Tuesday, August 12, 2014

X-ray diagnostics of the donor star in ultra-compact X-ray binaries

Fig. 1: An artist’s impression of an accreting Low Mass X-ray Binary. The donor star fills its Roche lobe and its material overflows the inner Lagrangian points and accretes on the relativistic star (in this case a black hole). Due to the large angular momentum of the infalling material an accretion disk is formed around the compact object. Credit: ESA 2002/medialab

Fig. 2: A sketch of the innermost part (~1000 gravitational radii) in a low mass X-ray binary in the so called hard state. The inner part of the accretion flow is filled with hot and tenuous, optically thin plasma. Comptonization of the low frequency radiation in the plasma cloud is the main mechanism of the spectral formation in this state. Some fraction of this radiation illuminates the surface of the accretion disk and of the donor star. It is reprocessed by the material of the accretion disk and of the donor star giving rise to the so called ‘reflected component’, depicted in Fig.3. Credit: Gilfanov M., 2010, in Belloni T., ed., Lecture Notes in Physics, Vol. 794, The Jet Paradigm. Springer-Verlag, Berlin, p. 17

Fig. 3: The spectrum of the reflected component for an accretion disk of solar abundance. Superposed on top of the reflected continuum produced by Compton scatterings on electrons, are absorption edges and fluorescence lines of various elements. Also shown is the Comptonized continuum produced by the hot plasma cloud in the vicinity of the compact object (see Fig.2). An observer near the Earth will observe the sum of the two components

Ultra-compact X-ray binaries are a small but fascinating subclass of low-mass X-ray binaries, in which the donor is a white dwarf - a remnant of a moderately massive normal star. In order to understand the formation and evolution of these systems, it is critically important to identify the nature of the donor, which can be made of either helium or carbon and oxygen. MPA scientists have recently proposed and tested using - XMM-Newton observations - a principally new method to answer this question by the means of X-ray spectroscopy. 

Low mass X-ray binaries (LMXBs) are stellar systems consisting of two stars, one of which is a relativistic object - a neutron star or a black hole - and the other is a normal low-mass star, like our Sun, for example (Fig.1). If the separation between the two objects is comparable to the size of the normal star (which is hundreds thousands to millions of times larger than its relativistic companion), it may overfill it’s Roche lobe - the region of space where dynamics of matter are dominated by the gravitational attraction of the star. Consequently, it will start losing its outer layers under the gravitational pull of the second star. Material is predominantly lost through the so called inner Lagrangian point - the point on the line connecting the two stars where the forces of gravity and the centrifugal force balance each other out. The material of the donor star will flow through this point and will fall into the gravitation potential well of the relativistic star, initiating the process which is called accretion. Due to its large angular momentum, the infalling matter will form an accretion disk around the relativistic object (Fig.1). The classical theory of accretion disks around black holes and neutron stars was developed by Nikolai Shakura and Rashid Sunyaev in 1972. Due to the small size of the relativistic object (~15 km for a neutron star) the gravitational energy released during accretion constitutes a significant fraction of the rest mass energy of the accreting material, typically about 5-20%. This makes these systems very luminous sources of X-ray emission. 

There is a small but fascinating subclass of low-mass X-ray binaries, called Ultra-compact X-ray binaries (UCXBs) in which the donor star is a white dwarf - a remnant of a moderately massive normal star. These systems are extremely compact (hence their name) and have orbital periods shorter than 40 minutes, the fastest one having a period as short as 11 minutes. 

An interesting feature of these systems is that the chemical composition of the donor star is dramatically different from the composition of the donor star in ‘normal’ low-mass X-ray binaries. While donor stars in normal LMXBs have chemical composition similar to our Sun, i.e. are made of mostly hydrogen and helium with small admixture of metals, UCXBs feature donors that are depleted of hydrogen. They can be made of the ashes of nuclear burning of hydrogen (mostly helium and nitrogen), of helium (mostly carbon and oxygen) or carbon (mostly oxygen and neon). 

Depending on the particular evolutionary path through which UCXBs form, they may have a variety of donors ranging from non-degenerate helium stars to white dwarfs. It is critically important to distinguish between these possibilities, in order to understand the processes that lead to UCXB formation and control their evolution. So far this task has been performed using methods of optical astronomy, with various degrees of success. 

