CZT Imager of AstroSat measures first phase resolved X-ray polarisation of Crab pulsar

X-ray polarisation: All types of electromagnetic radiation, like X-rays and optical light, are bundles of energy called photons defined by an electric vector, and an orthogonal magnetic vector. The electric vectors are mostly random in orientation, but quite often, they are aligned to a particular direction depending on the conditions in the source of these photons. For example, optical light scattered in the sky are aligned, or polarised,  in the plane of the scattering and you can take a simple optical polariser to look at the sky and determine the direction of the incident photons (prior to scattering) from the source – in this case the Sun. The polarisation properties of electromagnetic radiations have been regularly used by astronomers to study the conditions of the cosmic sources emitting these radiations. For example, the direction of the magnetic field in our galaxy, Milky Way, can be precisely mapped using polarisation measurements.

Strong X-ray emission from astrophysical sources often signifies the presence of exotic compact objects in the universe: neutron stars and black holes. The X-ray emission from these objects traces the regions of particle acceleration and understanding the conditions of particle acceleration can tell us about several exotic phenomena:  for instance, whether the black hole is spinning, does the strong gravity near black holes obey Einstein’s equation, or if neutron stars  are made up of ordinary matter or strange matter and so on. It has long been thought that X-ray polarisation properties will tell us more about the mysteries of these strange objects.

Immediately after the birth of X-ray astronomy in 1962, due to the serendipitous discovery of an extra-solar X-ray source in a rocket flight, there was a flurry of activity in all aspects of X-ray astronomy, including polarisation measurement. Scientists at Columbia University flew a simple scattering based X-ray polarimeter in a rocket flight in 1969 to look at the pulsar in the Crab Nebula: they could not detect any polarisation – less than 36% of X-ray photons had their electric vectors aligned. They sent another rocket in 1971 with an improved version with graphite crystals to look at X-rays of specific energy at 2.6 keV and discovered X-ray polarisation in Crab: about 20% of the photons at this energy show aligned electric vectors. Buoyed by this success, they sent a polarimeter in the 8th Orbiting Solar Observatory (OSO-8) in 1975: it confirmed the X-ray polarisation in Crab and put quite stringent (5 – 10%) upper limits on the polarisation of several bright X-ray sources.

Exploring X-ray Polarisation using AstroSat 

AstroSat, being India’s first dedicated astronomy mission, was a massive effort involving the participation of many ISRO centres, academic institutes, and universities. Tata Institute of Fundamental Research (TIFR), Mumbai led the scientific effort by shouldering the responsibility to deliver three of the five major payloads of AstroSat.

Inter-University Centre for Astronomy and Astrophysics (IUCAA), Pune, founded for the explicit purpose of promoting “the nucleation and growth of active groups in astronomy and astrophysics at Indian universities’’, was an indirect participant in the AstroSat instrument development: Dipankar Bhattacharya (IUCAA) designed the specialised `coded aperture masks’ for two instruments of AstroSat.  Since AstroSat is supposed to be an observatory class satellite catering to all users, including those from the universities, it was thought that a formal participation of IUCAA in several allied tasks like software development, student training, etc. would enhance the utility of AstroSat data.

Arthur Holly Compton had discovered, in 1923, that X-rays can deposit part of their energy in a material and undergo `Compton’ scattering. The CZT detectors are pixellated (and hence, have position information) and at certain energies, X-rays will interact primarily by Compton scattering. If the instrument can be made sensitive to measure the incident X-rays as well as the scattered X-rays, then the distribution of the scattered X-rays in the neighbouring pixels of the CZT detector will show some tell-tale signs of the polarisation properties of the incoming X-rays. Such pixellated X-ray detectors are thought to work as X-ray polarimeters, but experimentally no one, so far, had demonstrated the polarisation properties of pixellated X-ray detectors. Since X-ray polarisation measurement is a long sought after experimental technique in X-ray astronomy, examining the utility of CZTI as an X-ray polarimeter looked attractive.

