5. OH maser emission from IRC+10420
The observations of OH maser lines from IRC+10420 were carried out with the Nançay radiotelescope on April 16th 1997.
The instrument has a RAxDec beam of 3.5' x 19'. The system noise temperature was 45 K; the antenna temperatures were converted to flux densities using the efficiency curve of the radio telescope which is 0.9 K/Jy for point sources at zero declination. An autocorrelation spectrometer consisting of 4 banks of 256 channels with 0.57 kms-1 resolution was used to observe the 1612 MHz, 1665 MHz and 1667 MHz lines in both left and right circular polarizations. The observations were made in the frequency-switching mode with a total velocity coverage of 146 kms-1. The spectra shown on Figs. 7-9 have been Hanning smoothed and the final resolution is 1.14 kms-1. The rms noise level is 0.1 Jy. One may argue that the OH radio and FIR line observations are not contemporaneous. However, the 1612 MHz OH emission of IRC+10420 is regularly monitored at Nançay and the last profile was taken on 12 October 1996 at about the same periods as the ISO observations. Within calibration uncertainties (), there is no evidence for a significant flux variation over a 6 month period, at least at 1612 MHz.
5.2. Model calculations
The absorption of 34.6 µm and 53.3 µm photons leads to the inversion of the 1612 MHz transition. A consequence of this pumping cycle is that the other satellite line at 1720 MHz is anti-inverted. The occurrence of main line masers is more difficult to explain. Bujarrabal et al. (1980) proposed the overlap of FIR hyperfine lines of OH as the main mechanism for the inversion of main lines. The 1612 MHz satellite line is also enhanced by the overlap effect. In the model of Bujarrabal et al. (1980) the Doppler shifts are limited to a maximum velocity (2 kms-1). This limit is probably valid for optically thin Mira variables or if OH masers arise from clumps occupying a small fraction of the envelope. But OH-IR objects usually have higher expansion velocity and hence higher velocity limits are expected. These authors also used local physical conditions to evaluate the source function and the optical depth of the FIR overlapping lines, which are spatially separated from the overlap region. In spite of these limitations, the calculations of Bujarrabal et al. (1980) did reveal many important implications of the FIR line overlap on the inversion of OH main line masers.
Recently, Collison & Nedoluha (1993, 1994, 1995) built a more sophisticated model of the circumstellar OH maser taking into account the variation of the physical conditions in the circumstellar envelopes. Their results are qualitatively consistent with those obtained by Bujarrabal et al. (1980), but when the limitation of the Doppler shift was dropped they found that the effect of the FIR hyperfine line overlap was much smaller. They were unable to explain the dominance of main line masers over the satellite maser in the context of their model for the circumstellar envelopes. They proposed the overlap between OH near-infrared (NIR) vibrational line and a water line at 2.8 µm as a possible mechanism to explain the main line masers. Although the NIR overlap seems to be a possible pumping mechanism, so far no attempt has been made to check quantitatively this effect. One major improvement was implemented by Collison & Nedoluha (1995). They solved directly the equation of radiative transfer for maser lines in the envelope instead of relying on the Sobolev assumption.
Using a non-local radiative transfer model with the physical conditions of the envelope which fit the OH rotational line intensities, we obtain solutions for the populations throughout the OH shell in two cases. In model 1 we limit the Doppler shift to 2 kms-1 and in model 2 all overlapping hyperfine pairs, which correspond to the expansion velocity ( 35.0 kms-1) of IRC+10420, are included. The local linewidth (FWHM) is assumed to be 1.0 kms-1 in the envelope. The emergent OH maser profiles are given in Figs. 12-15. There is a clear tendency for the 1612 MHz satellite maser to dominate the main line masers. This is qualitatively consistent with the observations. Although in model 1 the two main line masers are present, the 1665 MHz maser is however much weaker than the 1667 MHz maser. The observations towards IRC+10420 indicate that the 1665 MHz line is stronger than the 1667 MHz line. As already noted by Collision & Nedoluha (1995) we also find that the main line emission comes from the region which is coincident with the unsaturated core of the 1612 MHz maser. But the behaviour of the main line masers are very different (see Fig. 10). The reason is as follows: the 1667 MHz maser is partially saturated while the 1665 MHz main line is not inverted in the inner and outer parts of the envelope. Therefore in the radial direction, some 1665 MHz maser photons are reabsorbed by OH molecules, thus reducing its intensity. The calculated total flux of the 1667 MHz maser is much greater than that of the 1665 MHz maser. In model 2 when all overlapping pairs are included, the main line emission is quenched while the 1612 MHz satellite maser is strongly enhanced. Model 2 fails to reproduce the OH main line maser emission. See Table 2.
Table 2. Calculated and observed fluxes of the OH masers.
5.3. Maser pumping efficiency
The pumping efficiency of the OH masers is an important parameter to constrain the models. Previous estimates of the pumping efficiency suffered from the absence of FIR line data (Evans & Beckwith 1977; Nguyen -Q-Rieu et al. 1979; Epchtein et al. 1980). They were based on the continuum observations around 30 µm. The detection of the 34.6 µm line by ISO allows us to reconsider this issue. Our model predicts that the 53.3 µm line is also in absorption. As a result, the maser pumping efficiency of the 1612 MHz maser can be defined as the ratio of the number of maser photons to that of FIR photons:
The envelope of IRC+10420 is a type II source which emits preferentially the 1612 MHz line. Owing to the uncertainties in the shell parameters, the fluxes determined by our model do not fit closely the observations. The pumping efficiencies for the 1612 MHz maser line estimated from formula ( 18) are 0.04 in model 1, 0.12 in model 2. Since the 53.3 µm line has not been detected we redefine the observed pumping efficiency as the ratio:
The observed pumping efficiency is then equal to 0.09. The corresponding calculated efficiency obtained from model 1 is 0.05. Our calculations also demonstrate that saturated masers convert more efficiently the pumping photons into maser photons than the non-saturated masers. In model 2 the 1612 MHz maser is strongly saturated, thus the pumping efficiency is much higher (0.17). This characteristic of saturated masers is predicted by Elitzur (1992) using a phenomenological model of two-level maser.
© European Southern Observatory (ESO) 1998
Online publication: February 4, 1998