Homonuclear N2 has no permanent dipole-moment, which excludes the infrared detection
of transitions between ground-state vibrational and rotational levels. The first electronically- excited level occurs 50 000 cm−1 (6.2 eV) above the ground-state. However, neither this nor many other higher-lying states are accessible from the ground-state by electric-dipole transitions because of various symmetry selection rules, and the entire visible-range spec- trum is suppressed. Nonetheless, the weak Lyman-Birge-Hopfield and Vegard-Kaplan band systems involving excited states a1Πg and A3Σ+u, respectively, have been observed in absorption from the ground state [111, 163, 166]. Further electronic states have been accessed by electron spectra [68, 76], and in some cases [75, 183], laser-based excitation of higher-lying states is facilitated by a preliminary excitation to the metastable state
a′′1Σ+g, which is located 98 840 cm−1 above the ground state.
Significant absorption is finally observed for energies of 100 000 cm−1 (12.40 eV) and above, in the XUV region. The dipole-accessible excited-states are discussed in Sec. 3.3 and are plotted with low resolution in Fig. 3.1. This shows an electron energy-loss cross-section of N2 between 96 780 and 177 440 cm−1, measured with sufficiently high scattering-energy
so as to approach an equivalent optical spectrum. This spectrum shows the onset of the dipole-allowed transitions, and resolves a progression of vibrational bands with erratically variable oscillator-strengths up to the beginning of the ionisation continuum.
Direct dissociation following an electric-dipole transition may occur 116 300 cm−1(14.42 eV) above the ground state, although predissociation occurs for even the lowest dipole-allowed excited states. The mechanism leading to predissociation requires the simultaneous pertur- bation of multiple zeroth-order electronic-states, discussed in Sec. 3.4, and is accidental in nature and highly erratic. The first ionisation limit is reached at 126 200 cm−1 (15.65 eV) and resonance structure of the ion has been observed [78], as well as super-excited neutral
Figure 3.1: Oscillator strengths of N2 determined from measurement of∼0◦ scattering of 8 keV
electrons. The resolution of the instrument is 0.01 eV FWHM (490 cm−1FWHM). After Chan
et al.[21].
levels which are significantly autoionised.
Despite the ubiquity of predissociation, the majority of rotational lines in the extreme ultra-violet have Doppler limited lineshapes. The room temperature Doppler width of
14N
2 is 0.23 cm−1FWHM at 100 000 cm−1 (2.9×10−5eV FWHM at 12.4 eV), an example
of a Doppler limited observed absorption cross section is shown in Fig. 3.2. This width may be reduced by cooling the N2 target gas by means of liquid nitrogen or supersonic
expansion. Such cooling will also suppress the appearance of high-rotation transitions, leading to decongested absorption spectra and simplifying the assignment of rotational lines.
Molecular nitrogen’s three isotopomers 14N2, 14N15N and 15N2 exist in the ratio 1 :
0.0073 : 0.000013 in the Earth’s atmosphere. The natural abundance of 14N15N is suf-
ficiently high that it is sometimes observed simultaneously with 14N
2 in high-pressure
experiments, and studies have also been conducted with enriched 15N2. A purified sample
of14N15N will rapidly gain an admixture of14N2and15N2when exposed to XUV radiation
because of its high dissociativity and the resultant association of nitrogen atoms.
The majority of experimental data concerning the dipole-allowed spectrum of N2 have
arisen from the direct measurement of photoabsorption and emission. All such experiments are seriously hampered by the need to generate and manipulate radiation in the XUV range, for which there is no known transparent medium suitable for constructing windows and beam splitters. For the strongest bands, low pressures of N2 are required in order to
avoid the saturation of absorption spectra, in which case the radiation source, target gas, and detector must be contained inside a common vacuum system.
§3.1 Overview of experimental techniques 55
Figure 3.2: Photoabsorption cross section showing the b1Π
u(v′= 5)←X1Σ+g(v′′= 0) bandhead
region. Measured in absorption during the KEK series of experiments, discussed in Chap. 4. The observed rotational lines have negligible natural width with respect to room temperature Doppler broadening of 0.23 cm−1FWHM, and the instrumental resolution, 0.65 cm−1FWHM. After Stark
et al.[155].
essary ultraviolet radiation from classical discharge lamps, most frequently employing the helium emission continuum, spanning 60−100 nm. Dispersion of this continuum was achieved by means of gratings, and its detection made photographically. Many of these experiments were capable of resolving individual rotational-transitions, and allowed for the qualitative observation of line strength and predissociation broadening. The utilisa- tion of synchrotron radiation in the mid 20th century provided a source of radiation with significantly improved brightness and spectral profile. Also, the photographic detection of ultraviolet light has been completely replaced by photosensitive electronics, enabling the quantitative determination of the strength and lineshapes of individual rotational lines. The spectral resolution of functioning grating spectrometers has remained largely static, or perhaps decreased, but has recently been surpassed by XUV Fourier-transform spec- trometry. This new technique also allows for an absolute wavelength calibration and rapid data acquisition.
Recently, frequency multiplied lasers have been developed that are capable of generat- ing XUV radiation with a very narrow frequency profile. When combined with a source of nitrogen cooled by supersonic expansion, these experiments have enabled the most precise measurement of natural linewidths. Laser-based radiation sources are frequently unstable so careful monitoring of the beam brightness must be made simultaneously with absorp- tion measurements. More seriously, these experiments may not be scanned over large wavelength ranges due to the complexity of the various optical elements.
Multi-photon excitation may preclude the need for XUV radiation by accessing high energy levels via an intermediary excitation. Additionally, these allow for the probing of optically-forbidden transitions and the attainment of sub-Doppler resolution.
Electron scattering experiments employing low or threshold collision energies are freed from the optical-dipole transition selection rules, and so allow the probing of transitions that are optically invisible. High-energy small-angle electron scattering, however, closely approximates the direct absorption of radiation and allows for simpler measurements of
equivalent photoabsorption cross sections. The principal disadvantage of electron scat- tering techniques lies in the poor energy discrimination of electron sources and analysers when compared with optical methods. However, the magnitude of electron excitation cross sections may be fairly reliably determined at low resolution; whereas, absolute absorption cross sections may only be determined from optical spectra if all resonant features are completely resolved, as discussed in Sec. 2.8.6.
Additional information regarding the decay mechanism of excited N2has been obtained
from photoabsorption and electron excitation experiments after detecting the subsequent fluorescence, or any ionisation and dissociation fragments.