C.R. Brazier et al- Fourier Transform Spectroscopy of the A^3-Pi-X^3-Sigma^- Transition of NH

8/2/2019 C.R. Brazier et al- Fourier Transform Spectroscopy of the A^3-Pi-X^3-Sigma^- Transition of NH

1/22

JOURNAL OF MOLECULAR SPECTROSCOPY I?& 38 -4021986)

Fourier Transform Spectroscopy of the A311-X32- Transition of NHC.R. BRAZIER,R.S.RAM,AND P.F.BERNATH

Department of Chemistry, University of Arizona, Tucson, Arizona 85 721The AII-X32- transition of NH has been observed using a high-resolution Fourier transform

spectrometer. The first three vibrational levels in each state were observed and the vibrational,fine structure, and rotational constants obtained. 0 1986 AGUI.WGCsr IN .

I. INTRODUCTION

The A3H-X3Z1- system of NH has been known for a very long time. The bandswere first described by Eder (I) in 1893 and have been the subject of many subsequentinvestigations. The main branches of the O-O and 1 1 bands were assigned from emis-sion spectra by Funke (2) in the 1930s. The analysis was greatly improved by theobservation of the 1-O and O-O bands in absorption by Dixon (3). His use of a roomtemperature source made it possible to observe most of the satellite branches, whichdrop rapidly in intensity with increasing rotation. Dixon was thus able to correctlyassign his measurements together with those of Funke, and obtained reasonably ac-curate rotational and fine structure constants for the ground and excited states.

Additional measurements on the O-O and l- 1 bands were made by Murai andShimauchi (4) while the weak 0- 1, l-2, l-0, and 2- 1 bands were recorded by Malicetet al. (5). The A311-X32- system of the isotopic species ND has been studied byShimauchi (6) and by Bollmark et al. (7).

The accuracy of all these measurements was limited to, at best, +0.03 cm- by theuse of grating spectrographs. The development of CW dye lasers around 1970 did nothelp as the transition occurs near 3360 A-well into the ultraviolet. However, therecent development of intracavity frequency doubling has made it possible to recordlaser-induced fluorescence spectra in the W. Ubachs et al. (8) recorded the laser-induced fluorescence spectrum of the A3H-X3Z- system of NH in a molecular beam.Their linewidth was sufficiently reduced that they were able to resolve the hypertinestructure of both nuclei. While they were able to measure hyperfine splittings accurateto +0.00007 cm-, they did not attempt an absolute frequency calibration of therotational lines and hence did not determine any rotational or fine structure constants.

We have remeasured the A311-X38- system of NH using a Fourier transform speotrometer with a precision of ~0.0002 cm- for the strong, unblended lines in theO-O band, an improvement of more than two orders of magnitude over the previousbest measurements.Several high-resolution studies of the ground state have been made, measuring bothpure rotation and vibration-rotation transitions. The far-infrared laser magnetic res-onance spectrum was recorded by Radford and Litvak (9) and by Wayne and Radford

381 0022-2852186 $3.00CwyrQhl 0 1986 by AcademicPress, nc.Au rightsof reprcducticmn any form rcmved.


2/22

382 BRAZIER, RAM, AND BERNATH(IO). With the development of tunable infrared lasers additional experiments becamepossible. Van der Heuval et al. (II) recorded the zero field pure rotational spectrumin the far infrared, while Bemath and Amano (12) observed the vibrational fundamentalin the infrared. The infrared mndamental was observed by matrix isolation spectroscopyby Mill&an and Jacox (13). Several vibration-rotation bands have been observed byFourier transform emission spectroscopy at moderate resolution (14, 15) and, recently,at high resolution (16, 17).

There are four known singlet states of NH, aA, bZ+, cII, and dZ+. All fourallowed transitions between these states have been studied. The c-u system at 3240A, just to the blue of the triplet system, has been extensively analyzed. The papers byPeat-se (18), Dieke and Blue (19), and Nakamura and Shidei (20) in the 1930s allidentified this system at about the same time and gave comparable results. Morerecently this system was recorded at higher precision by Cheung et al. (21) and byRamsay and Sarre (22). The cII-u A transition was also observed in our Fouriertransform spectrum. The higher precision of these measurements made possible thedetermination of the A-doubling interaction in the a A state. Dymanus has also de-termined the A doubling in a A by the high-resolution laser-induced fluorescenceexperiment in a molecular beam (23). Our data on this system will be publishedelsewhere (24).

The &I-blZf system at 4502 A has been studied by Lunt et al. (25) and by Whittaker(26). The dZ+-cII transition at 2530 A is also well known and has been studied byLunt et al. (27) Whittaker (28), and Krishnamurty and Narasimham (29). The otherallowed transition dZ+-b2+ is in the vacuum W at 1620 A and has been observedby Graham and Lew (30). The a*A vibrational fundamental of NH has been studiedby Hall et al. (31) by color center laser kinetic spectroscopy. Very recently Leopoldet al. (32) observed the far-infrared laser magnetic resonance spectrum of the a Astate of NH and ND.

An important advance in the spectroscopy of NH came with the observation of thebZ+-X32- intercombination band by Masanet et al. (33). From this emission thelocation of the lowest singlet aA could be determined relative to the ground state.The singlet-triplet splitting was also determined by photoelectron spectroscopy ofNH- by Engelking and Lineberger (34). The electron affinity of NH was establishedto be 0.381 eV in this photoelectron experiment. More recently high-resolution ob-servations of the bZ+-X32- system have been made (35) as well as detection ofa A-X3Z- emission (36). The b-X measurements of Cossart (35) result in a singlet-triplet splitting of 12 688.39(10) cm- for aA (u = 0, J = 2) - X32- (v = 0, J = 1,N = 0) using the line positions of Refs. (24, 26, 35).

There have been many photochemical studies involving NH (37-45) produced bylaser photolysis. Using precursors such as ammonia, hydrazine, and hydrazoic acid,the internal energy distributions, radiative lifetimes, and quenching rates for the dif-ferent states of NH initially populated have been determined. Similar results have alsobeen obtained using electron impact dissociation (46). Cooling of spin-orbit com-ponents and population inversion in the A311state has been observed by Carrick andEngelking in a corona excited nozzle expansion (47).The transition probabilities for several bands of NH have been determined. Lents(48) published a review of the molecular constants for all the known states of NH.


3/22

&I-X32- TRANSITION OF NH 383The absolute absorption intensities were determined by Harrington et al. (49) in ashock tube, The lifetime of the A3H-X3Z- system has recently been remeasured byFairchild et al. (50) for several vibrational bands. They comment that the Franck-Condon factors computed from RKR potential curves are not very accurate becauseof the poor quality of the spectroscopic constants for the higher vibrational levels. Wehope the new constants determined in this work will make these calculations moreaccurate. This group has also recently measured quenching rate constants for the A311state (51).

Smith et al. (52) studied the variation of lifetimes with rotation and vibration forboth the A311 and cII states. They discovered predissociations in both states, causedby crossing of a repulsive Z- state. The observed intensities in the bands which wehave studied are consistent with their results. The observation of rotational predis-sociation by Graham and Lew (30) and Zetzsch (53) has resulted in an accurate de-termination of the dissociation energy L$ = 3.46 + 0.01 eV. Foner and Hudson (54)have measured the ionization potential by mass spectroscopy to be 13.47 +- 0.05 eV.The dipole moment has been measured by Scar1 and Dalby (55) to be 1.38 D in theground state using the high field Stark effect, where the electric discharge which producesthe molecules also provides the Stark field. The standard heat of formation of NH is85.2 + 0.4 kcal/mole (56).

