Reopening the Window of Astronomical Soft X-Ray Polarimetry

Reopening the window of astronomical soft X-ray polarimetry

Hua Feng 冯骅 (Tsinghua University) on behalf of the PolarLight collaboration

1 01 Science with X-ray Polarimetry

PPT模板下载：www.1ppt.com/moban/ 行业PPT模板： www.1ppt.com/hangye/ 节日PPT模板：www.1ppt.com/jieri/ PPT素材下载： www.1ppt.com/sucai/ PPT背景图片：www.1ppt.com/beijing/ PPT图表下载： www.1ppt.com/tubiao/ 优秀PPT下载：www.1ppt.com/xiazai/ PPT教程： www.1ppt.com/powerpoint/ Word教程： www.1ppt.com/word/ Excel教程： www.1ppt.com/excel/ 资料下载：www.1ppt.com/ziliao/ PPT课件下载： www.1ppt.com/kejian/ 范文下载：www.1ppt.com/fanwen/ 试卷下载： www.1ppt.com/shiti/ 教案下载：www.1ppt.com/jiaoan/ PPT论坛：www.1ppt.cn

02 Technique for X-ray Polarimetry 目录

CONTENTS 03 PolarLight

04 CubeSats in Space Astronomy

2 PART ONE

Science with X-ray Polarimetry What can we learn from X-ray polarimetry?

• Information about the magnetic field – Synchrotron radiation (PWNe, Jets) – Plasma polarization (Magnetized plasma) – QED vacuum birefringence (Neutron Stars) • Information about the scattering medium – Thomson/Compton/InverseCompton scattering – Geometric symmetry (accretion flow, BH spin, Sgr B2, etc.)

4 Science cases

• Synchrotron radiation and B-fields

Blazars GRB Jetted TDEs

(Marscher et al. 2010) Pulsar wind nebulae

5 Science cases • Accreting Pulsars

(Meszaros et al. 1988)

