSFZ r.
<ioe3.'~ito.i lcrnc rniincnn ihrcnri
0. fo tne:;Q(-:~ ra(sti3n+ T
0~ 9it terutaot.etreeerenst adAt finctn
S.0
AA -A
0.9 0.r 0.9 1~ 1.0 . 04 1r 0 1.08 r . 1.1 1 . 4 )
C */'*?*Cre.ROV
23 1r a ir n -5.to i: ris ,c om n ih rc n-:-,
f r tn
we.tte t ft5.he xeijet o- qarai uci,.o
2:-,~' t d
IR (~ .'3 8 :3.7 .2 . I I
-. --w"%'% W ~ - ,I I - N~ -' - rvl % -- ,- w ivI. 'fh. J-IV '. I _19T -d-
Uising this fnctional form f .r the relative transition moment, we determie/ a
smoot ed set of branching ratios for emission from a given upper state as
being
ER - %'v" viv"' _v'vII relv' v: 3 ( 2 %46'-'.' qv'v"1 V "IR e )r]2jd
V1 rel
The E-nstain coefficient for soontaneous radiation from each band is tae proc-
uct of the branching ratio for the given band to the total radiative decay
rate of tne upoer state vibrational level. The average of nine apparently
relible determinations of the fluorescence lifetime of NOI(A, v' 0 ites a
value of (202 ± 14) ns (Refs. 74,76,84-90). Eight different determinations
for tne v' = 1 level give a value of (192 14) ns (Refs. 76,84,65,68,69,91-
* 93). :n both cases the error bars represent one standard deviation. Table 7
lists the Einstein. coefficients for each band.
Equation (47) relates the Einstein coefficient for a given v'v"
transition to tne absorption oscillator strength:
p -as a e u
;v 2 2 d, Aviv"
,wer me is the electron mass, e its charge in esu, the soeed of light,
the :rans:tion wavelength, and d and d, are the ele3ctronic degeneracies of
the uoer and lower states, respectivelv. These last quantities are 1 and 2
for At and X2t, respectively. Applying the Einstein coefficients to
Eq. (47) gives abscrotion oscillator strengths for the 0,0 and 1,0 bands of
4 (3.9 - .3) x 10- 4 and (8.2 t 0.6) x 10- 4 , respectively. These values agree
quit e 'well with literature measurements of (4.03 t 0.22) x 10- 4 for the 0,0
transition (five different experiments) (Refs. 94-98) and (8.26 : 0.48) x 10- 4
-: for the 1,0 transition (three different measurements) (Refs. 95,97,98). Ths,
the transition moment finction satisfies the important criterion that the
lifeti:ae and a7esor=ston .esre:3entS be consistent.
066
-4:,, C :e ff -c Jents f-cr '-(A , X
-~~, r'~ vdri: B ranc.;ag
I 3 o548 3 '569 I. 291 3.921429 *.'26
''9 .33o
4 1 0 .390505 1.33 .Z07 3:a8
3 85.17 3. 456916 *.3135 1.091448 0.338894 i 1.3: ~ . .:21,85 1.977 1.166a03 0.017717
82 2 309363 0.382l 0-5046 3.07673 3
8 23A.- ).203954 3'.36a4 1.34-1121 ).-03',66
9J 3--12 .9548 1.429015 3.0259
%N -536 2.2.2264; I .34ia 1.540i95 .3004653z.6 .C00253 3.a295 1.645876 3 001I83
1 ;.2. , 00 102 .. )177 1.-541 10 0. 23CC67 ' 33E
- ~ 4 ve 1 Va..) Sr~iI..ac ... -
-V I e
9 2.25. 134 1.3353 1.1364 0.962957 .45 179 5 .iz37
224.122 02:0295 1.1108 0.922,347 10.1131988 585-
I" '.30097 I 1.22a 0.?15801l O.C00925 4 -2
3.47231 1.3754 0.237 0.061794 2 '-~~ -5.446 , .332 I1.)551 0.9571.53 0.67
-0. 13257 1.3366 1.C04782 0 .1017-46 ZSC.4:6 1-197792 1 .0196 1.065i39 3.3732099 3.SC-
1 94.644 3.625 .,;039 1 . 35635 0 .:,44406 3 i~'09 3 '9.)90 3.333426 3. a888 1.215151 -3.32413005-
2: '6.723 : 2.32042 0.9745 ,1.301742 0.31204310 45 . 32- 3.23082i2 o.9609 1.394282 0.355 453
1'1 365.139 ".080 '398 .4912-16 I .302Z524 1.3.241 2 3.57.-12 3.01 '3.9357 1.592003 0.0084 5.64!Z-O^-
131 4'1.a7 0.:00753 3.3238 1.697225 03.203451
. 14; 429. z36 )-. 002 u.s 6 i1.803203 3.3201833
'15 40i .-9
-i 3.03 .3. )017 1.912,310 3.Oc00c 2'lz -
avser.g2 - R,- racln
4~~: j 32 2 246 264 .353901 0.:466:219--~. ~ I 3.017314 -.15,S 1.303423 3.-2247
I2I~ .8 0.15636 1194 I .932396 3.154996 433 3 '.372i9 120947 .925646 3.28696 5 I 3 s23'-SI-s~ .;-0 :4
-. 336 03 - .1 1 .2644 0.339840 3.:2936 E'..
61:345 440 .9838397 0.057946
2.2573 .0 1.339439 0.055914297E:
.30 j 3.62245 0.11951 i1.180487 0.338608 22
2.369 38476 13.9601 1. 2628 a1 3.323464I.2!9 ;
11 1 26 4432 1 89 _.967 1 1 .3506Z3 0.312995 7. !4:E. 413 7 2 5 .29 4 .4 l i589 3.9542 1 .4435-23 0.006707 SF sr--.:4
12.--s0.8991 0.9419 1.540195 3.303273 '5i4 ~i6 .061 0.3301 1.640572 0.301534 .3-2
1.~5 42.582.3 0239 0.3289 1.74276 1 j .063 3.3iZ-2'3!6 4-:.i43688 2.3281 1.3476830 :)c4
775: =er .'..i ec,:- -
67
Einstein coefficients have been calculated for v' = Z assuming the tran-
sition mcment variation of Eq. (45) and a radiative-decay lifetime for v' = 2
of (132 =13J) ns (.Refs. "16,84,85,88,89). 3ecause transition moent vari-Ation
V was extrap-clata-i to regions outside the fit, the transition probabilities from
P are less re!Liable. This may be reflected in the modest fs~ezn
between the absorction oscillator strength of the 2,0 band caiculated from the
tr3is-ition orobabil ities given in Table 7 of (8.1 ± 0.4) x 10-4 and the exzDer-
imental value of (0-.8 t 0.2) x 11- (Refs. 77,79,80).
Figuires 34 and 35 show a comparison between the observed and the
- ~ synthetic lbest fit spectra for the case of no assumed trainsition =Cmentt iafia-
tion, an the transiti.on moment variation determined in this work. Clearly
including the transition moment variation makes a significant: difference i'
the quality of the fit. A poor fit could also be the result of having anl
incorrec-t relat~ve monochromator response function. In tnat case analy!sis for
the transition moment function would also be invalid. This does not ajn)ear to
be a significant problem in this work, however, because of the accurate fit of
the 'Vegar:!-Kapian bands of nitrogen over the same wavelengzh region. In addi-
tion, the good =agreement between our own results and those of the twio othier
groups alluded to previously (R.efs. 71,77) confirms that the wavelength
respo- Se ftnction i's accur-ate.
3.1.4 The Kin.etics of the N-,(A) Plus NO Reactio-Cm te characteri-
zatizsr. of the energy- transfer reaction between N-2(A) and NO,
IsN (A Z ) + NOCX 2 1) NO(A ) + N WX (48a)2 U 2
other channels(4b
2 2
V ~ involves measuring both the rate coefficient for removal of N2 (A) by NO and
the rate coefficient for the exzitntion of the NO 'i-bands in the enery
63
0.0
>. 0.80-
0.60-
0.00 :
20C 220 240 260 280 300 320 340 360 380 401)
WAVELENGTH flfll
Figu.re 34. E%-:erimental (light line) and synthetic (heavy line) spectrumfor tt'e NO y-bands assuming a constant electronictranIsition =oment.
1.00-
v) 0.80-
0.60-I
0.,40o
0.20-
20 20 240 260 280 300 320 340 360 380 400
WAVELENGTH inmi
Figure 35. Experim~ental (light line) and synthetic (heavy line)_spectrumof -the NO *ibnsusing the electronic trans it ion-moment f-nc-ticn jetqrm~ined in this work. The major remaining areas ofdiscrepancy between the two spectra at 260, 276, 294, and313 nm result pr4.anarily from small additions to the expert-=ental apectruim from the 0,5 throu4;h 0,8 bands of the N2'/e ar -'aplan zyst" w iz ere not incluie-i ~nthe
syntheic f~t
transfer reaction. The rapid vibrational relaxation of N2 (A) ry molecil es
such as CF4 , CF 3H, and Ci4 , with no accompanying electronic quenching
(Refs. 7,66) allows us to alter the vibrational distribution of the N2 (A).
This makes state- tz-state measurements possible.
3.1.4.1 The 2uenching of "- (A 3Z,, v' = 0) by "CO--Measuremen:s
of the rate of removal of N2 (A) by NO are .not so easy as corresponding mea-
surements of N2 (A) quenching by other molecules. Ordinarily, one follows
N-(A) number density decays by monitoring the Vegard-Kaplan emission
(Refs. 7,50,51). The extremely bright NO y-band emission in the same region
of the spectrum however, masks the Vegard-Kaplan bands. The -v-band emission
is a sensitive tracer of the N2 (A) number density.
The differential equation describing the rate of change in the NO(A2 :-)
number density with time is
]ENOA)~ k E (A)][NO(X) - k 4NO(A)] (30)
dt 48N2 49
The NO(A) is in steady state in the observation volume because the lifetime
of NO(Aj is short compared to the time a molecule resides within the field of
view of the detector. Thus the intensity of the y-band emission is
= k 49[NC*; k 48- [N (A)][NO (51)
Jpon rearranging this equation, we relate the number density of N2 (A) in the
observation volume :o the ratio of the y-band emission intensity and tne NO
number density:
[N2 (A)] = ;NO (52)k48
The differential equation describing the decay of N2 (A) in the reactor is
d [N-(A)d[t = (A) k NO] k ) EN (A) ] (53)dt = (k48 (kwall "q 2
r,7
where is t,-.e first-order (pressure dependent) rate coefficient for Nz;.A)
quenching in wall :ollisionsa. Because the NO number density is tyPicall7y
several orders of magnitude greater than the Nn2(A) number density, it can be
assumed that th-e NO number d-ensity is a constant (the pseudo first-order
aooprox-mat-ion). --his acoroximatiJon leads to an analytical solution to
Eq. ( 53), viz ,
Zn LX -k Nj + A /V (54)N_ A,~ 48 ~ wall
where -we have replaced the reaction time by the ratio of the distance from
.low tube in'ector t,: observation volume, z, to the bulk flow velocity, in.
the rector, .Iserting Eq. (52) into Eq. (54) gives
Zn 0 48 [kNO] + k al z/V (55)P. IP.1 0*/NOJ0 8wal
NO
Equation (53) shows that measurements of the logarithm of the ratio of y---Inc-
intensity zo NO number density as a function of NO number density but with
fixed reaction t~me will give a linear relationship with a slope of -k48 z,'v.
Such measurements at several different reaction distances, under otherwise
-c s-nt zonditio;s --f pressure, temperature, total flow rate, etc. , will
correct for nonirnstantaneous mixing at the injector. The results must fartlier
A' be corrected by a factor of (0.62)-l to correct for the coupling of a radial
d-ensity': gradient iA N2(A) number density with a parabolic velocity proffile
(Refs. 99-106).
As shown previously (R.ef. 50) rate coefficients measured using a tracer
can be seriously in error i-t the tracer is sensitive to several different
N2(A) vibrational levels, each of which quenches at significantly different
rates, the N2(A) vibrational distribution has been relaxed to only v' 0
CF-li, CF4, and CH4 all were used to relax N2(A) to v' - 0. As expected, the
results were invariant with relaxation partner.
7I
Fiaire 36 s"ows a plot of the ratio of the natural log of t-he "-band
intensity to the NO number density as a function of tie N number density for
several different !istances between the injector and observation volume. The
lineari-.-y of t-.ese plots is quite good, extending over more than two orders ofmagnituoe. Figar_ 37 snows a olot of the s3oes of the lines in Fi;. .6 and
two other sets of data not shown plotted as a function of the reaction time.
The slope of this plot, when divided by the radial-profile correction factor,
0.62, gives the rate coefficient for quenching N2 (A) by NO. Note the nonzero
intercept, indicative of the finite time required for complete reagent
mixing.
A number of experiments spanning a range in total gas pressures from 0.7
to 3.7 torr and reaction times from 11 to 124 ms, and using several different
• NO/Ar gas mixtures all gave consistent results for the rate coefficient f;r
2(A v1 = C) quenching by NO of (6.6 ± 0.8) x I cm3 mclecule- s-. !he
quoted error estimate is one standard deviation in the averagina process. The
total experimental uncertainty, including estimates in the uncertainties in
the calibrations of the flowmeters, pressure gauges, etc. is about 15 per-
cent. A few decay measurements in which the N2 (A) was not vibrationally
Srelaxed gave decays only slightly larger (<5 percent) than zhose meas"ired
for t.e rIaxed N:;,A). It is inferred that NO quenches vibrationally e:.:cited
N2(A) a: a rate similar to that for quenching v' = 0.
The result fisagrees markedly with Dreyer and Perner's reported valie of
2.8 x 10-11 cm3 -olecule- 1 s-1 also for v'=0 (Ref. 47). We agree with t-ne
recent result of Shibuya et al. (Ref. 107), (6.9 ± 0.7) x I0- 1 1 cm3 molecule -1
s 1, and also quite well with early measurements by Callear and Wood
(Ref. 45), 8.0 x 0- 1 1, Young and St. John (Ref, 44), 7.0 x 10-11, eill t al.
(Ref. 42), 7.5 x 10- 1 1 and Piper (Ref. 43) at 196 K, (9 t 2) x 10" 1. Mandel
and Ewing's (Ref. 46), rate coefficient, 4.3 x 10- 1 1 appears to be discordant
with the rest of the literature. All the rate coefficients are in units of
% cin2 molecule-I -- 1S- 5-.All measurements excepting Creyer and Perner's used
tracer techniques, and were not state specific. As has been pointed out, how-
ever, our meas-rements indicate that the quenching of N2(A; oy NO does not
72
1 1 01 I if I ..
-2A
-25
10 c
0 -26-
202c
-327
0.0 1.0 2.0 3.0 4.0
[NO) 0 12 MOLECULES cnV- 3 i
Figur-2 36. Decay in the natural loo of the N-,CA, v' =0) number esv%:as a function off '0 number density at three diffferent e;n
mixi.ng distances.
1.5
I
C-1
0
Uj
0 0.02 0,04REACTION TIME
Figure 27. _4-A, v' =0) decay conantls in '10 as a3 function cf
reaction tim~e.
73
11 U . . .4*
Os
appear to show a strong dependence on the N2 (A) vibrational level. Cai!ear
and Wood Ref. 45) a rso reached this conclusion when they attempted to retax
N2 (A) vibration with large additions of helium to their flash photolysis
system.
3.1.4.2 The Excitation of NO(A 2 +, v' = 0) by N9 (A3, , v' )
We have determined the rate coefficient for excitation of NO(A 2 ,*, v' = 0) oy
measuring the increase in the intensity of several bands originating frz.,
NO(A, v' = 3) as a function of added NO number density but for constant N2(A)
number density. :f we note that the N2 (A) number density is the intensity of
the Vegard-Kaplan bands divided by the Einstein coefficient (Ref. 108) for
spontaneous radiat.ion, we can rearrange Sc. (51) to give the working equation
for the analysis.
IVK
1N0~ =48a A5)VK
One convenient feature of this analysis is that the absolute calibrations
for photon-emission rate measurements of the two intensities cancel, and only
the relative spectral response is important. Thus, the intensity measurements
do not introduce significant potential sources of systematic error. :n crac-
tice, total Vegar-Kaplan intensity was determined from a spectral fit to the
whole band system. We tnen measured the change in the peak intensity of one
of the bands of NO(A) as a function of added NO number density, belng careffJl
to keep added NO number densities below the range giving significant N2(A)
quenching. Multiplying the peak intensity by a correction factor gave the
total integrated intensity under that apecific band. Dividing the integrated
intensity by the appropriate branching ratio (Paragraph 3.1.3) determined the
total emission from NO(A). Observations were made on the 0,1, 0,4 and 0,5 v-
bands. Under the experimental resolution, the 1,5 and 1,6 bands overlap the
0,4 and 0,5 --bands and, thus, contribute to the observed emission intensity.
This small contribution was subtracted from our data. All three of the
" observed -f-bands gave excitat:on rate coefficients which were identical within
exper;.mental error.
74
Fig-zure 38 snows that the intensity of the 0,1 band increases l71neArli
with added NO numner density in accord wi.th Eq. (56). A number of suc , excer-
iments yielded a rate coefficient for exciting NO(A, v' = 0) by '12 (A, v' = )
of 9.0=2.7) x 10-'!c= 3 mcle:ui- a where the error bars represen te
total estimated statistical and systematic error. The major contributior. to
the uncertainty is in the 20 percent uncertainty quoted for the N2A) Zinste-n
coefficient (Ref. 108). Variations of greater than a factor of 5 in pressure,
and of more than an order of magnitude in N2 (A) number density gave consistent
results. The distance was varied between the NO injector and the oose-vation
region to insure that the NO was fully mixed. In addition using Xe* : "2 as
the N2BA) source, and using several different NO/Ar gas mixtures did not
change the results.
0 Relatively hich resolution scans over the 0,6 and 1,7 bands as a f-inctior.
of pressure betweeri 3.4 and 10 torr showed that the ratio of NO(A, v' = 1) to
NO(A, v' = 0) excitation by N2 (A, v' = 0) was 0.94 t 0.06, Fig. 39. Spectral
scans between 200 and 400 nm indicated that excitation of NO(A, v 1 = 2) and
NO(B, v' = 3) were both > 0.003 as compared to NO(A, v' = 0). Thus, the total
rate coefficient for NO excitation by N2(A, v' = 0) is (10 - 3) x 13- 1 cm3
molezule- -I
3.1.4.3 State-to-State Excitation of NO(A, v' = 0,1,2) by
N2(A, v' = 0,1,2)--A number of spectra of the NO y-bands and N2 Vegard-Kaplan
bands were scanned witn fixed NO number density, but with varying CF4 nunoec
density and, thus, varying No(A) vibrational distribution (Figs. 40 and 41).
The total N2(A) number density changed little over the series of experiaents,
but the vibrational distribcution changed from one in which more than half of
tne N(A) was vibrationally excited to one in whch well over 30 percent of tne
N2 (A) was in v1 - 0. These measurements therefore tracked how th.e NO(A)
vibrational distribution changed with changes in N2 (A) vibrational distribu-
tion. The observed intensity of a given NO(A) vibrational level can be
expressed by
= < [N.A. + 'A [N A] + 'K [N A] rNCi (57)v 2 1v 2 1 2v 2 2
75
@1%
*I I I Jj
3n
0/z
0 10 20 30
[NO] 10 9 MOLECULES cm-3,
Figure 33. Variation in the peak intensity of tne NO(A-X, 0,1) band as
function of added NO number density.
0.12
010
[4N o 0,\ 094, t o006
O0OO7 O9tOO
'V 008
i 0.06!I 0070
PRESSURE 'TORR 0
Figure 39. .a tio in the rate coefficients for exciting NO(A, v' -, t=NO(A, v' = 0) by NI(A, v' = 0) is a f'unction of irgon ;rssaure.
76
P P * p ~%
1I.00-
0.80-
0 0.60-
0.40
z
02 0.20-
200 220 240 260 280 300 320 340 360 380 400
WAVELENGTH nmi
4-: ?iure 40. Scectrum of :tO(A-X) and N-I(A-X) in the absence of CF4 . The
.. light line shows the experimental data, while the heavy
1.00 line shows the synthetic best fit to the data.I 1.00-
.:
.'- ' 0.80-
060-
O.AO
'V 0.20-
,z; ,. .,o
0.00-200 220 240 260 280 300 320 340 360 380 400
WAVELENGTH 'nm,
Figure 41. Spectrum of NO(A-X) and N (A-X) i-i ?resence of CF4, which
relaxes -most f the N-2-) vibrational ener3y. The ligntline is tfie experi:nentaL data -hile tne heavy line shows thesyntAet.c best fit.
