Laser Spectrometer for Stellar Astrophysics – LASSA

This project began as a collaborative investigation between P. Crozet (LASIM) and J-F Donati (LATT) initiated in the framework of the Programme Nationale de Physique Stellaire (PNPS) in 2007, aiming to address the laboratory requirements of stellar astrophysics, and in particular of stellar polarimentry imaging. We propose to investigate the electronic properties of transition metal hydride radicals, and in particular to determine the magnetic tuning of these species using optical spectroscopy, in this context. Simple molecules, particularly oxides and hydrides, are observed in the spectra of cool stars (and sunspots), and since, in the presence of a magnetic field, many molecular spectral lines exhibit the Zeeman effect, they probe cool magnetized stellar atmospheres. In relatively cool stellar objects (M dwarfs, starspots), atomic spectroscopy actually becomes ineligible because molecular spectra dominate. Taking constraints of wavelength (atmospheric windows), magnetic response and detection sensitivity into account, transition metal monohydrides emerge as particularly promising monitors of stellar environments, particularly since the resolution available with modern instruments such as ESPaDOnS allow their molecular Zeeman structures to be resolved. The difficulties arise from the chemical nature of transition metal complexes. If molecular Landé factors can be extracted from effective Hamiltonian-based analyses, where the Zeeman effect is proportional to (at least, in the scheme of Hund's case (a) coupling) Omega*(g_S*Sigma +g_L*Lambda)/[J(J+1)], from a formula analogous to that for the atomic Zeeman effect. But this formula breaks down for transition metals and rare earths, which assumes that Sigma and Lambda represent orbital and spin angular momentum in the molecular radical. Neither of these are strictly 'good' quantum numbers in the context of transition metal containing species, and the apparent effect is that the Landé factors g_S and g_L deviate from their expected values of 2.0023 and 1.0 respectively, varying with rotational quantum number J. Additional difficulties arise as the coupling of angular momenta in the molecule changes also as a function of J, and of magnetic field. To invert Zeeman spectra from astrophysical sources, reliable Landé factors are needed for selected quantum levels. Laboratory investigations therefore need to focus on the electronic states observed in stellar spectra, either by investigating exactly these electronic systems, or by devising alternative routes to observe the two states independently. Magnetic fields of the order of 3 kGauss encountered in sunspots are accessible in a laboratory environment; but the temperatures > 3000 Kelvin are not. This means that the intensity distributions recorded in the laboratory will differ significantly from the stellar spectra, but by working with lines in the wings of a laboratory distribution, the characteristics of the molecular energy levels can be clearly established. The requirements are (1) to produce the radicals which are intrinsically unstable in appreciable quantities, (2) to study them at sufficient resolution to distinguish the field-induced splittings, and (3) to provide a model with which to interpret the data. Although the methods and experiments we propose can be, and will be, readily applied to other systems, iron monohydride and nickel monohydride have been selected as primary targets. The planned laboratory Zeeman experiments require an intersection between laser beam and molecular source in a strong and uniform magnetic field. A prototype device built in LASIM (2000 turn solenoid accepting currents up to 2 Ampères) already provides a magnetic field of 0.14 Tesla. A threefold increase in field strength will suffice to match the stellar-type magnetic fields (0.4 T) we wish to reproduce. The laser source required to probe these species (continuous wave Ti:sapphire) is the major part of the equipment to be funded by this project. In parallel, the partners based at LATT plan to investigate stellar spectra for NiH and FeH in appropriate wavelength regions on the THEMIS instrument. THEMIS offers resolving power of 300 000, which matches the Doppler width of MH lines at temperatures of the order of 3000 K, but across a narrow (7 Å) wavelength window. Preliminary work in Lyon has already identified prime spectral regions for observations. Appropriate code will be written to analyse the spectra, using an effective Hamiltonian approach. Interpretation the results of such analyses will then rely on quantum chemical models, which take account of configuration mixing (particularly through spin-orbit mixing in these cases). This aspect will be addressed in the later stages of the project by Mme S. Magnier (Université de Rennes).