ELECTRONIC STATES OF NITRITEION
June 5 , 1957
CROSS-SECTIONS AND
TABLE VI11 RATECONSTANTS FOR THERMAL REACTIONS Extrapolated experimental
Theoretical (ref. 13)
ka$a°K.
k
(298OF.) cm. Reaction
X 1016
(cm.a/ moleculesec.) X 109
(cm.3,' moleculesec.) X 109
CzHn+ CnHz + (C4H4 +) Cz+ CzH2 + C4H+ f H CH+ CnHz -* (C3H3')
251 256 282
2.1 2.1 2.5
1.3 1.3 1.6
f.
+ + +
previously.' In studying reaction rates a Consolidated Electrodynamics Corporation (CEC) Model 21-620 cycloidal focusing mass spectrometer was used. In calibrating the ion source for pressure, the ionization cross section of acetylcm.l in reasonably ene was determined as 5.60 X cm,* reported good agreement with the value of 4.98 X
[CONTRIBUTION FROM
THE
2669
by Massey and Burhop.14 Appearance potentials were measured on a Westinghouse Type LV mass spectrometer. The acetylene used in these studies was commercial tank material obtained from Matheson Chemical Company and purified by repeated condensation a t liquid nitrogen temperatures. The 1,3-butadiene was Phillips research grade having a stated purity of 99.51% and the 1-butyne was an API standard sample of 99.87% purity. Both were used without further purification. The vinylacetylene was obtained from the du Pont Company and was used without treatment since a mass spectrum indicated it to be of satisfactory purity.
Acknowledgment.-We wish to express our appreciation to Mr. B. L. Clark for his help in making the measurements and calculations. (14) H. S. W. Massey and E. H. S. Burhop, "Electronic and Ionic I m p a c t Phenomena," Oxford University Press, London, 1952, p. 265,
BAYTOWN, TEXAS
DEPARTMENT O F CHEMISTRY, THEUNIVERSITY O F ROCHESTER]
Electronic and Vibrational States of the Nitrite Ion. I. Electronic States1 BY JEROME W. SIDMAN~ RECEIVED AUGUST16, 1956 The lowest absorption transition of the nitrite ion has been studied in the crystalline state a t 77'K. and a t 4"K., using single crystals of NaNL4O2and N a h P 0 2in polarized light. The corresponding fluorescence transition also has been studied. The spectra are sharp, and can be analyzed in considerable detail to yield vibrational and lattice frequencies for both the ground and excited electronic states. From the intensity dichroism and vibrational structure, it is concluded that the for NaNL5O2 a t 4°K.) is allowed by weak absorption transition of NaNI4Ozwith origin a t 25977 cm.-' a t 4°K. (25977 symmetry and is polarized perpendicular to the plane of the NOz- ion ('Bz -+ IAI). Comparisons with theory and with previous work are made wherever possible. The effect of environment on the electronic spectrum of NOz- is discussed. The NOz- spectra also are compared with the spectra of other 18-valence electron molecules, such as 0 3 , Son, HCOz-, ONC1, etc. The simplest LCAO-MO treatment appears to give a qualitatively correct description of the excited electronic states of NO*-
Introduction Within the last few years, considerable progress has been made toward understanding the electronic spectra of polyatomic molecules. Although much attention has been given to large aromatic hydrocarbons, a survey of the literature reveals that relatively few triatomic molecular electronic spectra have been completely analyzed. The vapor spectra of triatomic molecules are sometimes sharp enough to permit analysis of the rotational fine structure, from which the transition moment direction and the geometry change can be deduced. However, even in such cases, there may be so many closely spaced bands present in the spectrum that the vibrational analysis is made difficult. Furthermore, if the molecule is very different from a symmetric top, the rotational analysis becomes quite difficult. I n the cases in which the bands are broadened by predissociation, less information can be obtained. The study of crystal and mixed crystal spectra has provided exact information about symmetry properties and vibrational structure of the electronic transitions of large, complex molecules. In this paper and in the following one, the results of such a (1) This research was generously supported by t h e Office of Ordnance Research of t h e United States Army, under Contract DA-30-115 ORD-728 with t h e University of Rochester. (2) Department of Theoretical Chemistry, Cambridge University, England. Post-doctoral fellow, 1955 t o 1956, under a Erant by t h e Shell Fellowship Committee t o the Department of Chemistry of the University of Rochester.
