Measuring stellar magnetic fields from high resolution spectroscopy of near-infrared lines

Zeeman splitting of otherwise degenerate levels provides a straight-forward method of measuring stellar magnetic fields. In the optical, the relative displacements of the Zeeman components are quite small compared to the rotational line broadening, and therefore observations of Zeeman splitting are usually possible only for rather strong magnetic fields in very slowly rotating stars. However, the magnitude of the Zeeman splitting is proportional to the square of the wavelength, whereas rotational line broadening mechanisms are linear in wavelength; therefore, there is a clear advantage in using near-infrared spectral lines to measure surface stellar magnetic fields. We have obtained high resolution ( R>=25000) spectra in the 15 625-15 665 A region for two magnetic chemically peculiar stars, viz. HD 176232 and HD 201601, and for the suspected magnetic chemically peculiar star HD 180583, as part of a pilot study aimed at determining the accuracy with which we can measure stellar magnetic fields using the Zeeman splitting of near-infrared lines. We confirm that in principle the magnetic field strength can be estimated from the magnetic intensification of spectral lines, i.e. the increase in equivalent width of a line over the zero-field value. However, due to line blending as well as the dependence of this intensification on abundance and field geometry, accurate estimates of the magnetic field strengths can be obtained only by modelling the line profiles by means of spectral synthesis techniques. Using this approach, we find a 1.4 kG magnetic field modulus in HD 176132 and an upper limit of 0.2 kG in HD 180583. The very weak infrared lines in the spectrum of HD 201601 are consistent with a 3.9 kG field modulus estimated from the splitting of the Fe II 6149.258 A line seen in an optical spectrum. Finally, we would like to draw attention to the fact that there are no sufficiently detailed and reliable atomic line lists available for the near-infrared region that can be used in high resolution work; a large fraction of the features observed in our spectra remains to be identified.