All of these experimental results can be compared with those obtained from theo-retical calculations by, for example, Kouba and Ghrn (57) Hay and Dunning (58)and Meyer and Rosmus (59).The frequent observation of NH in flames and other energetic environments hasresulted in much interest in the spectrum. The species NHi have significant effects onthe production of NO, in fuel-air combustion (60, 61). The A311-X32- transition canbe used to monitor the concentration of NH in flames (62, 63).

NH was first observed in a nonlaboratory source by Fowler and Gregory (64) inthe spectrum of the sun, although at that time (19 18) the 3360-A band was incorrectlyassigned to ammonia. Since then NH has been found in stellar atmospheres (65, 66)and in comets (67, 68) by the spectroscopic observation of the A311-X3X transition.With the current interest in infrared astronomy, the vibration-rotation spectrum ofNH has also been observed in stellar atmospheres by Fourier transform spectroscopy(69, 70).

II. EXPERIMENTAL DETAILSNH molecules were produced in a standard copper hollow cathode discharge, op-

erated at a current of 430 mA with a continuous flow of 4.5 Torr of He. Small amountsof nitrogen and hydrogen,40 and 120 mTorr, respectively, were added. These param-eters were adjusted to optimize the production of NH. The emission spectrum wasrecorded through a copper sulfate titer with the Fourier transform spectrometer as-sociated with the McMath Solar Telescope of the National Solar Observatory at KittPeak. A total of 15 scans were coadded in one hour of observation. The unapodizedinstrumental resolution was 0.042 cm-.

The National Solar Observatory is operated by the Association of Universities for Research in Astronomy,Inc., under contract with the National Science Foundation.


4/22

384 BRAZIER, RAM, AND BERNATHThe signal-to-noise ratio for the strong lines of the O-O band was more than 1000

to 1. The l-l band was almost as strong at about 500 to 1 while the other observedbands 2-2,0- 1, l-2, l-0, and 2- 1 were all much weaker. A sample of the P-branchregion of the Av = 0 sequence is shown in Fig. 1. The linewidth varied to some extentwith rotational and vibrational quantum number but was typically 0.18-0.2 cm- forthe strong O-O lines. As a result the relative line positions could be determined to aprecision of 0.0002 cm- (the linewidth divided by the signal-to-noise ratio) for strongunblended lines. The accuracy of the measurements is less than this and is limited bythe absolute wavelength calibration. This was rather difficult to achieve as only a fewHe(I) and He(I1) lines were observed with high signal-to-noise, and these had notpreviously been measured with very high accuracy.

Only one He line, 4P-2S, proved suitable for calibration because of a symmetriclineshape and a signal-to-noise ratio greater than 500. The previous measurement25 215.270 f 0.005 cm- for this transition (71) was improved by comparing aHe/Ti hollow cathode emission spectrum with an Ar/Ti hollow cathode emissionspectrum. Norlen (72) has made excellent line-position measurements for argon. Ti-tanium atomic lines were used as a transfer standard to give 25 215.2824 f 0.003cm- for He 4P-2l5. The estimated error of kO.003 cm- means that our absoluteaccuracy is still a factor of 10 larger than the precision of the strong lines. Any im-provement in the accuracy of the He 4P-2S line measurement or a direct measure-ment of a strong NH transition by laser methods would bring the estimated accuracyof our work closer to the precision.

III. ANALYSISThe interferogram was transformed by standard methods to yield the spectrum,

which covered the range from 23 800 to 35 600 cm-. This range covers the Au = 0NH A%-X32-

qw 00)i= 123

pi051 0.1)i=

I I I 129160 29190 cm-FIG. 1. A segment of the NH ,4%-X32- spectrum in the P-branch region. Tripled rotational lines of theO-O. l-l and 2-2 vibrational bands are labeled with N.


5/22

A&XZ- TRANSITION OF NH 385and Au = f 1 vibrational bands of the A311-X32- system of NH. In addition, it coversthe cII-a A system, and our analysis of this system will be reported elsewhere (24).Some problems were experienced with bands of NZ and N: which occur throughoutthe region. There were many areas with strong bands due to NZ and N: but generallythese did not overlap the NH lines. There was also a weak background throughoutmuch of the region of interest which limited the signal-to-noise of the NH transitions.

The frequencies of the NH lines were obtained from the spectrum using DECOMP,a data reduction program developed at the National Solar Observatory. This programcontains both a line list feature which provides the frequencies of all lines above agiven intensity, and a least-squares fitting procedure which uses Voigt lineshape funotions to determine the true positions for all the components of blended lines.

The width (FWHM) of the lines varied from 0.17 cm- for the low-J O-O band to0.23 cm- for the high-J O-O band to 0.25 cm- for the high-J 2-2 band. This is partlydue to a range of temperatures found within the hollow cathode. However, the lastthree or four lines in each band show a rapid increase in linewidth from -0.200 to-0.250 cm-. These lines are affected by predissociation by the repulsive ?- state.Application of the uncertainty principle to the N = 3 1, o = 0 level of A311 with theknown lifetime of 96 nsec (52) results in a broadening of only 1.6 MHz (0.00005cm-). This is much less than the observed excess broadening of this level (0.227-0.200 = 0.027 cm-). A plausible explanation is that the levels affected by predisso-ciation are very pressure sensitive, but further work is required to confirm this spec-ulation.The observed line positions of the A311-X32- O-O, 1- 1, 2-2, l-0, 2- 1 O- 1, andl-2 bands are given in Tables I-VII, respectively. The unblended lines have a precisionof linewidth/signal-to-noise while blended lines are somewhat less precise. Virtuallyall of the Q-branch lines are blended to some extent while the P and R branchesexperience overlap from satellite lines at low J and a crossing of the three spin com-ponents at higher J.

The A311-X3X system of NH is a good example of a Hunds case (b) to Hundscase (b) electronic transition at high J. (See Her&erg (73) for an explanation of couplingcases, and branch structure and notation for a 311-3Z- transition.) Each vibrationalband has 9 strong transitions, PI, P2, P3 ; Qi , QZ Q3 and R, , R2, R3, with the triplingof the P, Q, and R branches due to electron spin. These are the only branches allowedin a pure case (b) to case (b) band, however, at lower Ja transition to case (a) couplingin the A311state occurs and weak satellite branches become allowed. There are a totalof 18 possible satellite branches of which 14, 11, and 9 were seen for the O-O, l-l,and 2-2 bands, respectively. In the O-O band Dixon (3) observed the same numberof branches, but we were able to follow them to higher J, while only a few l-l satellitelines were seen by Murai and Shimauchi (4). The 2-2 band of NH has not previouslybeen observed at high resolution. No satellite branches were measured in theAv = + 1 bands, although several weak lines were observed. The signal-to-noise ofthese satellite lines was at best 5 to 1, and the other data from diagonal bands wassufficiently good to predict their positions much more accurately than they could hemeasured.

