We present multi-epoch high-spectral resolution observations with VLT/CRIRES of the OH doublet $^2Pi_{3/2}$ P4.5 (1+,1-) (2.934 $mu$m) towards the protoplanetary disk around HD 100546. The OH doublet is detected at all epochs and is spectrally resolved while nearby H$_2$O lines remains undetected. The OH line velocity profile is different in the three datasets: in the first epoch (April 2012, PA=26$^{circ}$) the OH lines are symmetric and line broadening is consistent with the gas being in Keplerian rotation around the star. No OH emission is detected within a radius of 8-11 au from the star: the line emitting region is similar in size and extent to that of the CO ro-vibrational lines. In the other two epochs (March 2013 and April 2014, PA=90$^{circ}$ and 10$^{circ}$, respectively) the OH lines appear asymmetric and fainter compared to April 2012. We investigate the origin of these line asymmetries which were taken by previous authors as evidence for tidal interaction between an (unseen) massive planet and the disk. We show that the observed asymmetries can be fully explained by a misalignment of the slit of order 0farcs04-0farcs20 with respect to the stellar position. The disk is spatially resolved and the slit misalignment is likely caused by the extended dust emission which is brighter than the stellar photosphere at near-infrared wavelengths which is the wavelength used for the pointing. This can cause the photo-center of HD 100546 to be mis-aligned with the stellar position at near-infrared wavelengths.
Download