. ,,
i ,7
where the subscri2 cs on the k's represent the vibrational level of the :; 2 LA;
and NO(A), respectively. CF4 vzibrationally relaxes N2 (A) in jv=1 transitions,
and relaxes v' ) 2 much more efficiently than it does v' I 1 (Ref. 64). Thus,
for small CF4 add:tions, the v' = 1 number density stays relatively constant,
and prtmarily v' > 2 is quenched. For moderate to nigh amounts of vibrAtional
ralaxation, therefore, only N2(A) v' - 0 and I remain in the reactor and
further relaxation beyond that point changes only the ratio of v' = 1/v' = 0;
thus Eq. (57) can be simplified and rearranged to give,
NOW CN,A]V 1 1+ l -I N ['40] (58)
,.:N2 A kov Iv N2 A] O)
The ratio of the slope to intercept of the linear plot implied by Eq. (58) will
give the ratio of the rate coefficients klv/kov. Figures 42 and 43 show that
0this linear relationship does indeed obtain. Using the results for v' = 1
excitation derived from the moderate-to-high relaxation data, we can subtract
out the contribution to observed excitation from v' = I for the data showing
little relaxation and thereby probe contributions from v' > 2. The working
equation then becomes
NOW [N2A] 2 2k [NO] k + k [NO] (59)N 2" Av N 2 v 2v [N2 A 0
'' 2' 0 kv[2A0 2
Figures 44 and 45 snow the linear relationship implied by this equation, and
the rato of !'he slope to intercept from these plots gives the ratio v
Because excitation of NOkA, v' = 2) and NO(B, v' - 0) were such minor cnan.el3
*&.,'t I. only their contributions to the total excitation were estimated by zeas.:r'-:
the integrated intensity under the 2,3 gamma band and the 0,7 zeta nar_
several spectra in which the N2 (A) was vibrationally excited and sev.er:-
in which it was relaxed. Thus, only an excitation rate coff - . -
for excited and nexcited N2 (A) for these two states. Table 1 . -
relative excitation rate coefficients for the state-t:-stj:. -.
"O(A, v' = 3,1,2) and NO(3, v' = 3) by N2(A, v' -
::'-
AD-A193 122 CONAN; CHEMISTRY OF NITROGEN-A NASCENCE(J) PHYSICAL 2SCIENCES INC ANDOVER NA L 6 PIPER ET AL JAN SBPSI-876/TR-593 AFUL-TR-86-95 F296$i-84-C-9976
UNCLASSIFIED F/ 7/2 U
EhEEEmmomhEmiEsmEohmhEEEEEEsmmhEmhhEmhmhhmEEmhEEmhEmhhEEohEohEEmhEEEIEohEohhhmhhmhE
Iv
P~ ~ ~ ~ I 'lnl
6.0
* 5.0
,0 04
4.0
0
z
0 Q2 0.4 0.6[N 2(A,v'- I)]/CN 2 (A,v'.0) ]
Figure 42. Excitation of NO(A, vl - 0) as a function of the ratio of
N2(A, v' 1) to N2(A, v - 0).
p Ic I = I I E
00
0L
UJ .4 J.o
[.%2(A .'V , ) ]1 % [.2 (A.V'.I)l
Figure 43. Excitation of NO(A, vl - 1) as a function of the ratio ofN-(.;, vI - 1) to N2(A, v ° - 0).
79
ClC
- In
ojA VI t -A.V 1
-- 0
A A
2 -
W
0 2 0.4
["2(A,V' -2)']I[ 2(A,V'-).
. Figure 4. VEito i h xcitation of NO(A, v' 1) as a function of gertooth !i f)k v' 2) to N(A, v' 0).
08
2(NY)/E 2A, .)
.d . N.Av 2)t A ' n _~
K,,
, UD
A -a
0.2 0.'
[N 2 (A,v,6 2)]/([N 2 (A,v', O
! = Figure 45. Variation in the excitation of NO(A, v 8 - 1) as a function of
~~the r3tio of N (A, v' - 2) to N-1(A, v' - .
80
ivi
TABLE 8. State-to-State Relative Excitation-Rate Coefficients.
N2(A) v' NOW v' 1 NO(B, v' - 0)
0 1W 0.094 ± 0.006 0.003 0.0032 t 0.0007
1 1.11 ± 0.07 0.22 ± 0.030.024 0.033 ± 0.007
2 .29 ± 0.07 0.32 ± 0.03
S*1 - 9.0 x I0-11 cm3 molecule-1 s-1
3.1.4.4 Discussion--If the energy transfer between N2 (A) and NO
*proceeds only through exit channels of radiating NO states, then the rate
coefficients for N2(A, v' - 0) quenching by NO (6.6 x 10-11 cm3 molecule-1
s "1) and for NO(A,B) excitation by N2 (A,v' - 0) ((10±3) x 10-11 cm3 molecule "1
" 1) ought to be the same. The lack of congruency between the two measure-
ments is somewhat disturbing, even though they do overlap slightly at the
extreme limits of their respective error bars. An attempt was made to cross
check data carefully, and to vary the experimental conditions over a fairly
wide range to find systematic trends which might explain the discrepancy.
None were found. The conclusion, therefore, is that the Einstein coefficientfor the '2(A-X) transition is in error by about 30 percent (it should be
smaller).
The experimental determination of lifetimes on the order of 2s is
extramely difficult and have manifold uncertainties. The accepted value of
the Einstein coefficient for the N2 (A-X) transition rests upon absorption mea-
surement3 by Shemansky in the vacuum ultraviolet (Ref. 108) and his reanalysis
(Ref. 109) of Carleton and Oldenberg's absorption measuretaents of N2(A) in a
b . discharje (Ref. 110). His analysis requires a long extrapolation of the
61
J'ILW
transition-moment function with r-centroid from the region encompassed by his
absorption measurements into the region of r-centroid sampled by the strong
transitions from the v' - 0 level. He tied this extrapolation to the lifetime
for the v1 - 0 level derived from the Carleton and Oldenberg reanalysis.
Carleton and Oldenberg (Ref. 110) attempted to measure simultaneously the
absolute photon-emission rate of the 0,6 Vegard-Kaplan band and the absolute
number density v' - 0 level of the A state via resonance absorption on the 1,0
transition of the first-positive system (N2 B-A). Assuming that the experi-
mental observations of Carleton and Oldenberg are accurate, and that
Shemansky's re-analysis of their observations is correct, then their derived
lifetime for N2(A, v1 - 0) depends directly upon the accuracy of the lifetime
of the v' - 1 level of N2 (3). While the recent lifetime measurements of Eyler
and Pipkin (Ref. 111) on tne radiative lifetimes of N2(B, v1 - 5-12)indicate
that the transition probabilities of the first-positive system given by
Shemansky (Ref. 112) are essentially correct for v' > 3, we do not feel confi-
dent that Shemansky's transition probabilities for the three lowest levels are
necessarily accurate. The transition probabilities for these three levels
depend predominantly upon an extrapolation of the electronic transition-moment
function which Shemansky derived from relative intensity measurements of bands
with r-centroid values between 1.35 and 1.6 A out to r-centroid values as
small as 1.0. This is generally a risky procedure. The recent ab initio cal-
culations of the :ransition-moment function by Werner et al. (Ref. 113),
Yeager and McKoy kRef. 114), and Weiner and Ohm (Ref. 115) all show a much
slower increase in the transition moment to smaller r-centroid than is given
by 3he-ansky's extrapolation (Fig. 46). The lifetimes Werner et al., calcu-
late from their transition-moment function are consistently 16 percent larger
than the lifetimes measured by Eyler and Pipkin (Ref. 111), but the relative
variation of their calculated lifetimes with vibrational level matches that of
Eyler and Pipkin quite well. They also match the relative variation in the
lifetimes measured by Jeunehomme (Ref. 116), and by Carlson et al. (Ref. 117)
and those calculated from Shemansky's transition probabilities for v' ; 4.
They deviate markedly from the experimental results, however, for the lowest
vibrational levels, with the calculated lifetimes of Werner et al., being
somewhat longer. If we reduce the calculated lifetimes of Werner et al. by
82
1.6 Iii
YEAGER AND McKOY ---- HRPA J1.1WERNER et al. ,SF
I IEMPIRICAL
N2 I POS. SYS.
S T9 - A
JAIN FROM TURNERAND NICHOLLS INTENSITIES
1.0
TRANSITION SHEMANSKYAND aU
MOMENT (0) 0.8 BROADOOT -'MENKO. -. et al.
. THEORE7ICAL UNIO AND
0. -REGION OF VALIDITY OF
SHEMANSKY' S TRANSITION- "c
MOMENT 1ARI A iON
0.2"
STRONGEST V' 2 0 0 0TRANSITIONS V -I N U U
1.0 1.1 . 1.2 1.3 1.4 1.5 1.6
INTERNUCLEAR DISTANCE (A)
?iure 46. ELectronic tr3nsition moments for N2(337, -A3:, ).K 83
16 percent to make them coincide with Eyler and Pipkin's measurements for the
high vibrational levels, a lifetime is obtained for vI - I of N2(B) of 9.5 Ls
in contrast to the value of 7.8 ,Is which results from Shemansky's transition
probabilities. This large a change in the lifetime of the B state will reduce
the transition probability for N2 (A, vI - 0) from Carleton and Oldenzerg's
experiment by 20 percent. This reduction would then bring our quenching- and
excitation-rate measurements into reasonable agreement. Taking the ab initio
transition probabilities at face value would result in a Vegard-Kaplan transi-
tion probability about 40 percent smaller than the currently accepted values,
and would bring the two measurements into almost perfect congruence. The
other theoretical treatments agree with Werner et al.'s calculations. The
lifetime of N2 (B, v' - 0) measured by Heidner et al. (Ref. 118) via resonance
fluorescence is also somewhat larger than given by Shemansky's extrapolation.
A reduction in the transition probability of N2(A,v'-3) state on the
order of 20-30 percent would still give a variation in the absolute transition
moment of the A--X transition fully consistent with the absolute measurements
of Shemansky that sampled smaller values of the r-centroid, and the relative
transition-moment measurements of Broadfoot and %laran (Ref. 119) whi:h sampled
larger r-centroid values, tnose sensitive to the Vegard-Kaplan transitions
from v' = 3 (Fig. 47). we feel that good experimental measurements of the
relative transi:ion-moment variation of the first-positive systlm which sample
smaller values of r-centroid are imperative to clear up this conflict. 'his
will require relative intensity measurements extending out into the infrared
* to 1.5 M.
The measurements on the vibrational-level dependence of NO excitation by
N2(A) show that N2 (A, v' - 1) excites NO(A,B) about 25 percent more efficient-
ly than does N2 (A, v'-0). N2 (A, v'-2) however, is somewhat less efficient at
A exciting NO transitions. The reduction in observed intensity of NO(A,3) from
excitation by 1;2 A, v' - 2) could result from one of three possibilities.
First, the iaenching efficiency could be smaller. Second, the more hi;hly
excited N2 (A) can access higher lying levels of NO(A,B) which might be colli-
sionally coupled into other states of NO which do not radiate or emit outside
84
' II I
0 ABSORPTION MEASUREMENTS
0.10 (1.)A FROM BROADFOOT AND MARAN
BAND (5,0) BAND
UPPERLIMIT-. 0
a CHANGE IN-0.10 - (0,S) BAND EXTRAPOLATION"I RESULTING IN
C) \25% LONGER4 LIFETIMEX,Z -0.20 - A
VARIATION OF ELECTRONIC
-0.30 TRANSITION MOMENT (R,(F))
S-0.3012)"."BAND A 'L
I. I
1.0 1.1 1.2 1.3 1.4
R-CENTROID (Al
Figure 47. Variation of electronic transition moment forN,.(A .J7-
the spectral bandpass, such as the b4 Z- or a4 n states. The third possitility
is that some of the encounters between N2(A, v' - 2) and NO end up dissociat-
i.g the '1O. Only vibrational levels of N2(A) > 2 have sufficient energy to
iisscciata the NO. Atom production from this interaction has not been
attempted, but such measurements would confirm this possibility.
The difference between the excitation rates of N2 (A, v1-1) and
N2 tA, v' - 0) is not sufficiently great for us to observe significant changes
in decay-rate measurements involving vibrationally excited and unexcited
N2 (A). Given a typical v' - 1/v' - 0 ratio of 0.6, we compute that the effec-
tive decay rate would increase by only 10 percent when both N2(A) vibrational
levels were present. Within experiment error, this small enhancement is con-
3i3tent with observations.
S5
To investigate more completely the energy disposal in the reaction, we
scanned the 0,6 and 1,7 bands of NO(A-X) under moderate resolution
(IN a 0.20 rnm) at pressures between 0.4 and 9 torr. This resolution was ade-
quate to resolve partially the rotational structure. We tnen adjusted rota-
tional temperatures in the fitting program until the observations could be
monitored. At 0.4 torr, Boltzmann rotational temperatures of 1400 K and 80 K
fit the emission from v' - 0 and 1, respectively. At higher pressures, how-
ever, the band contours were decidedly non-Boltzmann. Collisions with the Ar
bath gas relaxed the lower rotational levels much more efficiently than they
did the higher rotational levels. For examples, at 4.0 torr a rotational
temperature of 600 K fit region around the heads of the 0,6 band quite well,
while the high rotational levels which are prominent in the short-wavelength
tail of the band followed an 1100 K Boltzmann distribution. Assuming a hard-
sphere model with a 40 12 collision cross section, we calculate that an
excited NO molecule will experience 1.5 collisions during a radiative lifetime
at 1 torr. Thus, at 0.4 torr most of the NO(A) molecules will not experience
a collision prior to radiation, whereas at 4 torr they will experience an
average of six collisions. Thus, the rotational relaxation of NO(A) by Ar is
a relatively efficient process, requiring only a few collisions to remove most
of the rotational energy. The efficient rotational relaxation of low 3 levels
of NO(A) by Ar and the correspondingly much smaller efficiency for high-.i
level relaxation has been studied in some detail by Ebata et al. (Ref. 120).
The efficient transfer of vibronic energy from N2(A) to NO may occur by a
Franck-Condon type of mechanism. Deperasinska et al. (Ref. 121) have calcu-
lated Franck-crndon factors for the transitions relevant to the transfer of
energy from N2 (A, v' - 0) to NO. The Franck-Condon factors for producing
NO(A) are three orders of magnitude greater than those for producing N0(3),
which they claim is reflected in the much smaller efficiency for producing
NO(B) relative to NO(A). Their calculated Franck-Condon factors, however,
would predict roughly equal probabilities for producing vibrational levels
v' 0 and 1 of NO(A). In contrast, observations show that NO(A,v'-0) is pro-
duced ten times more efficiently than NO(A, v' - 1). They have not performed
the relevant Franck-Condon cal:ulations for N2 (A, v' - 1, 2).
Z6,
The kinetics of the No(A) + NO energy transfer have been studied 'v
several other investigators. Callear and Wood (Ref. 45) estimated rate coef-
ficient ratios froa thei work of k0 1 /k 0 0 - 0.105 in reasonable agreement with
the value of 0.094 t 0.006 and k1 1A 1 0 = 0.53 in disagreement with the value
of 0.20 ± 0.03. Clark and Setser (Ref. 43) determined a population ratio for
NO(A, v' 0,1,2) of 1.0:0.15:0.014, respectively, from excitation by N2(A)
with the ratio v' = 1/v' - 0 of 0.61. With the same N2(A) vibrational distri-
bution, we calculate an NO(A) vibrational distribution from excitation rate
coefficients of 1.0:0.14:0.011, respectively, in excellent agreement witn.
Clark and Setser's observations. More recently Golde and Moyle (Ref. 122)
have measured vibrational distributions of NO(A) from N2 (A) excitation cf
1.00:0.083:0.002 for excitation by N2(A,v'-0) and 1.00:0.17.0.025 for excita-
tion by N2(A) with a vibrational distribution of 1.00:0.48:0.19:0.14 for
vI - 0-3 respectively. Rate coefficients would predict an NO(A) vibrationa.
distribution of 1.00:0.19:0.013 given the same initial N2(A) vibrational is-
tribution. To make this comparison, it was assumed v' - 2 and v' - 3 had thesame excitation rates. Golde and Moyle's data show a 7 percent decrease in
total NO(A) intensity for the vibrationally excited case whereas our results
would indicate that the intensities of the NO(A) produced from vibrationally
exci:ed and unexcited N2(A) would be within 2 percent of each other. Error
lizits encompass a range from a 6 percent decrease in intensity to a 3 percent
increase with some additional uncertainty added by our having treated N2(A)
vibrational levels 2 and 3 the same. They are therefore fully consistent with
5olde and Moyle's result.
3.1.5 Summar:--It has been shown that the transition-moment variation
with r-centroid for the NO(A--X) transition is significant, especially for
transitions with larger changes in v. The result of this variation is that
the transition probabilities of the NO(A--X) transitions with .'v < -3 ara sub-
stantially larger than had been believed previously. This means that the
optical gain of some of the redder of the y-band transitions will be larger
than anticipated (Table 9), so that some of these transitions will probably
have enough lain to support laser oscillation. The benefit of using these
transitions with greiter differences in vibrational level between the :;oer
and -:wer iatis results both from reducing the probability of Lower state
37
TABLE 9. Revised Optical Gain Predictions for NO(A-X) Transitions.
I WvNormalized Optical Gain
Transition Wavelengthv'' 1 (v"Previous Prediction Present Results
0,0 2269 0.73 0.660,1 2370 1.00 0.92
0,3 2479 0.78 0.77
0,2 2596 0.45 0.49
0,4 2722 0.22 0.27
0,5 2859 0.10 0.,3
0,6 3009 0.04 0.060,7 3112 0.01 0.03
bottlenecking and from working in a region where optical difficulties are
less severe.
The measurements on the quenching of N2 (A) by NO show that NO(A) excita-
tion is extremely efficient in this system, and that the system is scalable to
moderate pressures, at which the N2 (A) is predominantly vibrationally relaxed,
without significant reduction in the excitation efficiency. Indeed, resuits
show that some degree of vibrational relaxation is desirable since vibrational
levels of N2 (A; greater than or equal to to excite NO less efficiently than
do the two lowest vibrational levels.
3.2 N2(A 3 au+) ENERGY POOLING
3.2.1 introduction--Stedman and Setser (Ref. 123) first discovered
energy pooling in triplet nitrogen when they observed that the intensity of
nitrogen second-positive, N2 (C3Iu -B3q ) emission varied quadratically with
the intensity of the Vegard-Kaplan emission in their reactor. They estimated
the r te coefficient for energy pooling to form N2( 3 u) to be 2.1 x 1.)-11 :m3
molec:ue-1 3-1 Tn a series of investigations of time-resol'ea emiss;.cns in tne
afttr:,;w of i .ulsed nitroen discharge, Hays et al. (Refs. 124-12*) studied
I8-q ,V%---
N2tA) energy pooling and determined rate coefficients of 2.6 x 10- 1" for tor-
mation of N2(C 3 "u) (Ref. 126), 0.25 x 10-10 cm3 molecules -1 s-1 for the fo =a-
tion of N2 (C'3qu), and 1.1 x 10- 9 cm3 molecule-1 s- I for the formation of
N2(33g) (Ref. 125). In the case of the latter state, one would have to
assume that their number was a lower bound, because their detection system
could see only as far as v1 - 3 of the B state. Thus, if significant energy
from the pooling reaction flowed into v 1 = 0-2, the rate coefficient would be
somewhat larger. Subsequent work by Clark and Setser (Ref. 43) confir.ed a
rate coefficient of about 2 x 10-10 cm3 molecule-1 s- 1 for the production of
N2(C 3au) from N2 (A3 u+) energy pooling. They were unable, however, to see any
evidence for formation of the N2(C'37 U) state. Nadler et al. (Ref. 127) dis-
coverad in 1980 that the Herman infrared (HIR) system was populated by '2 (A)
energy pooling, and showed that the distribution among the vibrationdl levels
of that state depended strongly upon the vibrational distribution of the N2(A)
state (Ref. 128,. They estimated a lower limit for HIR formation of
2.5 x 10-11 cm3 molecule - 1 a-I which they have subsequently revised upwards to
7 x 10- 11 cm3 molecule-1 s-1 (Ref. 128). They also showed that some produc-
tion of the B stace did indeed occur, but were unable to estimate the rate
coefficient of B state formation because of that state's rapid quenching .ynitrogen and argon. Nadler and Rosenwaks (Ref. 129) have also shown recently
that the vibrational distribution of N;2 (C3ku) formed by energy pooling also
changes as a function oZ the N2 (A) vibrational distribution, but that the
total excitation rate of the C state appears to be independent of the vibra-
tional distribution of N2 (A).
Unpublished observations at PSI in the near infrared recorded the HIR
system in N2(A) energy pooling, but failed to detect significant populations
of N2(3) (Ref. 130). There is reason to be skeptical of the magnitude of the
pooling rate coefficient which Hays and Oskam claimed made N2 (S). The work of
Nadler et al. also showed convincingly that HIR production had to be similar
in magnitude to the production of N2(B). Thus, the present investigations
wera motivated in part by the desire to reconcile the conflict between Hays
and Oskam's report, and the observations of Nadler et al. and our own. We
83
have investigated energy pooling in some detail and have been acle to deter-
mine vibrational-level-specific rate coefficients for formation of N2 (Ctu,
.0-4), N2 (.33 g, v' - 1-12), and HIR vI - 2,3 by N2 (A3 u, v' = 0-2).
3.2.2 Experimental--Paragraph 3.1.2 describes apparatus and general
operating procedures in some detail. Thus, only a brief summary will be given
here. The studies all involved measuring spectra of the Vegard-Kaplan, first-
and second-positive and HIR systems of molecular nitrogen in a flowing after-
glow apparatus. The N2 (A) is produced cleanly in the apparatus in the energy
transfer reaction between argon and xenon metastables and molecular nitrogen
(Refs. 61,62). A hollow-cathode discharge produces the rare gas metastables.
The electrode has been fabricated from aluminum shim, but for some of the
studies here, a 0.002-thick-tantalum shim was used as the electrode material.
