Optical Counterparts of an Ultraluminous X-Ray Source X-1 in NGC 2500

TL;DR

This study identifies and analyzes potential optical counterparts of the ultraluminous X-ray source X-1 in NGC 2500 using HST data, suggesting possible stellar types and accretion disk reprocessing effects.

Contribution

First identification and characterization of optical counterparts of ULX X-1 in NGC 2500 using archival HST data, including spectral and age estimations.

Findings

Two faint optical candidates identified as possible counterparts.

Candidates are likely B-type main sequence stars with ages around 45 Myr.

Spectral energy distributions suggest X-ray reprocessing in the disk.

Abstract

We present the results of a search for optical counterparts of ultraluminous X-ray source (ULX) X-1 in the nearby galaxy NGC 2500 by using archival images taken with the Hubble Space Telescope ({\it HST}) Wide Field Camera (WFC3)/UVIS. Four optical sources have been identified as possible counterparts within the 2$\sigma$ error radius of 0.3 arcsec in the images. However, only two of them were investigated as candidates for counterparts due to their point-like features and their identification in various filters. These two faint candidates have absolute magnitudes of $M_{\rm V}$ $\approx$ -3.4 and -3.7. Also possible spectral classes of them were determined as B type main sequence stars. The ages and the masses of the candidates from Color Magnitude Diagram (CMD) were estimated as 45 Myr and 7 $M_{\rm \odot}$, respectively. The spectral energy distributions (SEDs) of two candidates were modeled by a power-law spectrum with a photon index ($\alpha$) $\sim$1.5. The spectra with such slopes could be interpreted as an evidence of reprocessing of the X-rays in the outer part of the disk that generates optical emission.

Tables

Table 1. Table 1: Log of HST /WFC3/UVIS observations of ULX X-1 in NGC 2500

Filter	ObsID	Date	Exp.
			(s)
F275W	ICDM38030	2013-08-28	2382
F336W	ICDM38040	2013-08-29	1119
F438W	ICDM38050	2013-08-29	965
F555W	ICDM38060	2013-08-29	1143
F814W	ICDM38070	2013-08-29	989

Table 2. Table 2: The properties of the possible optical counterparts of ULX X-1.

Source	R.A.	Dec.	F275W	F336W	F438W	F555W	F814W	Mv	${\left(B-V\right)}_0$
A	8:01:48.11	+50:43:54.58	24.95$\pm$0.18	25.64$\pm$0.31	26.75$\pm$0.27	26.26$\pm$0.11	25.72$\pm$0.19	-3.74	0.48
B	8:01:48.14	+50:43:54.52	-	-	-	26.91$\pm$0.19	25.55$\pm$0.17	-3.09	-
C	8:01:48.14	+50:43:54.61	-	-	-	27.07$\pm$0.21	25.22$\pm$0.13	-2.93	-
D	8:01:48.15	+50:43:54.66	24.88$\pm$0.18	25.51$\pm$0.26	26.61$\pm$0.24	26.59$\pm$0.14	26.87$\pm$0.51	-3.39	-0.36

Full text

††pubyear: 2019††pagerange: Optical Counterparts of an Ultraluminous X-Ray Source X-1 in NGC 2500–Optical Counterparts of an Ultraluminous X-Ray Source X-1 in NGC 2500

1 Introduction

Ultraluminous X-ray sources (ULXs) are variable X-ray sources located away from the nucleus of galaxies. These sources emit X-rays with isotropic luminosities in excess of $10^{39}$ erg s*-1*. This corresponds to approximately the Eddington limit for 10 $M_{\rm\odot}$ Black Hole (BH) (for recent review see Kaaret et al. (2017)). There are several current models to explain the observed luminosities of this particular class of objects. One such model suggests ULXs with a stellar mass black hole $M_{\rm BH}$ $\leq$ 20 $M_{\rm\odot}$ can be powered by either a super Eddington accretion rate and/or a strong beaming or any combination of both (Poutanen et al., 2013; King, 2009; Motch et al., 2014; Fabrika et al., 2015). Alternatively, ULXs can also be intermediate-mass black holes (IMBHs), $10^{2}$ $M_{\rm\odot}$ $\leq$ $M_{\rm BH}$ $\leq$ $10^{4}$ $M_{\rm\odot}$, which are accreting at sub-Eddington rates (Colbert & Mushotzky, 1999; Farrell et al., 2009, Mezcua et al., 2018).

