Spectroscopy of horizontal branch stars in NGC 6752 - Anomalous results on atmospheric parameters and masses

Context. The determination of fundamental parameters for horizontal branch stars in Galactic globular clusters frequently gives puzzling results, pointing to the lack of our understanding of their atmospheric structure and the inadequate approximations by the models. Aims. We wanted to measure effective temperatures, surface gravities, helium abundances, and masses for hot horizontal branch stars in NGC 6752 in order to compare the results with evolutionary predictions. Methods. We used the ESO VLT-FORS2 facility to collect low-resolution spectra of 51 targets distributed along the horizontal branch. We determined atmospheric parameters, by comparison with theoretical models through standard fitting routines, and masses by basic equations. Results. Results generally agree with previous works, although not always with the theoretical expectations for cooler stars ( T-eff <= 15 000 K). The calculated color excess is systematically lower than the literature values, pointing towards a possible underestimation of effective temperatures. Moreover, we find two groups of stars at T-eff similar to 14 000 K and at T-eff similar to 27 000 K that present anomalies with respect to the general trend and expectations. We suppose that the three peculiar bright stars at T-eff similar to 14 000 K are probably affected by an enhanced stellar wind. For the eight extreme horizontal branch stars at T-eff similar to 27 000 K that show unusually high masses, we find no plausible explanation. Conclusions. While most of our results agree well with the predictions of standard horizontal branch evolution, we still have problems with the low masses we derive in certain temperature ranges. We believe that Kurucz ATLAS9 LTE model atmospheres with solar-scaled abundances are probably inadequate for these temperature ranges. Concerning the group of anomalous stars at T-eff similar to 27 000 K, a Kolmogorov-Smirnov test indicates that there is only an 8.4% probability that these stars are randomly drawn from the general distribution in the color-magnitude diagram. This is not conclusive but indicates that these stars could be both ( and independently) spectroscopically and photometrically peculiar with respect to the general extreme horizontal branch population.