, The tantalum electrodes gave about 30 percent more metastabies and operated
better at high pressures. The energy transfer reaction between metastable
x enoand nitrogen produces N2(B v' 4 5 (Ref. 63). This eliminates the% %. xenonannirgnroue 2 1 g
possibility of contamination of the results on N2 (C) and the higher vibra-
tional levels of N2 (B) by scattered light from the rare-gas-metastable/
nitrogen mixing region. The HIR system was studied at relatively high pres-
sures (> 6 torr) with high partial pressures of nitrogen (I to 2 torr) in the
reactcr. This procedure virtually eliminated overlapping of tie HIR system by
the first-positive system. The consequence of this procedure, however, was
that the high partial pressures of nitrogen relaxed the N2 (A) vi_ration to
> 95 per:ent v' = 0,1, and thus the effects of higher vibrational levels could
not be studied easily. The nitrogen B-state studies used a neon carrier gas
to reduce electronic quenching (Refs. 131-133). Extrapolating measured popu-
lations to zero nitrogen pressure eliminated most significant quenching.
II All spectra were fit by a least-squares computer program which deterzined
the populations of all emitting states in the region of spectral coverage.
This procedure eliminates the uncertainties introduced by overlapping spectral
bands. Because the HIR system is unassigned, one cannot a pricri generite a
synthetic spectrum for this system. Therefore, experimental spectra (taken at
high pressures for the basis sets in the synthetic fits) was used. 5ecause
F,
N2 (A, v' 0) energy pooling produces only v' - 3 of the HIR system, a basls
set for that state could be generated cleanly. When only v' = 0 and 1 of
N 2 (A) are present, only v' = 3 and 2 of the HIR system are produced. Thus, a
. fitting basis set -was generated for v' = 2 of the HIR system by substrhcting
out the previously determined v' = 3 components from a spectrum containing the
two levels together. These basis sets were then used in analyzing the spectra
containing N2 (3) to eliminate confusion from HIR overlap. Energy pooling of
higher vibrational levels of N2 (A) do produce the two lowest vibrational lev-
els of the HIR system. This only confuses the fitting of the Iv = 2 sequence
of the first-positive system, and we were able to work around it.
The data analysis requires the measurement of absolute photon-emission
rates in the reactor. Subsection 2.4.2 describes these calibration procedures
in detail. In the case of the energy pooling reactions, the N2 (A) and the
N2(Z,B, HIa) have radial density gradients which are different. The ";2,A)
radial density gradient follows the form of a Bessel function of first order,
2N 7N, o'A) J (60)N2 A i ! - i 2 AJ ;]o -ro,
where ['!-)A)] ais the centerline number density, rO is the flow tube radius,
and ; = 2.405, the first zero of Jo(x). The field of view of the detector is
essentially a rectangular parallel piped across the center of the tube with
height, h, width, z, and length, 2ro (Fig. 48). it was assumed that variations
down the axis of the flow tube, across the field of view, could be neglected.
The av;arage number density ,zf N2(A) observed in the field of view is then
!@ . ro0iC r~dr
NN(A) > O r<N 14 A)> 2 -o o
2 ro (61)
f rGdr0
where [N2(A)] o is the N2 (A) number density in the center of the flow tube.
When r is less than or equal to h/2, 0 will equal -; but when r becomes
greater than h/2, ' will be given by sin 1(h/2r). Thus, each integral in
Eq. (61) becomes a sum of two integrals, one between the limits of 3 and h/2,
the other r-.nni,.: from h/2 to r For c.,e conditions of ro . Z.5 cm and
h = 1 cm, numeri:al integration gives
!9
;S
rr
Figure 48. Cross-sectional view of flow tube illustrating the geometrvgerrnane to the radial number-density gradient 2roblem. The
shaded area approximates the monochromator's field of view.
<[N 2 (A)]> - 0.601 [N2(A)] o (e2)
As shown below, :he number density of N2 (C,B,HIR) is proportional to the
*square of the N2 %A) number density. Thus
k-[N2Al2 o J2, k r rtdr2 o o1
,i ,BHR'> k-N(A)] 2 0 (63)<N 2 2 r
0 rdr
0
where k' is the proportionality constant.
Integrating this expression in a similar manner to that ;iven for <7N 2 A>
g ives
<rN 2 (C,B,HIR)]> - 0.45a k'[N 2 (A)]02 v64)
92Ie.
V. -,.,/, -,.,-,, ,, -, -, ... ,, .... , ,N' ,-__' * ,
L% Finally, using Eq. (62) for EN2 (A)]O , we find that
<[N 2 (C,B,HIR)]> = 1.267 k'<1N 2 (A)]>2 (65)
This correction must be made to the data on energy pooling to extract rate
coefficients. This correction factor increases as the ratio h/r increases
• beyond unity, reacting a maximum value of 1.446 when h/r equals two
(Fig. 49).
1.50
I40j
i.30
i.2-* 0 1.0 2.0
h/ r
*_ Figure 49. Variation in the correction factor for energy poolingp-. measurements with the ratio h/r.
3.2.3 N;(C37,, v' - 0-4) formation in N(2I,\,,+, v' 0-2) eneriy
* p~ini--The processes contro!lling the formation and destruction of N2(C) in
the eier~y poollng system are;
o?3-O * = -
N2 (,vI)+N2 (tV)----N2 (,v) + N 2 (X, vII') i66a)
N 2(C, V) ---- d N 2(3) + hv(6b
where the superscript v denotes the vibrational level cf t-he N2(') product
molecule, the subscript v's denote the vibrational levels of the N 2(A) mole-
cules, and krad is the radiative decay rate of N2(C). The li.fetime Of -;(C
is sufficiently short that electronic quenching of that staste (Ref. 36) =ay be
ignored. Because of its short radiative lifetime, N2(C) is in steady state in
the reactor so that its formation and decay rates can be equated. Thus, we
have
d' 2C ~' 1.267 k (AH v' 14 A Vl
-k Co ~N (C, v)l 0('7
ra (67)
For the case of only one vibrational level of N2 (A), Eq. (68) collapses t.D a
single term for wh;.ch the ~-V)number density var_ es linearly with the .square!
of the 112 (A) number density. The slope will equal the ratio of the eniergy-
0 pooling rate coefficient to the radiative-decay rate of N2 (C).
Figures 50 and 51 show representative spectra of the 220 to 400 na region
which encompasses most of the major emissions in the N2(A-X) and N2(C-B) ays-
tems. Comparison of the two figures shows how strongly the N2(C) intensity
varies with the N2(A) intensity. Figures 52 through 56 show plats of the
variation in the number density of N2(C, v'u3-4), respectively as a function
34
N2(C B - B g)
la"v=3 Av=2 %v=l Av=0 %v=-1 Lv=-2 Av=-2I n In rrrn r- rrr7- i-
V10 N2 (A 3Eu+ - X Ir
v 15 1 6 '7 18 19 '10 111 .0 0 - v '= l_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
0'5 V= 7 IS 9 '10. Ill 12'
V1 0.80- 'v0 7 81 '12
S0.60-
- 0.40-
0.20-
0.00-250 270 290 310 330 350 370 390
WAVELENGTH (nm)
Fig ire 53. Observed (heav= line) and computed best fit (light inie) -o
the spectral. region between 250 and 400 nm. Resolutirn
1.3 rim.N2( C 3Tfu - B 3 f1g)Av=3 Av:2Av~l4v=0 Av-- _Av=-3
I n M r--i m r-l-1
.0 'N 2 (A 3r&U - X kE)
v0 '3 4 5 '6 '7 18 '9 '10 111v"z3 '4 '5 V '7 '8 '9 '10 '1 12
S0.80- V"z16 '10 'S '12
) .50-
,-,
10
-
1 0
zzz
0.40-
200 220 240 260 280 300 320 340 360 380 400WAVEMfIGTH (NAMOR lf.PS)
Figure 51. Observed (hea-,7 line) and computed best fit (light lne) o the
sec -rl region between 220 and 400 nm. Note that the e con -
Post ve emissons are uch stronger in this spect-am relative to
I. the "egarJ-Kaplan emissions than is the case in Figure 5.
5Resoi,.zon 1.5 nm.
95
Mm
N20
-- 15 - -
0=
10
'-4
5 10 15 20[,'a'EN2 (A, v'=O)] 2 (1018 molecule-2 cm6 )
Figure 52. Variation in the number density of N-(C, v' 0) as a -:unction o .
the square of the number density of N,(A, v' 0).
25 I
_ o I ,20- 0 -
v 15-
A10-
00 5 10 15 20
.N2(A, v'0)]2 (1018 molecule2 cm"6 1
Fi:ur-i 53. "/zr4ation in the number density of NB)C, v' - I) as a 4inctior.0" the s:,uare of tae number density of 'JI A, v' - 3).
96
25k
'_ 20-
J)
.~15
ii10-
j5-
-
0 5 10 15 20
[N2 (A, v'=O)]2 (1018 molecule 2 cm 6)
Figure 54. Variation in the number density of N?(C, v' 2) as a function ofthe square of the number density of N?(A, v' = 0).
20
5 -
0 LC-- -1
0 5 10 15 20
2N(A. v'IO)]2 (1018 molecule2 cm- 6)
.-are -5. Variation in the number Jensit'v of N-(, v' 3) a a n ofthe 3quare of tne iumber density of N-I(A, v' a 0).
97
Mi
i-o-~--
1.00
0 510 15 20
£C2 (A. v'*)) 2 (1018 miolecule? cnV 6 )
Figure 56. Variation in the number density of Xj(C. v' 4) as a function ofthe aquare of tne number Jensity of N,)(A, v - 0).
of the square of the N2(A) number density under conditions where only v' 3
of the A state was in the reactor.
If only vibra:4ional levels of 0 and 1 of N2 (A) are present, Eq. (6E)
becomes
C'v
N 2(C, V)- 1.267 K:, 00-,A I.C)(
rad
01.3 L ' (,v - WI (A, ' -k 2. 3 -- -- N ( , L 2 L 2)
kred
* 1.267 J!L 'N (A, v9 -W
ra d
I mmM0RR = 1 11111 11
Dividing this equation through by [N2 (A, v' - 0)12 gives a quadratic e3uation
in tne parameter :N- A, v, = 1)]/N 2 (A, v' 0)):
*-N(' )].C,v kC,vI i00" k0v [N2(A v'=1),• - :.267 2.534
N2(A 2 C,v kC,V (N2 (A, vl=3)2rad rad
•"2 N2 'A, v'"j]2
a+ 1.267 !L j(0kC v N2 [A, v-0)' (70)rad
Figures 57 through 61 show data plotted in this fashion for each of t'e
vibrati.onal leve3.s excited i.n the energy-pooling reaction. The most interest-
ing thing these plots show is the absence of a significant quadratic ter:.
The intercepts of the plots, of course, give rate coefficients t.hat acree to
within 20 percent with those previously determined from studying .ust v'-3
pooling. Table IJ snows the results of N2 (C) energy-pooling studies. Our
. value for the total pooling into all vibrational levels of N2 (C) is about
70 perce..t lower for v'-O pooling than previous studies. 4 3 ,124,126 Part of
this difference results from the 27 to 45 percent correction for radial
density effects which have not been inzluded in other flowing afterglow
studies. Applying our measired rate coefficients to Clark and Setser~s: .. 71
p. V'uI/v'-3 A state distribution gives an overall rate coefficient of
1.28 x 10 " 10 cm! molecule-' s-1. This is about half of what they reported.
Our results indicate that their measurements may have been contaminated h/
scattered 1ight, and :herefzre too large because they observed roughly equalNp. populations in v'-0 and I wnereas for the same N2 (A) vibrational
distribution as they had, we observe a ratio of v'-0/v'-1 of 3.76. If
scattered lignt contamination is a problem, then it will enhance '12 (C,v=-Q)
more than the other vibrational levels.
3.2.4 HIR formation from N2(A) energy pooling--The HIR system was first
observed 30 years ago by R. Herman. 134 Subsequent spectroscopic work on the
system has faiied to 1lentllfy either the upper or lower states in the
99
Nin
& pr . D* .
I.D
EN.A v.0](.2A v-~
F'igure 37. Variation in the ratio of the number density of N-t-, vl zthe square of the number density of N-(A, v# 3 as a 'un:~of the ratio of the number densities of *vibrationally ec.~to unexcited ',2A).
& *0
030.2 0.4 0.6 0.8I i1.0
EN2(A, v0>)/CN2(A, v'-0))
7ijur-2 58. 7lriation In the r,:tio of tie n",-=.er :!en3- - ':, totr:-es:2are- ofl tne number Jensity of 'N i.A, v1 as a fnctioflof tne rio of the number densities of v~ao~L :i~to :,r.-xcited :4-,A).
Vw rq
0 -
01I E 0
0 0.2 0.4 0.6 0.81
6N2CA, v's1)),EIN 2(A. v0-))
Fi ur -359. Variation in the ratio of the number density of N7(C, y - 2) tothe squaIre of t-he number density of ',(A, v' - 0) as a functionof the ratio of the number densities of vibrationally exci.tedto unexcited N:-IA).
23
II 0
0 A0.0 0.2 0.4 0.6 0.8 1.0
F;IueLa. Variation in the ratio of the number density of N,(C, v' 3) to
the s uare off the naciber density of v' - 3) as a f-unctionof tlse ratio of t-.e number jensities of vibrationally excitedtouxi Z-)
U%
.1.
r 1.21
CO 0.6 oII3
0.4
0.4
0.01 I40.0 0.2 0.4 0.6 0.8 1.0
[N2(A, v'>0))/EN2(A, v'=0)J
Fi-Tare 61. 7ariation in the ratio of the number density of , - 4,1 t:)the square o! the number density of N_-(A, v' - 0) as a finctionof the ratio of the number densities of vibrationally excited
to unexcited N.)(A).
TABLE 12. Rate Coefficients for "2(C 3 .,.) Formation from
N2 (A3: "Energyv Pool ing.
C ,Ct V1 j C, V1
'2(,v') k 00 k' k 1
*0 2.6 = 0.1 3.4 i 0.7 < 1.04.1 t 0.2 .4± . < 2.0
2 4.1 t 0.2 3.3 t 0.8 '(1.03 2.8 t 0.2 2.2 ± 0.5 < 0.74 1.0 t 0.1 1.0 ± 0.6 < 0.6
Total N2 (C) 14.6 ± 0.8 15.3 ± 3.8 < 5.3
R~ate coefficients are in units of 10-11 cm3 molecules-1 3-1.
Error bars represent 2.1 statistical uncertainties.
102
trans-':ion, 125- 1:7 a-though Nadler and Rosenwaks1 2 9 have been ab!-e -: esta--
Il ish an upper liit to the upper state term energy of 12.02 eV. Nadler and
Rosenwaks have discussed the identification of the two states and concladed
lower state might be the G3- g,, the lower state of the Gaydon-Herman Green
system, and that the upper state was a 3, U state which is known only through
Michaels' calculations (Ref. 138). Gilmore disputes these identifications,
however, and suggests C'5.u and A'5"+ as the upper and lower states, respec-
tively (Ref. 139).
Figures 62 and 63 show the HIR system excited in energy pooling of N2(A,
vI - 3) and N2 (A, v' - 0,1) raspectively. The spectra were taken at 7.5 Torr
total pressure with a nitrogen partial pressure of 1.5 torr. Thus, nitrogen
first-positive emission is virtually absent from the spectrim. The ratio of
the HiIR intensity to the square of the N2 (A) number density does not vary with
pressure (Ref. 127). Thus t-he state is not quenched electronically under our
conditicns, and a steady-state analysis similar to that given above for the
N2(C) state applies here. We write
1.267 'r,3d IrHI., ' = = 1.267 k HIR 3 NA, v' = O'H I R rad L.00 ~2~(
F gure 64 shows data for v' - 3 of the HIR system plotted this way. For these
studies, CF, or Cli. was added to the reactor to relax the N2(A) vibrational
energy to v' = 3. From the slope of the plot, we find that the rate coeffi-
cient for producing v'3 of the HIR system in the energy pooling of two N2 A,
v, - 0) molecules is 8.1 x J1 1 cm3 molecule 1 s " I . Strictly speaking, this
figure is a lower limit, because other transitions of the HIR system from
vI - 3 might appear outside the bandpass of the detection system. The rela-
tive intensities observed of the four bands, however, indicate that both sides
of the Condon parabola have been sampled. Thus in all probability, the major
emissions from v' - 3 have been observed. When the N2(A) was vibrationally
excited, the ratio of the H:R v' - 3 intensity to the square of the N2 (A,
vI - 3) number density remains constant. Thus, vibrationally excited N2(A)
appears not to play any role in exciting v' = 3 from energy pooling.
103
0
N2(B 31 A 3EuA )
AV - 4 Av 3 AV - 2 = 1IrTrr Tr I I I II l I I ir I 7 -1
' 12 10 8 6 10 8 6 4 6 5 4 3 2 75 54
HERMAN INFRARED=3 AV = 2 AV - I .AV 0 1v -1
100" 1 r-- r--t i -r"-v' -3 32 321 3 2 103 2 1
~80_
60-
,- 40
i 20
600 650 700 750 800 4 0
WAVELENGTH (nm)
Figure 62. Spectrum of the Herman infrared v'=3 system excited in. the
energy pooling of N- (A, v'=0). Pt, tA = 7.5 torr, X. ., = 0.20.
N2(B 3a9 - A 3Eu*)
AV = 4 Av = 3 AV m 2 AV= 1Fm1rrl if I I t ( f I I I I I f
v' = 12 10 8 6 10 8 6 4 7 6 5 4 3 2 7 6 5 4
HERMAN INFRARED
100 AV 3 av- A a I Av , 0 AV = -1! 31 -- --- , i i I! -
v' =3 3 2 3 2 103 2 1
80'
50-
0 40
20
600 650 700 750 800 850WAVELENGTH (nm) ,,:
Fi-lire 63. Spectrum of the Herman infrared v'=2,3 svstens excited in the
energy poolin4 of N2(A, v'=0,1). P,. =l 7.5 torr, X 4 0.20.
104
~% %
200
• 150-
S100-", >
i 50-C,
0 5 10 15 20
-N2A, v,=0]2 (1018 molecules
2 cm- 6 )
Figure 64. Variation in the number density of N2(HIR, v'=3) as a function ofthe square of the number density of N(A, v'=O).
Under condttions where only v' = 0 and I of the N2 (A) are present emis-
sion from v' = 2 of the HIR system is seen. The lower two vibrational levels
of :he HIR system appear only when the number densities of v1 = 2 and 3 of the
'12 (A) become important. The HIR v' = 2 data can be analyzed in a similar
fash-on to our analysis for vibrational effects in N2 (C) production, but since
it is already known that two v' - 0 molecules do not pool to make an '--R
v- 2 molecule, the first term in the equation can be eliminated. Thus,
IHIR,2 2.534 kHIR,2 rN (A, v- O)'N (A, v' = 1)01 LN2 A 2
+ 1.267 k IRI N (A, v- 1] 2 (72)11 - 2
HIR,2 HIR,2A = 2.534 k
N ' - 1'.
L v,,2 A, " I
*~~ 1-16 7 kIt2
-2 (73)11 N, A, v' = 7
1 35
Figure E5 shows the data for the HIR v'=2 system. The intercept of the plot,
after application of the appropriate corrections, gives the value for the mixed
vI= /v'= 1 pooling rate coefficient, 9.9 x 0- 1 1 cm3 molecule- I s-1. .ithin
statistical uncertainty, the linear term is not significant, again implying
th.at energy pooling by two v'=i molecules is negligible. Table 11 sumarizes
our results on the Herman infrared system. Data further show that HIR vI = 2
is not formed by pooling of vibrational levels of N2 (A) higher than v' = 1.
Adequate data has yet to be collected on the formation of Hi:R v' = 0,! which is
formed in pooling of N2 (A, v' > 2).
, I
010 0.2 0.4 0.6 0.8 1
(N2A, V'=11/(N 2A, v'=0j
Fi;,;re 65. Variation in tne ratio of the N-iHIR, vl=2) number densitv to t.'e
oroduct of the r,!;mtpr densities of *; (A) v'=O and *i'=i as afunction of the ratio of the number densities of NI(A) v1=1to vl=O.
TABLE 11. Rate Coefficients for Herman infrared Formation fromn
N'2(A) Energy Pooling.
'Ir'
H:R, V, o k01 1
S)9.9 !O 1.0
?.ati! cf~icie-ntz are ir, inits of 10-11 cm-" moleo-ule-l 1 .Error bdars raepreser.t 2,: statistical uncerta-;nty.
106
p -"I
ed.