Each vibrational band was initially fitted separately using a nonlinear least-squaresprocedure. The Hamiltonian matrix for each state was set up using case (a) basis


6/22


7/22

-


8/22


9/22


10/22


11/22

A311-A?- TRANSITI ON OF NH 391functions and diagonalized, to produce a set of energy levels. The effective Hamiltonianfor a 32- state is given by Zare et al. (74), while Roux et al. (75) give explicit matrixelements to second order. The explicit matrix elements for a 311state were taken fromBrown and Merer (76).Initial attempts to reproduce our data for the O-O band using these matrix elementsproved unsuccessful. The lines were of very high precision (up to -+0.0002 cm-) andalso involved high rotational levels (up to N = 32 which has more than 16 000 cm-of rotational energy). As a result, many higher order distortion constants were required,for example, B, D, H, L, and M for the rotational energy, and q, qD, qH, and qL forthe A-doubling interaction. The higher order matrix elements were derived from theexisting ones by multiplying the matrices, making them Hermitian where necessary.For example,M = M(B/B)5 or QL= qtK(llq)(BIB)3 + WB)3W41/2.The results were checked using a symbolic algebra computer program available to us.The explicit matrix elements used are given in Table VIII, with x = J(J + 1).

Once the seven bands had been fitted separately, a modified version of the computerprogram was written to simultaneously fit all seven bands. This procedure is preferableto merging the results from the separate fits, particularly in a case like this, wherethere is a large variation in the precision of the measurements. In the separate fits theweights of the strong lines were adjusted so that the standard deviation of fit was closeto one. In each case these weights were close to the expected values. The strong O-Oand I- 1 band fits had standard deviations slightly higher than expected, probably dueto the presence of small systematic lineshape distortions from NH lines in naturalabundance and unresolved hyperfme structure (8). Only a relatively small proportionof the lines are completely unblended and thus most of the measurements were assigneda lower weight. Deconvolution of blended lines using DECOMP was accomplishedwith the aid of predicted spectra, based on constants determined from the more preciselymeasured lines. It was possible to extract reasonably accurate positions even for multiplyblended lines provided the total number of components was known, and their positionscould be estimated.The standard deviation for the global fit of all seven bands was 0.9, and there wasno evidence of systematic trends in the residuals, indicating that all 1370 lines wereadequately fitted. In addition to the 1240 lines measured in our Fourier transformspectrum of the A311-X32- transition, the ground state vibrational fundamental linesof Bernath and Amano (12) were included. Recently the l-0, 2-1, 3-2, and 4-3vibrational bands of the ground state were measured by Vervloet and Brion (26) usinga Bomem Fourier transform spectrometer. Their 1-O and 2- 1 lines were also includedin the final fit, but they had little effect except for u = 2 due to their generally lowerprecision and the very large number of lines from the electronic spectrum.

Our weighted nonlinear least-squares fit of the lines in Tables I-VII with the Ham-iltonian of Table VIII produced the constants of Tables IX and X. For unblendedlines the weight was determined from the spectrum, while the blended lines weredeweighted so that the precision was of the same order as the error (obs. - talc.) fromthe fit. The wavenumbers of vibration-rotation transitions of Bemath and Amano(12) and Vervloet and Brion (16) are not repeated here. These data were weighted


12/22

392 BRAZIER, RAM, AND BERNATHTABLE VIII

Matrix Elements for II and 2- Electronic StatesT (0.0) - 1 B (0.0)= x+*(1.1) 1 (0.1) -(Z&/2(2.2) - 1 (1.1) - x+2(3.3) - 1 (1.2) - - (2(x-2))1/2

I:::; - I (2.2) - x-2- 1 (3.3) - x(3.4) - -Z(x)/2A (0.0) - -I I:::; - x+2(2,2) - I -x

AD (LAO) - -(x+21 D(x/2) 1 (0.0) - -(x2+6s+4)(0.1) -(1.2) - -(2(x-2))1/2 /2 (0.1) - 2(2x)U2(.+2)(0.2) - -2(x(x-2))1/2(2.2) - x-z (1.1) - -(&+8x)(1.2) - 2x(2(x-2))1/21 (LO) - Z/3 (2.2) - -(.2-2x)(1.1) - -413 (3,3) - -(&+4x)(2.2) - Z/3 (3,4) = 4(.+1)(~)/~(3.3) * Z/3 (4,4) - -(x2+8.+4)(4.4) - -413 (5.5) = -9(5.5) - 213

lo (0.0) - 2(x+2) 43(0,l) - (2x)1/ I3 1 (0.0) - -2(O,l) - (x/Z)/2(1.1) - -4(x+2)/3(1.2) - (2(~-2))/~/3 I:*:; - -2((x-2)/2)1/2(2.2) - 2(x-2)/3 (3:3) - -1(3,3) - 2x/3 I?:; - (x)1/2(3.4) - 2(x)/2/3 - -2(4.4) : ;~U;+)/ (5:5) - -1(5.5)0 (0.0) =% (0.0) -(0.1) -OH (0.0) =(O,l) -(0.2) -P (0.0) -(0.1) -PO (0.0) -(0.1) -(0.2) -(1.1) -

i 17(x+2)*(x/2)1/2i(&+6x+4)+(zx)l/2~x+z)i(x(x-2))/2il*(x/z)/272(x+1 Ii(x/z) /2(x+3)7(x(x-2))/2/2ix

-3x-4(2x)1/2(x+6)/2-(x(x-Z))/2-4x-2(2(x-z))1/2h+2)/2-X+2-3x(x)/2(,+4)-4(x+1)-x-5x2 -8x(,)/2(.2+12x+8)-6x2-20x-8-x2

Note. The upper (lower) sign refers to a state with e(f) parity. Hunds case (a) basis functions were useda (0, 0) = (3Hol~13Ho)

0 = 1%)) 3 = IZie) x= J(J+ 1)1 = IQ,) 4 = 13Z6e)2 = I)&) 5 = 13Zif)

according to the estimated precision quoted and the quality of fit was found to beessentially the same as in the original work. A total of 78 parameters were requiredfor the three vibrational levels of the two electronic states observed. The parametervalues given in Tables IX and X have one standard deviation error estimates included.For the excited vibrational levels not all higher order parameters could be determinedand thus these parameters were constrained to zero. If both AD and y were allowedto vary in the AII state, then a very high correlation, 0.99999, was observed betweenthem and the values obtained did not appear realistic. For a II state it has been shownby Brown and Watson (77) that these two parameters cannot be simultaneously de-termined. However, they have both been measured in, for example, CuF (78) in itsb311 state. Clearly the problem is not a lack of data as a wide selection of rotational


13/22

AII-XZ- TRANSITION OF NHTABLE VIII-Continued

393

(0.0) - 7(3G+lox +4)P H g ;;; : :y:;;r$;;:;;)

(1.1) =72x(x+2)(IJ ) = t(z(x-z))~2x/z

9 (0.0) = il H(0.1) = *(2x)1/2(0.2) = i(x(x -2))/2/2(1.1) = 7x/2

WI (0.0) = 7(3x+2)(0.1) = +(x/2)1/2(3,+4)(0,2) = s(x(x -2))/2(x+2)/2(1.1) = 3(x+6)x/2(I,Z) = i(Z(x-2))/2x/2