study are reported for the lowest singlet-singlet electronic transition of the N02- molecule-ion. Although it has been known since 1934 that the spectra of crystalline NaNOz and KN02 are sharp a t low t e m p e r a t ~ r e sa, ~complete analysis has not yet been given. The recent polarization studies of Trawick and Eberhardt4 have provided additional information, which, together with the vibrational analysis presented in this work, lead to conclusions about the nature of the ground and lowest excited electronic states of this molecule. These results should also be applicable to other related molecules which possess the same number of valence electrons and similar geometry, and comparisons and predictions will therefore be made where possible. Experimental Mallinckrodt Analytical Reagent Grade NaN02 was further purified by recrystallization from water. Single crystals of suitable thickness were prepared readily by crystallization from molten hTaiY02 between quartz disks. The experimental arrangement used to record the absorption and fluorescence spectra of single crystals a t low temperature has been described in a previous publication . 5 In this work, a tungsten filament lamp was the source for the absorption experiments, and a medium-pressure Hg arc with a Corning 5860 filter was used to excite the fluorescence of NaNOz. All spectra were recorded on Kodak (3) G. Rodloff, 2. P h y s i k , 91, 511 (1934). (4) W. G. Trawick and W. H. Eberhardt, J . Chem. Phys., 22, 1462 (1954). ( 5 ) J. W. Sidman, THISJOURIAL, 7 8 , 4217 (19%).
2670
JEROME
W. SIDMAN
103-0 or 103-a-F plates. Fe arc spectra were used to calibrate the wave lengths. The NahP02 was prepared by reducing fused NaiVOs with powdered Pb.6.7
Results The polarized absorption and fluoresence spectra of crystalline NaN02a t 77'K. and a t 4'K. are shown in Figs. 1 and 2. A summary of the results of the vibrational analyses, which are further discussed in section (R), is given in Table I.
3
?,un:'
Fig. 1.-Microphotometer tracing of the absorption spectrum of N a P 4 02, tungsten strip filament lamp source: top, polarized spectrum of single crystal, 77"K., ac-plane; bottom, ac-plane, 4OK.
--+
9,cm:'
Fig. 2.-Microphotometer tracing of the fluorescence spectrum of NaS1402,uc-plane, 4OK. Lines in the source
obscure the 0 4 band. J. Turner, J . Soc. Chem. Ind., 34, 585 (1915). ( 7 ) HN'jOa was obtained from t h e Isomet Corporation, 118 Union Street, Palisades Park, N. J. T h e isotopic concentration of hY16was ((i)
99.5%.
Vol. 79
TABLE I SUMMARY OF THE RESULTS OF THE VIBRATIONAL ASALYSES OF THE ~ B z 'A1 ABSORPTION AND FLUORESCEXE TRASSITIOSS OF S a S i 4 0 2 A N D XaS15O2AT 77 A N D AT 4°K. +
Frequencies are in cm. -1. sa~l4o2
0-0, 'Bz 'Ai, '7'7°K. 0-0, 'Ba 'AI, 4°K. Vibrational frequencies in 'AI, 4°K. Vibrational frequencies in 'Bz, 4°K. Lattice frequencies in 'AI, 4'K. (&15cm.-l) Lattice frequencies in 'Bp, 4OK. (A15 cm.-l) +
+
25960 =I=S 25077 =I= 3 P I = 1325 i 4 V ? = 839 3~ 2 ~1 = 1018 =k 4 1'2 = 632 zt 4 45, 105, 165, 205, 270 65, 110, 155, 180, 218
NaNlsOi
25960 =I= 8 25977 zk 3 V I = 1306 & 4 vz 824 i 2 V I = 1006 i- 4 v p = 621 =I= 4 45, go(?), 105, 165, 205, 270 65, 110, 155, 180, 218
Discussion and Interpretation (A) Assignment of the Electronic Transitions.The determination of the symmetry properties of the combining states in an electronic transition of a polyatomic molecule is as important as the determination of the energy and transition probability. Until recently, the direct determination of this property has been quite difficult. It has often been difficult or impossible to determine correctly the electronic symmetry properties from polarized spectra of single crystals, due to intermolecular coupling between identical molecules in the c r y ~ t a l . ~Such ~ ~ effects are absent in the polarized absorption spectra of dilute mixed crystals, and considerable data are now available for the electronic assignments of the aromatic hydrocarbons from mixed crystal studies.I0 The crystals which have been studied previously are molecular crystals, in which the molecule and its nearest neighbors are identical. NaNOz is an ionic crystal, and the nearest neighbors of the NOZ- ions are Na+ ions, which shield the NO*- ions from each other. Consequently, effects due to intermolecular coupling, such as dichroic splitting oi the 0-0 band and/or formation of trapped exciton^,^^" should be much less prominent in NaNO2, which may be regarded as a concentrated (50%) mixed crystal in which there is order with respect to the position and orientation of both the solute (NO?-) and solvent (Ka+). Trawick and Eberhardt4 have studied the polarization properties of the electronic transitions of single crystals of NaXOn a t 77'K. They have found that the origin of the lowest absorption band a t X 3Xjl.4 A. is polarized perpendicular to the plane of the NOz- ion. The results of this investigation are in agreement with Trawick and Eberhardt's work. However, thicker crystals a t wave lengths shorter than X 3531.4 A. show that there is weak but definite absorption for light polarized in the plane of the NO2- ion (Fig. 1). This will be discussed in further detail in section (C). (8) D. S. XcClure and 0. Schnepp, J . Chem. Phys., 23, 1575 (19.55). (9) J. W. Sidman, P h r s . Rcv., 102, 96 (1956). (10) J. W. Sidman and D. S. McClure. J . Chem. Phys., 24, 757 (1956); J. W, Sidman, ibid., 25, 115, 122 (1956). (11) J. W. Sidman, THISJOURNAL, 79, 305 (1957).