14
, Fra no: v-I enery - -:arcn Xf.< 3' fr-:i :he~ nz. too'in '" f v-i' ,
=.... in a necn buffer has -::-
s u d . iS aweak quencher of N-(3), so that under c ndi:izns :-re -henumer iens.izies z o_., xenon, and ntrogen are minimized, :"-he bserved 3 z t--
Tu _, ate -.' ir re7.,resenzative 'or nascent stributins. The *est v
Of ',-') I') -, esu. a_ b r-cn and xenon were disca'arged - witn the neon.The n .. : rizr.z en was ded dowe - ream from the di scharg7e as usua!I. Fair-%- nodest
argon flow rates -mol s- in a total flow of 3500 4mol s-,) gave ade-
quate N2iA) produ;ction, so that the number density of argon was well below
that which would produce noticeable quenching effects. Nitrogen was still a
sign:f.cant quencher, so its quenching effects were experimentally de:armined
%" by repeatedly scanning at successively lower nitrogen flow rates. Modif-ying
Eq. (68) to reflect 3 state formation by N2 (A, v6 = 0) on-y and quenching of
the B state by molecular nitzogen gives:
- B~v kBV -
B , vj' N S, LN2'A, vI = 0)11 (74)'rad rad
Rearranging this equation shows that the ratio of the square of the A state
number density to the 3 sa3:e number density will vary linearly with the
number iensity of the added nitrogen:
-2' , -' - = 0.7 9 ~ra d 9. 8 '2 (75)
N. , VB,v kBIv-," " 0 00
-igures o6 and 67 show spectra of the region between 500 and 850 nm w1th
ni.rozen zartia pressures of 0.46 and 0.027 torr, respectively. Clearly, thefirbands incree strongly with the reduced nitrogen partial pres-
sures. Figures 63 through 71 show data on the formation of the B state froo
the pooling of N!(A, v' = 0) plotted according to Eq. (75). From these plots
and similar ones f. r the other vibrational levels studied, we obtained both
rate cofiinsfor B state formation from energy pooling and rate coeffi-
cients for quenching N2(B) by molecular nitrogen. The bulk of the experi7ents
167
N2(8 J!I - A 3E' )
,IV - 4 AV 3 AV V .rrrm m7-r-i-7n In I m--
12 10 8 6 10 8 6 4 7 6 5 4 3 2 7 6 5 A
HERMAN INFRARED
AV 3 Av - 2 Av A I Av - 0 Sv • -L100 1-= r r-rr -r-r
V, 3 2 321 3 2 1 03 2 1; 80
S60-
*40
0-600 650 700 750 800 3B0
WAVELENGTH (nm)
Figure 66. Spectrum of the nitrogen Herman infrared v' = 3 and first-Cositive
systems excited in the energy pooling of No(A, vl = 3) for a
nitrogen partial pressure of 0.46 torr. The experimentalspectrum is the light line while the heavy line shows thesynthetic best fit to the spectrum.
N2(8 9 " A 3U')
AV. SV -3 AV 2 AI Irr r r T 7 7 -I i r ' 1 'U ' ' 1 ' I "
. -12 108 6 10 a 6 4 7 6 5 4 3 2 7 6 5 4qERMAN I NFRARED
1 v 3 Ay - 2 av • I % - 0 av -
v 3 3 2 3 21 3 2 103 2
60-
0 F
- 40-
600 60 700 5o goo 850
WAYELENGTH (na)
Figure £7. Scectrum of the nitrogen Herman infrared v' = 3 and firsC-pos 1..ive% systems excited _,n the energy pooling of v-? .%, = J) f~ r a
nitro,.en artial pressure of 0.027 tDrr. The experlmentalspectrura is the light line while the heavy line shows thesynthetic best fit to the apectrum.
1Od
etL
-2p
5 10 Is
[92] (1015 -noleviles cm-3)
• Figure 63. Vriation in the ratio of .t-he sazuare of the rvcber densi-zv. of N, (A, vI t ) o that for N ,(B , v' 2) as a fu/nct Dn of
-. ',.the molecular nitrogen number density .
*'9
12
8s --
2-
1
41L o!I II I
L.OJ
44
0 2 4 1 8 0 12 1
-N2) (I015 molecules cm-3)
Figre 69. aratioin the ratio of the sqare of the number density ofNofA, v' V = ) t aat r r (, 4) as af-anction ft. e
K:tezur molecu a nronumbe r den ens.t.
II P5r
-4%,Z5 -r- -
IP6.
40
'J~;- 2 0 L.
0 5 10 150N2] (1015 molecules cm 3 )
Figure 70. Variation in the ratio of the square of the number densit; ofN-(A, v' 0) to that for N2(B, v' ) as a finction of tne
molecular nitrogen number density.
- -4
-- 5 -
0
.o" o I0 5 10 15 zo
CN2) (1015 molecules cm3 )
,igure 71. Varation in the ratio of the square of t.he number denS-=.tv ofNA, v' = J) to that for N.,(B, v' = 10) as a finction of tnemoleular nitr,-en nuacer Jensity.
110
r %U hl" ,, ,, " ',,w " ' ' %; ",,, "-, , , ,
were at 3.0 tort :tal pressure, but scans at 1.5 and 6 torr gave si--i.ar
vibrational distrizbtions, and similar ratios of B state number density to the
square of tie A state number density as the data just discussed (Fig. 72).Thus, at the !owes: ntroqen partial pressures studied, N2 (3) quenching by e
neon. argon, xenon, and C-i in the reactor did not appear to be signli-icant.
Therefore, a study was made of the variations in B state formation from the
energy pooling of vibrationally excited N2 (A) under conditions cocparabie to
those produciag the minimum 3 state quenching observed in the N2 (A, v' = )
studies.
_G-16
a',..
N • 1.5 torr
0 3.0 torr
A 6.0 torr
EXTRAPOLATED*1' , ' RAT.- COOUF ICE -''
RA-T7 'O~j~NI0-17
0 2 4 6 8 10 12VISRATICNAL LEVEL
Figure 72. Variation in '7-(3, v) excited from energy pooling of N-)(A, v'=3)wi:n neon pressure.
For these studies, the system was considered to be essentially two vibra-
tional leveis of N2 (A), i.e., v' = 0 and v' > 0. An analysis similar to th .at
given aove for tht- :42(':) state gives
i 1II
o , k B V k , ~
ON 2~N'Sv ^00N 2 (A, v > 3)2' . ' - rad -1.267 kO0 *.v -
2 9 - 2 534*S' '= ,2k,V 'B, N[ v' = -
V, V 3 2
kB v rN (A, > 0)-12
+ 1.267 v N( (76)kB,Vl 'N2 (A, =
rad
where we have corrected only for molecular-nitrogen quenching ( 10 percent
effect). 2uenchlng-rate coefficients determined in the N 2 (A, v 1 = 0) st.. as
were used. -:ore -3 throuzh 73 show how the spectrum 5e-50een 353 ani -72
chnnes as the c.e' ,: or vibraticnal excitation in the "' ,'A) Zhanges. raurs 7
through 78 shcw several different 3 state vibrational levels plotted accoriing
to Eq. (76). The intercepts of these plots gave results within 10 perzent of
these v' 0 results. These results also show that the pooling between a
v' = 0 and a v 1 > 0 is somewhat faster than the pooling between two v' =
molecules or between two v' > 0 molecules. Table 12 summarizes the rate coef-
ficients measured.
3.2.6 Discussion of energy-pooling results--Examination of the data
shows several interesting results. Close to half of the 3 st"ae forzatior
arises as a result of radiative cascade out of N2 (C). Furthermore, surnmii up
the rate coefficients for energy pooling into all three observed electronlc
states gires a value of 3.0 and 4.7 x 10 - 1 0 cm3 molecule-1 z-1 for enersv
pooling of two ;2lA, v' - 0) and of N2 (A, v' = 0) + N2 (A, v' = 1), respec-
tively. These numbers represent a lower limit to the total energy pooling
rate coefficient, because dark states might be involved, and also beca:use we
have not been able to evaluate the contributions of N2 (3, v1 = O) formation.
Since vibrational levels 1 through 8 are all formed at least 50 percent by
radiative cascade from the C state, we would expect N2 (B, v' = 0) to follow
suit. This would raise the total pooling rate coefficient by less than
10 per:ent. The formation of Jark states cannot readi.ly be assessed. A
I I1
112
SO '
9*i A 0 " 0 i
A N2 k3 3 g-A Eu)
' 12 108a 10 8 6 4
6 -
>- 0
= U
560G 580 600 6240 640 6650 680 7CC 720
WAVELENGTH (nm)
aF'.iur.i 73. azrlati: r, in !413 ,A) sptctrura produced in N-)(A) ener~v-
=oo2.jn,: with chan:es in nitr ) -n, and ar-.ori o>rtia-.
re. Vj, re
0 oo
rr C.0
3
IN0 >
0' CD£ '.0c-
(SI -o Nn AHi v AIS3N
r~5* VI
Coo
LO
0 00o 0 V2
-'I-
00:10
7*'1
03
('.,4
4 -
2 2
f01
0020.4 0.6 0.8A
CN2A, vW=1/[N 2A, v'=0J
Figu~re 76. 14'riation in the ratio of the number density of ',9(3, - v 2to th~e square off the number density of ',I-(A, vt 0)~ as a~function of !-he ratio of N,(A) v 1 to v' 0
CI 4 -I
%00
0.0 0.2) 0.4 0.6 0.8 .
£N2A, v'l1]/[N2 A v'=0)
F ~iur e 7 7. Varlation Inthe ratio of the number density of Nn(B, v -
to tne s'qui of the nuber density of N(,v 1 0 as afunction c.f te ratio )f NIA) v' 1 to vi -
NIf
m;21 F !
m ." 4
00 0 .2 0 1.4 0 1.6 0 1.8
[N2A, v'=1]/[N2A, Y'=O]
Figure 73. Variation in the ratio of the number density of "1Bv' "CmD*to tnke square of the number jensityr of N? (A, v' = ) as a-
f~uncz ion of the ratio of-\I (A) v' = I to v1 3
TABLE 12. Rate Coefficients for N2(B3.-,) Formation from- N2(A3-
Energy Pooling u
k ka
S I i 2.4 ! 0.6 3 6 ' --"' 6.6 ! 1.0 3 3 3.3 ! 1.O _.
, .-
2 3.9 ! 0. 2 2.3 1.6 ! 0.2 5.6 t 1.8 2 6 3.0 t 1.8 14.1 4.3
3 t 1.9 " 0. 4 !O. 0.1 5.8 -! 0.3 1.6 4.2 t- 0.3 "'
4 1.6 -! 0.1 1.1 0.5 -! 0.1 11.7 -t 1.0 1.1 1.6 t 1.0 3.8 : 2.5
i l 1 .5 ! 1 I 0 .8 10 .7 ! 0 . 1 2 .2 t 0 6 0 .6 1 .6 ! 0 .6 4 .5 ! .
n J I
| 6 1.1 ! 0.1( J. 10 6 -! 0.1 J.7 -! 0.2 0.4 1.3 ! .2 1.s5 ! o.5( .9 ! . 1 : 0 .3 1 .6 .! . 1 1 .6 , t o . , o . 3 1 . . ! o . 1 -
8 0 .6 t 0.! ') 2 0 .4 t- 0 .1 1.3 t 0 .1 0 .2 1.1 0 .11 --- ,
w 9 J.9 : 0.1 0.1 Q.8 0.1 1.0 ! 0.4 0.1 0.9 - 0.4 1.6 : . I A)
" 0I1 0 04 0. 0 2 1 .t .5
i =
N* I
N2 i 7.7 "- 1.0 21. - 6.2 28 - 11
NTA
"':E9F ' ioS n of 1t jNITS Of 10o11 CM3 ' LECUL " I to
TAL 2 ae ofiinsfo1 2 3hg Frainfo
EnryPoln
00.,.,.',"v" '" " ' , 01h € ".*" .'V "." , -k-,8 ± ,,-" , . "
availacle, most notably the W3_ U and B' 3u-tsates. We do see the 5-1 zand of
the B'-B system at 525 nm. In addition, a small amount of emission has been
seen in the 140 to 180 nm region which is probably from the Lyman-Birge-
- Hopfield system, hut the intensities are much too weak to resolve for an
adequate identification. The singlet states would not be expected to form a
significant exit channel even though their radiative lifetime is long, whicn
will in itself diminish the observed intensities of an otherwise strong exit
channel.
Clearly, the Hays and Oskam result for the formation of N2 (3) is iacor-
rect. Their studies were performed in the afterglow of a pulsed discharge of
nitroen. A number of metastable species in addition to N2 (A) will persist in
the afterglow of a nitrogen discharge, and we expect that one of these other
metastables is responsible for Hays and Oskam's observations. This hypothesis
Nwas tested briefly by diverting the nitrogen flow in our reactor to Xix with
the main argon flow upstream from the discharge. Figure 79 shows the result
-V 318 1 A ) E ' )
av-3 aw,2
'11) lot9 '8 ' 1 6 5 4 3 v''16 1S 14 13 I
3 2
p • 3.0 tOUR
IM DISCHARGED WITH Ar
M211 ADDED DOWdNSIREA
- - Or DISCIIARGE
6~j 60 61 see 7tJ ?g 7.V 78 no o* IA148DA (NkA11J4(1(SI
5iure 79. Spectra of H:? v' - 3 and N- first-positive systems when theNq(A) is formed by direct discharge of the N, with the arion
and when it is formed in the conventional manner by addingSthe '; downstream from the discharle.
117
*1j
gra phtai =a . Whi-a t_,soharzing the nitrogen with the aron nzreased t.e
N2 A) number densicy 'n the observation region by about a factor of 3, the
populations of the N-(B) vibrational levels grew by a factor of 70. -h
H7 3,1 -and which is relatively free from first-cositive overlap, and wh-hc
,s a o:ood zoni:cr on Lae increase in A state pooling, grew "sv 'nly a factor of1 4orme-d 4a th-e
as would be expected. Metastable N2(a'' u) molecules are
discharge (Refs. 65,140), and do have enough energy to excite N2 (3), -u: tnev
were not responsible for '2(3) formation. Adding H2 downstream from the f.-
charge, tut upstream from the observation region, completely eliminated N-'
emission at 171 nm but reduced the 3 state emission by less than 15 perent.
Other alternat._ves are being considered to the precursor of the enhancedJ E
state emission.
* Table 13 lists the rate coefficients for electronic quenching of Ni3,v%
determined in this study along with several other sets off values in the i4ter-
ature. Agreement with the measurements of Mitchell kRef. 141) and of
Shemansky (ef. 142), both of whom excited N2 (B,v) by electron impact, is fair
as is agreement with Gartner and Thrush's (Refs. 143,144) measurements ;-n a
recoabining N-atom afterglow. Agreement with Becker et al. 's (Ref. 145)
afterglow observations and the kinetic absorption spectroscopy determination
of rover and Perner (Ref. 146) following excitation of nitrogen by reletivis-
tic electrons is not good. :hese numbers are effective two-body quenching
rate ccefficients. The actual processes involved are much more coplex. Te
S_ 3a state of ni:rogen is coupled collisionally into a number of other hested
electronic states including the A3u ,3 3 W3, and perhaps variousI _u _U ) I
states off the singlet manifold including X (
Thus the quenching is actually a complex process which involves transferring
energy back and forth between the various states involved. Sadeghi and Setser
(Refs. 131, 132) and ?otem et al. (Refs. 133,147,148) have demonstrated this
coupling between the states unequivocally. These groups have excited specificvibrational levels of the 33 T state by laser pumping A-C ;_state moecules.
r '1 ' 'z-*-seuen: to the laser pumping they observe emission frou - -;a'
and lwer vibrational levels of the S3"g state, as well as pressure-dependent,
muil:exconential decays off tne fluorescence from the initially populated471:~.
Ne . 1~
-0
TALE 13. ?ate Ccefficientsa for N-(3) ,uencning cy N -.
Rotemb _a rtnoer 2oemIcThis Beker and and an/
% N B,v Work, I.i::ne:..I S aemansk7 et al. Rosenwaks Thrush ,Zerner
1 1.0-0.8 . 3 -1 1.2 )0.77 3.22:3.'2.
2 0.8=0.1 1.2 1.9:0.2 3.6 i1.89 3.32:].02;
I 3 3.0:0.2 2.1 1.710.2 9.0" 4 2.4_-3.1 1.9 1.8:0.2 4.3
5 2.6:0. I 1.3 2.3:0.2 6.8
6 2.1:0 .3 .5 4. i:0.47 6.4:3.3 2.4 6.6:0.4 -
a 8.0±2.4 7.2:0.4 7.5
9 7.0± 9.1.0 0.9 2.8
10 2.3-0.2 6.3:0.6 2.8 5.3 4.4:0.9
11 2.8 _0 . 2.0±0.3 7.2 -
a. Units 10-1 cm- molecule 1 s-.b. Based on two-state coupling model.
level. rn most cases the fluorescence decays can ze separated into :,o
pressure dependent components, one fairly fast, and the other quite s2,w. Therapi. a c .... en coupling into a reservoir state whi:n is
in eiuilizrium w4jh :ne ini..al, pumped level. 3enerally the W Nil s::te is
identified as the primar/ reservoir state.
Sade'ghi and Setser (Ref. 132) and Rotem et al. (Refs. 133,147,14:, tried
to describe their observations using the two-state coupling model given Oy
Yardley (Ref. 149. An atte.mpt has been made to derive a steady-state expres-
sion consistent with this model so that our observations coull be compare3u with
'4 those of Rotem and Rosenwaas (Ref. 133).
1.4
up%
kqO k W
The Iflowing reactions describe the model:
N 2 "(2,A) N2(3) + " 2(X) (77)
N (B) +N2 (X) BW N (W) + N (X) (76)
kWB
N (3) N (X) 2N (X) (79)2 2
k3
r2N (B) N2X) + h'()
kgN (W) NW(X) + 2N )2 2
14 (W) N 2(B) + hv 32
where we have ass:umed W as the reservoir state. This se: :f equations defines
the following rates:
-1 I kr -i kS3)
3 B
-= krwLN 2 5
R = kwBN2. (33)
3= W-r + kQrN2 i (37)
t
1201
4.'4
The rate of for=at:_on of state .W is
R = - ,R + R i 38dt 2 '-2 3'
Under steady-state conditions we obtain
%L[3[W] R-2 + R3
The rate of formazion of state B is
dEBa] - +(3)
d t "' 2 ' -2
,p Under steady-state conditions, with the inclusion of Eq. (39), Eq. (90)
becomes
R. + k + R3 2- 2
f 1 2 -2 3) 29_)"2,' FB]R_ + R3
a-2 3
The two-state coupling model of Yardley explains bi-exponential decay of state
B as
E[B2(t) = A ep A e 92?
Under conditions such that X >> X2 the following expressions apply:
X1 'R1 + R-2= k +3 "2(l+()
2 1 2 "+R 3 2 2
2 -2
B w B kWk< +k k yr r + --- +(95
-=1 k,4 A2a-K k AI
!11
,P,
Combining Eqs. (91 and (94) ;-ves
'f-- \ 2 = (97)
,-w [a. a ii)Applying expression (95) for N2 gives
3'w 3 W* a k r+ k r + k~
LaR 3 R 3
%2 2
This expression has a form similar to Eq. (75) which describes the quenching of
N2(3). The oroduct of the ratio of the factor multiolied by P to the zonstant
term in Eq. (97) imes k,3 gives the effective rate coefficient for quenchin;,
based upon rates determined experimentally from the laser-pumping experi-Ments,
i.e.,
EB w.: . .. Q -.. kwk ff kj B
k =, . f.
X +r
Rotem and Rosenwaks (Ref. 13) studied vibrational-level dependent decays as a
function of nitrogen pressure. The effective rate coefficient of their slcwvy
decaying component is
.3 .W,' . d , .- ,
k :Jj l "s :o I -y
This expression combines with Eq. (98) to give
k( 1+'(0keff Bs(I X) 3
2 . + .w r
We list values of based upon their measurements of and in Table 13.[ We used t'he radiative lifetimes of N23 n 2W cluated by Werner
et a!. (Ref. 113w'ere used to derive the keff values listed. The calcaiat-,n
122
assumes that the coupling is with the vibrationa level of --e st-te h
is closest in energy resonance to the vibrational level of B3 - cons e.rei
Clearly tnis approximation is somewhat simplistic and may account for tne
rather mediocre agreement between the measured, effective quenching rata coef-
ficient-, and those we calcu-ate based upon the two-state coupling mo-el and
Rotem and Rosenwaks' exper.mental results. Under the circumstances, pernaps a
factor of 2 to 3 agreement is acceptable. The coupling is probably too complex
to be explained e-tner bv an effective quenching rate coefficient, or 1y only a
two state model.
The microscopic details of the quenching processes strongly dep end upon
the quenching partner. Nitrogen appears to quench electronic energy out of all
levels at roughly ccnparable rates (Ref. 150). Argon, on the other hand, is
much less efficient than nitrogen at quenching B3 g manifold electronically,
but it rather efficiently alters the vibrational distribution. Figure 0
demcnstrates this point dramatically. The spectrum in Fig. 80a, for which the
nitrogen partial pressure is 0.21 torr, shows fairly strong first-positive
band attenuation relative to the unquenched HIR. band at 703 am. The vibra-
ticnal distribution, however, is quice similar to that in Fig. 73 which was
taken under cond':tions of sufficiently low nitrogen partial pressure that
electron1c quenching is minimal. Figure 30b shows that reducing the nmtrogen
partial pressure so as to eliminate N2 quenching effects results in a much
stronger first-positive system intensity. In argon, however, the vizrational-
level .istributlcn is drastically altered. Raising the argon pressure
furtner, as shown in Figs. a0c and d, does result in some reductioi in the
first-positive system intensity, but more noticeable is the continual shifting
of the vibrational distri:bution. Especially interesting in Fig. 80d, wnicn is
in 10 torr of argon, is that the vibrational distribution appears to hang up
in v' - S. The apparent electronic quenching by argon in Fig. 8Cc and d com-
pared to Fig. 8Cb might result only from shifting the B3 7g population into
vibrat-inal levels v' - 3-2 which do not emit in the spectral region observed.