98 (0,O) = ?(bx2+12x+4) L(0.1) = t(2x )/2(2x2+8x+4)(0.2) = r(x (x-2))/2(x*+6.+4)/2(I;I) = s(x3/2+8x2+8x)(1.2) = t(Z(x -z))~/2x(x+2)(2.2) = Sx(x-2)

a. (0,O) = i(10x3+40x *+36x+8)(0.1) = ~(2~)~2(5X3/2+2ox2+28x+8)(0.2) = ~(x(x -~))~~(x~/2+6~~+12~+4)(1.1) = i(x/2+15x3+40x2+16x)(1.2) = ~(2(.-2))/2(3,3/2+8~2~~)(2.2) = i(3x3-4i-4x)

(0,O) = ~~+12.~+24~+8(O,l) = -(2x)1/2$3x2+16x+8)(O,2) = (x(x-2)) 12(6.+4)(1,I) = x3+18x2+16x(1.2) = -(2(~-2))~/~(3~2+4x)iz;zj _.s_i,.(3.3) = x3+12x2+8x(3,4) = -(x)1~2(6x2+ZoX+8)(4.4) = x3+18x2+28.+8(5;5j = X3(0.0) = xr+20x~+80&+72x+16(0.1) = -2(2x)U2 2x3+2Gx 2+28x+8)(0.2) = (x(x-Z))1 I 2(12x2+24x+8)(1.1) = x4+32x+80x2+32x(1.2) - -2x(2(.-2))1t2(2x2+8x+4)(2.2) = x+4*3-0x2-Bx(3,3) = x4+24x3+48x2+16x1;;; = -4(x)~2(2x3+14x 2+16x+4)

= x+32x3+104x2+80xc16&) = x4

H (O,O) = x~+30x'+2002+344&+192x+32(O,,) = -(2.)~*(5x+80x3+216.2+160x+32)(0.2) = 4(x(x-2))1~2(5x3+20X2+18x+4)(1.1) = x5+50x+240x3+256x2+64x0.2) = -(2(~-2))~~(5,+40~+56X2+16x)(2.2) - x5+10x4-40x2-16x(3,3) = x~+40 @+160&+ 144&+3 2x(3.4) = -2(.)~~(5x+60x3+136 x2+88x+ 16)(4.4) = x5+50x4+280x3+416x~+208x+32(5.5) = x5

TABLE IXSpectroscopic Constants (in cm-) for the XZ- State of NH

v=o v=1 v-2

TV 0.0 3125.57292(25) 6094.87617(57)6, X.3432784( 45) 15.6964414(81) 15.050334( 26)

103X 0, 1.702786(35) 1.679548(64) 1.65803(30)10~ HL I. 23442( 115) 1.17513( 185) 0.9575(88)10~ L, -1.3974( 162) -1.4396(173) 0.010s H 4.18(80) 0.0 0.0

h 0.920063( 148) 0.921239(172) 0.91916(31)106X *o -9.09(142 ) -14.67(82 ) 0.0102x ly -5.4844( 22) - 5.1945(31) -4.9104(69)105x IW 1.5098( 75) 1.4731(117) 1.795(58)109x ,qV -1.366(89) 0.0 0.0

Note. One standard deviation error on the last digits is quoted in parentheses.


14/22

394 BRAZIER, RAM, AND BERNATHTABLE X

Spectroscopic Constants for the ASH State of NH (in cm-)constants v.0 v=l v=2

T 29761.1829( 1) 32795.9283 ( 3) 35633.7219(6)0

10% oy107x H

1oX L10~ M

A10% AD

A10% AD102x 110% 010% P10% PD10% PH10% q,10% qD10 % qH1ox 9L

01$X 0D10% 0

l&3214823( 55) 15.5762834(86) 14.798763(31)1.789698(43) 1.803235(69) 1.83399(41)1.0739(139) 0.9390(20) 0.3278( 137)

-1.6481(197) -3.0454(187) 0.0-2.757(99) 0.0 0.0

-34.61976(15) -34.64854(33) -34.68667(63)-8.14a -0.14a -8.14-0.19968( 22) -D.20068( 27) -D.19476( 49)-1.630(156) 0.0 0.0

2.9830(24) 2.8474( 35) 2.8332(68)-5.406(52) -3.410(133) 0.0

5.5222( 25) 5.1880( 41) 4.7483( 103)-1.7795( 151) -1.979(31) -2.358( 135)

2.124( 193) 1.61(59) 0.0-3.15870(40) -2.95597( 33) -2.73942(121)

1.3822(29) 1.3879( 20) 1.5275( 114)-1.861(66) -1.086(32) 0.0

1.95(46) 0.0 0.01.28447 (20) 1.22637(26) 1.15014(56)

-1.3440( 136) -1.703(31) -2.072( 99)-3.07(24) -6.30(71) 0.0

Note. One standard deviation uncertainties in parentheses. Fixed at computed value (see text for details).

levels in both A doublets of all three spin-orbit componentsVeseth (79) provides a formula relating AD to other constants:have been measured.

AD=2A_,= &4,+1 - Av)Dv=oB, - B,+l + 6B:fo;For I.J= 0 this gives AD = -8.14 X lo- cm -l. For the final fit, AD was constrained tothis value for all three vibrational levels.

The term values for the first three vibrational levels of the A311 and X32- states ofNH have been calculated (for the range of rotational levels studied in each case) andare given in Table XI. Any unmeasured lines can be determined from these termvalues. The term values have been extrapolated only to slightly higher J than wasobserved in the spectrum. The large number of high-order distortion constants requiredin the fit means that significant errors will soon occur in any extrapolation.


15/22

A%-XZ- TRANSITION OF NH 395IV. DISCUSSION

Relative IntensitiesEmission from u = 0, 1, and 2 in the A311 state of NH was observed. The O-O and1- 1 bands appear to differ in intensity by about a factor of 2 while the 2-2 band is an

order of magnitude weaker. This is shown clearly in Fig. 1. The relative transitionprobabilities of the three bands are approximately equal (50) and thus a purely Boltz-mann distribution of vibrational energy would imply a decrease of about a factor of4 for the 2-2 band relative to the O-O. Smith et al. (52) have measured the variationin lifetime with rotational quantum number for u = 0 and v = 1 of the A311 state ofNH. They observe predissociation in the v = 0 level beginning at N = 24 and inu = 1 from N = 13 onward. They find that the perturbing state, thought to be of %symmetry, crosses the A311 potential curve at 34 000 cm-. This is below the firstrotational level for v = 2, implying that all v = 2 levels are predissociated. The shorteningof the lifetime leads to fewer molecules fluorescing and hence a weaker band. Relativeintensities for the O-O, l-l, and 2-2 bands at low and intermediate N have beenmeasured from the spectrum and are given in Table XII. The ratio of u = 0 to u = 1intensity indicates a vibrational temperature of 5400 K. This predicts an intensity forthe 2-2 band of 0.21, relative to O-O, in the absence of predissociation. Other bandssuch as 3-3 and 4-4 would also be easily observable if there was no predissociation.The decrease in relative 2-2 band intensity at low N from the expected 0.21 to theobserved 0.068 implies a predissociation lifetime of 200 nsec, assuming a fluorescentlifetime of 410 nsec (50, 52). The overall lifetime is thus 130 nsec, for N = 3 or 4 ofA311,2)= 2. At N = 13 for v = 2, the decrease is much greater, implying a predissociationlifetime of 50 nsec and an overall lifetime of 45 nsec. The lines decrease rapidly inintensity with increasing N, and are unobservable beyond N = 18, indicating essentiallycomplete predissociation.Equilibrium Constants