Furthermore, recent discoveries of four ULXs exhibiting coherent pulsations which are powered by neutron stars (M82 X-2 Bachetti et al., 2014; NGC5907 ULX-1, Israel et al., 2017a; NGC 7793 P13, Fürst et al., 2016; Israel et al., 2017b and NGC 300 ULX1, Carpano et al., 2018). Hence, the debate on the true nature of ULXs is continuing.

Multiwavelength observations help us understand the nature of these binary sources. In addition to X-rays, their optical (including infrared and ultraviolet) and radio observations provide important information to constrain on masses, emission mechanisms and interactions with their environments. Determination of optical counterparts allow us to measure rotation curves of the companion stars. However, since ULXs are faint sources in optical band ($m_{\rm V}$ = 21$-$26) (Tao et al., 2011), it is difficult to measure their radial velocity curves, hence, determination of their mass functions. By making use of large and up to-date ground based and Hubble Space Telescope (HST), optical counterparts of about 20 ULXs have been identified (e.g. Grisé et al., 2008; Yang et al., 2011; Gladstone et al., 2013; Motch et al., 2014; Avdan et al., 2016b; Urquhart & Soria, 2016; Vinokurov et al., 2018). Such studies give important hints to estimate the origin of optical emission, which has strong contribution from the accretion disk as well as donor star. It is also possible to obtain information about various properties of the donor stars (e.g. their ages, masses, and spectral types) and their environments.

In this work, we present the results of a search for optical identification of one of the ULXs (X-1) of the galaxy NGC 2500 by using the archival images of HST/WFC3/UVIS camera with various filters. NGC 2500 is a spiral galaxy (type SB(rs)d) with a low surface brightness at a distance of $\sim$10 Mpc (Tully, 1988). Two ULXs were catalogued by Swartz et al. (2011) in this galaxy. X-1 is located 52*′′* away from the center with a Chandra coordinates, R.A. = $08^{h}$ $01^{m}$ $48^{s}.10$, Dec. = $+$ $50^{\circ}$ 43 $\mkern-4.0mu{}^{\prime}$ $54\aas@@fstack{\prime\prime}60$ having an unabsorbed X-ray luminosity of $5\times$1039 erg s*-1* in the 0.3$-$10 keV energy band. Besides the fact that no optical counterpart of the ULX X-1 has been identified so far, López et al. (2017) also reported that a candidate counterpart of the source could not be detected in the near infrared. They estimated the apparent limiting magnitude as >20.16$\pm$0.06 in the H-band. Although X-2 has been classified as a ULX, Gutiérrez (2013) identified that it is a background AGN at $z=0.5140$ by investigating its spectral features.

The paper is organized as follows: observations and details of data analysis are described in Section 2. Discussion of the various optical properties of the ULX with a summary are given in Section 3.

2 Observations and Data Analysis

Archival data of NGC 2500 from HST/WFC3/UVIS were utilized to search for the optical properties of ULX X-1. The observation log given in Table LABEL:tab1. The three color SDSS image of the NGC 2500 with the approximate position of ULX X-1 are shown in Figure 1.

