«In press in the ApJ Letters The Helium White Dwarf orbiting the Millisecond Pulsar in the halo of the Globular Cluster NGC6752 1 Francesco R. ...»
In press in the ApJ Letters
The Helium White Dwarf orbiting the Millisecond Pulsar in the
halo of the Globular Cluster NGC6752 1
Francesco R. Ferraro2, Andrea Possenti3,4, Elena Sabbi2, Nichi D’Amico3,5
Dipartimento di Astronomia Universit` di Bologna, via Ranzani 1, I–40127 Bologna,
Osservatorio Astronomico di Cagliari, Loc. Poggio dei Pini, Strada 54, I–09012
Osservatorio Astronomico di Bologna, via Ranzani 1, I–40127 Bologna, Italy 5 Dipartimento di Fisica Universit` di Cagliari, Cittadella Universitaria, I-09042 a Monserrato, Italy
Subject headings: Globular clusters: individual (NGC6752); stars: evolution – pulsars: individual (PSR J1911−5958A) – binaries: close.
1 Based on observations collected by using the Very Large Telescope, at the European Southern Observa- tory, Cerro Paranal, Chile, within the observing programme 71.D-0232.
1. Introduction PSR J1911−5958A has been discovered on 1999 October 17 during a search for Mil- lisecond Pulsar (MSPs) in Galactic Globular Clusters (GGCs) in progress at the Parkes Radiotelescope (D’Amico et al. 2001). It is a binary millisecond pulsar with a spin period of
3.26 ms, an orbital period of 0.84 days and very low eccentricity (e 10−5 ). Precise celestial coordinates (RA = 19h 11m 42.756, DEC = −59◦ 58 26. 91) have been recently obtained for s this source from pulsar timing observations (D’Amico et al. 2002). This position is far away (∼ 6 ) from the cluster optical center: indeed PSR J1911−5958A is the more oﬀ-centered pulsar among the sample of 44 MSPs whose position in the respective cluster is known, and it suggests that this object might be the result of strong interactions occurred in the cluster core.
Colpi et al. (2002) explored a number of possibilities for the peculiar location of PSR J1911−5958A: a careful analysis led to discard the hypothesis of a primordial binary (born either in the halo or in the core of the cluster) and to reject also the possibility of a 3body scattering or exchange event oﬀ core stars. Hence they conjectured that a more massive target (?, either a binary or a single intermediate-mass black hole, see also)]colpi03a,colpi03b could have provided the necessary thrust to propel PSR J1911−5958A into its current halo orbit at an acceptable event rate.
Moreover, beside PSR J1911−5958A, NGC6752 hosts also the second most displaced MSP ever seen in a globular (PSR J1911−6000C) and at least two (out the three) MSPs located in the cluster center (?, PSR-B and PSR-E, see Figure 2 in)]ferraro03 experience strong acceleration, implying an unusually high central mass to light ratio (M/L ∼ 6) (D’Amico et al. 2002; Ferraro et al. 2003). Ferraro et al. (2003) showed that NGC6752 is a dynamically evolved cluster probably undergoing a post-core-collapse bounce and investigated scenarios for simultaneously explaining both the anomalous acceleration of PSR-B and PSR-E and the ejection of PSR J1911−5958A and PSR J1911−6000C, concluding that the existence of a binary black hole of intermediate mass (Colpi et al. 2002) could be a viable possibility.
Optical detection of the companion to a millisecond pulsar in a globular cluster proved particularly helpful in assessing the origin and the evolution of the binary, besides supporting its cluster membership (?, see, e.g.,)]edmonds01,ferraro01a,edmonds02. In fact, unlike the systems in the galactic ﬁeld, the age, metallicity, extinction, distance and hence intrinsic luminosity and radius can be estimated from the parent cluster parameters. Hence, we undertook a systematic program (Ferraro et al. 2001a,b) devoted to the optical identiﬁcation of MSPs companion in GGCs; as a part of this project in this Letter we present the identiﬁcation of the optical counterpart to the PSR J1911-5958A companion.