From the molecular constants for v = 0, 1, and 2 of A311 and X32- reported inTables IX and X, equilibrium constants were determined (Table XIII). For the vibra-tional constants only AGllZ and AG3,2 are available so two constants, w, and w,x,,could be determined. The neglect of wd, introduces a large error into our estimatedvibrational constants in Table XIII [wcv, is 0.33 cm- for X3x:- (16)]. Our values forAG)I,zand AG$,z gree with previous determinations (12, 16).Equilibrium values for the fine structure and rotational constants may also be de-rived. The spin-orbit constant, A, was fitted to a linear expression in (V + l/2) because,although the deviation from linearity is much greater than the random error, there isa possibility of small systematic deviations. The spin splitting constant X for both theground and excited states shows no clear variation with 2).An exact fit to three parameters B, , a,, and -yewas made for the rotational constantsof both states. The errors quoted in Table XIII were computed by the customary rulesfor error propagation, the true error is probably an order of magnitude larger thanthis. The equilibrium bond length r, was then calculated and the values obtained areconsistent with, but much more accurate than, the previous values.


16/22

396 BRAZIER, RAM, AND BERNATHTABLE XI

Term En ergies (in cm-)x3x-

=O =l =2.J FI F2 F3 Fl F2 F3 Fl F2 F301 -0.00772 32.49733 97.71004 195.50315 325.74196 488.2594: 682.8508909.21329 1 1 6 7 . 2 4 4 9

1 0 1456.446211 1776.519612 2127.070913 2507.669414 2917.849315 3357.110216 3824.918317 4320.7072::

4843.87905393.805220 5969.8283

::6571.26217197.3932

23 1847.4021:: 8520.76439216.450926 9933.730127 10671.767628 11429.70*129 12206.675630 13001.774331 13814.089332 14642.687033 15486.615934 16344.9064

33.347990.6665196.5423326.8532489.4368684.0904910.5722

1168.60121457.85791777.98512128.58852509.23782919.46693358.77563826.63004322.46364045.67855395.64645971.70936573.18137199.34887849.47258522.70759218.50519935.813310673.8780

11431.843712208.834313003.954213816.288414644.903315488.847316347.150717218.8266

31.562697.55701 9 5 . 6 0 7 3326.0237488.6896683.4147909.96201168.0523

1457.36761777.55102128.20882508.91062919.19073358.54873826.45094322.33074845.59025395.60115971.70546573.21717199.42277849.5826*522.93209218.68239936.021310674.1148

11432.107512209.123213004.266113*16.621414645.255315489.216416347.534917219.223818103.2782

J Fl012 29770.61373 29842.31494 29943.98145 30076.71476 30240.92887 30436.70678 30663.95489 30922.470410 31211.9740

11 31532.124912 31882.530013 32262.749514 32672.299915 33110.656716 33577.255717 34071.495218 34592.736319 35140.304520 35713.490421 36311.550722 36933.708323 37579.153224 38247.042125 38936.501726 39646.622227 40376.463528 41125.051429 41891.377030 42674.396131 43473.026832 44286.147833 45112.595434 45951.1605

v=op2

29806.607529866.095629960.699430089.051030250.273630443.822830669-285030926.283231214.436931533.344331882.574232261.662032670.108633107.379133572.903634066.076634586.258035132.773435704.914*36301.940836923.077237567.517238234.422138922.920639632.109240361.051841108.778841874.286542656.535443454.448844266.910445092.761945930.799246779.7689

F22 9 0 2 6 . 9 3 0 029079.616629971.600630098.477330258.757230451.710130676.826730933.672531221.828431540.865231890.331432269.746632678.598533116.341833582.396834076.149634596.952435144.123635716.948036314.677736936.531937581.697638249.329738938.550839648.451440378.089241126.4***41892.640342675.498443473.980744286.965145113.287445951.737846801.057447659.9328

3125.5649 3157.62513156.7791 3220.35713219.4082 3314.35443313.3256 3439.49653438.3986 3595.62313594.4620 3782.53453781.3141 3999.99183998.7150 4247.71174246.3866 4525.39694524.0133 4832.67674831.2422 5169.16755167.6839 5534.44425532.9128 5928.04625926.4686 6349.47886347.8566 6798.21396796.5485 7273.69077271.9839 7775.31737773.5707 8302.47078300.6860 8854.49878852.6777 9430.72029428.8648 10030.426410028.5386 10652.881810650.9635 11297.3249

11295.3783 11962.969011960.9962 12649.003312647.0066 13354.593113352.5748 14078.880814076.8435 14820.986314818.9323 15580.007215577.9392 16355.0193

v=lFl F2

32839.721232803.6533 32896.427632872.3573 32986.598932969.5683 33108.989633096.3532 33262.753833253.1186 33447.352033439.9453 33662.369133656.7382 33907.424933903.2929 34182.135234179.3267 34486.093334484.4949 34818.862434818.3996 35179.9716351*0.594* 35568.913835570.5891 35985.144735987.8487 36428.083036431.7983 36897.109836901.8221 37391.568937397.2651 37910.766737917.4334 38453.972438461.5943 39020.417539028.9770 39609.296239618.7723 40219.764640230.1324 40850.940440862.1705 41501.902141513.9604 42171.688142184.5353 42859.295442872.8868 43563.678143577.9634 44283.745544298.6684 45018.359945033.8584 45766.3341

3155.83463219.24073313.40963430.65443594.86053781.84063999.36054247.14514524.88014832.21335168.75575534.08225927.73246349.21166797.99167273.51187775.18008302.37358854.43979430.697610030.438310652.926311297.399811963.0724

12649.132813354.746614079.055914*21.1*0615580.218116355.244217145.3127

6094.0670 6125.63216124.7932 6185.78036184.8415 6275.90336274.8876 6395.88196394.8001 6545.55796544.4159 6724.73376723.5354 6933.17306931.9214 7170.60137169.2988 '7436.70637435.3546 7731.13817129.1393 8053.51058052.0662 8403.40118401.9129 8780.3523a778.0221 9183.87219182.3016 9613.43519611.8261 10068.483210066.8377 10548.427010546.7469 11052.6465

11050.9338 11580.492611578.7495 izi3i.288212129.5168 12704.3295

6123.84286184.66116274.95236395.03046544.78256724.02366932.52207170.00537436.16217730.64348053.06308402.99898779.99349183.55479613.157210068.242910548.222311052.4753

11580.352812131.177612704.24571329*.*2*0

8332860.283332910.123932997.625633118.499133271.290633455.261433669.899433914.766234189.438634493.402534826.440435187.825335577.117035993.760136437.163536906.699537401.703937921.475238465.275239032.328339621.821640232.904240864.686941516.241342186.598842874.749643579.640944300.175245035.208045783.545146543.9400

Flv=2FT

3 5 6 3 9 . 6 5 0 535705.223235797.784935918.361436067.348036244.820636450.678936684.711636946.627337236.070437552.629437895.841938265.197738660.1408390*0.070539524.342839992.271140483.126840996.140141530.5003

35675.79423 5 7 2 9 . 6 1 5 935015.169235931.338836077.313036252.556036456.648436689.202536949.824537238.096137553.566937895.749738264.118538658.107539077.110739520.481639987.533440477.538940989.731341523.304042077.4113

