New Insight into Magnetar Evolution and Concluding Remarks ���

～10 Current Issues with Magnetars ���(with personal prejudice and preference)～

K. Max Makishima University of Tokyo / RIKEN

2012/9/1 Magnetar Conference at Rikkyo 1

Congratulations, Shibazaki-san. I am glad I have 29 joint co-authorships with you.

1. Are Magnetars Magnetically Powered?

2012/9/1 Magnetar Conference at Rikkyo 2

0.01

1

100

0.1 1 10 100 1000

Lx / E

dot

Characteristic Age (kyr)

Crab

10-4

PSR 1509

Geminga

1E 1547.0

SGR1900

1E 2259

SGR1806

SGR 0526

Kes 751E 1048.1-5937

4U 0142

XTE J1810J1708

J0100

1E 1841 J1647

Probably Yes

SGRs AXPs High F. Pulsars Active Pulsars

☆Evidence (1): their Lx ≫ their spin-down lum. ☆Evidence (2): Magnetars are more luminous than ordinary cooling NSs; internal heat is insufficient to explain their emission (Yakovlev)

2012/9/1 Magnetar Conference at Rikkyo 3

☆ Evidence (3): absence of binary companion. Accretion from, e.g., a fall-back disk, ���would be unable to explain their characteristic two-comp. spectra.

☆Evidence (4): their dipole B (estimated from P and P-dot) decreases with their char. age.

☆Evidence (5): mag. energy with B=1014.5 G, liberated in 10 kyr, can sustain Lx~1035 erg/s.

12

13

14

15

0.1 1 10 100 Char. Age (kyr)Lo

g Surf

ace F

ield (

G)

Kes 75

Char. Age (ky)

1

10

0.1

0.01

!F! (

a.u.)

IGR 16195 Her X-1Cyg X-1

Cyg X-1

1 10 1004U 0142+61

Aql X-1 (high)

Aql X-1

(low)

The Crab Nebula

V2400 O

ph

SGRs AXPs High F. Pulsars Active Pulsars

A compilation of Suzaku spectra

2. How Old Are Magnetars?

•  Characteristic age τc≡P/2Pdot ; a good approx. of true age t0 only when the braking index is n ~3, and B is constant.

•  If B∝ t -1/α⇒ τc ~[α/(α-2)] t0 ; ���a significant over-estimate��� (Colpi+00; Nakano+11).

2012/9/1 Magnetar Conference at Rikkyo 4

12

13

14

15

0.1 1 10 100 Char. Age (kyr)

Log S

urfac

e Fiel

d (G)

Kes 75

Char. Age (ky)

•  Observation; B∝ t -0.45⇒ α~1/0.45 = 2.2 ⇒ τc ~11t0

•  Magnetars can be much younger than their τc .

Possibly much younger than considered so far

2012/9/1 Magnetar Conference at Rikkyo 5

Magnetars are inferred to be extremely young objects, ���in agreement with their tight Galactic-plane concentration. ☆ Implications -  Magnetars are born as many as the B~1012 G NSs. -  Magnetars become quickly undetectable, because of their slow

rotation and exhaustion of magnetic energy. -  Many aged magnetars lurk in the Milky Way.

☆ Supporting evidence -  Mangetar-like objects lurking among ordinary pulsars (Hadasch). -  Many new magnetars with low dipole-B being discovered (Zane).

☆ Search for the predicted aged magnetars -  Candidates from ASCA Galactic plane survey -  Pilot survey with Suzaku AO7; soft BB spectrum, pulsations -  Hard tail search with ASTRO-H

3. How Are Magnetars Produced?

2012/9/1 Magnetar Conference at Rikkyo 6

Magnetars NSs B~1012 G NS CCOs (anti-magnetars) Black Holes

Core collapse���Supernovae

What causes this differentiation?

☆ X-ray diagnostics of ��� SNRs hosing magnetars ROSAT image of CTB109 (Sasaki+04), the SNR hosting the rather old magnetar X2259+586

2012/9/1 Magnetar Conference at Rikkyo 7

☆Suzaku study of CTB109 (Nakano+11)

A typical SNII, without peculiarity: •  T1=0.3 keV, T2=0.6 keV •  Abundance patterns

typical of SNeII •  Explosion energy

(1.7-7)×1051 erg/s

However, the estimated SNR age, ~20 kyr, is ≪ than τc=230 kyr of the central magnetar X2259+586. ���The age over-estimation of magnetars reconfirmed.

☆More detailed diagnostics to be conducted with the ��� ASTRO-H microcalorimeter.

4. Are There Magnetars in Binaries?

2012/9/1 Magnetar Conference at Rikkyo 8

SFXTs (Supergiant Fast X-ray Transients) and Long-Period (>103 s) X-ray pulsars can be promising candidates (Bozzo+08; Sasano+11)

100.

1no

rmali

zed c

ounts

s ke

V

Energy (keV)2 10

• Magnetar activity lasts < 105 yr, but what if they are in binaries? • SFXTs, with super-G companion, show violent flares, low quiescent Lx (<1032 erg/s), and often long (>500s) pulse periods. •  A popular interpretation, invoking stellar-wind clumps, is unlikely because NH does not necessarily increase during flares.