ELECTRONIC STATES OF
June 5 , 1957
267 1
NITRITE I O N
pg
+.
The polarization of the crystal absorption spec2 trum does not in itself lead t o an unambiguous electronic assignment, since it is possible that the electronic transition might be forbidden by symmetry and that its appearance might be due to a viX brational or to a lattice perturbation, or both. I-Iowever, an unambiguous determination often can be c .N CRYSTAL ::c Imm made if the same electronic transition appears in a ~ 3 . 5 5 A fluorescence. Excitation of crystalline NaNOz with b=5.56ff the X 3660 A. Hg group produces a weak but detectable fluorescence spectrum, which nearly is the C = S.30h L. mirror image of the absorption spectrum.12 Although lines from the Hg arc interfere in the region near X 3851.4 A,, the vibrational analyses of the absorption and the fluorescence spectra leave no doubt that the lowest absorption band a t X 3851.4 %A, A. is the 0-0 band of an electronic transition Fig. 3.-Crystal structure of PiaSOz and molecular strucwhich is allowed by symmetry and which is polarized perpendicular to the plane of the NOz- ion. ture of NOz-. The clear circles are a t a = 0, whereas the Since the molecular point group, molecular site dark circles are displaced by a/2. The crystallographic axes group in the crystal, and crystallographic space are labelled according to Ziegler's n0tati0n.l~ The bond group are all C ~ Vthe , polarization results of the distance and bond angle of ?io*-are from Carpenter's crystal spectra should be directly applicable to the work." free ion. the n- and n-orbitals of the NOz- molecule is shown The crystal structure of NaNOz has been ex- in Fig. 4, and orbital assignments of the three electensively and carefully investigated.l3-I5 A dia- tronic transitions which oare observed a t wave gram of the crystal and molecular structure is lengths longer than 2000 A. are also shown in the shown in Fig. 3. The crystallographic axes are diagram." It is assumed that the a-bonding orbilabeled according to Ziegler's n0tati0n.l~ If a totally-symmetric electronic ground state is assumed, AI), the polarization results lead to the 'Bz + 'A1 electronic assignment for the lowest absorption transition of N02-.16 The orbital assignment proposed by Trawick and Eberhardt is n3, which is in agreement with the fact that nN the transition is allowed by symmetry b9t is nevertheless weak (Emax 25, Xmax 2 3550 A."). The no + n* transitions of carbonyl compounds are also weak, even though the transition may be formally allowed by symmetry, as in glyoxal1*and bia~ety1.l~This is to be expected, since the n- and r-orbitals do not simultaneously possess large amplitude in the same regions of space, so that the integrand in the transition moment integral J(n) *R(n)dT is everywhere small, and consequently leads n, to low intensity for n + IT*transitions. Furthermore, n -+ n* transitions which are allowed by symmetry possess a transition moment which is necessarily perpendicular to the nodal plane of the n-orbitals, since the n-orbitals and the n-orbitals Fig. 4.-Schematic diagram of the one-electron n- and are, respectively, symmetric and antisymmetric mnolecular orbitals of SOe-. The orbital transitions which with respect to reflection in the plane of the mole- may correspond t o the observed electronic transitions of cule. This has been verified experimentally in gly- KO2- are shown in the diagram. oxal from the analysis of the rotational structure.l8 Thus, the IB2 + AI, nx n3 assignment for the tals are lower than X I , and the a*-antibonding orlowest transition of NO?- appears to be reasonable. bitals are higher than n3. This is reasonable, since A qualitative orbital energy level diagram for the hybridized AOs which form the a-MOs overlap with each other more than the p-AOs which form (12) J. W. Sidman, THISJOURNAL, 78, 2911 (1956). (13) G. E . Ziegler, Phys. Rev., 58, 1040 (1931). the a-MOs, so that the energy difference between (14) G. B. Carpenter, Acta Cryst., 5, 132 (1952). the u- and a*-MOs is greater than the energy dif(15) hI. R. Truter, ibid., 7 , 73 (1954). ference between the n-and n*-MOs. The n-MOs are (16) T h e group-theoretical notation is t h a t of H. Eyring, J. Walter and G. E. Kimball. "Quantum Chemistry," John Wiley and Sons, essentially AOs (which may be hybridized), since New York, N. Y.,1943. they are formed from orbitals which overlap very (17) H. L. Friedman, J . Chem. Phys., 21, 319 (1953). little with the u-orbitals. Since 0 is more electro(18) J. C. D. Brand, Tyans. Faraday Soc., SO, 431 (1954). negative than N, the no orbitals are placed slightly (19) J. W. Sidman a n d D. S. hfcClure, THISJ O U R N A L , 77, 6461, 6471 (1955). below the nN orbital4. The two no orbitals should