35696.642735743.527935826.337435940.945636085.910136260.488636464.162136696.483636957.019637245.323737560.924337903.317438271.963138666.283739085.662639529.443439996.930140487.386641000.037341534.066942088.620942662.8059

The centrifugal distortion terms D and H were each fitted to two parameters usinga weighted least-squares fit of the three available points. The results obtained for 0,and He can be compared with those calculated from the Kratzer and Dunham-Birge


17/22

&I-I-XZ- TRANSITION OF NHTABLE XI-Continued

397

Ah f levelsv-o v=l =2

&J Fl F2 F3 Fl p2 p3 Fl F2 F3012 29770.60183 29842.22894 29943.73785 30076.23226 30240.13237 30435.5274a 30662.32899 30920.3387

IO 31209.281411 31528.820212 31878.566013 32258.082914 32666.891515 33104.471616 33570.263417 34063.669618 34584.056019 35130.753020 35703.056221 36300.227222 36921.494123 37566.052324 38233.064325 38921.660426 39630.938127 40359.962328 41107.764129 41873.340330 42655.652131 43453.623232 44266.137833 45092.037734 45930.1192

29807.275329867.147829962.032330090.669630252.217630446.141930672.031330929.508631218.191531537.675831887.527532267.278732676.426833114.433133580.723734074.688734595.683835143.029835716.013936313.889836935.878237581.167238248.9127

32840.336432897.410432987.849933110.510533264.579933449.528733664.944233910.446134185.648134490.141334823.485935185.207635574.795435991.701436435.339736905.087137400.282437920.227038464.184939031.382139621.007240232.210540864.103841515.759342186.209442874.444643579.412744300.016445035.111745783.5048

32862.780032911.945632998.972733119.469133271.901333455.497233669.730133914.155434188.347534491.872334824.273435185.065935573.732435989.720936432.444136901.278337395.563637914.603438457.664539023.976539612.731740223.085140854.153141505.013342174.703242862.219043566.514444286.498045021.031745768.927946528.9463

35676.348435730.516235816.322835932.744536079.001636254.56813 6 4 5 9 . 0 2 6 636691.989036953.059237241.816337557.806537900.538538269.482338664.067239083.682139527.675239995.353640485.984240998.793441532.967842087.6543

29829.553729881.500929972.993630099.466530259.367230451.922730676.6084

35698.982335745.257835827.622635941.875936086.503436260.732236464.029736695.943436956.037837243.867337558.961537900.819438268.904438662.643039081.422939524.593039991.462540481.301540993.340441526.770842080.745242654.3777

32803.641532872.276632969.341733095.904833252.378233438.848733655.226133901.310534176.823234481.423734814.718035176.264135565.575135982.121436425.332336894.597137389.265637908.649238452.0205390X8.614439607.627440218.217540849.504141500.567142170.446042858.138643562.599944282.739945017.4214

35639.638635705.148035797.576235917.949636066.668736243.815236449.293836682.897736944.340037233.269437549.278937891.910938260.659938654.97523 9 0 7 4 . 2 6 1 739517.881039985.152740475.354840987.724141521.4573

31220.625631539.106131887.974132266.751432674.928333111.962633577.278034070.264734590.278835136.643135708.647136305.547436926.568337570.901838237.707838926.1142

38938.2381 39635.216939648.2347 40364.078840377.9606 41111.729941126.4409 41877.166041892.6663 42659.347142675.5919 43457.196343474.1352 44269.596944287.1747 45095.390345113.5462 45933.372345952.0404 46782.289246801.3982 47640.8334

relations (73), respectively (Table XIII). In all four cases the experimental and calculatedvalues are in good agreement.A-Doubling and Spin-Rotation Interactions

The Adoubling and spin-rotation constants can be calculated assuming pureprecession (80) between the A311 and X32- states. Contributions to p, q, and y ariseonly from triplet states, of which A311and X3Z- are the only two known. Theoreticalcalculations by Kouba and &rn (57) predict weakly bound II and 3Z- states nearthe A311state, but due to the large difference in equilibrium bond lengths, the Franck-Condon factors are likely to be small. There may also be high-lying Rydberg states,but these will have comparatively little effect due to the large energy separation. Inthe case of pure precession between a 32- and a 311 state the value of y in the 32-

TABLE XIIRelative Intensities for Selected Lines of the AII-X32- Transition of NH

Lines PF(N) =O v-1 "=2

b(4) 251 112h(5) 355 158P2(13) 220 89Pl(14) 240 100

17244.85.2


18/22

398 BRAZIER, RAM, AND BERNATHTABLE XIII

Equilibrium Constants for AII and X32- States of NH (in cm-)constant A% XS* W2 3 0 3 4 . 7 4 5 4 ( 3 ) 3 1 2 5 . 5 7 2 9 ( 3 )A 6 3 1 2 2 8 3 7 . 7 9 3 7 ( 7 ) 2 9 6 9 . 3 0 3 3 ( 6 )we 3 2 3 2 ( 2 ) 3 2 8 1 ( 2 )we 9 8 ( l ) 7 8 ( l )Be 16.681963(E) 16.666970( 7)*e 0.712680(35) O-647567(31)Ye -0.016160(24) 0.000365(20)

lOX0, 17.822(28) 17.1431(31)10*x6, 0.146(31) -0.2309(36)lO%, 17.78P 17.195aA, -34.6036( 23) _--A - 0.0314(26) _--lO%i, 11.6(E) 12.7(3)loxa - 1.6(E) -0.7(3)108XHe 11.6b 12.6bpe 1.036751 1.03722A

Note. Estimated one standard deviation error in parentheses.a Computed using the Kratzer relationship.b Computed using the Dunham-Birge expression.

state is equal and opposite in sign to p for the 311 state. For NH in the 2, = 0 levelsthe values are yz = -0.0548 andp = 0.0552.Expressions for o, p, and 4 in terms of matrix elements connecting electronic statesare given by Brown and Merer (76) [Refs. (81,82) are also helpful]. The ground stateof NH has a 1$2$3$1** configuration and the AII state is 1~?2~?3als~. In orderto evaluate the interaction matrix elements a total of four electrons in 3a and l?rorbitals must be considered. The matrix elements can be evaluated if 3a and 1~ areconsidered to be nitrogen atomic 2p, and 2p,, orbitals, respectively. The Slater de-terminant representations of the 311 and 32- components are given in Table XIV.The required matrix elements were evaluated using the determinants of Table XIVand the pure precession hypothesis (80) with i = 1. The results are

p = 2J-B/(En - Ez) = -4AB/(En - Ez)q = -4B2/(En - Ex)

where {is the atomic spin-orbit coupling parameter as defined by Tinkham (83). Themolecular and atomic spin-orbit parameters are related by A = -l/2. As a test of thepure precession hypothesis, the atomic t = 73.3 cm- (84) results in a predicted A of-36.6 cm- compared to the observed A0 = -34.6 cm-. The pure precession po andq. values are 0.0759 and -0.0358 cm- compared to the observed 0.0552 and -0.03 16