IGR 16195-4945 with Suzaku (Sasano+11)

(Rampy+09) Flare

Quiescence

2012/9/1 Magnetar Conference at Rikkyo 9

0.01

1

100

104

106

109 1010 1011

100 200 1000 500 Free fall time (sec)

Radius from the NS (cm)

Pressure (dyn/cm2 )

10-7 M0/yr

10-5 M0/yr

Mdot=10-6 M0/yr Ram pressure

0.01110010410610910101011

Co-

rotation

P=100 s

P=1000 s

Flare component

Persistent component

Magnetic pressure

Stellar wind

☆ Hypothesis (Sasano, Makishima+12): -  SFXTs contain magnetars. -  Co-rot. radius (rc)~Alfven radius (rA). -  Mag. gating produces violent variations. -  If density↓è rc＜rA èaccr. inhibition

due to propeller effects (flip-flop)

☆ With ASTRO-H: -  Search for ECR features

in 50-150 keV region. -  Pulse-phase dependent

Doppler shifts in the Fe-K line energy è estimates of rc~ rA

5. How is the Magnetic Energy Dissipated?

2012/9/1 Magnetar Conference at Rikkyo 10

Braithwaite 09

☆Need to understand basic MF configuration inside/outside the NS. •  Is the internal field configuration mainly toroidal or poloidal? •  Is the outer MF configuration mainly dipole or multipole? •  How large is the B-twist (how large is rot B) inside/outseide? •  Are the bursts triggered inside or outside the NS?

☆ Heat transport calculation is important (Yakovlev)

Still uncertain, but we already heard from Zane…

6. Do Microbursts Form Persistent Emission?

2012/9/1 Magnetar Conference at Rikkyo 11

☆Conjecture: Persistent magnetar emission is formed by numerous ��� μ-bursts (Thompson & Duncan 1996; Lyutikov 2003; Nakagawa 2007) LogN-LogS relations are too uncertain to answer this Q, but the increased persistent emission during burst activity, as well as spectral similarities between burst and persistent spectra, suggest Yes Persistent (left; Enoto+11) and stacked short-burst (right; Nakagawa+11)

spectra of SGR0501+4516 observed��� with Suzaku during the 2008 activity. ���Both exhibit clear hard-tail emissoin.

2012/9/1 Magnetar Conference at Rikkyo 12

Three broad-band Suzaku spectra of 1E1547.0-5408 in its 2009 activity (Enoto+10, +12) all show similar shapes è Possibly a common origin, involving sudden magnetic energy release.

Persistent

Stacked dim bursts A bright burst

Red/blue

Red/(blue-soft.comp)

☆ With ASTRO-H: -  Better comparison of

wide-band spectra -  Studies of still weaker

short bursts to better constrain LogN-LogS

-  Variability study of the persistent HXC.

7. How is the Hard Component Emitted?

2012/9/1 Magnetar Conference at Rikkyo 13

1

10

0.1

0.01

!F! (

a.u.)

1 10 100Energy (keV)

SGR 1806-20 (0.22 kyr)

1E 1547.0-5408 (1.4 kyr)

RXS J1708-40 (9.0 kyr)

4U 0142+61 (70 kyr)

☆ Suzaku wide-band persistent spectra of magnetars (Enoto+10) -  Characteristic 2-comp. spectra, ���

SXC + HXC (e.g., Kuiper+06). -  HXC is unusually hard; Γ= ���

0.5~ 1.5; no easy explanation. -  Older objects have lower HXC���

vs. SXC ratios but flatter Γ.

1

０

0.1 1 10 100 Characteristic Age τ(kyr)

2012/9/1 Magnetar Conference at Rikkyo 14

Positron infall

(Enoto+10, Shibata+11)

cutoff expected above 511 keV ���a wonderful subject of ASTRO-H

strong field 511

•  HXC and SXC are emitted with similar angular patterns ��� ∵) Lhard/Lsoft is specified uniquely by τc with little scatter. •  Both components probably emitted from similar locations, NS

surface or magnetosphere ∵) Similar pulse profiles. •  HXC has a high radiation efficiency ∵) some show Lhard≫ Lsoft

e+e- pear production in magnetosphere èannihilation on the NS surface è repeated “splitting” of the 511 keV photons in the MF

8. Why the Soft Comp. has Two T’s? (Nakagawa)

2012/9/1 Magnetar Conference at Rikkyo 15

☆ Suzaku spectra of SGR 0501+4516 SXC of the persistent emission TL=0.25 ± 0.03 keV TH=0.57 ± 0.03 keV TH/TL=2.3 +- 0.3

A strong short burst TL=3.3 ± 0.5 keV TH=15.1 ± 2.5 keV TH/TL=4.6 ±1.0

Enoto+10 Enoto+09

2012/9/1 Magnetar Conference at Rikkyo 16

At least a few possibilities..

1. The 2T property is an artifact caused by deviations from a pure blackbody spectrum.

1a. Effects of Compton scattering in the magnetosphere (Zane).

1b. Is the “blackbody” spectrum still valid when the emitting/absorbing electrons can move along 1-Dim?

2. The 2T results represent real effects,

e.g., different photospheres between O-mode and X-mode photons (double refraction; 複屈折)è polarization measurements essential

The two temperatures are in good proportion to each other (Nakagawa+07)

9. Do Magnetars really have B>Bc ?

2012/9/1 Magnetar Conference at Rikkyo 17

☆ Primary evidence -  MF strengths estimated from P and Pdot ☆ Circumferential evidence for very strong MF -  Tight clustering of their pulse periods (2—11 sec) -  Sporadic emission of bursts, often with super-Eddigton Lx -  Very unusual 2-component X-ray spectra -  … ☆ Decisive evidence -  Detection of proton cyclotron resonance features - Any firm evidence for QED effects..

10. How are the Magnetic Fields Held ?

2012/9/1 Magnetar Conference at Rikkyo 18

☆ Some electric currents (protons and/or electrons)?

☆Nuclear ferromagnetism due to neutron’s

spin alignment (Makishma+99; Tasumi)? ☆More sophisticated idea – pion-condensed

domain walls (Hashimoto+12) ?