KQ Puppis 1 is known as a bright VV Cephei variable. It was given the classification M2ep Iab + B by Bidelman (1954), who also listed it along with VV Cephei. Some ten years later Jaschek & Jaschek (1963) classified the stars as M1 Iab and B2V with the classification of the hot companion based solely on He I (3819 Å). A long term spectroscopic study of the system by Cowley (1965) established the period to be 27 years and the inclination, 45o. The orbital elements determined by Cowley (1965) are used in more recent papers, e.g. by Che & Reimers (1983), who determined the mass loss rate to yr. Later Rossi et al. (1992) have proposed a tentative model of the system and thereby revised the classification of the hot companion to B0Vp. The work by Rossi et al. (1992) provides further basic stellar data on KQ Puppis, such as mass ratio, interstellar extinction and distance.
Spectral analyses of KQ Puppis have been reported in several papers (McLaughlin 1948; Bosman-Crespin & Swensson 1956; Jaschek & Jaschek 1963; Cowley 1965; Swings 1969; Altamore et al. 1982; Che & Reimers 1983; Altamore et al. 1992; Rossi et al. 1992; Muratorio et al. 1992).
The early observations (UV and visible) are summarized by Cowley (1965) and Swings (1969). They propose the presence of a shell spectrum (1947-49) with sharp absorption and emission features of ionized iron group elements such as Sc II, Ti II, Cr II, Fe II, and Ni II. The spectral lines have intensities that vary with time, increasing towards periastron. However, the intensity periodicity is suggested to be less pronounced than in the case of VV Cep. The forbidden emission lines of KQ Puppis, e.g. [Fe II], are observed to have more or less constant intensities. These lines are therefore suggested to originate in a tenuous, gaseous region enveloping the stars. Broad and hazy absorption features associated with the hot companion were difficult to evaluate in these early observations.
Altamore et al. (1982) published a list of line identifications based on high resolution spectra from the International Ultraviolet Explorer (IUE). The spectra which were observed in 1979 and 1980, cover the region 1190 - 3228 Å, and coincide with the resonance region of singly ionized iron group elements (Johansson & Cowley 1988). Altamore et al. (1982) confirm the findings of earlier observations and conclude that at least two line systems exist, viz. a broad absorption line system of multiply ionized species from the vicinity of the B star and a sharp line system of neutral and singly ionized elements originating from an extended envelope. They also found that absorption and emission features of the sharp line system have a radial velocity difference of about 40 km/s, and claimed the presence of Fe II emission lines originating from levels as high as 13 eV above the ground state. Che & Reimers (1983) discuss the Fe II spectrum in KQ Puppis in some detail and observe emission in a number of UV multiplets, see discussion in Sect. 3.
Three sequential papers following up the line identification list of Altamore et al. (1982) appeared in 1992. 1) "An atlas of the optical and the ultraviolet spectrum" (Altamore et al. 1992); 2) "A possible model" (Rossi et al. 1992) and 3) "A discussion of the Fe II spectrum, level population and spectral synthesis" (Muratorio et al. 1992). Muratorio et al. (1992) present a model for Fe II in which they use different Boltzmann excitation functions for levels having excitation energies below and above 8 eV. However, the conclusion in the paper is that Fe II levels below 8 eV are populated according to a mean excitation temperature of about 9000 K whereas the population excess of levels above 8 eV could be due to dielectronic recombination. Muratorio et al. (1992) claim that the main source of scatter of the observational data in their model is caused by uncertainties in the atomic data.
In the present work we have looked through high resolution IUE spectra and we discuss excitation processes for all identified Fe II emission lines. We use the designations PCR for photoexcitation by continuum radiation and PAR for photoexcitation by accidental resonances. PCR is used for a process where photons of the continuum radiation give rise to selective photoexcitation in a species (Johansson & Hamann 1993). PAR is a designation for the Bowen mechanism, and was introduced by Kastner & Bhatia (1986). It is used for a process where a strong emission feature from one species coincides in wavelength with a transition in another species. If the lower level of this later transition is populated, photons can be absorbed, and the photoexcited level decays in one or more fluorescence channels. Both processes normally result in a flux redistribution towards longer wavelengths.
© European Southern Observatory (ESO) 2000
Online publication: January 29, 2001