6 aeo hneof change of of component rate polarization linear the determine oaiaineouini ilcrcbrfign medium, birefringent dielectric Shaviv and a Heyl in evolution polarization ioa emty h oaiainsae vleadiabatically evolve if states polarization the geometry, dipolar antcdpl oeto h eto tr and star, neutron the of moment dipole magnetic ciiglna oaiain,tedrcinof direction the polarizations, linear scribing h antcfil ieto hog h equation the to through polarization direction photon field the magnetic of evolution the the couple will gence the below account; frequencies into at only taken optical important is be will electrons this plasma however, the with interaction ⌬ pulsars radio of the study from terminology borrowing radius, polarization-limiting nl ewe h ioeai n h ieo sight; of line the and axis dipole the between angle where where stenraie tksvco,and described. vector, previously Stokes vector normalized the is where changes, around iiigrdu hrfr osntdpn ntert at rate the on depend not does significant which a therefore polarization develop or radius coupling modes length limiting The two physical them. other between the the difference In problem, phase over change. the that of variables scale requires physical adiabaticity the words, which over scale refraction, of aiiyt odis hold to follow baticity to able of be amplitude not the NS, the from enough aiainlf beyond left larization direction in change EEYS ELADNRJ HVVPYIA EIWD REVIEW PHYSICAL SHAVIV J. NIR AND HEYL S. JEREMY n foeasmsta h edsronigtesa a a has star the surrounding field the that assumes one If sn h euto uoadNagata and Kubo of result the Using hscly h daai eiei prpit sln as long as appropriate is regime adiabatic the Physically, ls otenurnstar, neutron the to Close • k r • Շ x l r l ͓ ⍀ տ 3 ˆ ⍀ stesaelnt ftemgei field. magnetic the of length scale the is 15,16 r stedsac rmtecne ftestar, the of center the from distance the is ˆ ⍀ stedsac ln h ieto fpropagation, of direction the along distance the is pl ˆ ,where 1, hs stedrcino h antcfield magnetic the of direction the as Thus, . Ϸ ϵ hne ieto.Tedirection The direction. changes ilrtt and rotate will ϫ 1.2 ͩ ͔ k . 45 ␣ ͑ stewv etrand vector wave the is l ϫ sin ͑ ͓ ⍀ ␯ c ⌬ ˆ ͓ 7 10 7 ͒ ͪ ⌬ ͔ ␤ k ͔ 1/5 ildtrietecrua component. circular the determine will 7 of ͒ n aesonta hsvcu birefrin- vacuum this that shown have ͒ ⍀ ͩ ϭ ͓ ˆ 2/5 r ͩ 15 stedfeec ewe h indices the between difference the is 10 pl ⍀ ץ B ͉ ˆ cm, ץ n ogr h odto o adia- for condition The longer. any ⍀ x Since . ˆ ͔ QED ␮ s 30 . 3 il oee,dtrietepo- the determine however, will, / s ϭ ͑ ⍀ ␮ ˆ ٌ ilflo taibtcly Far adiabatically. it follow will cm G sin ⍀ aeo hneof change of of component rate polarization linear the determine oaiaineouini ilcrcbrfign medium, birefringent dielectric Shaviv and a Heyl in evolution polarization ioa emty h oaiainsae vleadiabatically evolve if states polarization the geometry, dipolar antcdpl oeto h eto tr and star, neutron the of moment dipole magnetic ciiglna oaiain,tedrcinof direction the polarizations, linear scribing h antcfil ieto hog h equation the to through polarization direction photon field the magnetic of evolution the the couple will gence the below account; frequencies into at only taken optical important is be will electrons this plasma however, the with interaction ⌬ pulsars radio of the study from terminology borrowing radius, polarization-limiting nl ewe h ioeai n h ieo sight; of line the and axis dipole the between angle where where stenraie tksvco,and described. vector, previously Stokes vector normalized the is where changes, around iiigrdu hrfr osntdpn ntert at rate the on depend not does significant which a therefore polarization develop or radius coupling modes length limiting The two physical them. other between the the difference In problem, phase over change. the that of variables scale requires physical adiabaticity the words, which over scale refraction, of aiiyt odis hold to follow baticity to able of be amplitude not the NS, the from enough aiainlf beyond left larization direction in change EEYS ELADNRJ HVVPYIA EIWD REVIEW PHYSICAL SHAVIV J. NIR AND HEYL S. JEREMY ˆ slre and large, is ln n ϫ ␤ foeasmsta h edsronigtesa a a has star the surrounding field the that assumes one If sn h euto uoadNagata and Kubo of result the Using hscly h daai eiei prpit sln as long as appropriate is regime adiabatic the Physically, ls otenurnstar, neutron the to Close • ͉ ⍀ ⍀ ͪ s ˆ k 3 ˆ , r 2/5 • ͪ ͉ Շ x l r l si h - ln de- plane 1-2 the in is ͒ 2/5 ͓ ⍀ տ ͉ 3 ˆ ⍀ stesaelnt ftemgei field. magnetic the of length scale the is 15,16 r stedsac rmtecne ftestar, the of center the from distance the is տ ⍀ ˆ ⍀ ˆ ⍀ ˆ ͩ stedsac ln h ieto fpropagation, of direction the along distance the is pl ˆ ,where 1, hs stedrcino h antcfield magnetic the of direction the as Thus, . l ͓ 0.5 10 Ϸ ϵ hne ieto.Tedirection The direction. changes ϳ 17 ilfl and fall will stebirefringent the is ilrtt and rotate will ϫ 1.2 ͩ 17 r ͔ ͔ s k ␯ . 45 ␣ ͑ odsrb the describe to stedistance the is oae quickly rotates stewv etrand vector wave the is l ϫ Hz sin ⍀ ͑ ͑ ˆ ͓ n aeof rate and ⍀ ␯ c ⌬ ˆ ͓ 7 10 s 7 ͒ ͪ ͪ ⌬ ͔ ␤ k at ͔ 1/5 hl the while ildtrietecrua component. circular the determine will 1/5 7 of ͒ n aesonta hsvcu birefrin- vacuum this that shown have r ͒ ␤ ␮ ⍀ ͩ ϭ ͓ ˆ 2/5 r pl r ͩ 15 stedfeec ewe h indices the between difference the is 10 pl pl ⍀ ץ B ͉ sthe is ˆ s sthe is cm, ץ sthe is n ogr h odto o adia- for condition The longer. any ⍀ x Since . ˆ ͔ QED ␮ s 30 will will . 3 il oee,dtrietepo- the determine however, will, / ͑ ͑ ͑ s ϭ ͑ 2 4 3 023002-2 ⍀ ␮ ˆ ٌ ilflo taibtcly Far adiabatically. it follow will s cm G ͒ ͒ ͒ sin ⍀