Tne observations doul have to be extended into the infrared to clarify tais
pcint. Clearly N-(B c ) quenching is an extremely complex process, and simple
mccesi; su:h as !zol by 7-. (75) or even Eq. (98) are not completey
123
P~r -147 trr Ar = .47 crr
SPN2 =0.210 torr 0. PN2 0 .0:15 't o r
S-0 J . 0.4 I
0.4-U. 0.6 4' t
S0.2 ~ .2
600 650 700 600 650 7CO
WA'/ELENGTH (nm) WAVELEINGTH (.nr)
(a) (b)
PAr = 3..00 torr 1. PAr=101tr
~ .3 PN2 = 0.023 torr 0. PN2 - 0.014 ton'
S0.5-08
0.5 0.6
_ .4. I 4-0.2 I 0.4-
L"FiN
0.0 ____ ____ ___ ____ ____ _ _ 0 .0~
600 650 7CO 600 650 700oWAVELENGTH (nm) WAVELENGTH (nm)
F~ire 3J. Variation 4n NJ-i3,A) spectr-wlr produced in ()enrv o1n-*wi::- crnanjes in nitrogen ind argon partial pressu;res.
124
*6 A'Mp
* -------------------- -
"" ' 3.3 ACTIVE NITRCGE' ITAI _Z OF :F
3.3.1 :ntroduction--Two years ago it was observed that adding iodine
monofluoride to a stream of active nitrogen resulted in strong excitation of
IF(3 '-O) (Re'. ). The effective excitation rate of the IF(B) appearad to
correlate with the number density of (3. g) in the observation zone. The
N 2 (3) could be ruled out as the precursor of the IF(B) excitation, however,
because an excitation rate coefficient several orders of magnitude greater
than gas kinetic would have been required to explain the magnitude of the
!F(S) fluorescence observed. The N2 (3) appeared to act as a tracer of a dark
specis which itself was resoonsiole for the excitation. The excitation-rate
measu rements indicated that the product of the excitation-rate coefficient and
.0the number density of the exciting species was approximately 1 s- 1. This
* implies that the number density of the exciting species is at least 101 0 -ole-
cule cm-a 3 , given that the maximum expected excitation rate coefficient is
roughly 10- 10 cm3 molecule "1 s-. Various arguments indicated that most of
the comon metastables in the active nitrogen flow, such as N(2 D,2p),
N2ta 1 "g), N2 (a1 :. -, N2(A 3 Zu+), and N2(W 3 ju), were all present in too low a
concentration to ce responsible for the excitation, or else varied in number
density with changes in conditions in a fashion incompatible with the coserved
changes in IF(i; excitation rate. It was concluded, t-herefore, that the ost
likely species in the reactor responsible for the excitation was ground-
electronic-state, molecular nitrogen in relatively high vibrational levels.
Observations would, therefore, be clacsified as observations of vibrational-
to-electronic (V-E) energy transfer. Several cases of this phenomenon exist
in the literature (Refs. 151-153).
:he observations of fairly large number densities of N2 (3) were inter-
esting in that the the vibrational distribution and intensity of the observed
first positive bands were incompatible with formation from N-atom recombina-
tion. Tne indication of this is that adding molecules like 02, CO, and CO2
drastically reduced the first-positive intensity, by more than an order of
Ma-nitide. Under the conditions of our experiments, however, none of these
mclec'iles reacts at any si.niflicant rate with atomic nitrogen. Thus the
125
,..
reduction in intensity was not a result of N-atom removal. Flrthermore, the
res4iual first-positive intensities did begin to show the spectral distribu-
tion characteristic of t-he "-atom-recombination-produced, Lewis-Rayleigh
afterglow. Since these measurements were all made 10 to 2C :s downstream from
the exci:ing discharge, scme species in the active nitrogen, with a if-
against quenching and radiation in excess of tens of milliseconds, must be
collisiona!ly coupled to the N2(3) (and upon addition of IF, the IF(B)) in the
observation volume. The contention again was that this collisional couoling
is effected through high vibrational levels of the ground-electronic state of
molecular nitrogen. The kinetics of these coupling processes are discussed in
some detail in Rafs. 154,155.
% - The purpose of this task was to try to quantify our observations more
caref .lly than in the past, and to try to provide some additional support for
the contention that Nn(v) was the energetic species responsible for the :F
excitation. Some time was spent: trying to develop a relatively clean source
of N2(v) which would have reduced number densities of nitrogen atoms and vari-
ous atomic and molecular metastables; trying to implement and characterize .a
diagnostic for vibrationally excited nitrogen in the flow reactor, and addi-
tionally to study more fully the transfer of energy between active nitrogen
and :7.
.3.2 evelooent cf N-Av) source and diaqnostic
3.3.. ntrcducticn--A source of vibrationall'. excited, grouno-
ate nitrogen that as free from atoms or atomic and molecular metastables
would prove invaluable in these studies. Morgan and Schiff showed that vibra-
tionally excited nitrogen, which was created in a microwave discharge thnrough
nitrogen, was easily removed with only a slight reduction in the number den-
sity of nitrogen a-ons when a glass wool plug was placed in the flow reactor
some~where downstream from the discharge (Ref. 156). This tecniue develoned
froa their observation that the vibrationally excited nir -nW.3s moreez-
cient'v .remove from their floa reactor by wall collisions than dere the
nitrogen atoms. if one could reverse the order of wall-reccibination
12
efficiencies, then one coul-d enision removing the nitrogen at:zms -- zr at
least drastically reducing them -- -while retaining the vibrationally excited
nitrogen. Because wall collisions remove electronic metastable species 'witn
almost -,nit efficiency, one would then be left with a flow t:hat was
vibratio nally excited, ground-electronic-state nitrogen. Wire screens.were
added to the flow tube downstream from the discharge, hoping that they would
remove the atoms eff-iciently, while leaving the vibrationally excite d r.itrogen
larg:ely unaffected4. This idea stems largely from t-he work of Reeves and'
coworkers (Refs. 157,158), and KZenner and Ogryzlo (Ref. 159), who have
observed elactror, :: excitacion following catalytic recombination of at. Mns o)n
vario-us wire screens. most of the work used screens made f1rzm nick<el w-_ra.
The first idea foer developing a diagnostic for vibrationally ex=c:eo
nitrogjen was to extend the pioneering work of SchmelItakopf et a!. (Ref. 160))
and Young and ::orn ('Refs. 161,162). They relied on the vertical nature
Penning-ionization transitions from ground-electronic-state, neutral ntoe
to ratic N+(B v)distribution characteristic of the ground-sIt di-
tributions. >lixin-g metastable -ieli-= atoms with a flow of molecular litrogen
results in strong arissicn of the nitrogen first-negative system, N-kB3 +
x- Since the Penning-ionization process follows a Franck-Condon ex:ica-
tion pathway, --"e vibrational distribution in the neutral, ground state will
determine the distribution observed in the upper, ionic state. One probl_,em
with thi4s approach is that care must be taken not to have any He"' or Hie-,
thne flow of metasctle helium. Both of those soecies also excite N,,3, quita
strongly in charze-tranSfer reactions, but with an *i2 %3) viorationa.dst
buton, th-at is -decidedly non-Franck-Condon (Ref. 163). The ma~or pro'_lem in
applying this dia:.nostic to our studies of energy transfer to 1F is that tne
dianosicis most sensitive to vibrational excitation of t.ne first few .ri*-ra-
tional levels in the ground state. The excitation of IF(B), and N2(3), On the
other hiand, requires vibration~al excitation of about 10 and 30 vioratior;nal
quanta respectively to effect the excitation. We had hoped to be able to
estimate the overall vibrational distribution of the ground-state nitr:z'en by
comoninn our exerimental determination of the vibrational distribution in
*the f~rat few levels with modIeling calculations on tne temocoral _deve.' 'Mn
f o l1- .' t e -anif) ,1
1 11IN
3.3.2.2 Exerimental--The apoaratus used in these exzeriments has
been described in most respects in Subparagraohs 3.1.2 and 3.2.2. T hi sara-
graph 4-escrites the =odifications required for the measurements related to
N2 v). Fiure S, sncws the conficuration of the flow reactor. A microwave
- discharge at the -.istream end of the flow reactor, throutgn a flow of
and nitrogen, created the active nitrogen. The active nitrogen entered a sac-
tion of 2-in-diameter Pyrexl containing three side arms. Small amounts of SF6
flowed :hrough the first side arm to attach electrons created in the Penning-
ionizat.on reactions. A rod, to which the metal screens were attached,
extended throuzh the second side arm. This allowed the screen to be rotated
.ether -ara.. eal to, or normal to, the gas flow so as to vary: the effectiv:e
diffusion :enz:h between the -as species and the -dire surfaces on the screen.
.t. .Thus, tne efficiency could be varied with which the screen scavenged lct:'-e
0 species from the fl.:w. The metastable helium atoms entered throu,,h tne t.'-ri
side arm. A second length of 2-in flow tube seoarated the .- (v) preparaticn
and characterization section from the stainless steel viewing region. This
.,
me C 4
IL:P, r[ ,.'
I-r
EFL -u~~kT. CO E t 1 ,
I "-- 4 1a U
Fi 7ure -20 Flow reactor for studies on the vibrationei - e r -z co n e- f* active ~nitrog en an-' of IF excitation by active n~>e.
% .% % % %
section was :eflor. coated and c;ontained an in'ectr for . We h'ave z'.....
and characterized the IF source previously (Ref. 6). The 0.5-" =onochrz7ezr
is mounted on rails so that it could Ze moved toward the upstream end of the
reactor to observe the fluorescence created in the Penning-ionization reac-
tion, cr back to thea stainless steel viewing region to observe the interaction
of active nitrogen with IF. A photomultiplier/interference fil'er comination
(580 _ 5 nm) provided photometric measurements of the intensity of the
nitrogen first-pocsitive bands in the viewing region.
A hollow-cathode, dc discharge through a flow of purified helium crevated
the .meastable helium atoms. Flowing the gas through a molecular-sieve trao.
at liquid nitroq:en temperature upstream from the discharge removed most of the
* impurities in the helium, including nitrogen and oxygen, with the excezn
some residual neon. The small amounts of neon (01 ppm) have no effect cn
observatios. A number of years ago, it was shown that operating the hollow-
zathode discharge at high voltage and high current tended to produce snlf- -
cant number densities of atomic and molecular helium ions (Ref. 163). These
ions are an anathema to the Penning-ionization measurements because they .ro-
duce highly non-Franck-Condon distributions in the N2 (B). Thus one wou' e
led to false conclusions regarding the extent of ground-state vibratio.n
excitazon. The discharge was operated at 350 V with a current, limite- 4: a
200 '-. resistor, below a milliamp. The absence of significant ionic numoer
densiti-es was confirmed by failing to observe -2+(B, v' > 2) when the active-
nitrocen jischarge was off.
3.3.2.3 Theory behind the Penning-ionization measurements--.hne
:enninz-ionization between metastable helium atoms and molecular nitroen is a
vertical process. One can calculate the vibrational distribution in the final
state, therefore, knowing onl Z the vibrational distribution in the lower state
and the Frenck-Condon factors that couple the two states together. Thus
.4 VI N f LJZ)V1 v" v' V'V "
0.
vow or
wnere v' and v" :euresenz the v'braticnal levels of the upper and lower
states, respect ively, and fv'v. is the Franck-Condon factor coupling them. :n
theory, if one measured the vibrational distribution in the upper state, it
could ze determined in thne lower state by simply inverting the Franck-Ccncn
matrix, and Luyt"o-,;ing this inverse matrix on bcr.h sides :f Eq. (102). his
procedure does not work in practice because the measurements of the uoper-
* state populations have some uncertainty associated with them, and the uncer-
taintles become jreatly magnified by tFne matrix multiplication. Jolly at al.
(Ref. 166) apparently did use this approach to analyze :12 (v) distribu tions
created in a glcw discharge. We do not understand their success in light of
our lack of it. Our spectra' fitting approach for determining -;2' ' nummer
zlensities ought to be more accurate than their graphical integration of one
branch from each band. Our data was analyzed using a model to describe t.e
N h) vibrational distributions to predict ;24 (3 ) vibrational dstri u-ions
and :-en ccmoarinz these predictions with the observations.
A set of Franack-Condon factors was calculated over the range of grsund-
electroni--state vibrational levels of 0-18, and N2 (B)-state vibrational
4-v. levels of 0-9. The calculational procedure was previously discussed for
-deterL'n:ng Franck-Condon factors (?.ef. 167). razle 14 snows tne resu:s of
these 3rlatiors. given the Franck-Condon factors in Table 14, tne ex. ectez
';2 k)-state "-izrational zoPulatons were calculated relative to that n
= ) for a number of values of vibrational temcer -.ure of gron- -
state nitrogen etdween ,- and 0,00 K, assuming a Soltzmann o~srauin
among the levels. Figure E2 shows the results of tihese calculations. If the
, ground-state vibrational level populations follow a Boltzmann distribution,
one can use experimentally derived ratios of the populations of the various
vibrational levels of N2 (B) relative to that for 42'(B , v' = 0) and Figure a2
to find the vibrational temperatures of ground-state nitrogen which corre-
zron -- to those excited-state poculation ratios. The excite--state cop,.ation
ratios, of course, result from analyzing the N-,)3-X) spectrim excitei in the
- "..-.on z ati n reaction.
:ALE 14. Franck-Condor. Factors of N-* (B'u +, v) - Nn{X - g, X,.
r io b J.3 4 )i .. ,
Ivi*Ii j.,,45 .107u .012
3 1L. . 1 1
- i 4 . .1 7 2. .J33
0.06 104 -3z .08 3 44 081 )09 .J5 XC121
.2i .3.la .309 5 . 4C.9 .399 ).3940 . 17 . 05 J.01i
)I t
'0, .40 4"3 J
1.05, 2994 :56 51 J 1 13d I Q.
1 13 3..3 0 -5355 37 0 C .5 4 2.
.
3Z Ra i of'."+ 3 '. o N * v, )cr =" in "4e" :enn--.:--n-.'-tzoq of 'N-(X, v". ii a fuanction of :,0.r -__- . __on__ _ .
I~* I l t.-- m H Nm III i l i l Il .i- -
%f th-e ground-electronic-state vibrational levels do not fol a
Boltzmann distribution, one can still compute expected excited state Ccula-
"ions fr-_m Eq. (,31), if the ground-state distribution is known. Effler.ts
from nt-4rgen olscharges are known to have nonezuilibrium vbZrazional /istri-
butiz:ns, w1tn t-e higher vibrational levels generally more strongly -. guatea
than would be predicted on the basis of a Boltzmann distribution. Capie.i
and Dilonardo (Ref. 163) provided an analytical approxi.zati-n which is reasor.-
ably accurate for ground-state vibrational levels below about v" = !I. For
vibrational levels below the Treanor minimum, v*, tne distribution 's that
given by 7reanor et al. (Ref. 169):
N 1 . 4 388 - 2e k.4 1VI " = - v r e e e (v- ). '3)
-
where [ is -ie 3cltzmann vibrational temperature referenced o *" = I, - Is
the a2bient gas temperature, and 'e and AeXe are soec:roscooic constan:.s in
units of cm The Boltzmann vibrational temperature is gven ny
'!'I 2,e x-GI < n N v,,=1 1/v,=01
For vibrational levels above the Treanor minimum, we have
V v-1 *2 e x:4- = -- exp - (v) -0.5 '
whe _re v is
. e ev I + 0.5 16)- 2.,- x -5
a e e 1
The Pennin7-ionization spectrum calculated from a ground-state distrib3UtIon
based uPon Ecs. (132), (103), and (134) is significantly different froa what
* ne woul calm from a Boltzmann ground-state distribution. The 1ennin:-
ionizatin ie, therefore ought to differentiate between the two 3rouna-
state Oistr u't3sn rather easai.v.
W
132
.0 . _ PI.P A . N 2 ,-
3.3.2.4 Results of the Penning-ionization measurents--F=cures
33 and 34 show a portion of the nitrogen first-negative spectrum with the
active-nitrogen discharge off and on, respectively. Clearly, t-he vibratioRal
J" develooment of the emission is greatly enhanced by discharging the gas. 'e
determined vibrational populations in the upper state by fitting the spectr:u
4n the manner already described in Subparagraph 3.1.2. Early in this measure-
ment program, it was discovered that moving the monochromator slightly down-
stream from the injector, t..rough which the He* entered the reactor, resulted
in a much d .t_-%-ent vibrational iistribution in the upper state. There
appeared t.) be no reason why the vibrational distribution in the active nitro-
gen should change over short distances. A diffuse emission could be seen
extending somewhat hownstrean from the well-defined flame created by the"NJ' Penning-ionization. Several centimeters downstream from the inlector, tie
PennnZ-4onization flame disappeared, and only the diffuse emission re-,ained.
-Fiure 35 shows this sOectrum. Cn the assumption that the emission was cause-
by energetic electrons exciting N2+(X), a trace of SF6 was added to act as a
scavenger for tne electrons. This addition eliminated the diffuse flame, and
also resulted in Ni (3) vibrational distributions which did not change with
the location of the 2onochrcm.ator relative to the He* injector. Presumably,
the electrors created in the Penning-ionization reaction pick up some extra
kinetic energy either from stray microwave fields which have penetratei down-
stream, or else from collisions with energetic species in the active nitrogen.
The aincunt of ener-y must oe fairly ccnsicerable, because exciting N2"-) from
S"Z ,"; re-ires =ore than 3 eV. All the results given in the following oar3-
zraon were obtained in the .nresence of small traces of SF6.
The first-negative spectra were fit to obtain populations for levels 2-3,
alchough, for the most part, only vibrational levels 0-7 contributed any
significant intensity to the spectra. A number of different and conflicting
-Sj, sets of --instein co-efficients for the N,(3,u fill the iterature
-(Eels. 3,112,170,171). in oarticular, the constancy of the electronic tran-
% sit iorn oent w tn r-centroid is unsettled (Refs. 172,173). For the analysis
of data ta'.3n for this report, a set of Einstein coefficients have been cacu-
lated base,' uopon the relative electronic transition moment variation 2i';en b'.
ar~wn Jn-L., :,.efs. 72, 174, 175), tue Frinck-Condon factors of
% 4.
VIBRATIONALLY COLD N ,
, 42 *1 2 1: 544Ez842464
(2,4)
-
WXI I(,~
4r 4 446AINAL 45- T 4z2 24E6 44o7
.0 eiAvELENG2H (nm)
S F .g r,,! 3 . Ser- of tc -v r -2 saquence of the N2+(3 2> + +,£yt.
excited in the ernn-on4 atio-n of active nitro,4en bv
134
Al I
ex tji enn-o a iono iiento nb
pe ,. -r-snpad bec o 2 ,
13
.ri~o et a!.-Ref-- j3-------rditiv .----- o n
[ (Ref. 3-:). Tacle 15 lists thne calcu-a-.ed Einstein coefficientsa. T his issue
R • of trans;.tion-mczenz variation is oeing investigated further. !he i4n-er-
" '=% ole=!:Iar : otenti:al constants givier in Lofthus and Kruoen-'e "Ref. --a" were' .% found to 'be I nad :uate for oredictin3 badpstosacura tev h tin
band -ci~n ac - ' " ttr
-]1
TABL .54fr,.132Alz rgra- :es a..th ~ nrraite reeIptnia osat ofin Gottch eM . ('=ef. 1 n.
). ABLE "5 linst-n Coeffci~ens foEinstein Coe(ffiie-en. ts u
'of v,, I vi o i bein st e r te i;e 2-
0 o1 i3 .,.3 0.-886 .15 4 .0 C 215
1 5.47 ; .75 .3 . 4 1.71 .454 0.097 0. 0. 8 i
3 .rcgr .2 7 9se "1. 0een poenia 2.3nt o.9f 8 0 3 6h et al. 0 .. 37
4?. . 0, 15 4 3.54 7.29 0.086 1.34 2.09 126 0.5 6 0.16,[r '.5 0.253 4.32 6.59 t0.327 0.662 1.74 11.31 ;0.62210.2271 1
0.01 021 3Ill'*--'0 ,. 2.7I. 8 0 .1 00.O26154 6.9 .35o75 .0 .5
3 0.066 0.12 6.32 8.32 .37241.2.29 100.38 5 3
a. Units of 1.6 s- I
Analyzing the N2 '(B,v') vibrational distributions with the helo of
.ur- 32 ;ave ifferent ground-state vibrational temperatures ce~enc.: : .ipon
.. .... e-statc vibrational level was considered. The aa:a f el into t'ree
. groups. The teiperature ietermined from v' = I was only 63 per:ent of tnat
0deter:lined from ,' = 2 and 3, which were themselves reasonably corsistent.
Vibratlonaal temperatures determined from v' = 4-7 also were reasonaily consis-
tent with each other, but were 30 percent larger than temperatures determined
from vibrational levels 2 and 3. Clearly, a Boltzmann model for the ground-
state vibrational distribution is inadequate.
136
L-%- 116 j
'
Fiting our _tributor,ns to the analytical model given by :api.el_: and
Di-onarto was more successful. Figures S6 and 87 snow two examples. -learly,
the nonequiiibrium model does a reasonable job of fitting observations out
through v' = 6. The sudden fiscrepenc'y at v'=7 is not clearly understood.
The d0str.bution is shown which would be predicted were the ground-state to be
determined by a Boltzmann distribution. This shows graphically the inadequacy
of that model for tne ground-state levels. We hope to use an imoroved model
to predict the population distribution up to higher vibrational levels.
