Particle acceleration in active galaxies the Xray view

1
Particle acceleration in active galaxies the
X-ray view

• Martin Hardcastle (U. Herts)
• Thanks to many co-authors including Ralph Kraft
  (CfA), Judith Croston (Herts), Diana Worrall
  (Bristol)
• X-ray Universe, Granada, 28th May 08

2
Overview

• Motivation
• Types of radio galaxy
• Key role of X-ray synchrotron
• Results from low-power radio galaxies
• Results from high-power radio galaxies
• Particle acceleration mechanisms

3
Motivation (1)
We observe synchrotron radiation, implying
acceleration of high-energy leptons we want to
relate what we observe to energy transport
mechanisms.
4
Motivation (2)
We observe high-energy cosmic rays implying
high-energy baryonic acceleration this must
always be accompanied by acceleration of leptons
so by studying one we can understand the other.
5
Types of radio galaxy

• In the radio, FRIs have centre-brightened
  structures often dominated by bright jets
• FRIIs have edge-brightened structures often with
  prominent hotspots.
• FRIs have low radio luminosity. FRIIs have high
  radio luminosity.
• FRI/FRII difference implies different jet physics.

6
Hotspot
Jet
Core
Lobe
Hotspot
Plume
FRI
FRII
7
X-ray synchrotron

• For radio and even optical synchrotron radiation
  we cannot distinguish between particles that have
  been accelerated upstream and advected to where
  we see them, particles genuinely accelerated
  where we are looking.
• For X-ray synchrotron the loss timescales are so
  short that particles can travel only a few pc
  from their sites of acceleration so effectively
  for the distances involved X-ray synchrotron
  emission tells us where particle acceleration is
  happening now.

8
FRI jets

• FRI jets started showing up in large numbers soon
  after the launch of Chandra (Worrall et al 01
  MJH et al 01).
• X-ray spectra mostly consistent with
  extrapolation of radio-optical gt synchrotron
  origin generally assumed.

9
FRI jets

• X-ray emission is diffuse gt can no longer
  sustain a picture of a single acceleration
  location.
• X-ray ( optical) spectrum is steep, Ggt2.0 not
  consistent with the Heavens Meisenheimer
  continuous injection model used for hotspots.
• High-energy particle acceleration appears to be
  associated with bulk jet deceleration.

10
Jet deceleration
Laing et al 2002a, 2002b MJH et al 2002
11
Particle acceleration process

• Derives its energy from the jet deceleration
  process (no problem with energetics).
• Distributed throughout the jet
• Averaged over the jet, produces a flat radio
  spectrum and steep X-ray spectrum, with a break
  in the IR/optical.
• We need to be able to resolve the particle
  acceleration process on the loss spatial scale to
  see whether it is genuinely diffuse or just
  distributed. Only possible in the nearest FRI,
  Cen A.

12
Cen A (Chandra)
720 ks of Chandra data, including a Chandra VLP
(PI Ralph Kraft). See Kraft et al 2002, MJH et al
03, MJH et al 06, Kataoka et al 06, MJH et al 07,
Jordán et al 08, Sivakoff et al 08, Worrall et al
08, Kraft et al 08 for some Chandra results.
13
Key results on Cen A jet
MJH et al 2003

• Strong point-to-point radio/X-ray ratio variation
  particle acceleration efficiency varies
  spatially
• Compact X-ray emitting knots are stationary in
  the radio maps could be shocks?

14
Key results on Cen A jet

• 3) Diffuse X-ray emission comes to dominate at
  large distances from the nucleus.
• 4) X-ray spectra of knots are flat X-ray
  spectrum of diffuse emission gets progressively
  steeper ending at very high values X-ray surface
  brightness falls off faster than radio.

MJH et al 2007 ApJL
15
Particle acceleration processes

• We suggest that the spatial and spectral
  differences between the compact knots and
  diffuse emission means that there are two
  acceleration processes going on in Cen A.
• The compact knots may be shocks producing
  X-ray-emitting electrons by first-order Fermi.
• The diffuse emission surrounding them is probably
  something else! return to this later.

16
FRII radio galaxies

• Hotspots in FRII radio galaxies are the physical
  manifestations of the jet-termination shock, so
  we expect first-order Fermi acceleration at the
  hotspot. Consistent with early work on broad-band
  SEDs.

Meisenheimer et al 1989 Blandford Rees 1974
17
X-rays from FRII hotspots

• Early work on X-ray detections of hotspots
  focussed on objects that radiate by the
  synchrotron-self-Compton (SSC) mechanism (e.g.
  Harris et al 1994, MJH et al 01). Wont discuss
  this here.
• Increasingly its become clear (MJH et al 04
  Kraft et al 05) that some hotspots X-ray
  emission cant be explained by an inverse-Compton
  model but must be synchrotron instead.

18
X-rays from FRII hotspots
Colours are radio emission, green contours show
X-rays
19
FRII hotspots
Colours are radio emission, green contours show
X-rays
20
Problems with the standard picture

• So we can use X-ray synchrotron to locate
  particle acceleration in FRII hotspots too.
• But the results deviate from our expectations in
  three ways
• 1) When we see X-rays coincident with
  radio/optical hotspots, although we sometimes see
  the sort of smoothly steepening spectra that we
  expect from radio-through-optical observations,
  we often dont

21
3C33
Kraft et al 2006
22
3C33
23
Problems with the standard picture

• (continued) i.e. if we want this emission process
  to be synchrotron emission, we have either to
  have an electron energy spectrum that turns up at
  high energies, or we have to abandon our one-zone
  model of the electron population.
• We often see spatial offsets between the peaks of
  the radio/optical/X-ray emission

24
3C227
MJH, Croston Kraft 2007
25
Problems with the standard picture

• 2) (continued) this more or less requires us to
  abandon the one-zone picture, but what do we put
  in its place?
• 3) We often see diffuse X-ray emission, implying
  distributed particle acceleration, throughout
  bright hotspots completely inconsistent with
  the idea that particle acceleration is taking
  place at localized shocks.

26
3C390.3
MJH, Croston Kraft 2007
27
Hotspot consequences

• Problem 1 means that either the hotspots are not
  homogeneous, or they are not accelerating
  particles in the expected way, contrary to what
  the radio/optical spectra told us.
• Problem 2 means that if particle acceleration is
  localized at shocks, the shocks are, at least
  some of the time, not where the peak of the radio
  emission is, contrary to what everyone has always
  assumed.
• Problem 3 means that at least some of the
  particle acceleration is not localized at shocks
  anyway, contrary to the standard picture.

28
Acceleration mechanisms

• Requirement for a diffuse acceleration mechanism
  (3) is interesting because it may be related to
  the one seen in FRIs.
• What diffuse acceleration processes are there?
• Turbulence (second-order Fermi acceleration) and
  shear provide possible sources for particle
  acceleration in jets (e.g. Stawarz Ostrowski
  2002). Unfortunately in resolved FRI jets like
  Cen A there is no evidence for systematic
  edge-brightening or flatter spectra in the
  diffuse emission at the edge.
• Magnetic field line reconnection (e.g. Birk
  Lesch 2000) is promising but makes few testable
  predictions about the spectrum of high-energy
  particles.

29
Summary

• The traditional picture that all particle
  acceleration in radio-loud AGN happens via a
  first-order Fermi process at shocks needs major
  revision.
• In FRIs, shocks may contribute but dont dominate
• In FRIIs, where we know shocks should be present,
  the model fails to explain much of what we see.
• In particular we need a distributed particle
  acceleration mechanism which can extend over tens
  of kpc.
• We need testable predictions from existing models!