401

HIPPARCOS, VLA, AND CCD OBSERVATIONS OF CYG OB2 No. 5

1 1 2

M.E. Contreras , L.F. Ro driguez ,M.Tapia ,

3 3 3 3

D. Cardini , M. Badiali , A. Emanuele ,P.Persi

1

Instituto de Astronom a, UNAM, Ap do. Postal 70-264, 04510 M exico, D.F., M exico

2

Instituto de Astronom a, UNAM, Ap do. Postal 877, Ensenada, M exico

3

Istituto Astro sica Spaziale, CNR, C.P. 67, 00044 Frascati, Italy

uxes and mass loss values for 10 ob jects and deter- ABSTRACT

mined accurate radio p ositions. In particular, their



sample included Cyg OB2 No. 5 BD+40 4220; V729



Cyg which is known to b e a contact binary system

Cyg OB2 No. 5 BD+40 4220; V729 Cyg is known

formed bytwo O7I Torres-Do dgen et al. 1991.

to b e a contact binary system formed bytwo O-typ e

For this ob ject they detected emission at the optical

stars. For this ob ject a 6-cm radio emission has

p osition as well as an additional weak radio source

b een detected at the optical p osition as well as an

lo cated at  0:9 arcsec to the north east of the con-

additional weak non-thermal radio source lo cated at

tact binary. They suggested that this radio comp o-

 0:9 arcsec to the north east of the contact binary.

nentwas p ossibly a third stellar ob ject in the system.

It has b een suggested that this radio comp onentwas

This was the rst time that a radio companion was

p ossibly a third stellar ob ject in the system. This was

observed to be asso ciated with the contact binary

the rst time that a radio companion was observed

system.

to b e asso ciated with the contact binary system. In

this work we present new radio, CCD and Hipparcos

Persi et al. 1985 reobserved this source at 2, 6 and

observations that provide the basis for a plausible ex-

20 cm. They found that the 6-cm ux density had

planation of the origin of the non-thermal radiation

increased by a factor of 3, with resp ect to the 1980

emitted by the `third' comp onent of this system and

observations made by Abb ott et al. 1981, and have

its nature.

found a very at sp ectral index of 0.2 0.1.

Later, Persi et al. 1990 continued monitoring Cyg

Key words: Radio continuum: general and stars;

OB2 No. 5 and found again that the emision had

early-typ e stars.

changed, this time decreasing backto the 1981 lev-

els. Persi et al. 1990 de ned what they called the

`high' and `low' states of emission. In the `high' state

the 6-cm measured ux densitywas of '5 to 7 mJy

1. INTRODUCTION

with a sp ectral index of ' 0.2 while in the

26cm

`low' state the 6-cm ux was of '1 to 2 mJy and the

sp ectral index had increased and b ecame consistent

One of the more reliable metho ds for the determina-

with the classical 0.6 value exp ected for a thermal

tion of the mass loss rate in massive stars is via the

wind. However, even in the low state of emission the

observation of the free-free radiation emitted by their

sp ectral index derived from the 6 and 20-cm uxes

ionized winds at radio wavelengths. The rst radio

had a negativevalue indicating the presence of a non-

survey was carried out by Abb ott et al. 1980. In

thermal comp onent. The presence of b oth thermal

this study they rep orted mass loss rate values for 15

and non-thermal radio emitting comp onents in Cyg

O, B and A-Typ e stars observed at 6-cm. Later stud-

OB2 No. 5 resembled the case of other Cyg OB2 as-

ies con rmed the detection of thermal radio emission

so ciation memb ers Cyg OB2 Nos. 8A, 9 and 12 and

at 2 and 6-cm Abb ott et al. 1981, 1984, 1986, Bieg-

some Wolf-Rayet stars WR 140 and WR 147.

ing et al. 1989 from OB stars. However, some early-

typ e stars show time-variable, non-thermal emission

Miralles et al. 1994 observed the source continuing

of p o orly understo o d origin Abb ott et al. 1984, Persi

the monitoring for ab out four more years. The source

et al. 1985, Williams et al. 1990, Miralles et al. 1994.

showed again its radio variations b etween the `high'

This synchrotron radiation can be comparable or

and `low' states with ux densities and sp ectral in-

even larger than the exp ected free-free emission and

dices similar to those found by Persi et al. 1990.

this non-thermal contamination limits the accuracy

The radio emission seems to switch between these

of the metho d.

two states in a p erio d of ab out 7years. They have

mo deled the emission as due to a synchrotron enve-

lop e created by a p erio dic pro duction of relativistic Abb ott et al. 1981 selected the most luminous stars

electrons in the wind. Miralles et al. 1994 con rmed in the , most of their sample residing in the

the coincidence b etween the radio maximum emission Cyg OB2 asso ciation. They rep orted 6-cm radio

402

and the optical p osition for the binary system and re-

vered the weak radio companion at 3.6 and 6-

disco CYGOB5 3.6-cm Pesado=NA Haz=0.3x0.3 arcsec

cm. They determined a separation of  0:8  0:1 arc-

sec to the NE of the main comp onent and supp orted

the idea that this source was a third comp onentofthe

and not a background ob ject. The sp ectral

system 41 08 05.0

index derived for the companion was = 0:3  0:5.