Observations show that effective vibrational temperatures, ZI, as Jetar-
[ % mined froo fits of tne data to Eqs. ('03) and (105) rather than what one woulf
.- calculate from _7. (104), tended to be smaller with larger nitrogen mole frac-
t-ons flowing ttrough the discharge (Figure 88) and also at higher total rres-
* suras. Placing a nickel screen in the reactor, downstream from the discharge,
only sligttly reduced the effective vibrational temoerature of the nitrogea
(about 5 percent). The Penning-ionization technique appears to orovi'-e a
rasonably accurate monitor of the vibrational distribution of ground-
electrjnic-state nitrogen, at least for the lower vibrational levels.
Extending the model to include higher vibrational levels would need
experimental confirmation. An %42 (v) diagnostic is being developed wnicn
shouli prove sensitive to higher vibrational levels under another Contricz.
was uncertain as to how well the populations of the more ener~etic
vibrational levels of Ni2 (v) could be deter.ned using the tachniyie jast
jescrized coupled witn model calcullations, we sought a diagnostic for :2,
containing energies of several eV. These are the levels whicn would be
responsible for the observed excitation of iF(B) and N2 (B). One possibility
is to search for tail tbands of the first-negative system. The tall bends are
bands from h lying levels of ";2(B); e.g., v' = 12-16, which are redI _%!de sraded and which appear to the red of the normal, blue degr3aded seluences of
-te first-negat:.ve system. Levels up through v' = 18 are enerjetic'a;
',% acce~ssile from 0 of '; 2 (X), but vertical transitions t- these levels
% -nwould nos: i ccu r frm v" = 15-20 which lie 4 to 5 e' anove v"
S.7
N0N
-e r J. %.? %~ %I % C ~ C C ~ p
5, .'-
-- NON-_Q IL ..!UM "
I 'N DISTRIBuT:CN
SBOLTZMANN- -~ ~. DISrRIBUr EON -r '
-42 0.046
VIBRATIUNAL LEVEL
Figure 36. Comparison between experimental and calculated vibrational distri-butions of N'(B3) created in the Pennina-ionization of active
nitrogen by metastabie helium atoms for a nitrocen mole fractionof 0.011 (p = 1.5 torr, transit time from discharge = 11 ma).
-- 8DISTRIBUTION
-£J 8U L T Z A N N N N
-4 -X2 .1 ISTRIBUTION IN 'N
0X
4-4
,,'[-'% V12RATIONAL LFE EL
- 9',
1AFi~ure 3 Compa.:.rison etwe2en exzeri~nental and calcuated v'iorationaldir-~~butions of '^{ created in the Penning-ionization of active
nitroler, by ,-3:-table hnelli:ir atoms f.'r a nitrogen al rc,~,:f :. 4t, p , tcorr, _transit ti-ne .4rz,m s h r e = 1 =s
138
Bu'N
5CCC-
3CCO L
0.0
Figuire *B8. Effwictive vinrationa. temoerature from the nonequilibriumS ~~~mocel-versus aitrogen mole fraction througn the ds:~:e
Another possibility is to excite the nitrogen second-negative system,Z.2 _- X 2 ) . This system can be excited by collisions bter
-N ( 1: z-
metastable helium and molecular nitrogen only if the nitrogen has at lest,
3.8 eV of internal energy. Because the N24 (C) potential curve is somewhat
5- of fset from that of N2 (X) , vertical transitions connecting the two states
would arise 'from ground-state levels around v" -10. Observing the secor.4-
negative 3ystam, th-erezore, would provide unequivocal evience of highl'!
'4excited gro:und-stata nitlrotgen in the reactor.
I 5IN I I 1 1 1 1 1 1 1 ' l 1 1 1 1 " ' 1 1I .I ' l ' , , 1 1
A quartz piece ,;as built for the flow reactor so that a search coal -e 7..ace
for the second-nega 'ive system in the ultraviolet between I15 and 100 .L.
Fi ure 89 shows thne spectral region between 138 and 208 am with the metastable
heliuim off and on. -n both cases, the active-nitrogen discnar-e is on. -ne
V. two featres cearin: at 191 and 199 7= are the ,v = -6 an- sezuences,
respecti;ely, of the nitrogen second-negative system. The lv = -3 sequence at
206 rm is masked by the 2,0 band of the NO(A2I+ - X 2 ) system. This system is
excited by interactions between N2 (A3_u+) created in the active-nitrogen zis-
char-ze, and some impurity NO also created in the discharge (Ref. 177). Scans
with the active-nitrogen discharge off, but the metastable-helium dischar;e on
resulted in no observed emissions. This latter observation is imocrtant
-e because it esta'lishes that He+ is absent from the metastabla heliom flow.
NW Charce-transfer between He+ and N2(X) is a well-known source of the secoria-negative system (mefs. 163,178).
[ " I I I NO (A 2-, + ,, = 2 X.>2 , , " C
U." o - - ,
.4,r -~ X2- I)N2+( C
2Z u+ - L I
0- v c -7 -
., g I
S" t METASTABLES
il!.,% O FI
MEATA L EN1
• .. OF5 20 25
195 'II- ' rAVELEIUT h (nun)
.•
7'iure 29. Lhectra frcm active nitroten between 164 and 203 nm in theabsence ana presence of metastable ielium ats::s.
140--t
I~~ %, %.,,-% %,, '
Observing t.-.e nitrogen secornd-negativil systeam from t -, iateractzr.
between metastable helium atoms and active nitrogen establishes the cresence
in the afterglow of metastable nitrogen containing at least 3.8 eV. r.e pos-
sibility for these metastables might be N2(A 3 :u+) tr N2 (a'1 2u- ) both of wnih
have beea observed in active nitrogen afterglows. Clean sources have oeen
made of these two metastables, therefore, and ocserved the resulting spectra
when they were mixed with helium metastables. The i2 (A3ku ) is. produced
cleanly by adding N2 downstream from a hollow-cathode, d.c. discharge through
a mixture of helium and argon. Scans of the Vegard-Kaplan bands between
220 and 400 nm established that the N2 (A) number densities from this source
were cf comparable magnitude to those obtained from the active-nitrogen
discharge. Turning the metastable nitrogen on and off in the presence of
metastable helium had no effect on the emissions in the region between
185 and 200 nm. N2 (a') is made along with slightly enhanced number densities
of N-2(A) - about a factor of 3 - when the molecular nitrogen flows thr)'gh the
dc discharge with the He/Ar mixture. Again, turning the metastable nitrogen
on and off had no effect on the emissions between 185 and 200 rim. It was con-
cluded tnat interactions between metastable helium and N2 (A) and N2 (a') do not
,.roduce N2+(C). The appearance of N2 (C) from the interaction between meua-
'.% stable helium and active nitrogen, therefore, appears to be the result of
,eui -Penning-ionization of ground-electronic-state nitrogen containing at least
3.8 eV ot vibrational energy.
The first-negative bands in the presence and absence of the molecular
nitrocen :etastables were monitored, to see if the hollow-cathode sources oro-
"duced any significant levels of vibrationally excited nitrogen in lower vibra-
tional levels. The ratio of the populations of v' 1 1 to v' = 0 of N, +(3)
remained unchanged in the presence or absence of N2 (A) or N2 (a'), and we saw
no evidence of any higher levels of N2 +kB) except for the trace of v' - 2
which accounts for 0.2 percent of the Penning-ionization of 300 K N-. We con-
clude that the vibrational temperature produced in the hollow-cathode dis-
charies is below 100 K and that this discharge produces no significant levels
of vibrationally excited nitrogen, i.e., > 96 percent v' = 0.
141
1.'il lijwj:, il.
:n summaro f :(v) by metastable helium ato*s
excites N-'(3) u;p to at least v' - . Analysis indicates the ground-st.ate
nitrogen vibraticnal iistribution is highly non-Boltzmann, with effect've
vibratl:nal temoera'tures of : to 6,000 K at times up to 20 ms 4ownstream from
the aczJ.ve-nitrcgen ischarge. The diagnostic is c licated -"the resence
of free electrons :reated in the Penning-ionization. These free electrons
absort energy from the active-nitroGen medium and pr6duce further excitation
of Ni"(3). Adding traces of SF6 removes the free electrons from the reactor,
and thereby, eliminates their interfering effects. Observing emission frou
N 2 (), from Penning-ionization reactions in the afterglow, demonstrates the
presence of N2!v) containing at least 3.8 eV internal energy.
2.3.2.5 Obser-:ations of nitrogen first-oositive e=ission--
Another ossible diagnostic for the presence of N2(v) with relatively larze
amounts of internal energy is to monitor the emission of the nitrogen first-
positive bands, N2(3g - A37 +). These bands emit strongly in the afterglow
many tens of milliseconds downstream from the discharge (Figure 90). Some of
the observed emission, of course, is 4ust normal Lewis-Rayleigh afterglow
which is excited from the recombination of atomic nitrogen. Such emission,
however, has a very characteristic vibrational-level distribution, peaking
strongly in vibrational levels 10-12 at low pressures in nitrogen or helium
* bath :as (Fiure P). :n addition, for the N-atom number densities present -n
the reactor ani the total ressures in the system, such emission is relatively
weak. The pr'nar-' emission ;bserved has a vibrational distrIbution which
pea-cs at low .'brational loyels, is relatively strong at low pressures, and
quenczhes readily upon the addition of molecular species such as 02, CO, and
CO2 . The quenching is much too efficient to be either electronic quenching of
the first-positive emission itself, or to result from reactions of the mole-
cules themselves with atomic nitrogen. Thus, a long-lived metastable state of
nitrogen travels Jown tne flew reactor many milliseconds, and then is coupled
collis4onally into the E-:2 state in the field of view. :n this way, the
nitr ,,en first-:osit).ve e::isS.on can act as a tracer of tnis metastazle
species.
•14 2
- | ~
* IJ
71 :1 ), from aci
91-' -
510 60 0 66 640 660 720
* nAVELENGTH (nm;
iure 9. -i rs t-pos t ve spectrmrsltn o N~( -a oin) fraomb a ;r.* ~~~~~~~n~oe .ihlu p=15trX .1
'V -L k n '-,L
..e .... .. three .nown e.lectr n i ca l v excite d nitrogen me as.:azes .:,
suffci-entlv long lifetimes. to travel tens of milliseconds down the flow reac-
t'r -are -'" , -.na1_u, vI = I), and N (3u, v' = 0). Snergy-poolin
reizt-,.:ns of -; LA; are, indeed, a solirce of N'AB) excitat n. "4easurements -n
3... 2, hweer, hae lemonstrated that energy ooli ;--. a-
*. number ensities of 109 to I10 molecules cm- 3 produce Nn(3) number zens::ies
o on-. 1,
to i 4 molecules cm 3 . The active-nitrogen source of N"'B). , ,on
-Q the otner hand, zenerates number densities three to four orders of maan::uae
greater than this. Molecular hydrogen readily quenches -(a') (Ref. 14).
Addfng small traces of molecular hydrogen to the active nitrogen afterlow,
, showever, has little effect the first-Positive intensities at number zens;ties
a.,, sufficient to remove all the the N-(a'). The N-2('4, v'=0) state lies 73.3 cm-
abve N-,3, v' = 3) (Ref. 179). Several ,orkerS have shown that the 5- and
W-sa-, _- ar o,-'-d: uite efficiently (Ref. 132,133,14 . Collision s ;i:n
the heli'am bath 3as would couple the N2 (W, v' = 3) to N2 (3, v' = wnere it
would lecay radiatively on the time scale of tens of microseconds. it is con-
cluded, therefore, that high vibrational levels of the ground-electronic state
of nitrogen couple with N2 (3) in the reactor.
?lacing the Ni zcreen in the reactor does attenuate the -(B) ntenst-
ties, not oes not change the observed vibrational istribution (Figure 322.
By rotating the screen either normal to or parallel to the gas flow, we Gouli
vary t:e first-cositive intensity by up to a factor of 30. S;m.'ar chan:es inal,. -
screen ositzcn re.uced effective nitrogen vibrational temperatures by only
only 5 to 10 pertent. The N2(S) number density does vary linear> itn the
number density of N2 (X, v ) 5), as determined from vibrational temperature
measurements (F:;. 93), but placing a Ni screen in the flow reduces tie NtB)
number density drastically but has only a small effect on the N2 EX, v ) 5)
number density. Thus, the correlation is only ;ualitative, and no quantita-
tire relationships result from it. The measurements should be repeated using
2"(:7 as the - v) diagnostic.
144
P[ .
92d. -iA
217.2 iA
-.
,-p-
"2. )A-
Figure 92. Elbrational distribution of N37;)in active nitro,3en-or various ittenuatiors '-y a3 Ni screen down~stream from
the 3ischar~e, but upstrea~a from the observ;ation rezir).
For a num-ber of the measurements to be discussed in the next paragraph, a
pnot: meter centered at 580 rim was used to measure relative N-(3) number .iensi-
ties. Care must be taken in using this diagnostic that the relative vibra-
tioral listributi;on of the N-i(B) remains constant. If the vibrational distri-
butior, does not chang~e, Fig. 4 shows that the tphotoraeter does give an azu-
rate Alet'ermiatIs~ of the N-'3) number density.
% %I., -
ettIC4 .- 0 : .5 n
% S
141
SCPEEN CLOSED-
fi. X, V,, -~j (1014 molecules c~n-(
Figure 93. Variation in N, (B) n'Qmber density withi number iensi,;, or,vibrationallv-excitel, ground-electronic-sta ta ni4tro:en-.
-- ,A
w
%
.3 .he -v az.ve nitrogen
3.3.3. Scectrcscopv of active N- plus :F--Figure 95 snows t.-.e
spectr- m of active itro.en between 430 and 780 nm in the absence and presence
of ailed :?. To zharacter:ze the spectrum better, we normalized the tvo spec-
tra to the 2,0 through 4,2 first-positive bands between 740 and 7"3, nm, and
then suotracted the active nitrogen spectrum from that with active nitrocen
plus IF. Figure 96 shows the result. While some of the features in t-e scec-
trum are the 7F(3.0_--X 17?), a number of the spectral features belong t)
bands have not been identified as being known bands of IF or similar Mole-
cules. :-wo such systems have been identified which are labeled 1 and 2 in
,OP Fig. 96. Figures 97 and 98 show Plots of the transition energies for systems
1 and 2, respectively, plotted against the number of the peak as it accears on
the spectrum. Assuming that the two observed sequences result either fron
transitions from a coamon upper-state vibrational level to a number 3f
INI _Iifferenz lower-state vibrational levels, or else transitions from a nucmer of
different uoper-state vibrational levels to a common lower-state vibratioral
level, the slopes of these plots give e for the state of varying quantu.n
. number. The resulting sloes ive Values of 667 3 and 587 t 13 cm-' ocr
systems I and 2, respectively. Table 16 shows the values for the now~n
Seotr-niz states of -F 3ni similar molecules which might possibly be in tne
active nitrogen. flow. _ystem 2 is not too greatly different from the e
'va.ues for the round-electronic states of IF and IN. In that case, .-oweer,
the observed system would be a new one since the observed bands do ot l-e uc
with any of the known transitions in 7F and IN which terminate on the cr.,und-
g electronic state. -he *0e value for system 1 corresponds more closely i-_.i
that for the ground-electronic state of 10, but here again, the observed scec-
ctr m does not match with known transitions of 10. it was concluded that two
new band systems of IF or a similar molecule had been discovered. We hope toobtain better scectra in the future to aid spectral identification.
.ome of the observed s.)ectral features are from the well Known !F(_33- --
XZh sIts . -eca":s the .entit/ of the new syste:as is unknown, t.e
1 47
e1
20-
- IF ONS + IF OFF
16 1--,
N14-
'-' 12r- I
8 2i
024 I
t 460 520 560 600 640 680 76- 760
.-WAVELENGTH (nm)
Figure 95. Scectrum of active nitrogen between 4B0 and 30 n= in
the absence and presence of iF.
14 _ 2 2
212 -4
-+
2- 2
2-
4U3., 4 0 520 E+60 6co 640 6 0 720 760
[;.'.' :WA VE LENGT H (nm )
g- r• : ':e 96. Se ctr, m between 430 and 680 nm 4rom active ni~r,,en pl.is I
wp t. f7l se ta- i i d; 2 spcrlf-ue snrce :%
1431
[-w,2,
,,t ,:,q v
21 1
201
,' . 19 -- e 66 3" - 1--
! -
17'- -
1.41
0 2 4 6 8
PEAK NUMBER (ARBITRARY START)
P ~Figure 97. .p plot for IF system 1.
18
. 16
C-...-.I
L
14
5 7 9 12
PEAK IJbUER (A.'?BITRARY START)
Fi;ure 96. p. plot f)r IF svstem3 2.
"49
TXBLE 16. -3ectrosc:ci,: Para::eta3rs of :F an~d Similar 'oecu--s.
I olcIl State .e(m
IF D' 307D' 411 192ESA' -250A 381 15896
x 610 305
10 A 514.6 21815X 681 .5 341
704 13726X 603 331
System 1 667-3
'.8: . System 2 587-10
observed spectrum cannot fit by the normal least-squares approach. We .iave
determined the contributions to the spectrum from IF(3) by fitting indivi-dual
IF(3 ) casis functions by trial and error until an adequate agreement i4
obtained wit- the obvious !F(B) features. Figures 99 tnrough 1d1 illustrate
this .:rocedure over three different wavelength regions. The observed spectra
already have had the N2 first-oositive contributions subtracted out ,-f zne .
*. The spectral regi:rn between 453 and 5 a3 am contains mostly bards of i3) froze
vibrational levels v' = 3-3. The populations of these vibrational levels,
therefore, can be determined fairly cleanly in this spectral region. The con-
tributions of these levels to the spectrum at longer wavelengths then will be
reasonably well determined. Over the wavelength region between 500 and 600
nm, the major vibrational levels of IF(B) contributing to thie spectrum are
vI = 0-2. Fairly isolated bands at 533, 551, and 56a rim determine the v' = 1
population reasonably well. in the spectrum shown, the v' = 0 and 2 popula-tioros lvnbye it
tions can be "etermined ,rlv by fitting shoulders of bands. Certain cond:-
tions in the reactor favor the IF(B) over systems I afJ 2, so that so:-e r-ec-
tral fits give -;uite good populations for v' = 0. Figure 12 summarizes the
A5
--- -----------
rAD USTE) lF (3-X; -
z50-C
Wa.'. 2 aO- 4
-n0
C.,," 80,
20
IAI
• =- 0 1r'~#-
450 470 490
WAVELENGTH (nm)
Figure 99. Fit-ina of the IF(B 3 Thi--Xlr + ) bands between 450 3nd 5,- nto the soectruin excited by adding IF to active nitrogen.
0.9L OBSERVEDF(BX
I + ALUSTED IF (B-X)
z 0.76
C 6L
0-.
0 o
500 520 540 560 580 600
., WAVELENTH (nm)
cFiqur-e 13,0. F'ttinij of the :F(32--XIz ) bands between 5,)0 and n m nt' the srectrar: ex<it-d bv adding :F to ,c:ive r,:,.
151
O-
7 - OBSERVED I -"
+ ADJUSTED IF (B-X)S1.5 L
rnIr
.- 1 .0
FA 1
, 0 0
-- 450 500 550 600 65 0o ~ ~WAVELEN GTH(n,
Ef
SFigure 101. Ftigof the !F(B3.-,()+--Xlz+) bands between 450 anj 650 n
to the spectruIm excited by addini :F to dct,-ve nitrocen.
7. 7y E0 <
x
. C i5o0 zr.GG z..co z.¢ c.- 0 C "C -
-i -u e 1 . Vibrational listribution of !F(337n ) excited ty
152
---- -- - .. .. .. - - - W
popula:i.on distributions determined from fits to several spectra. The vlzra-
tional distribution fits lulte well to that given by an 1150 K Boltzmann
vibrational tempera:ure. The significance of a Boltzmann distribution in
terms of the dynamics of the energy transfer is unclear. in addition, such
districutions are somewhat decendent on the conditions of the measuremoent
since F(B) vibrational relaxation by He, Ar, and N2 is efficient (Ref3. 6,7).
Figure 103 shows the residual spectrum of the two new band systems when both
the active nitrogen and the IF(B) contributions have been subtracted out.
V Figures 104 and 105 show spectra between 450 and 650 nm taken under cordi-
tions to enhance system 1 and !F(B) features, respectively, relative to t.he
other two systems.
3.3.3.2 IF Excitation Rates in Active N2 -- To characterize t.-e
*energy transfer between active nitrogen and IF, some IF * excitation rates havB
been measured. The :F will be in steady state in the observation vol'ame so
tha: its formation and destruction rates can be equated. Nealecting 7:*
,, quenching gives
I = I * = k IF (iJ~7)-.IF* rad-I * ex 2 -
where 'ex is the excitation rate coefficient for 1F excitation by active NZ,
and krad is the iF* radiative-decay rate. Equation (107) saows that without
kncwing the identity of the precursor for tne IF excitation, the product of
the excitation-rate coefficient and the number density of the precursor
species can ce det3rmine .
* Figure 106 shows the variation in the intensity of the 5,0 band of !F(3)
as a function of number density of added IF for a number of different initial
N2,B) number densities in the reactor. The 12 (3) number density was varied by
changing the angle of rotation of a Ni screen in the flow, and was monitored
with the 580 nm photometer. Figure 107 shows that the excitation rates deter-
mined from the slopes of the lines in Fig. 106, corrected for absolute photon
eMisSion rates and for the fraction of total IF(B) emission appearing in the
5,j ,an., vary linearly with the measured N2(3) number densities. f ';*23
.*. were tne species responsiole for the IFtB) excitation, the slope in Fig. 1'7
. - 153
.e, t,., ,. . , .. , , , , ,,,. ": "' " "-''"' ":"-"'v:"' " + ' , ', ',¢
2.0 II I III~
1.5 -
0.5
-I,- _I
= 1.0. -
a-
',' 0.5 -
- o
-0.1,,,-0.2 ' I I I I I I I :
450 500 550 600 650
WAVELENGTH (nm)
-' ;.Figure 103. Residual spectrum of active nitrogen plus :F after spectrafeatures belonging to the nitrogen first-positive and
IF(Bm- -I ) systems have been subtracted out.