2.1 Astrometry and the Optical Counterparts

We used the HST/WFC3 F555W image (see Table LABEL:tab1) and the Chandra image (ObsID 7112) to obtain the astrometric correction for the precise source position. The required relative astrometric corrections between images were performed by the alignment of common reference sources in both images. Source detection steps were carried out using daofind in iraf for HST and wavdetect task in ciao for Chandra. Among the detected sources, two point-like sources were adopted as the reference objects for the calculation of the relative shift between HST and Chandra images. One of the reference sources is the previously cataloged QSO (R.A. =$08^{h}$ $01^{m}$ $57^{s}.85$, Dec. = $+$ $50^{\circ}$ 43 $\mkern-4.0mu{}^{\prime}$ 39.\mbox{!!{}^{\prime\prime}}80, (Flesch, 2015)) and the other is the center of the galaxy. These sources have X-ray counts (with statistical error radii) of 22 ($0\aas@@fstack{\prime\prime}08$) and of 7 ($0\aas@@fstack{\prime\prime}12$), respectively. The shifts between HST and Chandra images were derived $-0\aas@@fstack{\prime\prime}04$ for R.A and $0\aas@@fstack{\prime\prime}16$ for Dec. Then the corrected position of X-1 is determined as R.A.= $08^{h}$ $01^{m}$ $48^{s}.14$, Dec.=$+$ $50^{\circ}$ 43 $\mkern-4.0mu{}^{\prime}$ 54.\mbox{!!{}^{\prime\prime}}55 with a 2$\sigma$ positional error radius of 0$\aas@@fstack{\prime\prime}$3. Four possible optical counterparts were identified within this error circle. The corrected position of X-1 on HST/WFC3/UVIS images together with the possible optical counterparts are shown in Figure 2. These candidates are labeled as A, B, C and D according to increasing R.A. coordinates. Similar multiple optical counterparts for ULXs were also reported for various other galaxies (Kaaret, 2005; Feng & Kaaret, 2008; Wang et al., 2015)

2.2 Photometry

The Point Spread Function (PSF) photometry was performed with dolphot (v2.0) package with WFC3 modules. The data files (*flt.fits and *drz.fits) were retrieved from the HST data archive111https://archive.stsci.edu/hst/search.php. Data reduction steps were followed as given in the dolphot manual Dolphin (2000). The wfc3mask task was used to remove pixels flagged as bad and multiplies the pixel values by the pixel areas to convert to number of electrons on the raw images (designated .flt or .flc). Splitgroups and calcsky tasks were performed to split the UVIS image files into a single file for each chip of WFC3 and to create a sky image, respectively. Then dolphot task was used to detect the optical sources, their photometry, and their photometric conversion. The dolphot zero points and aperture corrections were applied to the results as default. As a result of the processes the magnitudes were calculated in the VEGA magnitude system. The galactic extinction towards NGC 2500 is $E(B-V)=0.037$ (Schlafly & Finkbeiner, 2011). The coordinates, reddening corrected VEGA magnitudes, absolute magnitudes and colors of candidates are given in Table 2. Candidates B and C have not been clearly detected in F275W, F336W and F438W filters, moreover, they have extended features. Hence, these candidates were not considered as possible optical counterparts. At about 70′′ southwest of the galaxy center, there is the bright red star 2MASS J08014853+5042433. The effect of the illumination of this star is quite noticeable, especially in F555W and F814W filters. Therefore, these two filters are not used for further analyses.

The Color-Magnitude Diagram (CMD) was obtained for age and mass estimation of each possible optical counterparts. For this, PARSEC isochrones were downloaded from the web222http://stev.oapd.inaf.it/cgi-bin/cmd. These isochrones were selected with solar metallicity and HST/WFC3/UVIS wide filters photometric system. The galactic reddening and distance modulus (30.02 mag) were used to produce the CMD. The resultant CMD, F336W (U) versus F336W$-$F438W (U$-$B), for possible optical counterparts of X-1 with the field stars ($\approx$ 500) are given in Figure 3.

