I10: From evolving binaries to the merging of compact objects
More than 70% of massive stars experience a binary interaction at least once in their life (Sana et al. 2012). In the course of their evolution, one of the stars first becomes a compact object (white dwarf, neutron star or black hole), and, if close enough, attracts matter from its companion. The stars thus exchange both matter and angular momentum, through an energetic process called accretion: they become accreting, compact binaries (Chaty 2013). Such a pair of massive stars eventually evolves towards the merging of two compact objects.
This phenomenon, leading to the emission of gravitational waves, has been beautifully revealed on the 14th of September 2015 by the LIGO collaboration, arising from the merging of two heavy stellar mass black holes of ~30 solar masses (Abbott et al 2016ab). The two firm gravitational wave detections already announced likely constitute the tip of the iceberg: indeed, close binaries exist everywhere in our Universe, and should be detected when they merge and emit gravitational waves!
Most evolutionary models of binary stellar systems are based on the coupled evolution of two single, isolated stars. However, these evolutionary models are incomplete: while we do not fully understand the mechanisms governing stellar evolution, we know even less about the physical processes occurring in close binary systems, where both stars exchange matter and angular momentum. First, the common envelope phase, occurring very early in the evolution of a compact binary, is still both theoretically and observationally highly unknown. Second, the natal kick received at the supernova event is not constrained, especially for the black holes. Finally, the metallicity plays an important role in the strength of the stellar winds, which can cause the star to lose much of its mass. Therefore, as the full evolution of binaries towards merging is not fully understood, the current population synthesis models of binary systems in galaxies have a high degree of uncertainty, implying that the search to identify the merger progenitors is flawed.
In this interface project between AIM and APC, we propose to tackle this problem by computing the evolution of the current population of compact binaries known in our Galaxy, using new data obtained from the Gaia satellite, revolutionizing the field of astrometry by providing a totally new 6D view (position and velocity) of our Galaxy. Then, by comparing the latest evolutionary stages of compact objects with the predictions of current population synthesis models, we will be able to constrain the three biggest uncertainties of these models: the common envelope phase, the natal kicks, and the metallicity. We then plan to extrapolate our results to low-metallicity galactic environments, computing updated population synthesis models, to improve the predicted rates of compact object mergers, and thus of gravitational wave detections. In short, “yesterday’s binaries are today’s gravitational waves”!
POSITION NAME SURNAME LABORATORY NAME GRADE, EMPLOYER WP leader CHATY Sylvain AIM PR, Université Paris Diderot, USPC WP co-leader PORTER Edward APC DR, CNRS WP member CHASSANDE-MOTTIN Eric APC CR, CNRS WP member COLEIRO Alexis APC Post-doc WP member FOGLIZZO Thierry AIM IR, CEA WP member FORTIN Francis AIM PhD, Ecole Doctorale ED 560 WP member MARSHALL Douglas AIM MCF, Université Paris Diderot, USPC WP member MIRABEL Felix AIM IR, CEA
Summary of scientific program
More than 70% of massive stars experience a binary interaction at least once in their life (Sana et al. 2012). In the course of their evolution, one of the stars first becomes a compact object (white dwarf, neutron star or black hole), and, if close enough, attracts matter from its companion. The stars thus exchange both matter and angular momentum, through an energetic process called accretion: they become accreting, compact binaries (Chaty 2013). Such a pair of massive stars eventually evolves towards the merging of two compact objects. This phenomenon, leading to the emission of gravitational waves, has been beautifully revealed for the first time on the 14th of September 2015 by the LIGO-Virgo collaboration, arising from the merging of two heavy stellar mass black holes of ~30 solar masses (Abbott et al 2016ab). Since this first event, the handful gravitational wave detections constitute the tip of the iceberg: close binaries exist everywhere in our Universe, eventually detected when they merge and emit gravitational waves!
Most evolutionary models of binary stellar systems are based on the coupled evolution of two single, isolated stars. However, these evolutionary models are incomplete: while we do not fully understand the mechanisms governing stellar evolution, we know even less about the physical processes occurring in close binary systems, where both stars exchange matter and angular momentum. First, the common envelope phase, occurring very early in the evolution of a compact binary, is still both theoretically and observationally highly unknown. Second, the natal kick received at the supernova event is not constrained, especially for the black holes. Finally, the metallicity plays an important role in the strength of the stellar winds, which can cause the star to lose much of its mass.
To better constrain these three fundamental parameters, we propose to study the evolution of known binary systems in our Galaxy, taking into account new observational parameters obtained with the Gaia satellite (offering us a 6D-view –position and velocity– of our Galaxy, see Bailer Jones et al 2013), and compare the final products with the current output of population synthesis models. We will then use this template, with better constrained parameters (such as orbital separation, mass ratio, metallicity, timescale), to extrapolate the population of binary systems to low-metallicity environments, allowing us to better predict the rate of compact object mergers, and in particular of binary stellar-mass black hole mergers detectable by the LIGO-Virgo collaboration. The timeframe of our project is ideal, with 1. the 2nd Gaia data release in April 2018 ; and 2. LIGO/Virgo observatories obtained data during O2 (6 months in 2016-2017) and will resume observations during O3 (beginning in February 2019), likely detecting tens of additional binary black hole mergers and possibly also NS+BH and binary NS systems. Our project will thus actively contribute to the scientific exploitation and astrophysical interpretation of these new observations.
