Hard X-ray Polarimeter for Small Satellite: Design, Feasibility Study, and Ground Experiments
Kiyoshi Hayashidaa[1], Tatehiro Miharab, Syuichi Gunjic and Fuyuki Tokanaic
a Osaka University, 1-1 Machikaneyama-cho, Toyonaka, Osaka 560-0043, Japan;
b RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan;
c Yamagata University, 1-4-12 Kojirakawa, Yamagata 990-8560, Japan
abstract
We make a plan of a hard X-ray polarimetry experiment with a small satellite. Bright point-like sources in 20-80keV are prime targets, for which we will not use focusing optics. Comparing various types of polarimeters, we adopt a scattering type in which anisotropy in scattering directions of photons is employed. After optimization of the design is considered with simplified models of scattering polarimeters, we propose to use segmented scatter targets made of plastic scintillators, with which scattering location is identified by detecting recoiled electrons. Simulations show that recoiled electrons are detectable when incident X-ray energies are above 40keV, for which higher polarimetry sensitivity is obtained. We confirmed the performance of such a polarimeter in experiments at a Synchrotron facility and performed a balloon flight in which a proto type unit of the polarimeter was onboard. We finally discuss feasibility of a small satellite experiment in which many of the polarimeter units will be employed. Twenty five units of the polarimeter enable us to detect hard X-ray polarization of 5-10% for a hundred mCrab sources. Improvement in the sensitivity to detect recoiled electrons will significantly improve the polarimetry sensitivity. We also consider a low energy extension of our system down to below 10keV in order to cover wide energy range.
Keywords: X-ray polarimetry, scattering polarimeter, hard X-ray, scintillator, PMT
1. INTRODUCTION
Astronomical X-ray polarimetry has a long history as X-ray astronomy itself, but is still an unexploited field. The filed has been recently activated by developments of various new technologies for stellar X-ray polarimetry, and by new understandings of importance of polarization information. It may be partly due to a kind of saturation in other fields of X-ray astronomy, spectrometry and imaging, for which huge resources will be required to make a big step. In the workshop held at SLAC in this February, many proposals and current going projects were presented1. The authors of this paper themselves have been developing several types of X-ray polarimeters so far, some of them jointly and the other independently. Based on those experiences, we are going to make a plan for a stellar X-ray polarimetry project in near future. This is firstly motivated by discussions for the NeXT project2, a large size X-ray mission next to the Astro-E2 in Japan, but we are planning to make a small satellite experiment specialized for X-ray polarimetry.
In the next section, we explain the target of our experiment, which is polarization of bright point-like sources in hard X-ray band of 20-80keV. As shown in the same section, a polarimeter using Compton scattering is best for such a study. In section 3, we consider how to optimize the design of the scattering polarimeter. We finally propose to use a segmented target polarimeter as a unit of our experiment. In section 4, our developments of scattering polarimeters on the ground are introduced. We also briefly touch a balloon experiment recently performed by one of the authors. In the last section, we show feasibility study of a possible small satellite experiment. The experiment primarily aims at 20-80keV, but we show low energy extension below 10keV is possible by some means.
2. hard X-ray, non-IMAGING, scattering type polarimeter
As mentioned above, X-ray polarimetry is still an unexploited field. It means that there are many rooms and directions to plan a polarimetry experiment. The situation is very much different from the case for X-ray imaging or spectrometry experiments. In this section, we fix the baseline of the design for X-ray polarimetry experiment with a small satellite. In short, our plan focus on hard X-ray range from 20 to 80keV, bright compact sources will be primary targets, and scattering type polarimeters will be employed for it.
2.1. Targets of X-ray polarimetry and its energy range
One may set a specific group of targets for X-ray polarimetry for stellar sources, but it would be better to cover as many X-ray objects in different categories as possible, for the polarimetry at current circumstance. In fact, lots of theoretical prospects and models have been presented for the target of X-ray polarimetry1, including super nova remnants, clusters of galaxies, galactic compact star binaries, pulsars, and active galactic nuclei. They can be summarized by the mechanism to produce possible polarized X-ray emission. Synchrotron radiation should produce polarized X-ray emission unless magnetic field is perfectly random. If we can map the X-ray polarization degree and direction in diffuse source, the magnetic field in the source will be known. Compton scattering process will also produce the polarized X-ray emission depending on the geometry of sources and scatters. For example, we expect polarization in scattered radiation from an accretion disk around compact objects.
Most of these processes are related to non-thermal radiation usually with power law X-ray spectra. We consider such a power law component is generally emphasized in hard X-ray energies than in soft X-ray energies. This is one of the reason why we set the target energy range to be hard X-ray band, for example 20-80keV. However, it is apparent that number of photons is much larger in soft X-ray band than hard X-ray band. In fact, polarization was detected previously in the soft X-ray band, though positive detection was limited to the Crab nebulae and Sco X-1. In addition, several proposals (AXP, PLEXAS, SXP, etc) were presented for soft X-ray polarization experiments with small satellites1, while there is only one proposal XPE, though POGO, a proposed balloon experiment, has some overlap with 20-80keV energy range1. We think it is most important to realize a positive detection of X-ray polarization as early as possible, which has been paused nearly 30 years. Considering possible instrumentations for soft X-ray polarimetry and for hard X-ray polarimetry, the latter needs smaller amount of development work and resources for us. That is also an important reason why we adopt hard X-ray band. Of course, hard X-ray band is fascinating as it is an unexploited energy range in stellar polarimetry. There might be further discussions on hard X-ray or soft X-ray, but considering non-thermal spectra from the targets of polarization experiments, energy dependence of the polarization is important as already emphasized by authors. Therefore, hard X-ray polarization is important even after some of the soft X-ray projects succeed in detecting polarization from many sources. Finally, we note another merit of taking hard X-ray range as a target; we are able to perform balloon experiments prior to a small satellite experiment, not only for instrument verification but also for real observation.