Scienceˆ cases slre and large, is ln ϫ ␤ ͉ ⍀ ⍀ ͪ s ˆ ϭ eitdtgte ihtedrcinta h polarization the that direction is the infinity at with polarization apparent together the where depicted 1, Fig. in given oaiaino aheeeti o 0%btsmaller. but 100% not re- is be element would respectively each %, frequencies of 35–55 In x-ray polarization and right. for % the values on 6–10 to 70% the duced is NS, it net realistic while the 13% more comparison, is For a left axis. the dipole on the polarization with linear 30° of angle an makes latitude magnetic inmpicuigbrfignefrafeunyof frequency a for birefringence including polariza- vacuum the map the shows the neglects panel right and that tion The mode assumes QED. the ordinary by one depicts the induced birefringence panel if in left only directions The emits polarization NS. surface the of of map image observed apparent the on layed htntaetr ohi pctm n ntePoincare relativity the ͓ general on of bending and light context incorporate spacetime the in in the sphere calculates both one trajectory accurately, process photon the calculate To tions. net larger a exhibit will stars result smaller polarization. generally bundles, radii ray picture smaller smaller with heuristic in stars this because large that Furthermore, a predicts observed and results. passes little the polarization varies so also detected net little, the parts varies angle, different be solid field from small polarization magnetic eventually this the Over the will of angle. distance, direction solid that this small At a rays surface. through mag- stellar of the the of from bundle direction far the a field with netic from correlated is direction element polarization surface observed the magnetosphere, polariza- circular of amount the provides The ( ratio direction. tion beneath polarization axis the the major towards element point of to surface ellipses component minor the linear the of the axes from of major direction originating the and ray lines The light them. a of tion eann epniua otewv etr Additionally, field vector. magnetic dipole wave the the distorts plane, the trajectory GR to GR, the to perpendicular in normal fixed keeps the remaining present it to that rays respect way with light a orientation such bent in rotates the direction polarization along that showed 3 ˆ , 18 2/5 ͪ ͉ ␮ si h - ln de- plane 1-2 the in is ͒ 2/5 apersl fteitgaino h oaiainis polarization the of integration the of result sample A over- star neutron a of image polarized apparent The 1. FIG. hshuitcpcuei on u ydtie calcula- detailed by out borne is picture heuristic This the in birefringence vacuum of because Heuristically, ͔͒ ͉ 30 Ϫ տ ⍀ ⍀ ˆ ˆ ͩ s eod ehv ouetersl fPineault of result the use to have we Second, . 2 l 3 10 ͓

0.5 • Surface thermal emission from NSs 10 ϳ .I ohmp,telredse uvsaelnso constant of lines are curves dashed large the maps, both In ). 17 ilfl and fall will stebirefringent the is 17 17 r ͔ s ␯ odsrb the describe to z h lissadsotlnsdsrb h polariza- the describe lines short and ellipses The Hz. stedistance the is oae quickly rotates Hz

