A Search for SiO Masers in Globular Clusters Using the VLA at 43 GHz

M. Colleen Gino

Abstract

Thirty globular clusters were observed with the Very Large Array in an effort to detect SiO maser emission from possible long period variable stars in the clusters. Data analysis was distributed among 10 individuals of the Swinburne University of Technology’s Introductory Radio Astronomy course HET608, and this paper reports on the results from the measurements of M13, M12 and M72. No SiO maser emission was detected from these three objects.

Introduction

This paper presents a subset of a larger study of 30 globular clusters that were observed as part of a search for SiO maser emission from evolved stars in globular clusters. The importance of this search is that the detection of masers in globular clusters would be a key contribution in constraining the various models of galactic structure and formation presently in use.

Over the past 30 years there have been a number of unsuccessful searches for masers in globular clusters (Knapp and Kerr 1973, Cohen and Malkan 1979, Dickey and Malkan 1980), and at least one partially successful search (Frail and Beasley 1994). However, a search of the literature yields no record of an attempt to detect SiO masers in globular clusters.

SiO maser emission has long been observed in evolved stars, in particular Mira variables and long period variables (Buhl et al. 1974, Kaifu et al. 1975). Such stars are known to occur in globular clusters, although their frequency appears to be low with clusters containing such stars having on average only 2-4 members per cluster. This initial search for a detection of SiO masers was performed using the Very Large Array (VLA), which consists of 27 antennas of 25 meters in diameter arranged in a Y-shaped baseline. In the A-configuration the VLA has a synthesized aperture of 27 km.

This paper describes the results of the search for SiO maser emission from 3 globular clusters, M13, M12 and M72 using the VLA.

Background

A globular cluster is a densely packed association of stars characteristically arranged in a compact sphere, as can be seen in Figure 1A. These star clusters can be found in the galactic halo, the nearly spherical aggregation of dust, gas and stars surrounding our galaxy. Approximately 150 globular clusters are currently known to be associated with

the Milky Way Galaxy, and they are believed to have formed around the same time as the Galaxy itself. In addition, a number of globular clusters in the galactic halo may have been removed from other nearby galaxies and captured by the Milky Way’s gravitational influence. One important feature of globular clusters is that many of the globular clusters surrounding our galaxy are known to contain Mira-type and long period variables, which, at least in some cases, produce SiO masers (i.e. WX SER).

As low to medium mass (1-7 x solar mass) stars near the end of their lives, they experience a brief period of tremendous mass loss. Stellar masers are found in the circumstellar shells produced by this loss of mass from the dying star, and occur when the turbulent upper photosphere is exposed to the intense radiation from below. SiO masers exist quite close in to the photosphere of these evolved stars, which contain an abundance of heavier elements (see http://spacesun.rice.edu/~parviz/maserspots.htm).

Since globular clusters are located in the halo of the galaxy and orbit around the center of mass of the galactic bulge, the measurement of variations in their orbits is a sensitive indicator of the total enclosed mass and the local mass distribution, which may include dark matter. The relatively close proximity of the globular clusters to our solar system allows the position of the cluster to be measured using the parallax technique, provided that the position of the globular cluster can be measured to sufficient accuracy. In the simplest case, comparing the position of the globular cluster over time with that of a distant unmoving reference point such as a quasar or other bright distant astronomical object allows variations in the cluster’s orbit to be detected.

While the VLA is an excellent instrument to determine the presence of SiO masers (at a rest frequency of 43 GHz) follow-up observations with the Very Long Baseline Array, a system of 10 identical 25-meter antennas spread across the United States from Hawaii to St. Croix and controlled from a central headquarters in New Mexico, allows the position of a globular cluster to be determined with a very high accuracy. Such high angular resolution measurements, typically having 0.1 milliarcsecond beamwidth (Ulvestad and Goss 1999), provides fine enough resolution to determine the angular transverse motions and, for some of the nearest clusters, obtain accurate positions using parallax.

Observations

The observations were conducted over a five-day period beginning October 12, 2000 with the NRAO Very Large Array located outside of Socorro, New Mexico. Observations took place during the move from the D-configuration to the A-configuration, the period of time when the size of the array was being changed from the smallest baseline to the largest baseline of ~27 km. The measurements for this study were insensitive to this change, making it an ideal use of the non-standard and changing array configuration. Although a total of 23 of the 27 VLA antennas are equipped with 7mm (43 GHz) receivers, the number of antennas used varied from 19 to 21 over the course of the observations due to antenna maintenance and complex array reconfiguration issues.

