Neutron star studies

Dr Silvia Zane


Massive stars (about 10 times heavier than our Sun) end their life in a gigantic explosion, a supernova, and leave behind a compact remnant, either a neutron star or a black hole.

Life cycle of massive star

With a mass comparable to that of the Sun packed in a sphere about 10 km in radius, neutron stars are as dense as an atomic nucleus and are the densest objects known in the present universe. Moreover, neutron stars host the strongest known magnetic fields, typically ten billion times that of a common, hand-held magnet. This large magnetic field coupled to the star rotation results in the emission of collimated beams of radio waves which at regular intervals sweep the Earth. This produces the most common observational manifestation of a neutron star: a pulsar.

A schematic view of pulsar A X-ray image of the nebula around the Crab pulsar
A schematic view of pulsar.

A X-ray image of the nebula around the Crab pulsar (the pulsar itself is visible at the centre).


Normal neutron stars (like radio-pulsars) are per se extreme objects and provide a unique laboratory to probe the properties of matter under conditions which will be never accessible to ground-based experiments, and which are not to be found in other astrophysical settings. Study of neutron star holds the key to answer such fundamental questions as if the ground state of matter is made of hadrons or of free quarks.

The two projects below deal with even more peculiar types of neutron stars, the magnetars (which stand for magnetic stars) and the so-called X-ray Dim Isolated Neutron Stars (or XDINSs for short). These two particular classes are extremely interesting, since understanding of their physical properties promises to test quantum-electrodynamics in the strong-field regime (in one case) and to probe directly the Equation of State of matter at supra-nuclear densities (in the second one), so to push our knowledge of basic physics exploring conditions not reproducible at Earth!

1. Magnetars

Over the last two decades observations at X-/gamma-ray energies revealed the existence of two peculiar classes of sources, the Soft Gamma Repeaters (SGRs) and the Anomalous X-ray Pulsars (AXPs). Both are X-ray pulsars with periods of a few seconds and are characterized by the emission of short, strong bursts. In addition three SGRs emitted a hyper-energetic giant flare, the strongest transient phenomena ever detected in our Galaxy. Most of them are powered by an ultra-magnetized neutron star, with a surface magnetic field as strong as 1014-1015 G, about 100-1000 times more intense than in ordinary neutron stars, and well in excess of the quantum critical field, BQ = 4.4x1013 G above which several quantum effects become observable! Stunningly, and for the very first time, we have now discovered similar bursts and flares from a new source with lower field, which means they should be far more numerous then what we know now (see [1], [2], [3], [4]).

What makes all these flaring neutron stars peculiar is that, while the radio-pulsar activity occurs at the expenses of the star rotational energy, in these cases it is the huge magnetic energy which powers the source emission. This is why they have been called magnetars, which stand for magnetic stars!

An artist impression of a magnetar


An artist impression of a magnetar.

Though our understanding of magnetars greatly increased in the last few years, a number of important points still remain to be studied. One of the major open issues is how to explain the broadband spectral energy distribution of magnetar sources. The persistent spectrum of magnetars in the 0.5-10 keV band is currently thought to be produced by the up-scattering of thermal photons emitted by the cooling star surface onto electrons/positrons which flow along the closed field lines of a twisted magnetic field. The twisting of the external field arises because in magnetars the internal magnetic field is so strong that it can overwhelm the crust yield, causing a movement of the surface tectonic plates which in turn drag the field along. SGRs/AXPs have been found to be bright also in other spectral bands. Observations in the hard X-rays (20-200 keV) with the INTEGRAL satellite revealed that a sizeable fraction (up to ~ 50%) of the flux is emitted above 10 keV. In addition, SGRs/AXPs are known to be infrared sources. The cause of this high energy emission is still mysterious.

A neutron star with a dipole field A neutron star with a twisted dipole field
A neutron star with a dipole field.
A neutron star with a twisted field.