⍀ • QED effects (vacuum birefringence) & B-field geometry ͑ ˆ n aeof rate and s ͪ at hl the while ͑ 1/5 r eaae y1°.Teosre’ ieo sight of line observer’s The 15°). by separated ␤ ␮ • Tested in optical (RX J1856.5-3754; Mignani et al. 2017) pl r pl sthe is s sthe is sthe is will will ͑ ͑ ͑ I.RESULTS III. 2 4 3 023002-2 – SGRs & AXPs s ͒ ͒ ͒ ͑ si ie,freape yPage by example, for given, is as ϭ eitdtgte ihtedrcinta h polarization the that direction is the infinity at with polarization apparent together the where depicted 1, Fig. in given oaiaino aheeeti o 0%btsmaller. but 100% not re- is be element would respectively each %, frequencies of 35–55 In x-ray polarization and right. for % the values on 6–10 to 70% the duced is NS, it net realistic while the 13% more comparison, is For a left axis. the dipole on the polarization with linear 30° of angle an makes latitude magnetic inmpicuigbrfignefrafeunyof frequency a for birefringence including polariza- vacuum the map the shows the neglects panel right and that tion The mode assumes QED. the ordinary by one depicts the induced birefringence panel if in left only directions The emits polarization NS. surface the of of map image observed apparent the on layed htntaetr ohi pctm n ntePoincare relativity the ͓ general on of bending and light context incorporate spacetime the in in the sphere calculates both one trajectory accurately, process photon the calculate To tions. net larger a exhibit will stars result smaller polarization. generally bundles, radii ray picture smaller smaller with heuristic in stars this because large that Furthermore, a predicts observed and results. passes little the polarization varies so also detected net little, the parts varies angle, different be solid field from small polarization magnetic eventually this the Over the will of angle. distance, direction solid that this small At a rays surface. through mag- stellar of the the of from bundle direction far the a field with netic from correlated is direction element polarization surface observed the magnetosphere, polariza- circular of amount the provides The ( ratio direction. tion beneath polarization axis the the major towards element point of to surface ellipses component minor the linear the of the axes from of major direction originating the and ray lines The light them. a of tion eann epniua otewv etr Additionally, field vector. magnetic dipole wave the the distorts plane, the trajectory GR to GR, the to perpendicular in normal fixed keeps the remaining present it to that rays respect way with light a orientation such bent in rotates the direction polarization along that showed 18 ␮ apersl fteitgaino h oaiainis polarization the of integration the of result sample A over- star neutron a of image polarized apparent The 1. FIG. hshuitcpcuei on u ydtie calcula- detailed by out borne is picture heuristic This the in birefringence vacuum of because Heuristically, ͔͒ 30 Ϫ s eod ehv ouetersl fPineault of result the use to have we Second, . 2 3 10 .I ohmp,telredse uvsaelnso constant of lines are curves dashed large the maps, both In ). ͓ 17 20 ͓ 21 z h lissadsotlnsdsrb h polariza- the describe lines short and ellipses The Hz. ͔ ͔ ertestar. the near 66 ic h intrinsic the since , ͑ GR 023002 , ͑ eaae y1°.Teosre’ ieo sight of line observer’s The 15°). by separated ͒ is,we First, . ͓ 19 I.RESULTS III. ͑ ͔ 2002 ͑ who si ie,freape yPage by example, for given, is as ␯ ´ ͒

~10% ~70-80% Simulations by R. Taverna and R. Turolla ͓ 20 ͓ 21 ͔ ͔ ertestar. the near 66 ic h intrinsic the since ,

͑ 7 GR 023002 , ͒ is,we First, . ͓ 19 ͑ ͔ 2002 who ␯ ´ ͒ Science cases

8 PART TWO

Technique for X-ray Polarimetry Exploration of the new window since the 1960s

X-ray astronomy First attempt: 1968 started in 1962 Scattering polarimeter First precise measurement: 1975 Aerobee 150 Bragg polarimeter, OSO-8 P = 19.2% ± 1.0% � = 156.4°± 1.4° Weisskopf et al. 1976, 1978

First detection: 1971 Bragg polarimeter Aerobee 350 Crab nebula P = 15.4% ± 5.2%

First satellite: 1974 Flat Bragg crystal Ariel 5 10 X-ray polarimetry based on the photoelectric effect

Dsptt.tfdujpo!pg!qipupfmfdusjd!fggfdu

dσ 2 ∝ cos ϕ dΩ

position angle degree

max− min max+ min

11 Technical difficulties

• Short range for electrons of a few keV - 800 U, – in silicon: ~μm 800 400

200

– in gas: ~mm x m ron)

