Hinode / EIS is operating - is there anything on Fe in the EUV left to be done in the laboratory? Elmar Träbert

Ruhr-Universität Bochum, D-44780 Bochum, Germany, and Lawrence Livermore National Laboratory, Livermore, CA 94550, U.S.A.

in collaboration with P. Beiersdorfer, G. V. Brown, H. Chen, J. H. T. Clementson, D. B. Thorn, a. o.

German support by DFG; work at UC LLNL under US DoE Contract No. W-7405-Eng-48

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

MSSL, Dec 12, 2006

Fe in the solar corona; EUV 20 – 35 nm

Fe to Fe A matter of temperature and density

Properties of laboratory light sources - Laser-produced plasma - Tokamak - Electron beam ion trap (EBIT) - Foil-excited ion beams (beam-foil spectroscopy) - Heavy-ion storage ring

Examples of individual techniques or data combinations Time integrated vs. time resolved data Tokamak + LPP + BFS BFS prompt / delayed spectra Lifetime measurements (BFS, storage ring, EBIT)

Promising approaches for Fe

BFS Spectra of elemental purity BFS Delayed spectra Storage ring E1-forbidden decay rates EBIT High-resolution survey spectra

Time characteristics?

8+ 14+

Atomic level lifetime t = 1/S(Aki)

Transition probability Aki Multipole order E1, M1, E2, M2, E3, M3

Resonance lines, high nuclear charge Z: femtosecond lifetimes

E1-forbidden lines, not so high charge states: millisecond lifetimes

Ultrahigh vacuum: Collision rates of order 1/s

Level populations (line ratios) depend on excitation and deexcitation: Density diagnostic

Measure radiative decay rates of long-lived levels

Optical depth depends on the A-value

Extreme cases in astrophysics: All lines in absorption, except for those from extremely long lived levels

Density Solid / Laser-produced plasma 10 cm atomsGas (1 bar) 3 10 cm atomsTokamak 10 cm atoms 10 cm electrons EBIT 10 cm atoms 10 cm electrons Solar corona 10 cm electrons Planetary nebulae 10 cm electrons

Properties of laboratory light sources: temperature and density- Laser-produced plasma - Tokamak- Electron beam ion trap (EBIT)- Foil-excited ion beams (beam-foil spectroscopy) - Heavy-ion storage ring

23

19

13

12

5

11

8 - 11

4 - 8

-3

-3

-3

-3

-3

-3

-3

-3

Ion beam

Foil

GratingDetector

Slit

���� electrons

Faraday cup

v = - => t = -xt

xv

v Ion velocityx Distance from the foilt Time after excitation

Beam-foil spectroscopy

0

200

400

600

800

1000

Fe 20 MeV, at the foil

Cou

nts

0

50

100

150

20 25 30 35 40 45 50 55 60

Fe 24 MeV, 175 mm downstream (15 ns)

Cou

nts

Wavelength (nm)

Beam-foil spectrarecorded at the foil (prompt spectra) are very line-rich (excitation under high density conditions).

Delayed spectra much more resemble the spectra obtained of the solar corona (low density environment). The lines represent intercombination transitions.

Signal (log)

Time

t << t1 2

Problem: Long-lived levels ==> little signal per time interval

Beam-foil spectroscopyObservation at the foil: The spectrum is dominated by decays of short-lived levelsDelayed observation: Much fewer lines, almost no background

Signal (log)

Time

t << t1 2

Problem: Long-lived levels ==> little signal per time interval

Beam-foil spectroscopyObservation at the foil: The spectrum is dominated by decays of short-lived levelsDelayed observation: Much fewer lines, almost no background

Mg Al Si

Fe FeFe

The intercombination lines in Fe XIII and Fe XIV were recognized in beam-foil spectra on the basis of unidentified solar corona observations. The solar data had been recorded with ten times better resolution and accuracy, but it took BFS to tell the element and note the upper level longevity.

Beam-foil spectra of Fe top: at the foil (prompt)middle: delayed by about 1 nsbottom: delayed by about 10 ns

At the time of measurement, and for years after, even the strongest lines in each of the spectra remained unidentified.