2.2. Using focusing optics or not
We fix the target energy range to be hard X-ray band of 20-80keV. For this energy range, usual focusing optics such as X-ray grazing incidence mirror was not available, and mechanical collimator was the only option. Recently, multi-layer coated mirror, called super mirror is available for this energy range as planed for the NeXT satellite2. Collecting power of such a mirror is very attractive to gain the sensitivity with a small detector and to reduce background. In fact, if we place a handy size scattering polarimeter on the focal plane a super mirror on the NeXT satellite, we expect to realize hard X-ray polarimetry for ten mCrab level sources. However, for a small satellite experiment, resources and geometrical size does not allow us to use such optics. We thus have to use mechanical collimators to limit the field of view.
Another merit of using focusing optics might be its imaging capability. As mentioned above, X-ray polarization mapping of diffuse objects, such as super nova remnants and clusters of galaxies, may be attractive to see magnetic field in those sources. We estimated photon fluxes in hard X-ray band for many diffuse sources from the observations in soft X-ray band. Nevertheless, it is found that with an exception of the Crab nebulae, other diffuse (if we look at arc minute resolution) sources, such that Coma cluster, or Kes75, SN1006 etc., are expected to have at brightest mCrab / arcmin2. We therefore had better to concentrate bright point-like sources, unless hard X-ray optics with large effective area is available. That is our baseline of the design.
2.3. Polarimeter type
There are basically two types of hard X-ray polarimeters (for stellar polarimetry), a photo-electron track type and a (Compton) scattering type. The former utilizes anisotropy in the emission direction of photoelectrons, while the latter use that in the scattering direction of photons. Photo-electron track type polarimeters are realized with an X-ray CCD or with gas counters. In particular, polarimeters using gas imaging detectors have been extensive developed recently. On the other hand, scattering type is somehow classical but various configurations are proposed. General arguments between these two types are beyond the scope of this paper, but we focus on the polarimetry sensitivity as a clue.
In order to evaluate the performance of an X-ray polarimeter, we often use a modulation factor M, and a detection efficiency h. The modulation factor M is defined as a modulation of the signal (counting rate etc.) as a function of rotation angle, e.g. (Nmax-Nmin)/(Nmax+Nmin), where Nmax and Nmin are the maximum and minimum of the counting rate. The detection efficiency h is defined as the ratio between the number of events detected and employed in the polarization analysis to that of the incident X-rays. For photo-electron track gas counters, it is the same as the quantum efficiency of the gas counters. For CCD, we have to consider an additional factor, the ratio of number of multi-pixel events to the total number of events. When we consider a scattering type polarimeter, as described in the next section, it is multiplications of probabilities, the probability of incident photons are scattered in a target material, the probability of the scattered photons travel to the sold angle covered by detectors, and the X-ray detection efficiency of detectors. Both M and h are the greater the better, but the sensitivity of a polarimeter is quantitatively evaluated with Minimum Detectable Polarization degree (MDP). The MDP is defined as the lower limit of the polarization degree of a target source of which polarization is positively detected. Apparently it is a function of a source intensity, exposure time, effective area, and background level, but is inversely proportional to Mh1/2 if the background is negligible. Therefore, a polarimeter with larger M h1/2 is better.
As described below, scattering type polarimeters can provide, for example, M of 0.5 and h of 0.6 at the same time, which makes M h1/2 of 0.39. On the other hand, photo-electron track polarimeters with gas detectors are difficult to get large values of M and h at the same time. In order to get large M, scattering of electrons in gas must be suppressed, for example, by using gas with a small atomic number. According to the simulation by Pacciani et al.3, Mh1/2 is optimized to be 0.07 at 20keV. Although the efficiency h may be improved by stacking the detectors, to gain the factor of 5 difference in Mh1/2, the efficiency h must be increased by factor of 25. In conclusion, we adopt the scattering type polarimeter for hard X-ray, non-imaging polarimetry.
3. Optimization of hard x-ray Scattering polarimeter design
We have fixed to adopt a scattering type polarimeter. Although this type of polarimeters is classical, we reconsider its optimized design with a simplified model, a scattering target and surrounding detectors. After that, we propose to improve its simplified polarimeter by introducing segmented and active targets. Its polarimetry sensitivity is explored with simulations.
3.1. Design optimization with a simplified scattering porarimeter
We first simplify a scattering polarimeter with a cylindrical shape target and detectors surrounding the target as shown in Fig.1 (a). The highest M ~1 is obtained when we make the height of the target and detector very small and place detectors far apart the target. Note that the anisotropy of the scattering direction is described as (1-sin2qcos2f), the modulation is maximized at the scattering polar angle q of 90º. That is different from the anisotropy in the photo-electron emission where the same azimuthal anisotropy for any polar angles. The scattering polarimeter of model (a) is good for ground experiment, and in fact we employed such model to calibrate the polarization degree of incident X-ray beam4,5. However, the detection efficiency is too small to use it for stellar polarimetry. We have to make the height of the detectors larger in order to cover larger solid angle of the scattering. There is an optimized point where Mh1/2 is maximized for a given target. If we idealize the target has a large scattering cross section and a negligible absorption cross section so that a small height target is enough to work, Mh1/2 is maximized to be about 0.55 when we cover the scattering polar angle of about ±50º. This value should be considered to be a kind of theoretical limit of a scattering polarimeter. In reality, we have to make the target height longer, as long as the mean free path against Compton scattering in order to gain the efficiency to collect scattering events. (Stacking number of model (a) polarimeters is another option, but we will not consider here.)