-
=
8
-
-
,
2672
JEROME
W. SIDMAN
VOl. 79
be nearly degenerate, since the overlap of the XOs bations by the environment. The full power of the from which they are formed is so small. vibrational fine structure analysis in elucidating the molecular and electronic structure can be felt The no x3 transition(s) of r\'02-ohas been previously assigned4 to the Xmax % 3000 A., Emax = if the molecule is small and if isotopically substi1 3 band." The present work has not studied wave tuted species are available. The recent availalengths shorter than 3300 A., due to the complete bilityz0of isotopically pure N1j has enabled such a absorption by the double Pyrex glass Dewar ves- study to be performed for NOz-. The lattice fine structure which appears in the sels a t the shorter wave lengths. An examination of Rodloff's spectral tracings3 revgals that the electronic transitions of molecules in crystals and bands which are observed pear 3200 A. are broader mixed crystals provides information about lattice than the bands near 3500 X., but there is no defi- oscillations, and should be valuable in understandnite evidence that they belong to another electronic ing intermolecular forces in the solid state. So far, transition. The broadening of the bands which cor- extensive interpretation of the lattice structure respond to higher vibrational transition may be due has not been made, although lattice frequencies in to perturbations by the no -+ 7 3 , B ' Z +- lAl transi- many molecular crystals and mixed crystals have tion, as Trawick and Eberhardt have s ~ g g e s t e d . ~been determined from the analysis of the elecThe absence of the no -+ x3 transition in the crystal tronic transitions. lo The analysis of the fine structure of the n N -+ x3, spectrum is difficult to understand, but possibly it absorption transition of NOz- in cryscould be explained if the upper state were unstable. IB2 + Absorption due to this transition would then be talline NaN1400?and n'aS150za t 4'K. is quite nearly continuous, and would not appear unless straightforward. At T K . , the 0-0 band in both 8 cm.-l, whereas the crystal thickness were much greater than that NaNI4O2and SaX1j02is 25960 which is necessary for detecting the very sharp a t 4'K. the 0-0 band has shifted slightly to 25977 3 em.-' in both cases. It is possible that there is n~ --t ~3 transition. The intensity due to a very , ~ the diffuse transition a t any individual wave length a calibration error in Rodloff's ~ o r k since would thus be very much less than the intensity work of Trawick and Eberhardt and of the present due to a sharp transition in the same region, even author are in agreement with respect to the 0-0 though the integrated intensities over the corre- frequency a t 77'K. The most prominent feature sponding regions of absorption are similar, as the of the absorption transition in NaN1+02is the prosolution spectra suggest." This interpretation gression of a single frequency of ( 3 2 f 4 c m - ' ((321 k 3 cm.-' in Na?J1jO2)in which the intensity must be considered tentative. S t shorter wave lengths, a more intense band of maximum is in the fourth band of the progression. This is seen most readily in the microphotometer SOz-ois found, both in aqueous solution (Amax 2100 A, emax E 5500") and in crystalline NaNOz4. tracings in Figs. 2 and 5 of ref. 3, in which the The polarization studies and the moderately high in- source is a hydrogen discharge tube, which pro, 'A1 vides a more uniform background than the tungsten tensity are in agreement with the 7r2 ~ 3 'BI+ assignment suggested by Trawick and Eberhardk4 filament lamp employed in this research. Besides Friedman1' considers this transition to be due to an the main progression, additions of a frequency oi intermolecular electron-transfer transition from 1018 + 4 em.