The observations of the globular clusters were made at 43 GHz in spectral line mode 2AC (a two-IF mode consisting of an RCP-LCP pair) using 64 channels across a 6.25 MHz bandwidth. With the primary beam size of the VLA antennas (~1’) being larger than the half-mass radius of nearly the entire sample, adequate coverage was accomplished with a single pointing toward the core of each cluster. A 3 by 3 grid was then produced by phase rotating the data to search directions adjacent to the phase center, enabling the detection of maser emission around the edges of the cluster (Beasley, 2000 private communication). This resulted in a grid of 1 arcminute diameter sampled areas spaced at 30 arcsecond offsets.

In the entire project, thirty clusters were ultimately observed, with a series of 10-second integrations equaling a total of 30 minutes integration time for each cluster. A total of 20.5 hours of observations were carried out over four separate observing sessions. The data from the first 2.5-hour observing session was rendered unusable by a high percentage of moisture in the atmosphere. All but one of the clusters (Pyxis) observed during the first session were re-observed in later sessions. A second cluster (NGC6402) was not observed due to scheduling problems. Also included was an observation of the known SiO maser WX SER in order to confirm the telescope’s ability to detect maser emission.

Three criteria were met in selecting 27 of the 32 clusters chosen for this survey and are itemized below:

1. -20° < declination <70°
2. distance from the Sun < 25 kpc
3. Log (metallicity/solar) i.e. [Fe/H] > -1.9

Five additional clusters that were chosen failed to meet one or more of the above criteria, but were included in the search since they may be captured globular clusters and as such are scientifically interesting objects. The complete list of the 32 globular clusters originally intended to be observed is listed in the Appendix.

Results and Discussion

The observational results for the three globular clusters M13, M12 and M72 are presented in this paper along with the results for the known SiO maser WX SER. As this paper presents a null result for the detection of an SiO maser, the analysis for M13 will be presented in detail with the results for M12 and M72 shown only as a summary in Table 1. Table 1 lists the name, raster position, mean, standard deviation, minimum detection limit, maximum detection value, possible detection and detection confirmation status for all three globular clusters and WX SER for easy comparison.

The raw data for the globular clusters consist of 64 spectral channels measured with respect to the Earth’s velocity. The velocities were transformed using the barycentric velocities appropriate for the astronomical object. Position information within a cluster was obtained by phase shifting the data resulting in 9 measurements for each object located on a 30 arcsecond spaced grid. The position of the measurements with respect to the center of the cluster is shown in Figure 1B and the 9 positions are identified using raster ordering.

The data from M12 and M72 were reduced in a similar fashion using –42.200 and

–345.100 respectively as the barycentric corrections for these objects. Again, no masers were detected in the profiles for these globular clusters and the reduced data is presented in Table 1 for comparison with the M13 and WX SER results.

The results of the study show that in an examination of the three globular clusters M13, M12 and M72, we were unable to detect SiO maser emission using the VLA down to a lower limit of approximately 0.3, 0.2 and 0.15 Jy for M13, M12 and M72 respectively.

Acknowledgements

I would like to thank Dr. A.J. Beasley, Dr. M.J. Claussen and Dr. J. Wrobel of the National Radio Astronomy Observatory, and Dr. S.W. Teare of New Mexico Tech for the many useful discussions. In addition, I would like to express my appreciation to the National Radio Astronomy Observatory for the observing time, as well as for the opportunity to act as the Array Operator of the Very Large Array for several hours of the observations. I would also like to thank G. Clark and B. Duff for the use of the 14" TIE telescope at Mt. Wilson Observatory. Finally, I would like to thank Dr. Tony Beasley for giving me and nine other students of the Swinburne Astronomy Online Astronomy Master’s Program the opportunity to participate in this research program.

References

Buhl, D., Snyder, L.E., Lovas, F.J., Johnson, D.R. 1974, ApJ 192, 97.

Cohen, N.L. & Malkan, M.A. 1979, ApJ 84, 74.

Dickey, J.M. & Malkan, M.A. 1980, AJ 85, 145.

Frail, D.A. & Beasley, A.J. 1994, A&A 290, 796.

Kaifu, N., Buhl, D., Snyder, L.E. 1975, ApJ 195, 359.

Knapp, G.R. & Kerr, F.J. 1973, AJ 78, 458.

Ulvestad, J. & Goss, M. 1999, S&T V.98 No. 6, 36.