1. Covering Solar-Wind Charge-Exchange from Every Angle with Chandra
Brad Wargelin
Chandra X-Ray Center
Smithsonian Astrophysical Observatory

2. Heliospheric Solar Wind Charge Exchange

3. The Forgotten Atomic Physics
Charge Exchange:
The Forgotten Atomic Physics

4. Outline Astrophysical charge exchange Charge exchange X-ray emission
Solar wind charge exchange
Charge exchange X-ray emission
Solar Wind Charge Exchange (SWCX) X-rays
ROSAT
Geocoronal CX
Heliospheric CX
Soft X-Ray Background and SWCX
SWCX in the Chandra Deep Field-South
Future observations

5. Charge Exchange
Charge exchange (CX) is the radiationless transfer of one (or more) electrons from a neutral atom or molecule to an ion.
Molecular cloud chemistry: O+ + H  O + H+ (13.618, eV)
Solar wind proton-H CX: H+ + H  H + H+
Some SW protons (tied to B field) CX with neutral H from the ISM, particularly between the heliopause and bowshock, creating hot H atoms, or Energetic Neutral Atoms (ENAs).
Heliosheath is between term shock and Hpause.
H wall is between Hpause and bow shock.

6. Local Interstellar Cloud (partially neutral, 26 km/s)
Hydrogen
Wall
Figure credit: NASA/Walt Feimer?
Local fluff/G cloud (with neutral gas!) moving roughly outward from Galactic Center (to right)

7. IBEX and ENA Imaging
Energetic Neutral Atoms (ENAs) no longer tied to B field. These can be “seen” by the Interstellar Boundary Explorer (IBEX) to image the CX interaction region.
ENAs form from CX lots of place, sometimes several times following reionizations.
Most of them form, however, beyond the term shock.
SW ions (protons) CX with neutrals, escape from B field---they image the CX interaction region.
ENAs were also observed by probes around planets dating back to 1970s. Current TWINS mission around Earth.

8. Astrospheres and Mass Loss Rates
Excess blue-shifted/broadened Lyman-α absorption due to the hot atoms in the Hydrogen wall can be used to determine mass loss rates of other stars surrounded by partially neutral ISM (Wood, et al ).
Top line = intrinsic stellar profile (modeled)
Dashed = after ISM absorption
Hashed = excess absorption vs other lines
a Cen B Lya; Linsky & Wood (1996)

9. Ionization balance of metals in photoionized nebulae—CX cross sections are large and a tiny fraction of neutral H is all it takes to make CX more important than RR and DR (e.g., Dalgarno 1985, Ferland et al. 1997).
Emission of 511-keV annihilation line in the GC via positronium formation (nearly 100%). Roughly half of the positronium is formed via CX with the rest from RR. (2g emission with opposite spins, 3g with same spin.)
Churazov et al. (2005, 2011)

10. Charge Exchange X-Rays
O8+ + H  *O7+ + H+ O7+ + H + hn
Highly charged ions capture electrons into high-n levels (nmax ~ q 3/4) that emit X-rays.
CX is a semi-resonant process. During a collision, energy levels distort and overlap at “curve crossings.”
Cross sections are large (few × cm2).
35 eV for O7+ n=5 to continuum vs 13.6 for H.
(13.6 eV)
(35 eV)

11. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays? (Silk & Steigman 1969)
ASCA; Tanaka (2002) S Ar Ca Fe
Notes go here
ASCA spectra of GC, Sgr, Sct with identical model curves (other than norm, NH, 6.4-keV line).

12. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
No.
Watson (1976),
Bussard et al. (1978)…
Revnivtsev et al. (2006, 2007)
Notes go here

13. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Notes go here

14. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating (e.g., Rice et al. 1986)
Na10+ n=7-1
Wargelin et al. (1998)
Alpha, beta, gamma, delta, epsilon, zeta, iota, eta, theta, kappa…

15. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating
Supernova remnants (Wise & Sarazin 1989)
ASTRO-H will be perfect for this

16. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating
Supernova remnants?
ASTRO-H will be perfect for this
Cygnus Loop; Katsuda et al. (2011)

17. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating
Supernova remnants?
CX X-ray emission was known
Notes go here

18. Discovery of Solar Wind CX X-Rays
Key events:
Comet Hyakutake, ROSAT (Lisse et al. 1996)
SWCX explanation (Cravens 1997):
Highly charged ions in SW
+ neutral H2O, CO, CO2
What was ROSAT? RASS and pointed obs’s, PSPC and HRI.
PSPC energy bands.