This PhD project aims at providing an answer this question. The student will develop a model for the currents flowing in a twisted magnetosphere starting from first principles and including pair creation along the flux tubes. The spatial and energy distribution of the charges will then be used to compute in a self-consistent way the broad-band spectrum, from the infrared to hard X-rays. The effects of multiple scatterings in shaping the 0.5-200 keV emission are investigated by means of fully 3D Montecarlo simulations. The code will include QED and special relativistic effects, and will be quantitatively applied to observed X-ray data through an implementation in available fitting packages, as XSPEC. Knowledge of the charge distribution will also allow to evaluate the radiative losses due to curvature emission in the inner magnetosphere and test if these can explain the observed infrared/optical output from magnetar sources. All theoretical results will be developed at MSSL using numerical tools, and applied making use of data from the XMM-Newton, Integral, Chandra space missions and from the ESO telescopes.

2. Explaining cause of X-ray bursts from magnetars

On the observational ground, the most distinctive properties of magnetars is the emission of short, intense bursts of hard X-rays/soft gamma-rays.  In the magnetar scenario bursts are believed to originate from a hot, magnetically-confined pair plasma (a "fireball") which forms above the star surface because of the huge energy injected in the magnetosphere as a consequence of a crustal displacement (a "starquake").

During 2006 we observed, with the  Swift X-ray satellite an intense burst `"forest" from SGR 1900+14 which lasted for ~30 s, during which 7 intermediate flares (IFs) were detected. These events have properties in between those of normal bursts and giant flares, i.e. ~1 s duration and luminosity up to ~ 1E42 erg/s. This unique data set provided us with the most complete information to date on SGR bursts. Data were well represented by a two-blackbody spectrum with peculiar characteristics, that lead theoreticiens to suggest  we are observing for the very first time the two different photospheres of photons with different polarization properties (i.e. the so called ordinary and extraordinary waves which are expected in a strongly magnetized medium - Israel et al. 2008). This would be un umprecedented observational confirmation of our understanding of wave propagation in strong field!

swift

A picture of the Nasa Swift satellite with the different instrument on board. BAT and the XRT observes high energy burst in the gamma-ray and X-ray band with high timing resolution. 

BAT and XRT light curve

BAT and XRT light curves obtained simultaneously during the burst ``forest'' of 2006 March 29. Different energy ranges are shown: 1-4 and 4-10 keV for the XRT (panels X1 and X2, respectively), and 15-25 keV, 25-40 keV, 40-100 keV and >100keV for the BAT (panels B1, B2, B3, andB4, respectively). 

However, the suggestion must be corroborated by a detailed treatment of radiative transfer in the trapped fireball: the lack of this important analysis prevented up to now a quantitative comparison between theory and  observations (see e.g. Liubarsky et al. 2002), which is mandatory to constrain the properties of the fireball. The main goals of this PhD project  are: i) to build a complete, self-consistent model for the transport of  radiation in the trapped fireball, and ii) to apply results to fit  observed burst spectra in order to validate the model and derive physical parameters. The student  will develop specific competences in high magnetic field physics, numerical radiative transfer and the application of theoretical models to observations. As a first step, he/she will derive expressions of the relativistic scattering cross below resonance for a thermal distribution of electrons/positrons in a form suitable to numerical calculations. Secondly, a complete transfer code will be  developed, including non-conservative scattering.  Finally, a spectral archive (obtained varying the model parameters) will be produced and used to interpret all available/forthcoming bursts observations. This will be the first ever state-of-the-art model for burst emission spectra. 

BAT and XRT light curves obtained simultaneously during the burst ``forest'' of 2006 March 29. Different energy ranges are shown: 1-4 and 4-10 keV for the XRT (panels X1 and X2, respectively), and 15-25 keV, 25-40 keV, 40-100 keV and >100keV for the BAT (panels B1, B2, B3, andB4, respectively).  However, the suggestion must be corroborated by a detailed treatment of radiative transfer in the trapped fireball: the lack of this important analysis prevented up to now a quantitative comparison between theory and  observations (see e.g. Liubarsky et al. 2002), which is mandatory to constrain the properties of the fireball. The main goals of this PhD project  are: i) to build a complete, self-consistent model for the transport of  radiation in the trapped fireball, and ii) to apply results to fit  observed burst spectra in order to validate the model and derive physical parameters. The student  will develop specific competences in high magnetic field physics, numerical radiative transfer and the application of theoretical models to observations. As a first step, he/she will derive expressions of the relativistic scattering cross below resonance for a thermal distribution of electrons/positrons in a form suitable to numerical calculations. Secondly, a complete transfer code will be  developed, including non-conservative scattering.  Finally, a spectral archive (obtained varying the model parameters) will be produced and used to interpret all available/forthcoming bursts observations. This will be the first ever state-of-the-art model for burst emission spectra. 