The work packages of our project are:
- Build a 6D-catalogue of known compact binaries, taking into account the new observational parameters obtained with the Gaia satellite: position and velocity, along with distance, proper motion, spectral type;
- compute the evolution of known accreting binaries in our Galaxy, for which we have enough parameters, through the classical isolated binary evolution channel (using the public and adapted code MESA);
- compare the accurate evolution of these binaries with the current output of population synthesis models, to better constrain the 3 most uncertain parameters: common envelope phase, natal kick, metallicity;
- Extrapolate from this template, with better constrained parameters, to compute the population of binary systems present in low-metallicity environments, with the binary star evolution (BSE) model;
- Predict the number of compact object mergers, and the rate of binary stellar-mass black hole mergers possibly emitting gravitational waves and detectable by the LIGO-Virgo collaboration.
We report below on our work concerning the work packages 1 & 2 of our project, which are nearly completed (10/2018).
Work Package 1 (WP1): Build a 6D-catalogue of compact binaries
We built a complete catalogue of all compact binaries, known in our Galaxy (and in SMC/LMC), called 6D since it will include the position (distance) and velocity, in addition to spectral type and extinction, by taking into account the new observational parameters obtained with the Gaia satellite (Bailer-Jones et al, 2013). Some of these parameters have been released in April 2018, in an adequate timeline for this work to be performed. We had already built such a preliminary catalogue in the pre-Gaia era, that we have successfully used to compute the distribution in our Galaxy of high-mass accreting binaries, with position and fitted distance, correlated with stellar formation complexes (Coleiro & Chaty 2013). The new catalogue we built is to our knowledge the most complete to date, both in terms of binary systems (including LMXB and HMXB systems), and of stellar parameters.
Step 1: Building a sample of High Mass X-ray Binaries (HMXB) to be cross–matched with GAIA DR2
A first sample was built by cross-matching various existing catalogues (Liu et al. 2006, Bird et al. 2016) of X-ray binaries, that were detected at high energies (HEAO, ASCA, INTEGRAL), containing a total of 154 HMXB and candidate HMXB. The next step was to check the literature for up-to-date results from recent studies on candidate HMXBs: new X-ray localization (Swift, XMM-Newton, Chandra), optical/infrared associations and spectral identification. A second crossmatch was made, this time with the Virtual Observatory (VO Simbad) to gather most of the updated results. Because not all the sources had sufficient constraints on their position, this was done using their various identifiers (up to three per source) since they are consistent within the literature throughout the years. Then, a manual check was performed on remaining candidates to ensure that all the latest publications were taken into account. In particular, some sources needed a careful check on the high-energy and optical associations and on the spectral identification references (not always up-to-date in Simbad). This update allowed us to discard 12 HMXB candidates as other types of X-ray sources (cataclysmic variables, low-mass X-ray binaries, AGN or even close-by active stars). After this crossmatch and check work, we now have 108 confirmed HMXB, which both have hard and soft X-ray positions, along with an unambiguous optical/infrared counterpart identified through spectroscopy. This leaves us with 39 candidate HMXB for which we do not have a clear spectral identification or localization yet.
Among those 108 confirmed HMXB, 94 have a position accurate enough to have a proper GAIA DR2 counterpart. The remaining 14 have near-infrared counterparts and are, for most of them, confirmed obscured systems that are too faint to be seen by GAIA in the visible domain. Among the 39 candidate HMXB, 8 have a GAIA DR2 counterpart, but still need a proper spectral identification to confirm their HMXB nature. We thus end up working on a sample of 108 confirmed and 39 candidate HMXB, and among them 94 confirmed and 8 candidate HMXB that have a counterpart in GAIA DR2. These 108 confirmed and 39 candidate HMXB are reported in Figure 1.
Figure 1: Galactic map of HMXB (confirmed and candidates) from our cross-matched catalogue
Step 2: GAIA DR2 view of HMXBs as gravitational wave progenitors
On 2018, April 22nd, Gaia Data Release 2 (DR2) was published (https://www.cosmos.esa.int/web/gaia/dr2), containing 1.7 billion stars. For 76% of the sources in the catalog, a five-parameter (5D) astrometric solution was released, including the position, parallax and proper motion. We carefully cross-matched our HMXB catalog with GAIA DR2 database to find the counterparts of sources of interest for this project. From the 157 sources we found 106 GAIA counterparts of which 92 are of great quality, containing 5D information, which was previously unknown for this HMXB population. To optimally perform the work of this project, the LabEx postdoc F. García participated during February 2018 in a GAIA School organized by the Paris Observatory, with specific training regarding working with GAIA mission data. This included talks related to mission specifications, science tools available and hands-on session on data analysis.