1.,& = 1.2
1 -4
U-
-0.2-0.4
.9 -0.6450 500 550 600 650
WAVELENGTH (nm)
-iure 104. -cectrim of active nitroqen plus IF with the nltro,een
t-0rst-tos tive features subtracted ouvt t3ken underconditions which enhance :F system 1 relative to
other IF spectral fiatures.
154
-r'
*22 * 3 - i, AC.2ST-J ;F (&-x) I -
=- ' "=--
%
:"WAVELENGTH (nm)
Figure 105. Soectrum of active nitrogjen plus IF with the nitro(jen firs-
oositive fea-tures subtr-acted out taken under cniin hcenhance t.e IF(3-X) syvstem relative to other lF spectra l
lea tures.
":'i I ,. _
" - 1 . 10 7
A 1. 07 i0
[IF:~~~.2 107m~ uls m
0.. 107
-0.5~.*' 500 55 60 6 4
F igr 105. Soec t n n ate ntent of ~ the n 5C t n as -fostior feade IFtace ut ten/ fnor -;niios izh-
enhance he F(3 in sythe reltiveatootr Fsec
i AN n 'I 1.37' Il
0":
00. o.o. 1. -
.1'B Y'= 2 12 1 7 o-ulsc
i. 1 7 -
-0.05
° ,.0 0.5 1.0 1 .5
t '- N2B, V' = 2-12) (10/ molecules cn 3)
.'-,Figure 137. xctation rates of IF(B) versus N-I( ) nutber 2e nsit'.
would 3ive the excit:aion rate coefficient. The valise derve, however, s*1 .(5 -..4) x _ c:3 3 molecu2e " s "1 more than an order :: a;fi
greater tan gas kinetic, meaning that N-2(3) cannot be the e.cicatir. .ar::er.
The data show o.ite clearly that the excitation rates do var/ with i
n-mber density. A previous study indicated this was probably tr-e Ref. 6),
but scatter in that data prevented unequivocal confirmation of this correla-
tion. Figure 13 shows that the intensity of the series 2 zand at -526 n
varies linearly with added IF, and Fig. 109 shows that the ;eries I and 2 band
intensities also vary linearly with the N2(3) number density in th f eac
tor ihen the IF number density is held constant. We wish to stress that the''.5' N .- ; -oes no t icc ea r to be the a c tua l p re cu rs o r its e "f, ra t .-.e r th e :e .. i a;le
whih eA:ites the IF emissions apparently also is co'pled co xsionallv to
t'e "'2 ) state.
*:.
30
26 -
22 -7
S14-
10-
6
b t-: 2 -15--a
0 2 4 6 8
[CF3I]added (1011 molecules cm- 3)
Fiaure 138. -ntensities of the IF series 2 band at 636 nm as a functioor" of added 1F rumber density.
.iC
"" & Z - SYSTEZC.100
Ag
6'6 -:.-%P.E0F SYSTE.' 2-
11
* #4 5 .0SNZB ' 2-'2, 'C 6
'ncle. es cm* 31... A
,%
Figure 1)9. t tensiit of IF series 1 and series 2 emissions as af,:nct;rn of "-( 3) number -ena;t-i for constant IF
w- r~.zz=er .. , iz
rui-rtejt, 15.,
:n tne previzus study, the excitation rates of IF(S) were chosen to he
either too larze to be caused by the known long-lived metastables in the reic-
tor, or else they varied in a way which was incomoatible with exoected .arla-
tions in the number densities of other metastables in the reactor. 7ni soi..
still remains to 'e proven !inequivocally. In particular, si-u:neo *eta'-
mination of 7F" excitation rates and number densities of the atomic nitrogen
metastables, N(2D, :?) remains to be done. In addition, the N-,AS<-u) number
density in the active nitrogen has not been pinned down, althcugh the observed
V IFIB) vibratioral distribution is incompatible with excitation primar iv bW
N2 (A), and the previous studies demonstrated clearly that systems 1 and 2 are
not oroducts of thne interaction of N2(A) with IF (Ref. 7). The :FS) excita-
tion rates measured in this study are somewhat smaller than those dearninec
previously, but transit tines between the discharge and the observation re~ion
were somewhat shorter, so presumably the number density of the active species
has been quenched somewhat more extensively in the present work.
. .'.
.9.
I9.
h.'."
REFERENCES
. Cheah, C.T. Clyne, M.A.A., and Whitefield, P.D., "Reactions Forming[Elctronic~a--Excited :ree Radicals, I. Ground State Reac:t*ns ianvcilngNFI and NF ?ad 4a1s, C. S. Faraday :: 76, 711 (1960).
2. Cheah, C.T. and Clyne, M.A.A., "Reactions Forming Electroncally-Exc-4:edFree Radicals. II. Formation of N(4S), N(2D), and N( 2 P) Atoms in the
% H-N'F Reaction, and N Atom Reactions,' J.C.S. Faraday II 76, 1543
3. Cheah, C.T. and Clyne, >.A.A., "Reactions Forming Electronicalli-ExcitedFree Radicals. :I. Formation of Excited *Molecular States in the u_-:.TReaction," . Photo Chem. 15, 21 (1981).
4. Piper, L.G. and Taylor, R.L., "Investigation of Induced UnimolecularDecomposition for Development of Visible Chemical Lasers," PS: :--6,prepared for the Department of Energy under Contract 4EY-76-C-02-2960,
r *n 00 (1977).
i. . 5. Piper, L.G., Krech, R.H. and raylor, R.L., "Generation of N3 in the
%-. 7Thermal Decomcosition of NaN 3 ," 2. Chem. Phys. 71, 2099 (1979).
6. Piper, L.G., Marinelli, W.J., Green, B.D., Rawlins, W.T., Murphy, H.C.,Donahue, M.E., and Lewis, P.F., "Kinetics of Iodine %lonofluorideExcitation by Energetic Nitrogen," Final Report prepared for Air ForceContract No. F9631-3-2-3051, Physical Sciences Inc. Technical Report-
"-* No. TR-48C (: -1 )
7. Picer, L.G., arineili, W.J., Rawlins, W.T., and Green, 3.C., "TheExciSation o- '.O bY "2 (ASu), ". Chem. Phys. 853, 5602 0385 .
3. Cldne, M.A.A. and White, :.F. , Chem. Phys. Lett. 6, 465 (1970).
. ?erbelin, J.M. and Cohen, N., "The Chemical Production of Electroni:a:1yExcited States in tne Hi{NF2 System," Chem. Phys. Lett. 20, 605 (97':.
13. Ciyne, M.A.A. and Stedman, D.H., "Reactions of Atomic Hydrocen withHydrogen Chloride and Nitrosyl Chloride," Trans. Faraday Soc. 62, 2164
% (r 13(1 66).
11. Clyne, M.A.A. and Thrush, B.A., "4echanism of Chemiluminescent Reactionsi involving Nitric Oxide -!he H+NO Reaction," Discuss. Faraday Soc. 33,
,39 (1 2 .
12. Din , A., Karliw, j., and Weiss, J., "Production of H-Atom and 0-AtomSeas by Microwa-.3 Discharge Source," Rev. Sci. instrim. 48,102 1977,.
• 159
13. :wck, "I.A., Herbellin, J.M., and Cohen, N., "Coilisional uench_:.. 3. oRadiation :ecay Studies of NF(aI ) and NF(b )," Electronic ransitlor.
Lasers, MIT Press (1976).
14. ?¢ntin, A., Meyer, C.3., and Schiff, H.I., "Absolute ;:antu= "iel
*"!easurezents of the NO-0 Reaction and Its Use a Stanfarf forChemiluminescent Reactions," J. Chem. Phys. 40, 64 (1364).
15. Vanpee, M., Hill, K.D., and Kineyko, W.R., "Absolute Rate ConstantMeasurements for the Radiative Combination of Atomic Oxygen with NitricOxiie, " AAA J. 9, 135 (1971).
16. Golde, M.F., Roche, A.E., and Kaufman, F., "Absolute Rate Constant for
the O-NO Chemiluminescence in the Near Infrared," j. Chem. Phys. 59,
3953 (1973).
17. Golomb, D. and Brown, J.H., "The Temperature Dependence of tne t 'n-e
" Chemiluminous Recombination. The .RIC Mechanism, " J. Chem. 3hys. ;-5246 (1975).
13. Woolsey, G.A., Lee, P.H., and Slafer, W.D., "Measurement of :he RateConstant for NO-C Chemiluminescence Using a Calibrated Piston Soorze
of Light," J. Chem. Phys. 67, 1220 (1977).
19. Suzoh, M., Morioka, Y., and Nakamura, M., "Absolute Rate Constant for theChemiluminescent Reaction of Ato.aic Cxygen with Nitric Oxide, " 3. Cheu.
Phys. 72, 20 (1980).
23. Pravilov, a.,. and Smirnova, L.G., "Spectral Distribution of tneChemiluminescence Rate Constant in the O+CO(+M) and CC'i(e) Reaczions,"
Kinet. anra atal. 19, 202 (1978).
, 21. "auf-an, F . The Air Afterglow and its Use in the Stud.' of Soe
.eactions of Atomic Oxygen," Proc. Roy. Soc. (Lcrdon) A 247, 123 958j.
22. Kaufman, F., "The Air Afterglow Revisited," Chemiluminescence and
Biluminescence, M.J. Cormier, D.M. Hercules, and J. Lee, eds., 7p. 83-100S (1973).
23. Husain, 0. and Slater, N.K.H., "Kinetic Study of Ground State AtomicNitrogen, N(24S 3 /2 ) by Time-Resolved Atomic Resonance Fluorescence,"
J.C.S. Faraday II 76, 606 (1980).
24. Lee, J.H., 4ichael, J.V., Payne, W.A., and Stief, L.J., "Absolute Rate oftne Reaction of N(4S) with NO from 196-400 K with DF-RF and FP-:FTechniques," j. Chem. Phys. 69, 3069 (1978).
V, 25. Clyne, M.A.A. and McCermid, :.S., "Mass Spectrometric Ceterminatiorns oftne Rates :f Elementary Reactions of NO and N0 2 with Grouna State *4S
Atoms," j.:2.S. Faraday I, 71, 2189 (1975).
160
Pu
26. Slanger, T.G., "Genera:ion of 2 u, Au 3 +) from Oxygen A:cM
Recombination," J. Chem. Phys. 6), 4779 (1978).
27. L.G. Piper, unpublished results (1978).
28. Baulch, D.L., Drysdale, D.C., Horne, D.G., and Lloyd, A.C., Evaluated
i Kinetic Data for Hizn Temperature Reactions. !I. Homoceneous Gas 7taseReactions of the -N-O Systems. (Butter4orths, London, 1973).
29. Piper, L.G., Murphy, H.C., and Rawlins, W.T., "Development of Cochise UV
Absorption System," A.FGL-TR-81-0318 (1961).
30. Mitchell, A.C.G. and Zemansky, M.W., "Resonance Radiation and Excited
Atoms," Cambridge University Press, London U.K. (1971).
31. Rawlins, W.T. and Kaufman, F., "Characteristics of 0(I) and N(I)
Resonance Line Broadening in Low Pressure Helium Discharge Lamps," -
2uant. Spectrosc. Radiat. Transfer 1S, 561 (1977).
32. Kaufman, F., and Parkes, D.A., "Sources of Error in Using Resonance Light
Absorption to Measure Atomic Concentrations," Trans. Faraday Soc. 66,1379 (1970).
33. Tellinghuisen, J. and Clyne, M.A.A., "Role of Hyperfine Structure In
Atomic AbsorpDtion: Oscillator Strengths in Br and I," J.C.S. Faraday Ii
'2, 783 (1976).
34. Clyne, :.A.A. and Piper, L.G., "Kinetic Spectroscopy in the Far VacuiimVntraviolet, ?arta 3. Oscillator Strengths for the 3S, 4S, and 53S-Z
.2 transitions in Atomic Oxygen," J.C.S. Faraday II 72, 2178 (1976).
35. a. Clyne, M.A.A. and ".cnkhouse, P.B., "Energy Transfer in Collisions ofAr( 3P0 ,2 ) 'eastable Atoms with H(
2 S) Atoms. I. Total Rate Constant
and Mechanism," Chem. Phys. 28, 447 (1978).
b. Clyne, M.A.A., Heaven, %I.C., Bayes, K.D., and Monkhouse, P.B., "Energy
Transfer in Collisions of Excited Ar(3 PO,2) Metastable Atoms winh
H(2 S) Atoms. I. Lyman-j Emission Profile," Chem. Phys. 47, 173
(1980).
c. Piper, L.. and Clyne, M.A.A., "Determination of the Translational
Energy of 3(3s3S) Excited by the Dissociative Excitation of 02 and NO
by He*(2 3 S)," Chem. Phys. 63, 77 (1981).
36. Donahue, T.M., Anderson, J.G., Rawlins, W.T., Kaufman, F., and Hudson,R.D., "Apollo-Suyuz O(3P) and N(4 S) Density Measurement by UVSpectroscofy," Geophys. Res. Lett. 4, 79 (1977).
37. Malins, R.J. and Setser, D.W., "Rate Constants, Branching Ratios, andEner-y Disposa for NFkb,a,X) and HF(v) Formation for tne H+N' 2
Reaction," J. P-ys. Chem. 85, 1342 (1961).
161
I.%
.
38. Lof:.'us, A., and Krupenie, P.H., "The Spectr'im of zolecular Ni-rcen,';. Phys. Chem. Ref. Dat:a 6, 287 (1977).
39. Piper, L.G., Donahue, M.E., and Rawlins, W.T., "Rate Coefficients forN(ZD) Reactions," J. Phys. Chem., submitted (1386).
40. Koffend, J.3., Gardner, C.S., and Hidner, R.F., "Kinet'cs of theSystem," Aerospace Report SD-TR-85-55 (1985).
41. Burrows, M.D., Baughcum, $.L., and Oldenborg, R.C., "Cotically pumpedNO(A 2_ X2 -) Ultraviolet Laser," Appl. Phys. Lett. 46, 22 (1985).
42. Hill, R.M., Gutcheck, R.A., Huestis, D.L., Mukherjee, D. and Lorents,D.C., "Studies of E-beam Pumped Molecular Lasers," SRI Report No. %IP74-39under ARPA Contract No. N00014-72-C-0478 (1974).
43. Clark, W.G. and Setser, D.W., "Energy Transfer Reactions of N(A 3-u") V:ruenching by Hydrogen Halides, Methyl Halides, and Related *%1olecules,"J. Phys. Chem. 84, 2225 (190).0
44. Young, R.A. and St. John, G.A., "Experiments on N2 (A3T). II.Excitation of N4O," J. Chem. Phys. 48, 898 (1971).
45. Callear, A.B. and Wood, P.M., "Rates of Energy Transfer From 12 (A Lu) toVarious M.olecules. Initial and Final 2uantum States in the Transfer toN;O X27 and H2 (61S0 ) and Vibrational Relaxation of N2 (A3,) (V = 1) In
in Helium," Trans. Faraday Soc. 67, 272 (1971).
46. Mandel, A. and Ewing, J.j., "Quenching of N2 (A3 7u+) by :2 ,'" J Chem.
Phys. 67, 3490 (1977).
47. irayer, J..., Perner, D. and Roy, C.R., "Rate Constants for the 2uenchinof N,2(A3,u , vA = 0-8) Oy CO, C02 , ':i3, NO and 02," . Chem. Phys. 6',3164 (1974).
49. Rawlins, W.T., Piper, L.G., Caledonia, G.E. and Green, B.D., "COCHISEResearch," Physical Sciences Inc. Tecanical Report TR-298 (1981).
50. Piper, L.G., Caledonia, G.E. and Kennealy, J.P., "Rate Constants forDeactivation of N2 (A
3I u v1=0 ,1) by 02," J Chem. Phys. 74, 2888 (951).
51. Piper, L.G., Caledonia, G.E. and Kennaly, J.P., "Rate Constants for
Deactivation of N2 (A 3ru+ v'=0,1) by 0," J. Chem. Phys. 75, 2847 (1921).
52. Piper, L.G., "The Excitation of O(Is) in the Electronic Energy TransferBetween N-)(A 3 -u+) and 0," J. Chem. Phys. 77, 2373 (1982).
162
* uo Wv
53. Piper, L.G., Murpny, H.C. and Rawlins, W.T., "Development of CC.iSl JYAsorp-i-on Sae: ina Report," :R-292 (November 1981),
Air Force Geophysics Laboratory Report (1981).
54. Rawlins, W.7. and Piper, L.G., "Effect of Excitation Mechanism onLinewidth Parameters of Conventional VUV Discharge Line Sources, " Proc.Soc. Photo. int. instr. Eng. 279, 58 (1981). Also presented at therachnical svmposium East 'SI of the Society of Photo-OpticalInstrumentation Engineers, Washington, D.C. (April, 1981).
55. Krech, R.H., Diebold, G.J. and McFadden, D.L., "Kinetics of the O + F-
Reaction: A Case of Low Reactivity of Elemental Fluorine," J. Am. Chem.
Soc. 99, 4605 (1977).
56. Krech, R.H., "ESR Discharge - Flow Kinetics: The 0 + F 2 Reaction,"Masters dissertation, Dept. of Chem., Boston College, 1976.
57. Pak, S.J., Krech, R.H. , McFadden, D.L. and MacLean, D.I., "EPR Discharge-Flow Kinetics: The ? - CIF3 Reaction," J. Che.. Phys. 62, 3419 '1375).
58. Kaufman, M. and Kolb, C.E., '"Iolecular Beam Analysis of the Reaction
Bet-een Atomic Fluorine and Carbon Tetrachloride," NR 092-531, 6 (!971).
59. Berg, H.C. and Kleppner, D., "Storage Technique for Atomic Hydrogen,"Rev. Sci. instrum. 3 , 248 (1962).
6C. Lewis, P.F. and Green, B.D., "Computation of Electronic Spectra ofDiatomic Molecules," PSI TR-413 (1984).
61. Setser, D.W., Stedman, D.-i., and Coxon, J.A., "Chemical Applications ofMetastable Argr.n Atoms. IV. Excitation and Relaxation of Tri,)Iet Statesof N2 ," J. Chem. Phys. 53, 1004 (1970).
62. Stedman, D.H. and Setser, O.W. , "Chemical Applications of MetastableArgon Ato-,s :. A Clean System for the Formation of 142 (A - )," Chen.Phys. Lett. 2, 542 168).
63. Sadeghi, N. and Setser, D.W. , "Primary N2 (B) Vibrational Distributionsfrom Excitatior,-Transfer Reactions Between Kr( 3 P 2 ) or Xe(3 2) Ato.is andN2 ," Chem. Ptys. Lett. 82, 44 (1981).
64. Thomas, j.!. and Kaufman, F., "Production and Kinetics of N2 (A 3 -U
v43)," Photochemistry Symposium, Harvard University, August (1984)1.
65. Golde, M.F., "Vacuum JV Emission by Electronically Excited N2 : The
Collisional-Induced N2 (a 1 1g v=) - N2 (a' 1-u-, v'=0) Transition," Chem.Phys. Lett. 31, 348 (1975).
66. Thomas, J.M., Jeffries, J.B. and Kaufman, F., "Vibrational Relaxati:r. ofN-uA 3-, v'=1,2,3) -y CH4 and CK.4," Chem. Phys. Lett. 102, 50 193j.
163
67. Rozinson, :. and Nicholls, R.W., "intensity Measureent:3 on the3econd Negative, CO Angszrom and Third Positive an NO -, and 2 >olcular3and Systems, " Proceedings of the Physical Society :, 957 195).
68. Keck, 3.0., Allen, R.A. and Taylor, R.L., "Electronic Transition "1.entsfor Air 21-oIes, " i. uant. Soectrosc. Radiat. Transfer 3, -'376:
69. Callear, A.3., Pilling, M.J. and Smith, I.W.M. , Trans. Faraday Soc. i'Z,
2997 (1966).
70. Buzert, H., "Pooulation and Predissociation of Vibronic States of N-tricCxide," J. Ch.em. Phys. 56, 1113 (1972).
71. Srzozowski, J., Erman, P. and Lyyra, M., "Predissociation Rates andPerturbaticns of the A, S, B', C, 0 and F States in NO Studied UJsin7 Time
-esolved S ectroscopy," Physica Scripta 14, 290 (19765).
72. Kuz'menko, :'E., Kuznetsova, L.A., Monyakin, A.P. and Kuzyakov, Yu. Ya.,
"?robabilities for Electronic Transitions of Molecular Svstams ot
ian-Temoerature Air Components--II. The y and 3 Systems of NO and the-4+) System of C0," J. 2uant. Spectrosc. Radiat. Transfer 24, 219
73. Poland, H.M4. and Broida, H.P., "Fluorescence of the -e, and S Systems of
Nitric Oxide; Polarization and Use of Calculated Intensities forSoectrometer Calibration," J. 2uant. Spectrosc. Radiat. Transfer 1,
w 1886, (1971).