.200 -

8 -400 -

•600

• Electron tracks are not straight due to scattering -800

) ) ) IOOC -1000 -800 -800 -400)))) -200 0 200 400 000 800 1000 x axis (micron)

• Challenge for detector Figure 9. Simulated tracks (projected onto the detector plane) produced by 5 keV electrons in pure Ne, at pressure of 1 Atm (right). Starting directions are distributed according to equation (15), simulating photoelectrons emitted in – Require 2D imaging deviceresponse to a collimated, 100 % linearly polarized 5.9 keY photons beam. By means of a progressive- "zoom800 out" (left), it can be clearly seen that the modulation of the track directions is more and more blurred by CoulombU, scattering while – Resolution < 100 μm going away from conversion point. 800 direction with the principal axis of the released charge distribution. At higher energy, in fact, tracks400 are longer

and the probability for a large angle electromagnetic scattering to occur increases as well. On the200 other hand, a more sophisticated algorithm which isolates the initial part of the track (like the one we use) allows to efficiently overcome this limitation, as it is clearly confirmedx bym ron)experimental data. By means of our Monte Carlo code we have investigated the performance of a polarimeter with a 100 m readout pitch (which is feasible with currently adopted technology and will be soon available for tests).200 -for several

8 -40012 -

•600 70 C.) -800 o 60 ) ) ) IOOC -1000 -800 -800 -400)))) -200 0 200 400 000 800 1000 50 x axis (micron)

Figure 9. Simulated30 tracks (projected onto the detector plane) produced by 5 keV electrons in pure Ne, at pressure

of 1 Atm (right).20 Starting directions are distributed according to equation (15), simulating photoelectrons emitted in I ) ) ) I response to a collimated,200 400 100 %600 linearly800 1000 polarized 5.9 keY photons beam. By means of a progressive "zoom out" (left), it can be clearly seenDistance that thefrom convertion modulation point (micron) of the track directions is more and more blurred by Coulomb scattering while going away from conversion point. Figure 10. Modulation factor, as a function of distance from absorption point, evaluated from the distribution of primary ionization for 5 keV photoelectrons in pure Ne, 1 Atm. direction with the principal axis of the released charge distribution. At higher energy, in fact, tracks are longer and the probability for a large angle electromagneticProc. ofscattering SPIE Vol. 4843 to 391 occur increases as well. On the other hand, a more sophisticated algorithm which isolates the initial part of the track (like the one we use) allows to efficiently overcome this limitation, as it is clearly confirmed by experimental data. By means of our Monte Carlo code we have investigated the performance of a polarimeter with a 100 m readout pitch (which is feasible with currently adopted technology and will be soon available for tests) for several

C.)

o 60

20 I ) ) ) I 200 400 600 800 1000 Distance from convertion point (micron)

Figure 10. Modulation factor, as a function of distance from absorption point, evaluated from the distribution of primary ionization for 5 keV photoelectrons in pure Ne, 1 Atm.

Proc. of SPIE Vol. 4843 391 Gas Pixel Detector (GPD)

• First demonstrated by INFN-Pisa & IAPS-Rome (Bellazzini et al.; Costa et al. 2001)

Gas Electron Multiplier (GEM)

13 Detector assembly

14 Measured electron tracks

15 Angular modulation

2.67 keV 3.74 keV 5.33 keV

6.09 keV 7.49 keV

16 PART THREE

PolarLight Polarimeter Light (PolarLight；极光计划)

x-ray

collimator PolarLight

CubeSat window

photoelectron track GPD GEM

measured 2D track HV ASIC chip DAQ

18 Shocking test

19 Vibration test

20 Thermal vacuum

21 Launched into a low Earth orbit

October 29, 2018

22 Re-detection of X-ray polarization from the Crab nebula

23 Time variation of polarization

glitch

With pulsar emission Without pulsar emission significance level: 3σ

• Bayes factor • Bayesian posterior Magnetosphere altered after the glitch • Bootstrap

