FeH --- An Emerging Magnetic Probe for Sunspots and Cool Stars

Michael Dulick and Jeff A. Valenti

The McMath-Pierce Fourier transform spectrometer (FTS) enables the full potential of FeH to be used as a critical diagnostic of magnetic fields in cool stars. Wing et al. (1977, ApJ 216, 659) were the first to provide a convincing detection of FeH in sunspots and cool stars. Interest in the Zeeman response of FeH shortly followed, when the investigators commented on the unusual "square-like" profiles of the FeH rotational lines in the sunspot spectrum, implying possible unresolved Zeeman structure. Use of FeH as a magnetic probe was rekindled in the mid-1990s by stellar astronomers concerned with the existence of magnetic fields in very cool main-sequence and brown dwarf stars, and with the role that magnetic fields play in the coronal heating of G and K dwarfs.

Spectra of cooler stars (T<3500 K) are generally devoid of strong Zeeman-active atomic lines. Instead, the spectra are often characterized by their richness in molecular band structure. Among the diatomics present in these spectra, FeH perhaps offers the best hope of measuring magnetic fields in these stars. The prominent 989.6 nm band (a ⁴ - ⁴ 0-0 transition) is present in the region of this spectrum that is virtually free of atmospheric absorption lines. Moreover, the combination of wide-open rotational structure and large Zeeman broadening allows the Zeeman measurements to be performed with spectrometers at moderate resolving powers of 50000.

The possibility of using FeH to measure magnetic fields was advanced by the Wallace et al. sunspot atlas for the red and infrared (1998, NSO Technical Report #98-002). From an archived high-resolution sunspot spectrum taken during the 1981 solar maximum with the McMath-Pierce Fourier Transform Spectrometer (FTS), Wallace et al. were able to measure a fair number of partially resolved Zeeman-broadened rotational linewidths in the eight P and R branches of the 989.6 nm band. These measurements represented the only available Zeeman data for FeH. Unfortunately, the data set was just too complicated to extract any meaningful information with regard to Landé g_J factors, which allow magnetic field strengths to be deduced directly from measured Zeeman pattern splittings.

This setback was, however, short lived. In 1998, John M. Brown (Oxford) made very high-precision laboratory measurements of Landé factors in the ⁴ ground state. Since this state coincides with the lower state of the 989.6 nm band, the sunspot data could finally be interpreted. Valenti and Dulick proceeded to construct a ⁴ Zeeman Hamiltonian model which allowed the Brown data to be reduced to a single empirical formula. This in turn provided the opportunity to extrapolate the data over the full range of observed rotational levels in the sunspot data. With the measured Zeeman widths from the McMath-Pierce FTS sunspot data, a similiar formula for g_J in the upper state was also derived.

Much of this can be attributed to perturbations in both the lower and upper states by neighboring electronic states. More extensive laboratory measurements are needed to incorporate these perturbations into the model. Such measurements involving the McMath-Pierce FTS are planned for the upcoming year. Nevertheless, the quality of these preliminary results is sufficiently good to the extent that attempts at measuring magnetic fields using FeH are already underway for a small number of active K and M dwarf stars (Valenti, Johns-Krull, Piskunov).

Caption: The results of the analysis are summarized in the figure. The smooth curves represent the model results, while the open and filled symbols denote the experimental data. Overall agreement is reasonably good; however, the residuals do reveal, in certain instances, a tendency either to underestimate or overestimate the gJ's in the higher rotational levels.