74. Jeunenomme, A., "Transition-iloment Variation in the f System of Nc",J. Chem. Phys. 45, 4433 (1966).
75. lumma, '4.j., "Molecular Branching-Ratio Method for Intensity Calibrtiolof Octical Z'stems in the Vacuum Ultraviolet, " J. Opt. Soc. 2rner.__,1453 (1972).
76. 'Moal-ann, 5.R. , Van Sprang, H.A., Bloemen, E. and De 'Heer, F.J.,"Experimental Determination of the Electronic Transition .oment for t.he
* NO(A 2:- - X 2) System. Lifetimes of the NO(A 2,, v'=0,1,2) Levels,"
Chem. Phys. 32, 239 (1978).
77. McGee, T.J., Miller, G.E., Burris, J., Jr. and Mcliratn, T.J.,"Fluorescence Branching Ratios from the A 2 Z (v'=0) State of NO," J.uant. Spectrosc. Radiat. Transfer 29, 333 (1983).
78. Navati, B.S. and Korwar, V.M., "Electronic Transition M.oment Variationfor the A X - x 21 Transition of the NO Molecule," Physica 124C, 421( 1984).
164
.- ,
79. W:t, ., -ye, J. and :lnelm, N. and "Pocke :-bcrne "-easuremen-z
Sca :ted 3 nih- in :ne Mesopnere," J. Atmos. Terr. _.s. -'2,(1 76).
80. Slanger, 7.3. , Bischel, W.K., Dyer, M.J., "Nascent NO VibrationalDistribution from 2485 & NO2 Photodissociation," J. Chem. Phys 79, 2231
(1983).
91. Her, R. R. , Sul.ivan, .. and Whitson, M.E., "Nitric Oxide Vibra t onalExcitation from the N(43) + 02 Reaction," J. Chem. Phys. 79, 222:
(1983).
* 82. Nicholls, R.W. , "Franck-Condon Factors to Hign Vibrational 2uant'
Numoers IV: NO Band Systems," J. Res. NBS 68A, 535 (1964).
83. Albritton, D.L., Schmeltekopf, A.L. and Zare, R.N., "Diatomic Intensity
Factors," private communication via D.W. Setser (1985).
34. Zacharias, H., Halpern, J.B. and Welge, K.H., "Two-Photon Excitation ofNO(A 2Z+; v'=0,1,2,) and Radiation Lifetime and Quenching Measurements,"
Chem. Phys. Let. 43, 41 (1976).
85. Smith, A.J. and Read, F.H., "Measured Lifetimes of the A 2, _= nd
C 2" States of NO," J. Phys. B: Atom. Molec. Phys. 11, 3263 (1978).
86. McDermid, 1.S. and Laudenslager, J.B., "Radiative Lifetimes and
A Electronic uenching Rate Constants for Single-Photon Excited RotationalLevels of NO (A 2,+, v1=0)," J. Quant. Spectrosc. Radiat. Transfer 27,483 (1982).
37. Crosley, D.R., Zare, R.N., J. Chem. Phys. 49, 4231 (1968).
! 3. Brozozowski, J., Eiander, N. and Erman, P., Phys. Scr. 9, 99 (1974,.
39. Benoist D'Azy, 0., Lopez-Delgado, R. and Tramer, A., Chem. Phys. 9, 327(1373).
9K. -;esser, J., :. Chem. Phys. 48, 2518 (1968).
91. Bubert, H. and Froben, F.W., Chem. Phys. Lett. 8, 242 (1971).
92. German, K.R., Zare. R.N. and Crosley, D.R., J. Chem. Phys. 54, 4039
93. Weinstock. E.M. and Zare, R.N., "Lifetime Determination of the NO A
State," J. Chem. Phys. 56, 3456-3462 (1972).
94. Hasson, V., Farmer, A.J.D., Nicholls, R.W. and Anketell, J., "Applicationof Disperse-on Techniques to Molecular Band Intensity Measurements :1.Oscillator Strength of the (D,0) Band of .C-y1, 21 _ X 2-) S ,stPhys. B: Atom. Molec. Phys. 5, 1248 (1972).
165
95. Farmer, H asson, V. and Nicholls,.W. , "Absol::e 2 at:r
Strength .!ea Sre-en t. fhe (v";f v' =-3) 2ends of :'-e 2: -
,.-System of Ni-ric Oxide," J. ,uant. Spectrosc. Radiat. :ransfer ,2, 627t1972).
96. Pery-Thorne, A. and Banfield, F.P., J. Phys. B: Atom. '4olec. Phys. 3,1311 (1970).
97. Weter, Z. and Penner S.S., "Absolute Intensities .:r .e Ultrav' -1• -Bands of NO", J. Chem. Phys. 26 560 (1957).
98. Bethke, G.W. "Oscillator Strengths in the Far Ultraviolet. I. Nitr:.cOxide", J. Chem. Phys. 31, 662 (1959).
99. Ferguson, E.E., Fehsenfeld, F.C., and Schmeltekopf, A.L., "FlowingAfterglow Measurements of :on-Neutral Reactions," Advances in Atomic
and Molecular Physics V, edited by D. R. Bates, New York, Academic Press,(1970).
100. Bolden, R.C., Hemsworth, R S , Shaw, M.J., and Twiddy, . .
"Measurements of Thermal-Energy :on-Neutral Reaction Rate Coefficiets
for Rare-Gas Ions," j. Pnys. B 3, 45 (1970).
101. Farragher, A.L., "Ion-Molecule Reaction Rate Studies in a Flowing After-
glow System," Trans. Faraday Soc. 66, 1411 (1970).
102. Huggins, R.W. and Cahn, J.H., ",letastable Measurements in Flowing HeliumAfterglow," j. Appl. Phys. 38, 180 (1967).
103. Walker, R.E., "Chemical Reaction and Diffusion in a Catalytic Tubular
Reactor," Phys. Fluids 4, 1211 (1961).
104. Pcirier, R.V. and Carr, R.W., "The Use of Tubular Flow Reactors forKinetic Studies Over Extended Pressure Ranges," J. Phys. Chem. 75, 15.3(1971).
105. Cher, M. and Hoilingsworth, C.S., "Chemiluminescent Reactions of ExcitedHelium with Nitrogen and Oxygen," Adv. Chem. Ser. 30, 115 (1969).
106. Kolts, J.H. and Setser, D.W., "Decay Rates of Ar(4s, 3 P2 ), Ar(4s, 3P3)Kr<-s, 3 P2 ), and Xe(6s, 22) Atoms in Argon," J. Chem. Phys. 68, 4643
(1978).
107. Shibuya, K., Imajo, T., Obi, K and Tanka, I., "Formation and quenching of3 + ,22,Metastable N2A 3u in the Electronic Relaxation of 'dO C2. and D2 in an
NO/N 2 Mixture, J. Phys. Chem. 88, 1457 (1984).
• 108. Shemansky, D.E., "N-, Vegari-Kaplan System in Absorption," j. Chem. Phys.~51, 689 (1)63).
109. Shemansky, D.E., and Carleton, N.P., "Lifetime of the N2 Vegard-KaplanSystem", j. Chem. Phys. 51, 682 (1969).
166
:a -. rltn, N. ar~ *..._en e rg, . '@L i fe time o f the o.w eS t .x c-, -eve.
1111. Eyir-er, a nd Pipkin, F-..,"LftmMesrensothB ae
ofN,' Us.n Laer Excitati," 46 (1he7. s79364(93)
112. Shezansk7, D.E. and Sroadfoo~t, A.L., "Excitation Of N2 and 12 _ -Systems '_Y
'aPia. Trn'sfer 2,138; (1971).
113. Werner, N.J., Kalcher, 3. and Reinsch, E.A., "Accurate ab initio Calcula-tions of Radiative Transition Probabilit~es Between the A3 $ - 3-
W s'uj ; and 0 3..u States Of N12," 3. Chem. Phys. 81, 2420 (1964).
114.YeaerD.L an McoyV.,"Transition Moments Between ExcitedElectronic States of N2*,?s Chem. Phys. 67, 2473 (1977).
115 . Weiner, 3. and Ohm, n, "A Note on the Radiative Lifetimes of - e 33._-9State of N,,' J. Chem. Phys. 80, 5866 (1984).
116. Jeunehomme, M., "Transition Moment of the First Positive Band Systa.3 ofNitrogen," J. Chem. Phys. 45, 1805 (1966).
117. Carlson, T.A., Ourio. N., Erman, P. and Larsson, 74., "Collisional
Transfer to the 3 State in N2 ," Physica Scripta 19, 25 (1979).
118. Heidner, R.F., III, Sutton, D.G., and Suchard, S.N., "Kinetic Study ofN2 (B
3 11r v) 2aenching by Laser-induced Fluorescence," Chem. Phys. Lett.37, 249'(1976).
19. 3r.)a 4f :ot, ... and X~aran, S.P. , "Electronic Transition 'Momentr theaVegard-Kaplan Bands," j. Cnem. Phys. 51, 678 (1969).
120. Zbata, T., Anezaki, Y., Fujii, M., Mikami, N., and Ito, '4.,"Rotational Energy Transfer in NO0(A 2 Z+, v=0 and 1) Studied by No-ColorDouble-Resonance Spectroscopy," Chem. Phys. 84, 151 (1984).
121. :DeCe ras-inska, I., Beswick, J.A. and Trainer, A., "1A Distorted WaveCalc'iiation for Electrorlic Enerzy Transfer in Molecular Collisions.
*Application to N2 (A 3 Zu-) + C3(X Z ) + IN2 (X 1: +) -,CO(a 3 system,"
j. Chemn. Phys. 71, 2477 (1979).
122. Golde, M.E'. and Moyl.e, A.M., "Study of the Products of the Reactions of!i2(A3Zu*) The Effect of Vibrational Energy in N2 (A),"1 Chem. Phys. Lett.117, 375 (1965).
123. Stedman, D.H. and Setser, D.W., "Energy Pooling by Triplet Nitrogen(A3, ) molecules," J. Chem. Phys. 50, 2256 (1969).
124. Hays, 3.1., Tracy, C.J., Demonchy, A.R., and Oskam, H.J., "Production~of .4-(C3-,1 ~('-u) by Mutual Collisions ofN( 3 u)Mtastable
%Iolecules,' Chemn. Phys. lett. 14, 352 (1972).
167
0
125. :ays, 3.N. and 2skam, -.J., "Population of N2 ',3 by " (A
:uriag the Nitrogen Afterglow," :. Chem. Phys. 59, j507 1972).
126. Hays, G.N. and Oska.'m, H.J., "Reaction Rate Constant fcr 2N 2 (A 3 ,)':( 3 - u + 1;, ( 1 - + , "N;C.U) 1 , v > 0)," J. Chem. Phys. 59, 6068 (1973;.
127. atler, i., Setzer, D.W. , and Rosenwaks, S., "Producti.n. of -..e : -e- an
:nfrared Syszem by the Energy Pooling Reaction of -,I -) Metastacle
Nitrogen Molecules," Chem. Phys. Lett. 72, 536 (1980).
% 123. Nadler, I., Rotem, A., and Rosenwaks, S., "Observation of New Sands lnthe N 2 Herman Infrared System and Kinetic Study of its Formation in aPulsed-Discharge Apparatus," Chem. Phys. 69, 375 (1982).
129. Nadier, I. and Rosenwaks, S., "Studies of Energy Transfer Processes inA) Triplet States of N2). I. Energy Pooling by Vibrationally selected
, _ + v) Vlolecules," j. Chem. Phys. 82, 3932 (1965).
* 130. Piper, L.G., unpublished results (1980).
131. Sadeghi, N. and Setser, D.W., "Collisional Coupling oa
_u States Studied by Laser-Induced Fluorescence," Chem. Phys. Lett.
77, 304 (1981).
132. Sadeghi, N. and Setser, O.W., "Collisional Coupling and .elaxation ofN2(3Bg) and ' 2 (;_u) Vibrational Levels in Ar and Ne," J. Chem. Pkiys.
79, 2710 (1983).
, 133. Rctem, A. and Rosenwaks, S., "Laser-induced Fluorescence Studies of-olecular Nitrogen," Optical Engineering, 22, 564 (1983).
134. ;erm.an, R. "'Nonvelle Transition Interdite de la Molecule 4,'-. Acad. Sci. 2aris) 233, 738 (1951).
135. Cirr~ll, ?.K. and Sayers, N.D., "The Band Spectrum of Nitrogen: NewStudies of tlhe Triplet Systems, " Proc. Phys. Soc. (.ondon) 64, 1133(1953).
136. ?liter, D., "Spectral Observations on a Radio-Frequency Excited13, itrogen Jet," Can. J. Phys. 41, 1245 (1963).
137. Mahon-Smith, D. and Carroll, P.K., "Isotope Shifts and the 'Vibrational
Structure in Some Weaker Systems of N2 ," J. Chem. Phys. 41, 1377 (1964).
138. a. Michaels, H.H., in The Excited State in Chemical Physics,
J.W. McGowan, ed. , New York: Wiley, Vol. II, Ch. 3 (1981).
b . Micnaels, H.H., private communication to S. Rosenwaks (1965). C-ited
in Ref. 129.
168
0I
epTo&, .*
%'5
139. Gilmore, F.R., private communization (1966).
142. Piper, L.G. and DeFaccio, M.A., "'uenching Rate Coefficients forN( (a ), J. Chem. Phys. submitted (1966).
141. Mitchell, K.B., "F.uorescence Efficiencies and Collis-onal Deact:vaticnRates for N2 and N2- 3ands Excited by Soft X Rays," J. Chem. Phys. 53,1795 (1970).
142. Shemansky, D.E. "A 3 Molecules in the N2 Afterglow," J Chem. Phys.
64, 565 (1976).
143. Gartner, E.M. and Thrush, B.A., "Infrared Emission by Active Nitrocen.I. The Kinetic Behaviour of N2 (3'3ru-)," Proc. R. Soc. Lond. A. 346,
103 (1975).
144. Gartner, E.M. and Thrush, B.A., "infrared Emission by Active Nitrogen.".. The Kinetic Behaviour of Ne(B'3g)," Proc. R. Soc. Lond. A. 346, 121(1975).
145. Becker, K.H., Fink, E.H., Groth, W., Jud, W., and Kley, D., "N2Formation in the Lewis-Rayleigh Afterglow," Discuss. Faraday Soc. 53, 35
:%" (1972).
146. Dreyer, J.W. and Perner, D., "The Deactivation of N2 v=-2 andN 2a' lZu-, v=0 by Nitrogen," Chem. Phys. Lett. 16, 169 (1972).
147. Rotem, A., Nadler, I., and Rosenwaks, S., "Direct Observation of
Collision induced Transitions from N2(3 3 1g) to N U2('3 u)," 3. Chem.Phys. 76, 2109 (1982).
145. Rotem, A., Nadler, I., and Rosenwaks, S., "Laser-Induced Fluorescence
Studies of Collisional Coupling of N2(B 3 .g) with N2(W 3,u) and
N2(A3--X " Chem. Phys. Lett. 83, 281 (1981).
149. Yardley, J.T., Introduction to Molecular Energy Transfer, AcademicPress, New York, 1960.
153. Marinelli, W.j., Green, B.D., Piper, L.G., and Blumberg, W.A.M.,"Nitrogen First-Positive Excitation and Quenching," Manuscript inpreparation.
151. Krause, H.F., Fricke, J., and Fite, W.L., "Excitation of Na D-Line inCollisions of 3odium Atoms with Internally Excited H2 , D 2 , and N2," J.Chem. Phys. 56, 4593 (1972).
152. Rich, J.W., Bergman, R.C., and Paymonda, J.W., "Vibration-VibrationPumping of 'aroon ".onoxile initiated by an Optical Source," Appl. Phys.
Lett. 27, 363 (1975).
l1 &
153. S-_arr, W. "Excitatior of Electronic Levels cf Sodi,-z by vb ..Excited Nitrogen," J. Chem. Phys. 43, 73 (1965).
154. Piper, L.G. and Caledonia, G.E., "Long-Lived States in MiitroezenAf:arglows", 38th Annual 3asecus Electronics Conference, %4ontere>°
California(1965). Bull. Am. Phys. Soc. 26, xxx(1986).
155. Caledonia, G.E., Davis, S.J., Green, B.D., Piper, L.G., Rawlins, W.:.,Simons, G.A., and Wey:, G.M., "Analysis of 74etastable State Productionand Energy Transfer," Physical Sciences Inc. Final Technical Report
'' T.R-576, Prepared for AF7,AL under Contract No. F33615-85-C-2568 (1S6'
156. Morgan, J.E. and Schiff, H.I., "The Study of Vibrationally-Excited ';2Molecules with the Aid of an Isothermal Calorimeter," Can. J. Chem. 4',
903 (1963).
157. Harteck, P. and Reeves, R.R., Jr., "Formation and Reactions of the
Excited 02't-- u) Molecules," Discuss Faraday Soc. 37, -2 (1964).
138. Chu, A.L., Reeves, R.R., and Halstead, J.A., "Surface-CatalvzeaFormation of Electronically Excited Nitrogen Dioxide and Oxygen, ' J.Phys. Chem. 90, 466 (1986).
_e 159. Kenner, R.Z. and Ogryzlo, E.A., "Quenching of 02 (c 1-u- ) v=0 by 0(3p),
02(a1 ig), and Other Gases," Can. J. Chem. 61, 921 (1983).
% 160. Schmeltekoff, A.L., Ferguson, E.F., and Fensenfeld, F.C., "Afterglow
Studies of the Reactions He"' He(Z 2S), and O with Vibrationally-Excifed
Ni2 , ' J. Chem. Phys. 49 (1968).
161. Young, S.J. and Horn, K.P., "Measurement of Temoeratures ofVibrationally-Excited '42," Chem. Phys. 57, 4835 (1972).
162. Young. S.J., "Measurement of Vibrational Temperature ,f ZO and N- Usingthe He(2"S) Pnning-.onization Technique," J. Chem. Phys. 38, 16031973).
163. Piper, L.G., Gundel, L., Velazco, J.E., and Setser, D.W., "Excitationof Nitrogen and Carbon ".tonoxide Ionic Emissions by He(2 3 S), He", and
He 2+ " J. Chem. Phys. 62, 3883 (1975).
164. Caledonia, G.E. and Center, R.E., "Vibrational Distribution Functionsin Anhar.onic Oscillators," J. Chem. Phys. 55, 552 (1971).
165. Center, R.E. and Caledonia, G.E., "Anharmonic Effects in the VibrationalRelaxation of Diatomic Molecules in Expanding Flows," Appl. Opt. 13,I)795 (197;).
166. Jolly, J., Touzeau, '4., and Ricard, A., "Determination of Non-3oltz.,annSVibrational Level Populations of Excited Nitrogen Using the Penning-
:ni'zation Technique," J. Phys. 3 14, 473 (1981).
170
_O iij111
167. 'A arinelii W.,. and Piper, L.G. , "Franck-Concdon Factors arid Absc .Ite,ra-sitlon Pr~abi'i::es for the !F (B3 0-X) rnstin, .)n.
Scectrosc. Radiat. Transfer 34, 321 (1985).
168. 04ilonardo, *.. and Capitelli, M., "Non-Equilibrium Dissociatlon f
NMitrogen,"1 Rev. Phys. Appl. 13, 115 (1978).
169. Treanor, C.E., Rich, J.W., and Rehn, R. G., 'Vibrational ?eaxat_,:n ofAnharmonic Oscillators with Exchange-Dominated Collisions," i. Chem.Phys. 43, 1798 (1968).
170. Nichols, R.W., 'Einstaein A Coefficients, Oscillator Strang3tas andAbsolute Band Strengths for the N2 Second Positive and N- F 3
Negative Systams," J. Atmos. Terrest. Phys. 25, 218 (1963).
- ~ 171. Broadf )oc, A.L. , "Resonance Scattering by N2+*,"t Planet. Space Scl.. 15,
1,301,(16k
172. Brown, W.A. and Landshoff, R.K., "The Electronic Transition 'oment ofthe N2 Frs Negative System," j. '2uant. Spectrosc. Radiat. Trinsfer11, 1143 k1971).
173. Lee, L.C. and judge, D.L. , "The Electronic Transition M4oment of the N,+(32, +4.X2 ,g ') System," j. Phys. 3: 6, L121 (1973).
174. Comes, E.J. and Speiar, F., "The Optical Formation and the Collisional1
eactivato of te First Negative System of Nitrogen, " Chem. hsLett. 4, 13 (1969).
1A-,175 . Chang, 'R.S.F., Setser, O.W. , and Taylor, G.W. , "Assignment of Rate
Constaints to Exit Channel's from 2uenching of Hie(2 3 s) :Ietaatibia -Xt~cs,Chem. Phys. 25, 2C1 01978).
176. Gottscho. FR.A., Fieldi, R.W., Dick, K.A., and Benesch, W.4- C 4,. , !C
"C:eoerturbation of the .42 First Negative Group B2Iu--X2: " .)ZSpec. XX, 435 (1919..
0 177. ?ieL.G., Cowles, L.M., and Rawlins, W.T., "State-to-StateZxcitaticn of '40(A 2-,, v'=0,1,2) by N 2 (A3ku+, vl=O,1,2)," j. Chem.Phys. 85, 3369 (1986).
178. Inn, E.C.Y., "'Char~e Transfer Between He*. and N2 ," Planet. Soace Sc-:.,15, 19 (1967).
171/172
innw
fit 'i19
W- -W w w w w W .4W