2.3 Spectral Energy Distributions

Spectral energy distributions (SEDs) for two candidates constructed from HST photometry are plotted in Figure 4. The SEDs (reddening corrected) for counterparts A and D are adequately fitted with simple power-law models with Fν $\propto$ $\nu$α, where Fν in units of erg cm*-2* s*-1* Hz*-1*. The power-law indices ($\alpha$) are 1.56$\pm$0.17 for A and 1.40$\pm$0.17 for D. These fits are acceptable with $R^{2}$ of 0.96 and 0.94 where, $R^{2}$ is goodness-of-fit for a linear model. As aforementioned, we obtained a good fit by using only the F275W, F336W and F438W values to constrain the fits. We note that these indices are consistent with the most of the ULXs in the sample given by Tao et al. (2011) ($\alpha$ = -1 to 2). Such indices show the Rayleigh-Jeans tail of reprocessed disk emission with temperatures below about 6000 K (Gierliński et al., 2009).

3 Results and Discussion

We have studied the optical properties of ULX X-1 in the galaxy NGC 2500 for the first time using the archival data from HST/WFC3/UVIS. Four possible optical counterparts were identified within $\sim 0\aas@@fstack{\prime\prime}3$ error radius after the astrometric correction. However, two of them (B and C) were not investigated as possible optical counterparts of ULX X-1 due to their extended features and their absence in filters (F275W, F336W and F438W). Colors and absolute magnitudes ($M_{\rm V}$) of all possible candidates were obtained from the available data. It is noted that dereddened magnitudes of counterparts are quite faint when compared with the known ULXs optical counterparts. $M_{\rm V}$ values of these sources are consistent with the optical counterparts of ULXs in Table 4 of Tao et al. (2011), which lie in the range -7 < $M_{\rm V}$ < -3. Assuming the optical light is dominated by the donor star, the spectral types of A and D were determined as early B type main sequence stars using the intrinsic colors and the absolute magnitudes in the Schmidt-Kaler Table (Aller et al., 1982).

As given in Table 2, two possible optical counterparts (A and D) have been detected in all images, making them the most noteworthy candidates. In their nearby environment any star group or cluster have not been observed within a radius of about 5*′′* region. We obtained a CMD ((U$-$B) versus B) to estimate the ages and masses of the candidates by using the mean extinction value as $E(U-B)=0.034$. The magnitudes and colors derived from HST/WFC3/UVIS data, compatible with those used in the available PARSEC isochrones and thus the derived age and mass values of A and D are almost the same as 45 Myr and 7 $M_{\rm\odot}$, respectively. The dereddened (U$-$B) color value is around -1 for both counterparts. It seems that counterparts A and D are not clearly bluer than the field stars.

In order to examine possible nebulae and the star formation regions around ULX X-1 we have used the ground-based continuum subtracted $\mathrm{H}\alpha$ image (NGC$\_$2500$\_$I$\_$Ha$\_$d2009.fits) available at NED (NASA/IPAC Extragalactic Database). This image was obtained from 1.8 m The Vatican Advanced Technology Telescope (VATT) with an exposure time of 1800 s of NGC 2500. This image is given in Figure 5. Two bright regions are noticeable on the figure with a distance of 13*′′* $(1^{\prime\prime}=48.4$ pc) to the position of X-1. Using $\mathrm{H}\alpha$ flux of the galaxy given as ${\log}f=-11.60$ (Kennicutt et al., 2008) and the $\it{aper}$ function in IDL (Interactive Data Language) program, a total of 1.9$\times 10^{6}$ counts (ADUs) were obtained from the $\mathrm{H}\alpha$ image for the entire galaxy. The errors of the counts are about 2$\%$ and 1 count corresponds to 1.3$\times 10^{-18}$ erg cm*-2* s*-1*. The contours are also plotted on $\mathrm{H}\alpha$ image. The source ULX X-1 falls on the contour value of 1.14$\times 10^{-16}$ erg cm*-2* s*-1* which corresponds to a luminosity value of 1.40$\times 10^{36}$ erg s*-1* at a distance of 10 Mpc. HII regions from the catalog of Zaragoza-Cardiel et al. (2015a) are shown also on the image (Figure 5). An HII region is seen at about 1*′′* away from ULX X-1. The luminosity of this nearby HII region is given as 1.15$\times 10^{37}$ erg s*-1* from catalog (Zaragoza-Cardiel et al., 2015b). This $\mathrm{H}\alpha$ luminosity is approximately ten times greater than the contour level the nearest to the ULX X-1. This decrease in $\mathrm{H}\alpha$ flux is clearly seen on the figure. Considering the nearby significant emission region, we think that a strong $\mathrm{H}\alpha$ emission cannot be originated from this source.