19. SWCX X-ray Spectrum (for Slow Wind)
Model CX spectrum (C,N,O) with 6 eV resolution
Wargelin et al. (2004)

20. Discovery of Solar Wind CX X-Rays
Key events:
Comet Hyakutake, ROSAT (Lisse et al. 1996)
SWCX explanation (Cravens 1997)
First CCD spectrum of comet, by Chandra (Lisse et al. 2000)
LINEAR S4
C/1999 S4 (LINEAR)
Chandra/Lisse 2000
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….
Beiersdorfer et al. (2003)

21. Discovery of Solar Wind CX X-Rays
Key events:
Comet Hyakutake, ROSAT (Lisse et al. 1996)
SWCX explanation (Cravens 1997)
First CCD spectrum of comet, by Chandra (Lisse et al. 2000)
LINEAR S4
C/1999 S4 (LINEAR)
Chandra/Lisse 2000
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….
Two dozen comets and several planets to date by ROSAT, EUVE, BeppoSAX, Chandra, XMM, Swift, Suzaku.
Meanwhile……
Beiersdorfer et al. (2003)

22. LTEs & Discovery of Geocoronal and Heliospheric Emission
ROSAT All Sky Survey (RASS; 1990) revealed multi-orbit (“Long Term”) enhancements in the SXRB.
Cravens, Robertson, & Snowden (2001): temporal correlations between counting rate and SW flux
 LTEs are from geo/helio SWCX fluctuations.
¼-keV band, Galactic coords;
Snowden et al. (2009)

23. Geocoronal Emission
Geocoronal emission = SWCX in Earth’s exosphere, outside the magnetosphere (R > 10RE).
10 R_E corresponds to 2.5 minutes SW transit time

24. Geocoronal Emission
Geocoronal emission = SWCX in Earth’s exosphere, outside the magnetosphere (R > 10RE).
10 R_E corresponds to 2.5 minutes SW transit time
X-ray missions generally look out through the flanks.
Robertson et al. (2006)

25. Lunar X-Rays (on the Dark Side) are Geocoronal
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….
ROSAT, Schmitt 1990
Chandra, Wargelin et al. 2004

26. Moon, Chandra; Wargelin et al. (2004)
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….
Moon, Chandra; Wargelin et al. (2004)
HDF-N, XMM; Snowden et al. (2004)

27. Heliospheric Charge Exchange
Solar wind + H/He from ISM  100-AU halo
Heliospheric CX ~ 10x geocoronal CX
Model heliospheric emission from CX with H. Axis units in AU. LIC is moving to the right. Robertson et al. AIP Proc. 719 (2004).

28. Heliospheric Emission--looking down on ecliptic plane
More neutral H upwind
He focussing cone downwind
(AU)
(AU)
Pepino et al. (2004)

29. Slow vs Fast Solar Wind
At solar max, wind is a mix of slow and fast at ~all latitudes.
At solar min, wind is stratified, with slow wind near the ecliptic.
The fast wind is much less ionized and produces less CX X-ray emission.

30. CX Emission at Solar Min--view from Ecliptic Plane
Stratified wind: slow and highly ionized near ecliptic  higher CX emissivity.
Little emission in fast wind. AU
Pepino et al. (2004) AU

31. The Soft X-Ray Background (SXRB)
SXRB emission components:
Absorbed extragalactic (~power law)
Absorbed thermal Galactic Halo emission
Unabsorbed thermal from Local Bubble
Heliospheric and geocoronal SWCX
How much emission is from CX vs the Local Bubble? The answer strongly affects our models of the LB.

32. Modeling CX Emission Need to know H and He distributions
SW composition and density all along line of sight (LOS)
State-specific CX cross sections for all ions (and neutrals) as f(v)
Radiative decay paths and line yields
Local SW measured by ACE

33. Living in a Fog
We can try to observationally separate SWCX emission from cosmic components with differential measurements:
Spatially (using dark clouds to block distant emission)
Spectrally (some day, with high-resolution nondispersive detectors)
Temporal changes (periods of hours to Solar Cycle)
Observation geometry

34. The Chandra Deep Field-South
4 Msec of observations:
3 in Oct, Nov 1999 (-110 C)
9 in May, Jun, Dec 2000
12 in Sep-Nov 2007
31 in Mar-Aug 2010
RA,Dec 3:32:28, -27:48:30
Gal l,b , -54.4
Ecl lat,lon 41.1, -45.2
The CDFS is the only X-ray deep field conducted during Solar Max and Min and it has the greatest orbital coverage.

35. LOS is 45 down, into page.
2000
0.8 Ms in 2000 (solar max)
1.0 Ms in 2007 (solar min)
2.0 Ms in 2010 (solar min)
J. Slavin

36. In 2000, there is slow wind all along the LOS and most observations are from within the He cone.
From within the He cone, CX intensity is higher and more of the emission is from nearby, where SW conditions are measured.