By now we know that in the bottom spectrum, most lines are from Fe X and Fe XI. The strongest lines are from levels with lifetimes in the range 10 - 60 ns. The strongest lines in the solar corona (in this range) have upper levels of lifetimes closer to 600 ns.

0

25

Counts

50Fe

Ar

7 MeV

0

25

50Cl/S/P

S

S9 MeV

0

15

30

S

SSi

8 56 e eV 3 / 5 s ds; s t s .x2'=14.6" .0075"/CH

12 MeV

17.5 20.0 22.5 Wavelength / nm

25.0 27.5

0

15

30

P

P

S

P

Si

15 MeV

Beam-foil spectroscopy

The ion beam energy determines the average charge state reached. Lines from particular charge state ions can thus be enhanced or suppressed.

The delayed spectra shown here were recorded during a search for specific Fe IX (Ar-like, not all of them found yet) and Fe XIII (Si-like, successful search) lines.

0

25

Counts

50

FeDelay 5 ns

Ar

7 MeV

0

25

50Cl/S/P

Delay 5 ns

S

S9 MeV

0

15

30P

Delay 5 ns

S

SSi

12 MeV

0

15

30

P

P

Delay 5 nsS

P

Si15 MeV

17.5 20 22.5

Wavelength / nm25 27.5

0

25

5015 MeV

PDelay 14 nsP

PSi

S

The ion beam energy defines the charge state distribution after the ion-foil interaction.

Systematic variation of the ion beam energy can help to maximize individual spectra and to identify the charge state of origin.

ns

np

2

1S P

o

0

12E1ns

P1 3 o

01

M2 M1

hfs

Mg - likeBe - like

hfs+

picture credit: M. Grieser

Heavy-ion storage ring TSR at Heidelberg Circumference 55 m

from 12 MV tandem accelerator

The ring is mostly used for atomic physics.

Channeltrons

+-

Beam profile monitor

LiF filter

Heavy-ion storage ring

Injector

Photomultiplier

Mirror+-

Beam profile monitor

Beam current monitor

Detector of visible / near UV / UV light

Detection of VUV / EUV light

Passive atomic lifetime measurements at TSR Heidelberg

Atomic lifetime measurements at TSR have reached high accuracy: as little as 0.14% uncertainty.

Garstang & Shamey 1967Nussbaumer 1972Mühletaler & Nussbaumer 1976Lin & Johnson 1977Nussbaumer & Storey 1978Glass & Hibbert 1978Laughlin, Constantinides & Victor 1978Cheng, Kim & Desclaux 1979Glass 1982Cowan, Hobbs & York 1982

Curtis 1991

Chou, Chi & Huang 1994

Fleming, Hibbert & Stafford 1994

Froese Fischer 1994

Ral'chenko & Vainshtein 1995Ynnerman & Froese Fischer 1995

Smith et al . 1984

Kwong et al. 1993

Doerfert et al . 1997

Transition probability / s -1 70 80 90 100 110 120 130

Brage, Fleming & Hutton 1996

Kwong et al. 1983

Jönsson & Froese Fischer 1997

T

E

E

E

E

T

T

TT

T

T

T

T

TTT

TTTTT

T

x

Froese Fischer & Gaigalas 1997T

T Chen, Cheng & Johnson 2001

Intercombination transition in C III(Be-like)

The agreement of experiment and theory, achieved after so many years, is splendid - but not perfect.

10

15

20

25

30

35

40

45

50

0 0.05 0.1 0.15 0.2 0.25

Be sequence

(Z-2

)2 S

r

1/Z

Resonance

IntercombinationXe

Kr

Fe

ON

CB

Ne

Be

F

MgAl

Si

PS

Cl

Ne

O N CBKr

Fe

L J Curtis has derived a mixing-angle formalism that interconnects resonance and intercombination line strengths on the basis of spectroscopic data. The heavy-ion storage ring lifetime results on Be-like ions of B through O are internally consistent, but do not agree with the prescribed trend.