19/22

AII-XZ- TRANSITION OF NH 399TABLE XIV

Slater Determinant Representation of the Electronic Structure of AII and XZ- States of NH1%2> l//T (lo*+ i 1_ ISI TY+w_Y_I,13E1> = l/2 ~~~~+~+~_l~la~+~+~_l,~~l~~+~_~_l+l~~+~_~_l~~fn(J>= - --lm- ( .n+~+*_l*a+~_~_l)j3P1>= ---l/n- (IUZ ~+~_(*I.0 nt*_l )I%-ge> = w-2- (IG~+~_~+~oTn+~_~)

Note. The upper (lower) sign refers to e(f) parity.

cm-, respectively. The agreement for A and q is very good but p is predicted muchmore poorly. Considering that only two well-separated electronic states were considered,Franck-Condon factors were taken as 6,t and the extreme nature of the basic pureprecession assumption of 1 = 1, the overall agreement is satisfactory.For a II state interacting with a single Z- state, the spin-rotation and A-doublingparameters are related by y * pn/2 (76). The NH A311state has p0/2 = 0.0276 cm-compared with the observed 0.0297 cm-. This relationship neglects interactions withother electronic states (3A aEcts y but not p) and first-order contributions to y (thetrue spin-rotation interaction). Also since AD and y are badly correlated, our assumedvalue of AD also a&cts y.

There is an additional A-doubling parameter for a 311state, o. This contains a first-order contribution from the spin-spin interaction as well as second-order effects throughspin-orbit coupling. Horani et al. (85) have investigated this for a series of moleculesisovalent with NH. For those species where the spin-orbit interaction is large thesecond-order effects dominate, while for cases like NH where A is small the spin-spininteraction is most important. They calculated the spin-spin contribution to be 1.4cm- with small negative terms from the spin-orbit interaction, bringing the valueclose to the measured one.

The first centrifugal distortion correction to the Adoubling pD and qD can be esti-mated using the formulae due to Veseth (79):

po=2p A$?_;( 1

&=-4qS .0AD was not determined but the calculated value gives a negligible contribution. Fromthe equations the calculated values for 0 = 0 are

pD=-1.2x 1o-5 cf. -1.8 X 1O-5qo= 1.39 x 1o-5 cf. 1.38 X lo+.


20/22

400 BRAZIER, RAM, AND BERNATHThe agreement forp, is reasonable while the value for qD s so close that the agreementmust be considered fortuitous.

The correction to the spin-rotation interaction can also be estimated using theequivalent formula from Veseth (79):yD=2y $-; .( 1

Again neglecting the hrst term, the following result is obtained:yD = -6.5 x lo-5 cf. -5.4 x 10-5.

CONCLUSIONThe A311-X32- transition of NH has been analyzed with much higher precision

than in previous work. A total of seven vibrational bands were simultaneously fittedyielding 78 molecular constants. The improved data should prove useful to the manyworkers studying the NH molecule.

ACKNOWLEDGMENTSWe are grateful for the expert technical assistance of Rob Hubbard in acquiring the NH spectrum. Wethank Jim Brault for the helium spectra employed to calibrate our data, and his interest in this work Tbis

work was supported by funding from the Gfhce of Naval R e s e a r c h N O 0014-84-K-O 122).RECEI VED: April 3, 1986

1. J . M. EDER ,De&.&. Wien. Akad. 60, 1 (1893).2. G. FUNKE, Z. Phys. %,787-798 (1935); 101, 104-l 12 (1936).3. R.N. DIXON, Cunud. J Phys. 37, 1171-1186 (1959).4. T. MURAIAND M. SHIMAU~HI,Sci. Light (Tokyo) 15,48-67 (1966).5. J. MALICET,J. BRION,AND H. GIJENEBAUT, . Chim. Phys. 67, 25-30 (1970).6. M. SHIMAUCHI,ci. Light (Tokyo) 15, 161-165 (1966); 16, 185-190 (1967).7. P . BOLLMARK, . KOPP, AND B. RYDH, J. Mot. Spectrosc. 34,487-499 (1970).8. W. UBACHS,J. J. TER MEULEN, AND A. DYMANUS, Cm&. .f.Phys. 62, 1374-1385 (1984).9. H. E. RADFORDAND M. M. LITVAK ,Chem. Phys. Lett. 34,561-564 (1975).

10 . F. D. WAYNEAND H. E. RADFORD,Mol. Phys. 32,1407-1422 (1976).Il. F. C. VAN DE RHEUVEL,W. L. MEERTS,AND A. DYMAWS, Chem. Phys. Lat. 92,215-218 (1982).12. P. F. BERNATHAND T. AMANO, J. Mol. Spectrosc. 95, 359-364 (1982).13. D. E . MILLIGANAND M. E. J ACOX,J. Chem. Phys. 41,2838-2841 (1964).14. H. SAKAI,P . HANSEN,M. E~PL IN, . J OHANSSON, . F%LTOLA, ND J . STRONG,A&. Opt. 21,228-

234 (1982).15. B. D. GREEN AND G. E. CALEDONI A, . Chem. Phys. 77,3821-3823 (1982).16. M. VERVLOETAND J. BRION,p e r s o n a lcommunication.17. D. BOUDJAADAR,. CHOLLET$NDG. GUELACHVILI,o r t i e t hSymposium on Molecular Spectroscopy,June, 1985, Columbus, Ohio.18. R. W. B. WE, Proc.R. Sot. London, Ser. A 143, 112-123 (1933).19. G. H. DIEKEAM) R. W. BLUE,Phys. Rev. 45,395-400 (1934).20. G. NAKAMURAAND T. SHIDEI, apan. J. Phys. 10.5-10 (1935).21. W. Y. CHEUNG,B. GELER NT, ND T. CARRINGTON, hem. Phys. L.&t. f&,287-290 (1979).

R E F E R E N C E S


21/22

AII-XZ- TRANSITION OF NH 4 0 122. D. RAMSAYAND P . SARR E, . Mol. Spectrosc. 93,4454l6 (1982).23. W. UBACHS,G. MEYER,J. J. TE RME~LEN,AN D A. DYMANUS, . Mol. Sctrosc. 115,88-104 (1986).24. R. S. RAMAND P . F. BER NATH, . Opt. Sot. Amer. Opt. Phys. B. 3, 1170-l 174 (1986).25. R. W. LUNT, R. W. B. P EARSE, NDE. C. W. SMITH ,P r o c .R. Sot. London, Ser. A 151,602-609 (1935).26. F. L. WHITTAKER, . P hy s. B. 1,977-982 (1968).27. R. W. L I J N T ,R. W. B. PE ARCE , NDE. C. W. SMITH ,Proc. R. Sot. London,Ser. A 155,173-182 (1936).28. F. L. WHIITAKER, r oc. P h ys. Sot. L on do n 90,535-541 (1967).29. G. KR ISH NAMUR ~ AND W. A. NARASIMH AM,I. Mol. Spectrosc. 29, 410-4 14 (1969); Proc. Indian

Acad. Sci., Sect. A 64,97-l 10 (1966).30. W. R. M. GRAHAM AND H. LEW, Canad. J. Phys. 56,85-99 (1978).31. J . L. HALL,H. ADAMS, . V. V. KASP ER , . F. CUR L,ANDF. K. TILL, J . Opt. Sot. Amer. Opt. Phys.