However with such a large error it was not p ossible

establish if the radio companion was thermal or

to 04.5

non-thermal. Herbig 1967 had rep orted a of

magnitude 13 or 14 lo cated at ab out  1:5 arcsec

 

a p osition angle of 60 70 from the optical

with 04.0

p osition of the brightest comp onent. Miralles et al.

1994 suggested the p ossibility that this could b e the

DECLINATION (B1950)

optical counterpart of the weak radio companion.

03.5

In a recent study Contreras et al. 1996 rep orted two

sets of observations for Cyg OB2 No. 5 obtained in

03.0

1994 and 1995. They found that the ux densities

app eared to b e increasing b etween the two observing

runs at the di erent frequencies observed, 6, 3.5 and This tendency and the fact that the sp ectral 0.7 cm. 02.5

20 30 34.95 34.90 34.85 34.80 34.75

ed were decreasing was indicativeofthe indices deriv (B1950) Cont peak flux = 7.4066E-03 JY/BEAM onset of the `high' emission state predicted to o ccur in Levs = 1.8000E-05 * ( -4.00, 4.000, 5.000,

6.000, 7.000, 8.000, 10.00, 20.00, 30.00,

y Miralles et al. 1994. Besides, in their 1994

1996 b 60.00, 100.0, 200.0, 350.0)

observations they detected the radio companion at 6-

cm and 3.6-cm. With these two observed uxes they

derived an upp er limit value for the sp ectral index,

Figure1. Cleaned, natural-weight map of Cyg OB2 No. 5

= 2:4  0:6, indicating beyond doubt that the

from self-calibrated uv data at 3.6 cm. The crosses mark

companion source was clearly non-thermal at least

the position of the optical stars see text.

during that ep o ch. The separation b etween the main

comp onent and the companion was 0:8  0:1 arcsec.

Hence, separation values determined in the radio by

di erent authors Abb ott et al. 1981, Miralles et al.

2.2. CCD Observations

1994, Contreras et al. 1996 indicated that the sep-

aration between the two comp onents was less than

1:0 arcsec, apparently smaller than the value rep orted

by Herbig 1967 in the optical. In this work we

CCD images were obtained through the broad-band

present new radio and optical observations that de-

lters B , V and R Tek4 800800 CCD on the 1-m

termine this separation accurately and provide the

Johannes Kepler Telescop e JKT at the Observato-

basis for a plausible explanation of the origin of the

rio Ro que de los Muchachos in La Palma, Canarias,

non-thermal radiation emitted by the `third' comp o-

Spain, on the nights 6/7 and 15/16 Novemb er 1994

nent of this system and its nature.

as part of the Service Observing program. The plate

scale was 0.33 arcsec/pixel and the mean seeing was

around one arcsec during b oth nights. The exp osure

times were b etween 1{10 s but even with the short-

2. OBSERVATIONS

est exp osure, Cyg OB2 No. 5 was saturated in the I

frames.

2.1. Very Large Array Observations

A Gaussian Point Spread Function was tted to iso-

lated stars on each frame and was then subtracted, af-

New 3.6 and 6-cm observations were made with the

ter scaling, from the sp ot centred on Cyg OB2 No. 5.

1

Very Large Array VLA of NRAO . The 3.6-cm ob-

The residual image clearly showed the fainter isolated

servations were made during 1996 Decemb er 27 and

companion star. Its relative p osition was measured

28, while the 6-cm observations were made during

on all frames with an average value for the sepa-

1997 January 4. During b oth observing runs the ar-

ration to primary binary of 0.98  0.06 arcsec and

raywas in the A con guration giving resolutions of

p osition angle 61  7 degrees. The di erential pho-

 0:2 arcsec and  0:3 arcsec, at 3.6 and 6-cm, re-

tometry Cyg OB2 No. 5 minus companion yielded

sp ectively. The amplitude calibrator was 1328+307

V = 3:96 and B V =0:11.

and the phase calibrator 2005+403 in b oth cases.

The b o otstrapp ed ux densities of 2005+403 were

Assuming the values of the photometry for the com-

2.980.01 and 3.240.01 Jy at 3.6 and 6-cm, resp ec-

p osite system to be those given by Torres-Do dgen

tively. Editing, calibration, imaging, and analysis of

et al. 1991, the individual colours of the sp ectro-

the data was p erformed using the AIPS software of

scopic binary Cyg OB2 No. 5 are V = 9:18 and

NRAO. Cleaned maps made from self-calibrated uv

B V  =1:81, while those of the stellar compan-

data at 3.6 and 6-cm are shown in Figures 1 and 2.

ion are V =13:14 and B V =1:70. As the star

1

NRAO is op erated by Asso ciated Universities, Inc., under

was saturated, no photometry was p erformed in the

co op erative agreement with the National Science Foundation.