2. Observations and data analysis
The photometric data presented here consist of a set of high resolution images obtained by using the camera FORS1 mounted at the ANTU Unit Telescope 1 (UT1) of the Very Large Telescope (VLT) at ESO on Cerro Paranal (Chile) on three nigths in March, April and May 2003. All the observations have been performed in the High Resolution (HR) mode of FORS1. In this conﬁguration the plate-scale is 0. 1/pixel and the FORS1 2048 × 2048 pixel2 array has a global ﬁeld of view of 3.4 × 3.4. The data comprise eight 220 s V -band exposures, ﬁve 360 s B-band exposures and three 1500 s U-band exposures, roughly centered on the PSR J1911−5958A (hereafter MSP-A) nominal position. All the observations were performed in service mode under good seeing conditions (FWHM= 0. 5 − 0. 7).
A sub-image of 800 × 800 pixels centered on the nominal position of the MSP-A has been extracted from the original frames and carefully analized. All the reductions have been carried out using ROMAFOT (Buonanno et al. 1983), a package speciﬁcally developed to perform accurate photometry in crowded ﬁelds, allowing for a visual inspection of the quality of the PSF-ﬁtting procedure. In order to optimize the search for faint objects a median-combined image in each band has been obtained and the search procedure has been performed on the deepest combined image. Then the masks with the star positions have been adapted to each image and the PSF-ﬁtting procedure performed on each individual frame separately. The resulting instrumental magnitudes have been transformed to a common photometric system and then averaged. Thus, a ﬁnal catalogue with the coordinates and the average instrumental magnitude in each ﬁlter has been compiled for all the stars identiﬁed in the considered sub-image. Photometric calibration of the instrumental magnitudes (in the B and U band) was obtained using four photometric standard stars (Landolt 1992) secured under photometric conditions. Since no accurate calibration of the V ﬁlter was possible with the data secured in service, we calibrated the V magnitude by using ∼ 200 stars in common with the B, V catalog recently published by Ferraro et al. (2003). The stars in common between the two catalogs permitted also an independent check of the calibration obtained in the B band, displaying agreement within a few hundredths of magnitude.
The stars identiﬁed in the considered sub-image have been reported to an absolute astrometric system by using the 200 stars in common with the Ferraro et al. (2003) catalog (already astrometized). The details of the procedure adopted to derive the astrometric solution are described in other papers (Ferraro et al. 2001b, 2003). In short, the new astrometric Guide Star Catalog (GSCII) was used to search for astrometric standard stars lying in the considered ﬁeld of view (FoV). At the end of the procedure, the rms residuals were of the order of ∼ 0. 3 both in RA and Dec and we assume this value as representative of the astrometric accuracy.
Figure 1 shows the U, U − V and B, B − V Color Magnitude Diagram (CMD) for the stars (large ﬁlled circles) identiﬁed in the 80 × 80 FORS1 sub-image considered here. Most of the stars trace a clean and well deﬁned main sequence (MS) spanning almost 8 mags in the U band reaching U ∼ 26. Only a few sparce objects, showing a blue excess, are visible on the left side of the MS. The most extreme blue objects in the CMDs are CO white dwarfs (WDs): once the MS is matched, they nicely overlaps the position of a CO-WDs population (shown as small empty circles in Fig. 1) observed in this cluster by Renzini et al. (1996) and ?)][2003b, in preparation]ferraro97. In particular, the colors and the luminosities of the three CO-WDs found here and of the previously observed population agree with the theoretical cooling sequence for 0.5M hydrogen rich WDs (Wood 1995) (drawn as heavy dashed line in Fig. 1).
On the basis of the accurate astrometric positions obtained from the photometric catalog, we identiﬁed a blue object (hereafter COM J1911−5958A) lying at only 0. 1 from the nominal position of PSR J1911−5958A. Finding chart (in the B-band ﬁlter) for COM J1911−5958A is shown in Fig. 2. Only a few objects (∼ 10) are lying in the region of the CMD to the left of the MS: considering the FoV of the sub-image, there is a very small probability ( ∼ 5 × 10−3 ) of detecting one of the blue objects just by chance in a circular aperture of radius ∼ 0. 3 (corresponding to the uncertainty in the relative radio-optical astrometry) centered on the pulsar position.
