We report the results of a long-term spectroscopic monitoring of the A–type supergiant with the B [ e ] phenomenon 3 Pup = HD 62623 .
We confirm earlier findings that it is a binary system .
The orbital parameters were derived using cross-correlation of the spectra in a range of 4460–4632 Å , which contains over 30 absorption lines .
The orbit was found circular with a period of 137.4 \pm 0.1 days , radial velocity semi-amplitude K _ { 1 } = 5.0 \pm 0.8 km s ^ { -1 } , systemic radial velocity \gamma = +26.4 \pm 2.0 km s ^ { -1 } , and the mass function f ( m ) = ( 1.81 ^ { +0.97 } _ { -0.76 } ) \times 10 ^ { -3 } M _ { \odot } .
The object may have evolved from a pair with initial masses of \sim 6.0 M _ { \odot } and \sim 3.6 M _ { \odot } with an initial orbital period of \sim 5 days .
Based on the fundamental parameters of the A-supergiant ( luminosity \log L/L _ { \odot } = 4.1 \pm 0.1 and effective temperature T _ { eff } = 8500 \pm 500 K ) and evolutionary tracks of mass-transferring binaries , we found current masses of the gainer M _ { 2 } = 8.8 \pm 0.5 M _ { \odot } and donor M _ { 1 } = 0.75 \pm 0.25 M _ { \odot } .
We also modeled the object ’ s IR-excess and derived a dust mass of \sim 5 \times 10 ^ { -5 } M _ { \odot } in the optically-thin dusty disk .
The orbital parameters and properties of the H \alpha line profile suggest that the circumstellar gaseous disk is predominantly circumbinary .
The relatively low mass of the gainer led us to a suggestion that 3 Pup should be excluded from the B [ e ] supergiant group and moved to the FS CMa group .
Overall these results further support our original suggestion that FS CMa objects are binary systems , where an earlier mass-transfer caused formation of the circumstellar envelope .