3. XDINSs

The XDINSs are a small group of nearby X-ray pulsars powered, at variance with the radio-pulsars and the magnetars, by the release of the residual heath the star retains after his formation. Their emission is purely thermal and convincingly comes from the hot (~ 106 K) star surface. XDINSs play a great, fundamental role in neutron star astrophysics, since they are among the few sources in which we see the radiation coming directly from the surface of a neutron star! As such, they hold the potential to answer key questions regarding the surface properties of neutron stars: do they have an atmosphere, or is the surface solid? Which is the surface composition, mainly hydrogen, or heavier elements (the so-called silicon ashes) synthesized during the final stages of the progenitor life? What is the topology of the magnetic field (and, in turn, the temperature distribution) on the surface and inside the crust? Even more fundamental, the combined knowledge of surface temperature and of the source distance provides a measure of the star radius. This, if the star mass can be inferred, can be used to constrain the Equation of State (EoS) of neutron star matter, the holy grail of neutron stars research.

XDINS RXJ1856 The mass-radius relatioin for different EoSs
HST optical image of the XDINS RXJ1856 as it moves across the sky.
The mass-radius relation for a different EoSs (black lines).


A reliable determination of the star radius demands for a careful modelling of the thermal surface emission, which, in turn, depends on the physical conditions of the star outermost layers. Isolated neutron stars are believed to be covered by an atmosphere (as normal stars), although in this case the huge surface gravity (~ 1014 cm/s2) makes the atmosphere only a few cm thick. Despite this, the blackbody radiation which originates in the deeper layers gets distorted as photons propagate through the atmosphere, so that the emerging spectrum is not a blackbody anymore. It is now clear that none of the currently available atmospheric models can consistently reproduce the observed XDINS spectra. To some extent, this not surprising. Due to the numerical complexity of the problem, most of the available models are computed under the simplifying assumption that the whole NS surface is characterized by a single value of temperature and of the magnetic field strength/inclination. A further possibility is that the (relatively) low temperature (~ 106 K) and high magnetic field ( ~ 1013 G) produce a phase transition in the star outermost layers, resulting in the formation of a condensate. In this case the surface would be either in a liquid or solid state, and the neutron star would be bare (i.e. without an atmosphere).

This project will address the issue of surface emission from XDINSs by comparing spectra and pulse profiles from different emission models with observations from the ESA and NASA satellites XMM-Netwon and Chandra.

As for the modelling and interpretation, the student will proceed from the simplest case in which emission is a blackbody at the local surface temperature, and explore then both emission from a condensed surface (using realistic physical computations) or a light-element magnetized atmosphere. The surface thermal map will be either specified a priori, or computed from existing models of heat transport in the star crust, like, e.g., in the case of a poloidal+toroidal magnetic field confined in the crust.

The calculation will include relativistic ray-bending and a full treatment of line-of-sight and geometrical effects. Results will be systematically applied to X-ray data taken with satellites as XMM-Newton and Chandra. It is expected the student will work with fresh data and as well will apply and coordinate new observational campaigns with these X-ray satellites.

A further development includes a more detailed study of the emission from a bare neutron star, addressing in particular the long standing issue of the role played by ions in the dielectric tensor. These results will be used in conjunction to data taken with the Hubble Space Telescope and the ESO/VLT, and will be of crucial importance in assessing the low-energy behaviour of the model in relation to the emission of XDINSs in the optical band.