Polarisation Experiments

There followed a flurry of activity to demonstrate that CZTI can be used as an X-ray polarisation instrument. Rao ascertained from the supplier of the CZT detectors that the properties of incident and scattered X-rays are measured and retained, and this was quickly demonstrated by some simple experiments using radio-active sources of known energies. Santosh used a sophisticated code to simulate this behaviour and satisfied himself that polarisation measurements can indeed be made. Then followed a series of Varkari between Ahmedabad and Mumbai: Santosh and his student Tanmoy Chattopadhyay (currently working as a post-PhD researcher at Pennsylvania State University) made controlled experiments on the flight models of CZTI detectors. Several improvisations  were part of the testing: like the use of simple thermocol pieces as experimental set-up, use of very strong radio-active sources of X-rays and making them scatter from blocks of aluminium to polarise them - but using sophisticated simulations to precisely calculate the energy of scattered X-rays.

All these activities were done while the fabrication and testing of the flight models were going on ! First  a single X-ray detector, called a module,  was  bombarded with X-rays of known energies and its properties were measured. A rigorous modelling, based on these experimental results, showed  that CZT Imager as a whole, with 64 detector modules, will have  enough sensitivity to measure X-ray polarisation from bright cosmic X-ray sources. Then started a series of controlled experiments for the proof of concept:  beams of X-rays were generated, with or without polarisation, and  the polarisation signal in the detector was measured and compared with the results from the simulations. All the  tests were repeated for X-rays of a single energy, as well as continuum of X-rays (as expected from Cosmic X-ray sources). Finally, some precious time was readjusted from the very tight delivery schedule of the flight model to repeat these tests in a fully flight-like environment.  

This is only the beginning of the story. Because, all the tests in the lab are made with small beams of X-rays and in flight, while observing X-ray source situated at thousands of light years away, one expects parallel beam of polarised X-rays. To test with similar large area polarised beam of X-rays in the lab, one needed to build a humongous facility – perhaps taking several years of effort just to fabricate it.  The whole detector geometry was simulated and sophisticated simulations were carried out to predict the behaviour of the detector. These were compared with the controlled lab experiments of small beam, at different experimental configurations, to satisfy that the simulation results are indeed as observed in the lab. This gave a confidence for the prediction of the parallel beam of large area and the results indeed showed that CZT Imager is a very sensitive X-ray polarimeter.

Finally, on 2015 September 28, a fully ground calibrated X-ray polarimeter was flown into space.

The special source ‘Crab’ and its polarisation

When a new era of astronomy was ushered in by telescopes in the 17th and 18th century Europe, it was realised that the sky consists of several fuzzy objects, apart from point-like stars, and they were named ‘Nebula’. The Crab Nebula, in the constellation of Taurus, is the first object identified by Messier, a French astronomer, who meticulously catalogued such objects. Crab Nebula cannot be seen by the naked eye, but can be seen as a fuzzy crab-like feature (hence the name) by a simple binocular. Among such nebulae, supernova remnants form a large fraction. Soon, it was realised that the Crab Nebula is none other than the `Guest Star’, or the supernova recorded by the Chinese astronomers in year 1054.

Within a year of the discovery of pulsars (rapidly rotating highly magnetised neutron stars) in 1967, a pulsar with a period of 33 milliseconds was discovered in the Crab Nebula. It was, however, surmised that the explosion of the supernova that occurred about 1000 years ago has long lost its energy and currently, the nebula is powered by the monster sitting at its centre: the fast spinning neutron star.  Crab Nebula, along with its pulsar, became the darling of all observers: it is the only pulsar observed to be pulsating in all branches of the electromagnetic radiation, from radio to even ultra high energy gamma-rays. Since the pulsations are due to simple light-house effect of the rotating neutron star, the period should be, and indeed is, the same in all wavelengths - hence the Crab pulsar became a calibrating source to test the timing accuracy of any new window of observation. The emission mechanism is understood as due to the movement of particles in a magnetic field, the particles being accelerated to humongous energies in the polar cap of the highly rotating magnetic neutron star. Since the magnetic field and the rotation period are reasonably stable for several years, the emitted flux, in hard X-rays, is expected to be a constant : hence it also became a good flux calibrator for any new observations. The realisation that particles are accelerated to very high energies has made a strong case for such young supernova remnants as sources of the mysterious particles bombarding the Earth: Cosmic Rays.