-' (1006 =t 4 em.-' in SaN1jOn)to S O z - to H20. The fact that the energy of this each of the bands in the main progression are also transition is only slightly different in the crystal seen in the spectrum. Rodloff has previously idenDI, etc., in from the value observed in aqueous solution is tified the main progression ( 9 1 , BI, CI, Fig. 2 , ref. 3 ) , but he has not identified the 101s strong evidence against Friedman's interpretation. The simple orbital diagram in Fig. 4 thus ap- cni.-] interval. It is clearly seen in his tracing, Bj; BI,Cs; C1, D6; etc., in Fig. 2 of pears to be successful in accounting for the polari- however (AI, zation properties and the order of the excited ref. 3). It is clear that this identification is corsinglet states of Not-. In section (D) of this paper, rect, and that B j represents a vibrational addition comparisons between the spectra of NOz- and other to A1 and not to B1, since there is no band between A1 and B1 which corresponds to band B5 between B1 isovalent molecules will be given. (B) Analysis of the Vibrational and Lattice and C1. The analysis of the lattice structure in the abStructure.-In an electronic transition of a poly'itomic molecule in the crystalline state, fine struc- sorption spectrum yields the values G5, 110, 155, ture due to transitions involving simultaneous 180 and 218 cm.-' for the lattice frequencies of Xachanges in the vibrational and lattice states usually NOz in IBz a t 4°K. Due to the greater diffuseness appears in the spectrum. In order to resolve this of the lattice additions, the lattice frequencies may structure, it is desirable to study the spectra a t be uncertain by 1 1 3 ern-'. Within this limit of very low temperatures. If the vibrational fine error, the lattice frequencies are indistinguishable and ~ in LTaN1j02. structure can be analyzed, it is sometimes possible in S Z N ' ~ O The analysis of the fluorescence transition is also to determine how the arrangement of the nuclei and the bonding between them differs in the upper quite straightforward. The main progression in and lower electronic states. This leads to an un- NaN1*02is 529 + 2 em.-' (823 j= 2 em.-' in derstanding of valence and molecular structure Na,"v"jOz). The 1323 + 4 c1n-I frequency in l-11 which is deeper than that obtained either from the of NaNl4Ozcorresponds to 1018 f 4 em.-' in IBz. study of the vibrational spectrum of the ground Virtually all of the intensity in the fluorescence electronic state alone or from the study of an elee- transition is contained in the bands of the 829 Plrvs., 24, G 2 G (1R.K). tronic spectrum which is broad because of pertur(201 11.. Syiridei a n d T . I . Taylor, .i.
-
N
*
-
June 5, 1957
ELECTRONIC STATES OF NITRITEI O N
cm.-1 progression, in which the intensity maximum is in the fourth band. The bands corresponding to the 1325 cm.-' frequency and the lattice additions appear only very weakly in fluorescence. A correspondence between lattice frequencies in 'A1 and lB2 is also found. The lattice frequencies (*15 cm.-l) in 'A1 of NaN1402 and NaNl5OZa t 4OK. are 45, 105, 165, 205 and 270 cm.-'. I n addition, a lattice frequency of 80 cm.-' possibly appears in NaN1502. It is interesting t o observe that the highest frequency lattice mode in lBz and in 'A1 is different by an amount which far exceeds the experimental error. The absorption and fluorescence spectra, although very similar, do show some differences, so that they are not quite "mirror images'' of each other. The lattice structure is much more prominent in absorption than in fluorescence, as can be seen from an examination of Figs. 1 and 2. It is interesting to note that the bands of the 632 cm.-' progression (in absorption) are accompanied by very pronounced lattice structure, but that the lattice structure is far less pronounced, or is even absent, in the 1018 cm.-' additions to the bands of the 632 cm.-' progression. The detailed assignments of the vibrational states of NaNOz and the calculation of the force constants of NOz- are discussed in detail in the following paper.21 A summary of the analyses of the absorption and fluorescence spectra is given in Table I. Anharmonic corrections in the u~ progression are negligible (