I band.

403

2.3. Hipparcos Observations sho cks, giving rise to the observed synchrotron non-

thermal radiation.

A complete, detailed description of the Hipparcos

Assuming the absolute V magnitude and intrinsic

mission and of the derivation of the astrometric pa-

B V  colour of two O7Ia stars for the contact

rameters is given in the Hipparcos Catalogue ESA

binary primary and the same distance and redden-

1997. During the 37 months of the Hipparcos mis-

ing Torres-Do dgen et al. 1991, then we inferred a

sion, the ob ject Cyg OB2 No. 5, identi ed by the

sp ectral typ e B2V for the companion star.

Hipparcos Catalogue number HIP 101341, fell 113

times in the eld of view of the rotating telescop e,

Avery similar situation has b een recently found for

each time p erforming an apparent transit through

WR 147 by Williams et al. 1997.

the mo dulating grid of its fo cal plane. By analyising

this set of signals with a pro cessing algorithm de-

scrib ed in Mignard et al. 1995, it has b een p ossible

REFERENCES

to obtain the separation 0:948  0:043 arcsec and

 

the p osition angle 54  3 . These data are in go o d

agreement with those of the CCD measurements.

Abb ott, D.C., Bieging, J.H., Churchwell, E.,

Cassinelli, J.P. 1980, ApJ, 238, 196

Abb ott, D.C., Bieging, J.H., Churchwell, E. 1981,

3. DISCUSSION

ApJ, 250, 645

Abb ott, D.C., Bieging, J.H., Churchwell, E. 1984,

ApJ, 280, 671

In Figures 1 and 2 we have sup erp osed the optical

p ositions for the primary and secondary optical com-

Abb ott, D.C., Bieging, J.H., Churchwell, E. Torres,

p onents, as determined by Hipparcos, onto the radio

A.V. 1986, ApJ, 303, 239

maps. For this we have assumed that the bright-

Bieging, J.H., Abb ott, D.C., Churchwell, E. 1989,

est optical comp onent coincides in p osition with the

ApJ, 340, 518

brightest radio comp onent. As can be clearly seen,

Contreras, M.E., Ro dr guez, L.F., G omez, Y.

the radio `companion' is lo cated in b etween the two

Vel azquez, A. 1996, ApJ, 469, 329

optical p ositions, although closer to the secondary

optical comp onent. While at 6-cm the separation b e-

ESA, 1997, The Hipparcos and Tycho Catalogues,

tween comp onents is 0:79  0:03 arcsec, the separation

ESA SP{1200

measured by Hipparcos is 0:948  0:0043 arcsec.

Herbig, G.H. 1967, PASP, 79, 502

Mignard, F., So derhjelm, S., Bernstein H. et al 1995, A&A, 304,94

CYGOB5 6-cm Pesado=NA Haz=0.3x0.3 arcsec

Miralles, M.P., Ro dr guez, L.F., Tapia, M., Roth,

M., Persi, P., Ferrari-Toniolo, M., Curiel, S. 1994, A&A, 282, 547

41 08 05.0

Persi, P., Ferrari-Toniolo, M., Tapia, M., Roth, M.,

Ro dr guez, L.F. 1985, A&A, 142, 263

Persi, P., Tapia, M., Ro dr guez, L.F., Ferrari-

04.5

Toniolo, M. Roth, M. 1990, A&A, 240, 93

Torres-Do dgen, A.V., Tapia, M. Carroll, M. 1991,

04.0 MNRAS, 249, 1

Williams, P.M., van der Hucht, K.A., Pollo ck, A.M.T., et al. 1990, MNRAS, 243, 662 DECLINATION (B1950)

03.5

Williams, P.M., Dougherty, S.M., Davis, R.J., van

der Hucht, K.A., Bo de, M.F., Gunawan, D.Y.A.S. , MNRAS, submitted

03.0

02.5 20 30 34.95 34.90 34.85 34.80 34.75 RIGHT ASCENSION (B1950) Cont peak flux = 6.9363E-03 JY/BEAM Levs = 1.8000E-05 * ( -4.00, 4.000, 6.000, 8.000, 10.00, 15.00, 20.00, 30.00, 60.00,

100.0, 200.0, 300.0, 400.0)

Figure2. Cleaned, uniform-weight map of Cyg OB2 No. 5

from self-calibrated uv data at 6-cm. The crosses mark the

position of the optical stars see text.

These results indicate that the radio `companion' is

not a star, but most likelyabow sho ck pro duced at

the p osition where the winds of the contact binary

and the secondary star collide. Acceleration of elec-

trons to relativistic sp eeds can take place in strong