In many cases, there are possibilities of confusion between emissions from ULXs and background AGNs. Therefore, we also tried to discriminate such a confusion by estimating a ratio of X-ray to optical flux for A and D counterparts. Since simultaneous optical and X-ray observations do not exist, we used the available Chandra data (ObsID 7112) for X-ray and F555W data for optical. Assuming that the X-ray flux did not change at the time of F555W observation, the ratio was calculated by following formula log($F_{\mathrm{X}}/F_{\mathrm{opt}}$) $=$ $F_{\mathrm{X}}$ + mV/2.5 + 5.37, where mV is the extinction corrected magnitude in F555W band and $F_{\mathrm{X}}$ is unabsorbed X-ray flux in 0.3$-$3.5 keV enery band. To determine $F_{\mathrm{X}}$ as 2.0$\times 10^{-13}$ erg cm*-2* s*-1* , the Chandra PIMMS toolkit was used with an absorbed power-law index of 1.7 and avarage Galactic column density NH$=$4.7$\times 10^{20}$ cm*-2* (1990ARA&A..28..215D). The log($F_{\mathrm{X}}/F_{\mathrm{opt}}$) values for A and D were found to be 3.3 and 3.2, respectively. These are significantly greater than the accepted range between -1 to 1.7 for AGNs besides BL Lac objects, normal galaxies and normal stars (Maccacaro et al., 1988; Stocke et al., 1991). As we mentioned above, even though the bright red star’s illumination effect on F555W (V) is present. we used mV value to calculate flux ratio. However, when considering the flux corresponding to the V wavelength from the power-law model the specified ratio is $\sim$ 3.6 and it still exceeds the upper limit for the give range. The log($F_{\mathrm{X}}/F_{\mathrm{opt}}$) value of our two sources are similar to other known ULXs’s (Feng & Kaaret, 2008; Tao et al., 2011; Yang et al., 2011; Avdan et al., 2016a; Avdan et al., 2016b).

The optical emission from a ULX source might originate from its donor star or from the accretion disk around the compact object or both. The criterion to differentiate between X-ray binaries with high mass (HMXB) and low mass (LMXB) is defined by the formula $\xi=B_{\rm 0}+\log(F_{\rm\mathrm{X}})$ where B0 is the magnitude in B band and $F_{\mathrm{X}}$ is the 2$-$10 keV X-ray flux in units of $\mu$Jy the optical to X-ray fluxes. The $\xi$ values for HMXBs and LMXBs were found to be in the range of 12$-$18 and 21$-$22, respectively (Kaaret et al., 2017). The calculated $\xi$ of the possible optical counterparts are about 20 i.e., closer to values of LMXBs. Other known ULXs also have $\xi$ values quite similar to LMXBs. Therefore, our analysis indicate that ULX X-1 has an emission mechanism similar to an LMXB for which the optical emission mostly comes from its accretion disk.

The resultant SEDs were fitted with power-law models. Best-fitting parameters of the model indicate that their optical emission is probably generated from reprocessing of the X-rays in the outer part of the disk (Tao et al., 2011). In our case, we do not have data for an optical variability check, which would provide a further support for the disk origin of the optical emission.

As a conclusion, we see that there are still several uncertainties in our understanding of the nature of ULXs. Therefore, future high-resolution observations in multiwavelenghts need to be encouraged for better interpretations.

Acknowledgements

We acknowledge support from the Scientific and Technological Research Council of Turkey (TUBITAK) through project no. 117F115.