MPA scientists have recently proposed and tested a principally new method of diagnostics of the nature of the donor star in UCXBs by the means of X-ray spectroscopy. 

The method is using the phenomenon called X-ray reflection. A fraction of the emission produced near the compact objects illuminates the surface of the accretion disk and the donor star (Fig.2) and is reprocessed by this material. In the jargon of high energy astrophysics this reprocessed emission is called “reflected component”. An example of its spectrum is shown in Fig.3

On top of the continuum produced by the Compton scatterings off electrons in the accretion disk, the reflected component also contains a number of characteristic lines. These lines (called emission lines) are due to the different chemical elements present in the accreting material. They are produced by the process called fluorescence and have well known energies, unique for each chemical element. Their shape and relative strength carry information about the geometry of the accretion flow and chemical composition of the accreting material. 

The reflected component is heavily diluted by the primary emission, therefore the fluorescent lines of most of the elements are very weak and difficult to detect. Except for the fluorescent line of iron, which in the case of neutral iron is located at 6.4 keV. Thanks to the high fluorescent yield and abundance of iron, this is the brightest spectral feature in an otherwise relatively smooth continuum. All normal LMXBs have this line easily observable in their X-ray spectra. 

While the reprocessing of X-ray radiation by the accretion disc and particularly the shape and strength of the iron line has been thoroughly investigated since 1970s, all prior work concentrated on accretion disks of nearly solar abundance of elements, with only moderate variations of the element abundances considered in a few papers. MPA scientists have now taken the first step in modeling X-ray reflection off hydrogen poor material with anomalous abundances, as expected in the accretion disks in Ultra-compact X-ray binaries. The model developed using the Monte Carlo technique is the first simulation of reflection spectra of C/O, O/Ne/Mg or helium rich disks. 

Using these simulations, MPA scientists came to a paradoxical conclusion: The strongest and most easily observable effect of the hydrogen poor, C/O rich material is not an appearance of strong fluorescent lines of carbon and oxygen - as one might expect - but nearly complete disappearance of the fluorescent line of iron! This is caused by the screening of iron by the much more abundant carbon and oxygen. 

In a neutral material of solar abundance, the most likely process for a photon with energy exceeding 7.1 keV - the photoionisation threshold of K-shell electrons in iron (so called K-edge) - is absorption by iron due to the photoionisation of its atoms. Photoionisation of iron is followed in about one-third of the cases by the emission of a 6.4 keV fluorescent photon. Consequently, the majority of photons with energies above this threshold will be absorbed by iron and will, therefore, contribute to its fluorescent line. 

In the case of a C/O (or O/Ne) white dwarf though, the overwhelming overabundance of oxygen makes it the dominant absorbing agent even at energies far beyond its own K-edge, leaving only a few photons to fuel the iron fluorescent line. Although the fluorescent line of oxygen produced in the process is significantly boosted, it is still strongly diluted by the primary continuum and therefore is difficult to detect. A much more visible effect is the significant attenuation or complete disappearance of the iron line. 

Helium, on the other hand, is not capable of screening iron, due to its smaller charge and, correspondingly smaller absorption cross-section at the iron K-edge. Therefore in the case of a helium-rich donor reflection proceeds ‘as usual’ and the iron line has its nominal strength. 

This opens an exciting possibility to discriminate between helium and oxygen rich donors by means of X-ray spectroscopy. MPA scientists calibrated the method using extensive Monte-Carlo simulations, investigated its luminosity dependence and proposed observational tests of the picture. They used the data of XMM-Newton satellite to verify results of theoretical calculations using observations of UCXB systems with a donor star of known composition. Furthermore, they provided tentative identifications of the donor star in several ultra-compact binaries, where its nature remained so far unknown.

Filippos Koliopanos and Marat Gilfanov


1. Koliopanos F., Gilfanov M., Bildsten L., 2013, MNRAS, 432, 1264

2. Koliopanos F., Gilfanov M., Bildsten L., M.Diaz Trigo, 2014 MNRAS