B: 2,781-785 (1985).32. K. R. LEOPOLD, . M. EVENSO N, NDJ . M. BROWN,J. Chem. Phys. 85,324-330 (1986).33. J. MASANET,A. GILL =, AND C. VER MEI L, . Photochem. 3,417-429 (1974).34. P. C. ENGELKING ND W. C. LINE BER GER ,. Chem. Phys. 65,43234324 (1976).35. D. COGSART,. Chim. Phys. 76, 1045-1050 (1979).36. F. R O H R E RAND F. STUHL ,Chem. Phys. Lett. 111,234-237 (1984).37. A. HOFZUMAHAUSAND F. STUHL,J. Chem. Phys. 82, 3152-3159 (1985); 855519-5526 (1985); H.HAAKAND F. STUHL,J. Phys. Chem. 88,3627-3633; 88,2201-2204 (1984); U. BLUMENSTEIN,.

R O H R E R ,AND F. STIJ HL,Chem. Phys. L&t. 107,347-350 (1984).38. B. GELE RNT, . V. F ILSETH, ND T. CARRINGTON, hem. Phys. Lett. 36,238-241 (1975).39. J. MASANET,C. LALO,G. DURAND,AND C. VERME IL,Chem. Phys. 33, 123-130 (1978).40. Y. MARWAMA, T. HIKIDA,ANDY. MORI,Chem. Phys. Lett. 116,371-373 (1985).41. F. ALBER TI ND A. DO UGLAS,Chem. Phys. 34,399-402 (1978).42. E. N. TERESHCHENKOND N. YA. D ~DON OVA,Opt. Spectrosc. 47,339-340 (1979).43. A. M. QUINTONAND J . P . SIMONS,Chem. Phys. Lett. 81,214-217 (1981).44. N. WASHIDA,G. INOUE,M. SUZUKI,AND 0. KAJ IMOTO,Chem. Phys. Lett. 114,274-278 (1985).45. C. HE LLN ER , . T. V. GR A~AN, ANDM. H. R. HUTCHINSON,. Chem. Phys. 81,4389-4395 (1984).46. I. TOKUEAND Y. ITO , Chem. Phys. 79, 383-389 (1983); 89, 51-57 (1984); I. TOKUEAND M. IWAI,

Chem. Phys. 52,47-53 (1980).47. P. G. CARRICK ND P . C. ENGELKING, hem. Phys. Lett. 108,505-508 (1984).48. J. M. LENTS,J. Quant. Spectrosc. Radiat. Transfer 13,297-310 (1973).49. J. A. BARRI NGTON, . P. MODICA,AND D. R. LIB BY, . Quant. Spectrosc. Radiat. Transfer 6, 799-

805 (1966).50. P. W. F AIRCH ILD, . P . SMITH,D. R. C ROSLE Y, ND J . B. J EF FR IES,hem. Phys. Lett. 107, 18 -186

(1984).51. N. L. GARLAND, . B. J EF FR IES,. R. CROS LEY, . P. SMITH,AND R. A. COP ELAND,. Chem. Phys.

85,4970-4975 (1986).52. W. H. SMITH,J. BRZOZOWSKI, ND P. ERMAN,J. Chem. Phys. 64,4628 (1976).53. L. ZETZXH, Ber. Bunsenges. Phys. Che m. 85639-643 (1978).54. S. N. F O N E RAND R. L. HUDSO N, . Chem. Phys. 74,5017-5021 (1981).55. E. A. .%ARL AND F. W. DALBY, Ca n u d. J. Phys. 52, 1429-1437 (1974).56. S. T. GIBSON, . P. G R E E N E ,ANDJ . BEFXO~Z, J. Chem. Ph ys. 83,4319-4328 (1985).57. J. KOUBAAND Y. &RN , J. Chem. Phys. 52,5387-5394 (1970).58. P. J. H AY AND T. H. DUNNING, R.,J. Chem. Phys. 64, 5077-5087 (1976).59. W. MEYERAND P . R OSMVS, . Chem. Phys. 63,2356-2375 (1975).60. B. S. HAYNES,Cornbust. Flame 28,81-91 (1977).61. J. A. MILLER,Combust. FIame43,81-98 (1981).62. W. R. ANDERSON, . J. DECKER,AND A. J . KOTLAR, Cornbust. Flu me 48, 179-190 (1982).63. M. S. CH OU, A. M. DE AN, AND D. STE RN , . Chem. Phys. 76,5334-5340 (1982).64. A. FOWLERAND C. C. L. GRE GOR Y,Philos. Trans. R. Sot. London, Ser. A 218,351-372 (1919).65. R. W. SHAW,Astrophys. J. 83,225-237 (1936).66. J. L. !&%MI?T,Publ. Astron. Sot. Pac. 81,657-664 (1969).67. P. SWINGS,C. T. ELVEY,AND H. W. BABCOCK,Astrophys. J. 94,320-343 (1941).68. M. M. LITVAKAND E. N. R . KUIP ER , strophys. J. 253,622-633 (1982).


22/22

4 0 2 BRAZIER, RAM, AND BERNATH69. D. L. LAMBER T, . A. BRO WN,K. H. HINKLE,ANDH. R. J OHNSON, s t r oph ys. . 284,223-237 (1984).70. S. T. RIDGWAY,D. F. CARBON, D. N . B . HALL, AND J . J EWELL, m oph ys. . I SuppI. 54, 177-210

(1984).71. W. C. MARTIN, . Res. Na t. Bur. Stand., Sect. A 64, 19-28 ( 1960).72. G. NORLEN,Phy sica Scripta 8,249-268 (1973).73. G. HERZBERG,S p e c t r a of Diatomic Molecules, 2nd ed., Van Nostr an d-Reinhold, New York, 1950.74. R. N . ZARE, A. L. SCHMELTEKOPF,. J . HAmop , AND D. L. ALBRIITON, . Mol. Spectrosc.46,37-

66 (1973).75. F. ROUX, D. CORNY, AND J . VERGES,T.Mol. Spectrosc.94, 302-308 (1982).76. J. M. BROWNAND A. J . MER ER , . Mol . Spect rosc.74,488-494 (1979).77. J . M. BROWNAND J . K. G . WATSON, . Mol. Spectrosc.65,65-74 (1977).78. T. C. STEIMLE,C. R. BR AZIER , ND J . M. BRO WN, . Mol . Spect rosc.110,39-52 (1985).79. L. VESETH, . Phy s. B3,1677-1691 (1970).80. R.S.MULLIKENANDA.CHRISTY,P~~~. Rev.38,87-119(1931).81 . A. J . MERE R,D. N. MALM, R. W. M ARTIN ,M. HORANI,AND J . R OSTAS,Canad. J. Phy s. 53, 251-283 (1975).82. B. G. WICKE,R. W. FI ELD,AND W. KLEMP ERE R,. Chem. Phy s. 56,5758-5770 (1972).83. M. TINKHAM, Group Theory an d Qua nt um Mechanics, McGraw-Hill, New York, 1964.84. C. E. MOORE, Atomic En ergy Levels, Vol. 1, NSRDS-NBS 35, Na tional Bu rea u of Standa rds, Wash -

ingt on, DC., 1971.85. M. HORAN& . ROSTAS,AND H. LEFEBVRE- BRI ON,anad. J. Phys. 45,3319-3331 (1967).

Documents

C.R. Brazier et al- Fourier Transform Spectroscopy of the A^3-Pi-X^3-Sigma^- Transition of NH