37. In 2007, SW is stratified, LOS through fast wind, observations all outside He cone, looking downwind.

38. Expectations for observed SXRB in 2000 vs 2007:
Higher baseline level
More variability
Closer correlation with ACE-measured SW ion flux

39. … but 2007 can be VF filtered.
Obvious difference (partly due to 40% decrease in QE near 600 eV)

40. … but 2007 can be VF filtered

41. Compare O emission vs average ACE/SWICS O7+ flux
Many thanks to the ACE/SWICS team for their public data!

42. 2000:
Higher SXRB
Variable
Correlated with local SW
2007:
Lower SXRB
Nearly constant
Little correlation with SW

43. The goal is to accurately model and remove SWCX emission and obtain the true cosmic background.?

44. The Future
High-resolution spectra from microcalorimeters will help immensely. Astro-H launch in 2014 (E ~ 5 eV) .
CX spectra differ from collisional: enhanced high-n Lyman, He-like f...
Explore the 1/4-keV band (where ROSAT LTEs are strongest)
500 km/sec  E=1 eV at 600 eV

45. 100 s of SXRB from XQC rocket flight vs thermal model (McCammon et al

46. CX Spectra
High-n levels are preferentially populated.

47. CX Spectra High-n levels are preferentially populated.
H-like: enhanced high-n
H-like Fe (with N2 in EBIT)

48. He-like Fe CX Spectra High-n levels are preferentially populated.
H-like: enhanced high-n
He-like: enhanced triplet f and i
EIE at 15 keV
He-like Fe
CX with N2
10 eV/amu
Wargelin et al. (2008)

49. Astrospheric Charge Exchange
CX must also occur around other stars with highly ionized winds (G,K,M) residing inside clouds with neutral gas (LIC, G).
Imaging + spectra yields:
Mass-loss rate
Local nneutral
Wind velocity and composition
Astrosphere geometry
Coronal emission is ~104x brighter, though.
Need very large collecting area, good spatial and spectra resolution.

50. Reviews
“Charge Transfer Reactions” Dennerl, Space Sci. Rev. (2010) Astrophysical examples and broad historical review
“EBIT charge-exchange measurements and astrophysical applications,” Wargelin et al., Canadian J. Physics (2008) Astrophysical examples and lab spectra/atomic physics

52. Geocoronal emission responds to SW much faster than heliospheric.

53. Talk amongst yourselves….
Things to think about:
Roughly 10% of RASS SXRB is from unresolved Galactic point sources
RASS was conducted at solar max
R12 (1/4-keV) rate ~4 x R45 (3/4-keV) rate
LTEs most prominent in R12 band
No models or data for CX in 1/4-keV band

54. More detailed correlations:
Short-term X-ray variability vs ACE O7+ flux  fractional contribution of geocoronal CX emission
Account for Chandra orbital geometry
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….

55. More detailed correlations:
Short-term X-ray variability vs ACE O7+ flux  fractional contribution of geocoronal CX emission
Account for Chandra orbital geometry
Comets 1996, planets with Chandra
Moon/geocorona with ROSAT and then Chandra with E res to see O lines
Heliospheric with XMM and Chandra--current hot topic
But I was asked to NOT talk about those….

56. Diffuse X-Ray Spectrometer (DXS)
SXRB eV. Sanders et al. (2001)

57. Where does the observed SWCX emission originate?
From within the He cone, CX intensity is higher and more of the emission comes from nearby, where SW conditions can be measured.

58. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating
Supernova remnants (Wise & Sarazin 1989)
ASTRO-H will be perfect for this
Cygnus Loop; Katsuda et al. (2011)

59. ACIS BG 70% higher in 2007 than 2000

60. CX Spectra High-n levels are preferentially populated.
H-like: enhanced high-n
He-like: enhanced triplet f and i
He-like O (with CO2). Beiersdorfer et al. (2003)

61. Solar Max vs Solar Min
At solar max, wind is a mix of slow and fast at ~all latitudes.
At solar min, wind is stratified, with slow between latitude -20 and +20.
The fast wind is much less ionized and produces less CX emission.
Ulysses data. E.J. Smith et al., Science 2003

62. Living in a Fog
Observed geo/helio SWCX emission depends on CX emissivity all along the LOS, which means that it depends on:
SW flux for each X-ray emitting ion as f(t,x) (only measured locally)
Neutral gas density as f(t,x) (modeled)
Observation direction
Observation location
We can try to separate SWCX emission from cosmic components with differential measurements:
Spatially (using dark clouds to block distant emission)
Spectrally (some day, with high-resolution nondispersive detectors)
Temporal changes
Observation geometry

63. Expectations for observed SXRB in 2000 vs 2007:
Higher baseline level
More variability
Closer correlation with ACE-measured SW ion flux

64. CX X-Rays in Olden Times
Idea dates back to 1970s:
Galactic Ridge X-rays from cosmic rays?
X-ray lasers: population inversion in “hollow ions”
Tokamaks: plasma edges and neutral beam heating
Supernova remnants?
ASTRO-H will be perfect for this
Cygnus Loop; Katsuda et al. (2011)

65. IMP-8 in LEO

67. Compare O emission (500-700 eV) vs…
Model CX spectrum with 6 eV resolution

68. Conducted at solar maximum.

69. Merge 52 observations for source detection

70. Remove 400+ sources

71. Trim radius to common FOV with radius 7.7’.

72. l And, DK Uma are marginal detections.
Prox Cen M5.5 V
EV Lac M3.5 V
x Boo G8 V + K4 V
= III,IV
l And, DK Uma are marginal detections.
Xi Boo
Wood et al. (2005)