24 High energy emission from pulsars

11 Multi-Wavelength Polarimetry of Isolated Pulsars 287

Cheng et al. 1986 Muslimov et al. 2004 Kalapotharakos et al. 2012 Harding et al. 2019

Fig. 11.3 Location of radio (blue) and high-energy emission in the outer gap (yellow), two-pole caustic (red) and current sheet (green) models in the meridional plane containing the spin and magnetic axes of a force-free magnetosphere. The dashed black lines denote the light cylinder and the dotted red lines show projections of the null-charge surface 25

site of particle acceleration. The accelerated particles radiate γ -rays through cur- vature and inverse-Compton radiation, initiating electromagnetic cascades through one-photon pair production in the strong magnetic ﬁeld (Daugherty and Harding 1982;HardingandMuslimov2002). In this model, the high-energy pulses would appear near the phase of the radio pulses, would be very highly polarised and exhibit the “S-shaped” PA sweep of the RVM. Phase-resolved high-energy polarisation measurements were not available until fairly recently, so this property of the polar cap models could not be tested. However the Fermi Gamma-Ray Space Telescope, launched in 2008, discovered many γ -ray pulsars whose pulses appeared at phases signiﬁcantly different from those of the radio pulses (Abdo et al. 2013). In fact, the variety of light curves agree more with predictions of models where the high- energy emission comes from the outer magnetosphere near or beyond the light cylinder, RLC c/", where " is the pulsar rotation rate. Such outer magnetosphere models, such as= outer gap (Cheng et al. 1986;Romani1996), slot gap and current sheet models had been proposed earlier (see Fig. 11.3). Naturally, polarisation predictions also changed substantially from those of polar cap models and, as will be discussed in the following sections, can better explain the properties of the observed polarisation. Reopening the window

IXPE in 2021 (Weisskopf et al. 2016)

OSO-8 (1975) PolarLight (2018)

eXTP in 2027 (Zhang et al. 2019) 26 PART FOUR

CubeSats in Astronomy CubeSat 立方星

• The CubeSat standard – proposed in 1999 by Jordi Puig-Suari of California Polytechnic State University and Bob Twiggs of Stanford University – an educational tool for teaching students about spacecraft hardware, electronics and programming – Low cost 1U 3U 6U

28 Astronomical CubeSats funded by NASA

• ASTERIA (Arcsecond Space Telescope Enabling Research in Astrophysics) – to measure exoplanetary transits across bright stars with <100 ppm photometry – launched in August 2017, one of the first CubeSats enabled for astronomical measurements • PicSat – to observe in visible light the potential transit of the directly- imaged giant planet β Pictoris b • HaloSat – measure the soft X-ray emission from the hot halo of the Milky Way galaxy to resolve the missing baryon problem • CUTE (Colorado Ultraviolet Transit Experiment) – survey of exoplanet transit spectroscopy in the near-UV • SPARCS (Star–Planet Activity Research CubeSat) – the far- and near-UV monitoring of low-mass stars (0.2–0.6 M☉) • BurstCube – to detect and localize GRBs

29 A rapid growth

Launches of CubeSats “CubeSat” on ADS

30 The GRID network

l 10+ CubeSats in LEO l Scintillation detector, ~60 cm2 each l Localization accuracy for GRBs within 200 Mpc ü <1°(for an on-axis event, ~0.14 yr-1) ü 10°~ 15°(for a GRB 170817A like event, ~5 yr-1)

GRID (Gamma Ray Integrated Detectors)

31 Flight model & satellite

32 GRID - a student project

The first group Testing the detector Talking at COSPAR 2018

• Started in 2016 October • More than 50 Students from 16 universities • GRID-1 in orbit • GRID-2/GRID-3 will be launched this year

33 CubeSats in Astronomy • To demonstrate new techniques – Sounding rockets vs. Balloons vs. CubeSats • Highly customized science objective – Large missions: observatories THANK YOU! – Small missions: dedicated • Long-term monitoring of a single or a few targets • Student training – Project cycle: ~3 years – All-around skills: science + engineering + leadership