This allowed us to start one of the key packages of this project which consists in the study of the distribution of HMXB in the Milky Way Galaxy and Magellanic Clouds (substantially improving the previous work of reference in this domain, by Coleiro & Chaty, ApJ 2013). For this whole set of sources, we used the 5D information from GAIA DR2, by considering for the first time their peculiar velocity in the Milky Way, which allowed us to constrain the strength of the kick imposed by the first SN explosion occurring in these systems. Our preliminary results show that almost 90% of the population of HMXB show slow peculiar velocities compared to Galactic rotation, meaning that they only experience a low SN kick. On the contrary, roughly 10% of the population show a high peculiar velocity, not compatible with the Galactic rotation, and thus these systems might have experienced a significant kick during the SN event during which was formed the compact object. This part of the population includes binaries already classified as high-proper motion systems but, based on the proper motions derived from GAIA DR2 data, we were able to reveal several new sources. We are currently finishing our data analysis, and we have started to write a paper describing our results on this subject (García et al. 2018, in preparation). This paper will 1. contain an up-to-date version of our HMXB catalogue including GAIA counterparts, and 2. describe this powerful tool, by studying several properties of this massive binary population, which will serve as an input for parallel ongoing work package 2 involving binary evolution (see below).
Work Package 2 (WP2): Compute the evolution of Galactic accreting binaries
Based on this catalogue obtained in WP1, we aim to compute the accurate evolution of known accreting binaries of our Galaxy, for which we have enough parameters, through the classical isolated binary evolution channel, with the public stellar-evolution code MESA. In addition, we have already initiated a collaboration with Thomas Tauris, expert on the evolution of LMXB (Tauris et al., including Chaty ApJ 2017), and he agreed to help interpreting the results obtained when computing the evolution of particular binary systems.
MESA modelling of binary stellar evolution
We use MESA (Modules for Experiments in Stellar Astrophysics, http://mesa.sourceforge.net/) stellar evolution code to study the connection between massive stellar binaries, HMXBs, and GW progenitors. Currently, MESA is able to follow the evolution of binary systems composed of two normal stars or a normal star and a compact object, which is treated as a point source, considering for instance the impact of the stellar wind at different metallicities. Connecting different simulations, it is also possible to simulate the history of a system experiencing a SN explosion with an arbitrary kick, currently not modelled by MESA. Furthermore, current MESA releases do not include any treatment for the common-envelope phase, which is of main importance in the determination of the final fate of a compact binary as a GW progenitor.
We developed a set of subroutines under MESA that allow us to run the full evolution of a stellar binary from its formation until the SN explosion of both stars up to the formation of the double compact object which will be a GW progenitor, or its disruption which includes the self-consistent application of a natal kick imposed by a spherically-symmetric SN explosion that blows-up the stellar envelope, and converts the core mass into a compact core. This subroutine include a novel treatment for the common-envelope phase, based on the a-g prescription (see Ivanova et al. 2013 for a full review on the common-envelope phase), fully exploiting the capabilities of a stellar evolution code such as MESA which, contrary to binary population synthesis algorithms, computes the full structure of each star at each evolutionary phase.
To run our simulations, we make use of the ARAGO cluster facility at APC laboratory, which allow us to simultaneously explore a huge range of stellar parameters, in order to understand the formation history scenario of the GW events detected by LIGO/Virgo and its expected rates, in the context of stellar binary evolution in our numerical code. We have reached an accurate version of this subroutine that we are currently running on the cluster. In the next months we will analyze the results of our simulations, and based on this we will submit a paper on possible binary evolution channels to form GW events. We also plan to make publicly available our MESA subroutines (García et al. 2019, in preparation).
García, Federico; Fogantini, Federico A.; Chaty, Sylvain; Combi, Jorge A. (2018)
Spectral evolution of the supergiant HMXB IGR J16320-4751 along its orbit using XMM-Newton
A&A, 618, 61 http://adsabs.harvard.edu/abs/2018A%26A…618A..61G
F. Fogantini, F. García, J. Martí, P. Luque Escamilla, J. A. Combi, S. Chaty (2018)
Precessional evolution of Fe and Ni lines from the baryonic jets of SS433 as seen by NuSTAR
Barack et al. including E. Chassande-Mottin, E. Porter and Chaty S., 2018,
Black holes, gravitational waves and fundamental physics: a roadmap
CQG subm. (arXiv180605195)
Keivani A. et al., including Chaty S. and Coleiro A., 2018,
A Multimessenger Picture of the Flaring Blazar TXS 0506+056: : Implications for High-Energy Neutrino Emission and Cosmic Ray Acceleration
for the Astrophysical Multi-messenger Observatory Network ApJ, 864, 84
Fortin F., Chaty S., Coleiro A., Tomsick J.A., Nitschelm C., 2018,
Spectroscopic identification of INTEGRAL high-energy sources with VLT/ISAAC
A\&A in press (arXiv180809816)