Energy Crisis in Be star Radiation Emission?

Krishna M. V. Apparao
Department of Physics, University of Bombay
Santa Cruz East, Mumbai 400098, India

Received: 1998 May 12

This note is an attempt to draw the attention of the Be star community to the problem of energetically accounting for the H-alpha and infrared radiation from Be stars.

It is generally agreed that the emission lines observed from Be stars arise from the gas envelope around it. It is presumed that the ionization of the envelope is by the Lyman continuum from the Be stars. The ionized gas in the envelope is also considered responsible for the infrared excess observed. Some attempts have been made to quantitatively account for the H-alpha and infrared radiations from the disk (Hoflich 1988; Waters, Cote, and Lamers 1987; Kastner and Mazzali 1989; Kerkwijk, Waters and Marlborough 1994). Hoflich considered a model in which the gas envelope is an extension of the atmosphere of the Be star and spherical symmetry is assumed. He was able to reproduce the H-alpha equivalent widths by considering Balmer continuum absorption in addition to the Lyman continuum absorption. However, the spherical symmetry assumed is clearly not tenable in view of several observations. Waters, Cote, and Lamers (1987), Kastner and Mazzali (1989), and Kerkwijk, Waters, and Marlborough (1995) on the other hand assume a temperature throughout the envelope considered, in order to account for the H-alpha and infrared radiation. It needs to be shown that the assumed temperature can occur throughout the envelope due to the absorption of radiation from the Be star, before the results can be accepted.

We have been examining the energetics of H-alpha and infrared radiation from Be stars by considering the absorption of all the ionizing radiation from the Be star; this should give the maximum possible ionization by the radiation (Apparao and Tarafdar 1987; 1997a; 1997b). We have assumed the Lyman and Balmer continua given by atmospheric calculations of Kurucz (1979) for various spectral types of Be stars and considered their absorption for ionizing the Be star envelope. In the first publication we have calculated the energy in the H-alpha emission due to the absorption of Lyman and Balmer continua from the Be stars, and compared the results with observations. We found that the observations can be accounted for Be star spectral types up to B5 (see Table 1); the emission from later spectral types cannot be accounted for by the above process. We had suggested the need for additional ionizing photons.

In Apparao and Tarafdar (1997a) we suggested that in the case of late type Be stars, a He star binary companion to the Be star can supply the necessary ionizing photons (see Table 1). We had shown that the He star will not visible in the glare of the Be star optical emission. This will imply that all the late type Be stars which are observed to emit H-alpha emission have a He star binary companion, and for those without a He star companion, though having a gas envelope, H-alpha emission will not be detected. Some of the other consequences of our suggestion are given in the paper.

We have examined whether the observed infrared excess for Be stars (these considerations do not pertain to B[e] stars, where the infrared emission is believed to occur from heated dust) can be accounted for by free-free emission by the gas envelope ionized by the Lyman continuum of the Be star (Apparao and Tarafdar 1997b). The calculated values and the observed values are given in Table 2.

^a Highest values observed for the spectral type (Ashok et al. 1984).
^b Observed value for X Per, when it is not at its peak emission (Roche et al. 1993).

It is seen from Table 2 that the infrared emission calculated from the absorption of the Lyman continuum is inadequate to account for the observed infrared emission from Be stars of spectral types of B1 and later. In the case of spectral types greater than B4, the energy in the Lyman continuum itself is smaller than that in the observed infrared radiation.

How then can the infrared emission be energized? We had considered two possibilities: 1) Be stars emit a Lyman continuum more copiously than given by the atmospheric calculations of Kurucz (1979), or 2) additional ionizing photons are supplied by a companion star to the Be star (He star, white dwarf, or neutron star).

Observations of the star Epsilon CMa (Cassinelli et al. 1995) by the EUVE Satellite has shown that the EUV flux is about thirty times the flux given by atmospheric calculations. Modifications of the atmospheric calculations could not explain the observations so far and the origin of this flux remains obscure. However, the possibility exists that B and Be stars emit EUV flux more copiously than given by the atmospheric calculations. If this is true, then the observed infrared and H-alpha fluxes for Be stars (at least for the early spectral types) can be accounted for energetically. However, the problem still remains for the late spectral types.

The required ionizing photons to account for the infrared and H-alpha fluxes from Be stars can easily be supplied by the presence of a compact object (He star, white dwarf, or neutron star) in binary motion around the Be star. The formation of such binaries was discussed by Pols et al. (1991). The surface temperature of the He star is high enough to supply the requisite photons (Apparao and Tarafdar 1997a). If the compact object is a white dwarf or neutron star, matter accretion is needed to give out ionizing photons. The presence of a neutron star is indicated by an X-ray flux, but the presence of a white dwarf star is difficult to detect due to absorption of the radiation by the Be star envelope itself (Apparao 1991). The detection of white dwarf and He stars in the far and extreme UV is discussed in Apparao and Tarafdar (1997a; 1997b). As mentioned earlier, the presence of a He star will be difficult to detect in the visible region due to the brightness of the Be star.

There is enough energy in the Balmer continuum of Be stars to account for the observed infrared and H-alpha emissions. However, a process by which the energy can be utilized to provide the necessary ionization has not been found yet.

The suggestion that the ionizing photons to account for the infrared and H-alpha emissions come from a compact object implies that all the Be stars, at least the later spectral types, are binaries. Even though this is a possibility considered earlier by other authors (e.g. Harmanec 1982), there is no evidence to support this contention.

We suggest that at present the accounting for the energy needed for infrared and H-alpha emissions of Be stars is not done adequately. Further observations and theoretical work are needed to give an answer.

Acknowledgments: I thank the Indian National Science Academy for a Senior Scientist Fellowship. I thank the Director of Tata Institute of Fundamental Research for use of the library and other facilities. I thank Prof. S. B. Patel and other colleagues at the Bombay University.

References:

• Apparao, K.M.V., and Tarafdar, S.P., 1987, ApJ, 322, 976
• Apparao, K.M.V., and Tarafdar, S.P., 1997a, Bull.Astr.Soc.Ind., 25,345
• Apparao, K.M.V., and Tarafdar, S.P., 1997b, J. Astrophys.Astr.18,145
• Ashok, N.M. et al., 1984 MNRAS, 211, 471
• Cassinelli, J.P. et al., 1995, ApJ, 438, 932
• Harmanec, P., 1987, IAU Colloquium no. 92, 339
• Hoflich, P., 1988, A&A, 176, 93
• Kastner, J.H., and Mazzali, P.A., 1989, A&A, 210, 295
• Kurucz, R.L., 1979, ApJS. 40, 1
• Pols, O.R., Cote, J., Waters, L.B.F.M. and Heise, J., 1991, A&A, 241, 419
• van Kerkwijk, M.H., Waters, L.B.F.M. and Marlborough, J.M., 1994. A&A, 300, 259
• Waters, L.B.F.M., Cote, J., and Lamers, H.G.J.L.M., 1987, A&A,185, 206