10 5

0 500 1000 1500

Cou

nts

Time (ms)

10

10

10

10

1

2

3

4

Co XIII190 - 290 nm

150 ms

11 ms

1 ms 2.5 ms

2

4 o DS P 22

3p3s 3

1/2

3/2

3/2

3/2

5/2

P - like Fe

oo

124 nm / 135 nm

217 nm / 241 nm

290 nm - 357 nm

11+

(Co )12+

The line intensities in the ground configurations of N- and P-like ions serve as temperature diagnostics in astrophysics.

Measurement of M1 and E2 transition probabilities in the ground configuration of ions that are of astrophysical interest

Very few calculations predict all four level lifetimes in P-like ions close to the experimental findings.

-80

-60

-40

-20

0

20

40

60

80

24 26 28 30 32 34 36 38

P sequence 2D5/2

lifetime

Nuclear charge Z

Kaufman & Sugar 1986

NIST 2005

Garstang 1972; Smitt 1976

Smith & Wiese 1973

Huang 1984Mendoza & Zeippen 1982

TSR

3s 3p2 3

-80

-60

-40

-20

0

20

24 26 28 30 32 34 36 38

P sequence 2D

3/2 lifetime

Dev

iatio

n fr

om e

xper

imen

t (pe

rcen

t)

Nuclear charge Z

3s 3p2 3

Kaufman & Sugar 1986

Smitt 1976

Experiment TSR

Huang 1984

Mendoza & Zeippen 1982

NIST 2005Garstang 1972Smith & Wiese 1973

Biemont & Hansen 1985Biemont 2001

The predicted M1 and E2 transition rates in the literature or on the web (including the NIST data base) are not all as bad as in this case. The problem is how to find out which calculational results are good, and one finds out - by experiment ...

Isoelectronic trends can be used to ascertain the consistency of data sets.

Biemont 2001

EBIT Experiment

Biemont & Hansen 1985

-30

-20

-10

0

10

20

0 10 20 30 40 50 60

M1 transition rate in the He sequence

Dev

iatio

n (e

xpt.

- th

eory

) (p

erce

nt)

Nuclear charge Z

EBIT range

Beam-foil spectroscopy range

TSR

The measurements at the heavy-ion storage ring TSR and at the LLNL EBIT are in excellent agreement with relativistic theory

6000

7000

8000

9000

0 50 100 150 200

Cl XIII

Cou

nts

Time (ms)

20000

10000

e-beam on

e-beam off

magnetic trapping mode

fluorescent decay

ion production

electron trapping mode

Livermore electron beam ion trap - the archetypical EBIT

Excitation curve in the fast trapping cycle for an atomic lifetime measurement

Stovepipe for optical observations

Atomic lifetime measurement in the visible range of the spectrum

The basic technique is simple and robust.E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

10

10

0

0

104

10 5

0 20 40 60 80 100 120 140

Cl Sum - 17800

Counts

Time (ms)

1

0

2

3

12+ 575 nm

1.25

1.30

1.35

1.40

1.45

1.50

1.55

Line Strength (M1) B-like ions

S

TSR

LLNL EBIT

NIST EBIT

EST

LLNL EBIT

LLNL EBIT

TSRHD EBIT

1.25

1.30

1.35

1.40

1.45

1.50

1.55

17 18 19 20 21 22 23

Line Strength (M1)

S

Nuclear charge Z

F-like ions

LLNL EBIT LLNL EBIT

EST

EST

TSR TSR

16 17 18 19 20 21 22 23

1+ 2ap

n=2 M1 transition between the fine structure levels of the ground state. Predicted line strength S = 4/3.