The position of COM J1911−5958A is marked as a large empty square in the CMDs in Fig. 1, whereas absolute coordinates and magnitudes are listed in Table 1. Taking into account the expected variability of this object and the global uncertainty in the calibration of the photometric zero-point, we conservatively adopted an overall uncertainty in the reported magnitudes of 0.15 mag. While we were writing this Letter a not yet refereed paper by Bassa et al. appeared on the web (astro-ph/0307340) also presenting the optical identiﬁcation of this object. Though the identiﬁcation is consistent with that presented here, it can be considered, at most, a preliminary detection since the object is nearly at the detection limit in the exposures utilized in that paper. The deep VLT observations presented here allow us to measure the object with a quite high S/N ratio (70 − 120) and to derive the photometric properties of the star with much higher accuracy.
The location of COM J1911−5958A in the CMDs (∼ 1.5 mag bluer than the MS in the diagram) excludes it is a CO-WD whilst it resembles Uopt, the companion to PSR J0024– 7203U in 47 Tuc (Edmonds et al. 2001), which was suggested (Edmonds et al. 2001) to be a low mass Helium core WD (He-WD) from the comparison with theoretical models by Serenelli et al. (2001). However, the He-WD tracks used by Edmonds et al. (2001) were –5– computed for progenitors with high (solar) metallicity (Z = 0.02).
Here we have applied the same procedure taking advantage of a new set of tracks speciﬁcally computed (Serenelli et al. 2002) for globular cluster applications, i.e. assuming progenitors with low metallicities. In particular, we used the cooling tracks at Z = 0.001, which are the closest to the cluster metallicity ([F e/H] = −1.43 ± 0.04) as recently derived by Gratton et al. (2003). The cooling sequences for two He-WD masses (0.172 and 0.197 M, respectively) are over-plotted in Fig. 1 (light dashed lines). The models have been drawn by adopting a distance modulus (m − M)0 = 13.13 (?, from the homogeneous distance scale for 61 GGCs derived by)]ferraro99 and a reddening E(B − V ) = 0.04 (Harris 1996; Gratton et al. 2003). Note that the distance modulus adopted here is consistent both with the estimate obtained from the WD cooling sequence ((m − M)0 = 13.05 ± 0.1) by Renzini et al. (1996) and with the most recent determination ((m − M)0 = 13.12 ± 0.08) derived from MS ﬁtting by Gratton et al. (2003).
Inspection of Fig. 1 reveals that luminosity and color of COM J1911−5958A well agree with its being a He-WD of mass in the range 0.17 − 0.20 M. In particular, from the Z = 0.001 models of masses 0.172M and 0.197M by Serenelli et al. (2002), we derive the following estimates for the properties of COM J1911−5958A: a temperature in the interval Teﬀ = 10, 000 − 12, 000 K, a gravity log g = 6.12 − 6.38, a luminosity L = 0.03 − 0.04 L, a radius R = 3 − 4 × 109 cm and a cooling age in the range 1.2 − 2.8 Gyr.
We have identiﬁed the optical companion to the millisecond pulsar PSR J1911−5958A located at 3.3 half mass radii from the center of the globular cluster NGC6752. U, B and V magnitudes (and related colors) of the optical source are compatible with its being a Helium white dwarf hosted in the globular’s halo.
From the pulsar mass-function (?, 0.0026887 M,)]damico02, it results a minimum companion mass (corresponding to a system seen edge-on, i.e. with orbital inclination i = 90◦ ) MCOM = 0.185 M for a pulsar mass of 1.35 M (Thorsett & Chakrabarty 1999). Given the upper limit (MCOM ∼ 0.2 M ) of the range of masses of COM J1911−5958A inferred from the cooling tracks of Serenelli et al. (2002), we can constrain the inclination i to be larger than ∼ 70◦. Adopting a larger value for the pulsar mass would result in a even larger