In recent years, Crab Nebula and pulsar are the subjects of diverse observations and theorising to understand the particle acceleration mechanism in fast spinning neutron stars.  Measurement of polarisation as a function of pulse phase in optical and radio wavelengths and detailed theoretical modelling are hinting that particle acceleration happens outside the `light-cylinder’ of this source and scientists across the globe are pondering over the nature of such acceleration

Crab polarisation using AstroSat

Crab Nebula  was the first source AstroSat stared at post switch ON and verification, and, it kept looking at this source quite often for a periodic calibration of the X-ray instruments, as well as to measure its polarisation. There followed a very meticulous analysis. Every recorded photon was examined, stringent criteria were imposed on them to be validated as `Compton-scattered’ events, their distribution among their neighbours were examined, for consistency, among different detectors in the instrument. Regions in the sky were identified and the position of the satellite was manoeuvred such that the instrument looks at a `source-free’ region in the sky at similar satellite orientation with respect to Earth. The behaviour of the detector was studied in these `background’ regions and the resultant profile was subtracted from the source observations. To establish that these `Compton-scattered’ photons were really from  Crab, their arrival times were investigated and found to be beating at the standard Crab clock of 33 milli-seconds.

Yes, X-ray polarisation was clearly detected even in individual observations lasting about half a day and when all the eleven observations were added together, the measurement is clearly the most precise hard X-ray polarisation measurements till date The results were submitted to the British journal Nature.

It is true that measuring polarisation with the highest accuracy possible is a technical challenge: but what is the new thing we are learning ?  After a few weeks of intense debate among the researchers, it was decided to collect more data and attempt a very precise measurement of the variation of the polarisation properties when the pulsar beam sweeps across us, the observers. For this, data from different observations spanning across 18 months had to be added. Do we know the pulsar period accurately for this ? The Crab pulsar was observed in radio wavelengths from the Ooty radio telescope every day and also many times during these months from the Giant Meter-wave Radio Telescope (GMRT) at Khodad, near Pune. The pulse of pulsar was accurately measured, the X-ray photons were assigned a pulse phase and the change in the polarisation property as a function of the pulsed beam emission was studied.

When we think that the pulsar beam is shining elsewhere, the remainder of the beam has sufficient X-ray emission, with the polarisation increasing and also showing a sharp change. What it means is that the light house is leaking and emitting high energy X-rays all the time. Most magnetospheric theories predict that the polarisation of X-ray radiation will show changes only during the emission of a pulse, but not at other times.  The new observations thus support the view that the particle acceleration is happening outside the conventional boundary of the magnetosphere, in a region where the charged particles generated by the pulsar are spiralling out in the form of a wind.  The surprising observation by CZTI of a sharp change of polarisation in the “off pulse” region  is clearly a big challenge to theorists.  

 This result is published in Nature Astronomy on November 6, 2017, publication doi: 10.1038/s41550-017-0293-z

        

Left panel: The grey line shows the brightness of the Crab pulsar as observed by AstroSat CZTI. The horizontal axis (phase) represents time expressed in units of the pulsar’s spin period. Phase 0.0 to 1.0 stands for the full rotation cycle of the pulsar. The same result is shown repeated between phase 1.0 and 2.0, for a clear demonstration of the periodic pattern. Colored bars indicate how strongly polarized the observed radiation is.  Sharp variation of polarization when the brightness is low is the surprising discovery by AstroSat.

Right panel: The angle of X-ray polarization measured by AstroSat CZTI shown superposed on a composite optical and X-ray image of the Crab nebula, taken by NASA’s  Hubble and Chandra telescopes respectively.  The white arrow represents the projected spin axis of the pulsar located at the center of the nebula. Other arrows display the orientation of the observed polarization. The color of an arrow indicates the range of phase it belongs to, being equal to that spanned by bars of the corresponding shade in the left panel.