Tentative identification of U93.174 as the molecular ion N2H/+/

Abstract

On the basis of a self-consistent field calculation of its rotation constant and hyperfine structure, it is suggested that the molecular ion N 2 H + is the carrier of the unidentified interstellar triplet reported by Turner in the preceding Letter. The calculated frequency of the/ = 1-> 0 transition of N 2 H + is 92.2 ± 1.8 GHz, in good agreement with the observed 93.1 GHz. Similarly, the calculated value of the quadrupole coupling constant for the outer nitrogen is-5.3 ± 0.5 MHz, while the coupling constant deduced from the observations is-5.7 MHz. Equally important to the assignment, the quadrupole coupling for the inner nitrogen is calculated to be only-1.2 ± 0.5 MHz, making its contribution to the hyperfine structure too small to be resolved in the sources so far observed. Detection of this additional structure in sources with very narrow lines would provide additional confirmation of the assignment. Subject headings: molecules, interstellar-radio lines Turner (1974) reported in the preceding Letter the discovery of three closely spaced new interstellar lines near 93.174 GHz. The structure of this triplet is characteristic of the hyperfine structure (hfs) produced by the electric quadrupole interaction of a single nitrogen nucleus in the / = 1-» 0 rotational transition of a closed-shell molecule, and is in fact virtually identical to that of the nearby 88.6-GHz / = 1-» 0 line of HCN. No molecule with a single nitrogen, however, has been found with the right spectroscopic constants to fit the observations. The purpose of this Letter is to point out that the molecular ion N 2 H + is an excellent candidate for the carrier of Turner's lines, although at first glance it might appear that the two nitrogen nuclei would give rise to more structure than observed. Simply by taking the bond lengths and hyperfine constants to be equal to those of similar molecules, one finds that the rotation constant and the hfs of the outer nitrogen are in substantial agreement with Turner's observations, whereas the hfs from the inner nitrogen is too small to have been resolved in the sources examined. These conclusions have been confirmed by an ab initio calculation of the structure of N 2 H + , putting the identification on a somewhat more secure quantitative foundation. 'NJ1 + is isoelectronic with N 2 , CO, HCN, HNC, and HCO-1-, and its electronic ground state is expected to be an extremely stable closed-shell configuration. An estimate of its rotation constant can be made by assuming that the length of the NN bond is the same as in N 2 + , about 1.11 Â, and that the length of the NH bond is about 1.00 Â. The result is vw = 2B = 92.2 GHz for the lowest-frequency rotational transition-in reasonably good agreement with Turner's frequency. Furthermore , the hfs can be estimated by comparison with the isoelectronic isomers HCN and HNC. For HCN the nitrogen hfs has been measured, and the coupling constant is eqQ =-4.709 MHz (DeLucia and Gordy 1969). Although eqQ has not been measured for HNC, an ab initio calculation (Pearson et al. 1973) predicts a coupling constant of only +0.93 MHz, with an estimated accuracy of better than 50 percent. Another estimate of the hfs for the outer and inner nitrogens can be obtained from measurements of CH3CN and CH3NC, which give eqQ =-4.4 MHz for the outer and \eqQ\ <0.5 MHz for the inner nitrogen (Townes and Schawlow 1955). To check this reasoning we have done an ab initio calculation of the structure of N 2 H +. The equilibrium geometry, the electric field gradients at the nuclei, and the permanent electric dipole moment have been calculated in the self-consistent field (SCF) approximation. The molecular wave function must be approximated by expansion in a finite basis set, and we have used Slater-type functions centered on the nuclei. It is easy to employ enough functions in a molecule as small as N 2 H+ to essentially reach the infinite-basis Hartree-Fock (HF) limit with current computational capability. Because the HF wave function is not exact, however, it is sometimes preferable to employ smaller basis sets and take advantage of the cancellation of errors known to occur when certain molecular parameters are calculated. The most accurate calculations of molecular geometry, for example, are obtained from a double-zeta basis set which contains two functions for each nl shell occupied in the separated atoms (Schaefer 1974; Pople 1973). Reliable predictions of one-electron operator properties, such as the electric dipole moment and field gradient, on the other hand, require near HF basis sets. To obtain the equilibrium geometry of N 2 H + we have calculated the SCF energy as a function of r(NN) and L89