EST Electrostatic ion trap (Kingdon trap) TSR Heavy ion storage ring at HeidelbergEBIT Electron beam ion traps at NIST Gaithersburg, Livermore, and Heidelberg

1

10

100

1000

104

105

0 20 40 60 80 100 120

Cou

nts

Time (ms)

Fe XIV

The "green iron line" decay curve as seen in the LLNL EBIT

46 48 50 52 54 56 58 60 62

Fe XIVKrueger & Czyzak 1966

Warner 1968

Smith & Wiese 1973

Kastner 1976

Kafatos & Lynch 1980

Eidelsberg et al. 1981Froese Fischer � & Liu 1986Huang 1986Kaufman & Sugar 1986Biemont et al. 1988Bhatia & Kastner 1993Moehs & Church 1999

Beiersdorfer et al. 2003

Transition rate (s )-1

ab initio

ab initio

ab initio

ab initio

Vilkas & Ishikawa 2003

Storey et al. 2000

Smith et al. 2005Crespo Lopez-U. et al. 2006

The "green iron line" in the solar corona has found plenty of interest

54 56 58 60 62 64 66 68 70

Fe XKrueger & Czyzak 1966

Warner 1968

Smith & Wiese 1973

Kastner 1976

Kafatos & Lynch 1980

Eidelsberg et al. 1981Huang et al. 1983

Kaufman & Sugar 1986

Bhatia & Doschek 1995

Moehs & Church 1999

Träbert et al. 2004

Transition rate (s )-1

Texas A&M / RenoTSR (extrapolated)

ab initio

767472

Kohstall et al. 1999Dong et al. 1999 ab initio

ab initio

Biémont et al. 1988

}

In Ar-like Kr ions, two lifetimes have been measured of levels with major M2 decay branches.

In Ar-like Fe IX, several of the "slow" ground state transitions are still being sought in terrestrial light sources.

1

2

0

3s 3p2 6

0

2M1

53s2 3p 3d

4

3

2

Ar - like Fe 8+

1S 3P D1 3F FP 3 11D

400 000

550 000

500 000

450 000

E / cm-1

0

600 000

1

1 3

32M1

M1M1

M2

E1

E1

E1

M2

M2

T / 10 cm3 -1

700

100

200

300

400

500

600

0

0J 4321 5

3s 3p

3s 3p 3d

3s 3p

2 4

5

2 3

Fe XIMany ions have a few levels of particularly high J that cannot easily decay (at least not by E1 transitions). These levels act as population traps and are important for charge state distributions.

3s 3p2

1S

1

3s

Si - like Fe

3p3

2

0

3P

2

0

D

2

1

E2

M1

E2

3s 2 3p 3d

3F432

M2

12+

M2

o

In delayed BFS observations, the 3s 3p 3d F level decay has been seen.

The decay of the 3s 3p 3d F level has been detected via its influence on an M1 decay curve in the ground configuration.

o3

4

2

2

3

3 o

-20 -15 -10 -5 0 5 10 15 20

Fe ion lifetimes

Deviation from best data (percent)

Fe XI

Fe XII

Fe XIV

Fe XIII

TSR

TSR

TSR

TSR

TSR

TSR

TSR

EKT

EKT

EKT

EKT

EKT

EKT

EBIT - I

EBIT - I

12D

12D

D

D

P

P

P

2

2

2

2

2 o

o

o

o

o5/2

3/2

1/2

3/2

3/2Fe XII

Fe XII

Fe XII

HD-EBIT oo

Electronmode

Magnetic mode Off

On

Magnetic field, drift tube voltage

Electron space charge, magnetic field,

drift tube voltage

Ion-neutralcollisions

Electron-ioncollisions

Iontrapping

X-rayproductionmechanism

Electronbeam

Major operating modes of electron beam ion traps

n=3 –7

X-raysproducedby chargeexchange

2 4 6X-ray energy (keV)

Tim

e (s

)

0.0

0.5

1.0

1.5

Ele

ctro

nm

ode

Mag

netic

mod

e

Ka

Excitation vs. decay curve Prompt vs. delayed emission

Charge state q0 18161412108642

E (eV)ion

0

600

500

300

100

400

200

1000

900

800

700

Ionization energies of Ar ions

X

X

X

X

X

X

X

X

X

XXX

XXXX

q+

It is easy to reach any charge state of ions in an electron beam ion trap, by adjusting the electron beam energy.

In fact, it is too easy to reach any low charge state - there is little selectivity. EBIT does better with higher charge states.

In the electron beam ion trap (EBIT), the electron beam energy determines the highest charge state that can be reached.

Systematic observations of this kind have been done at Livermore in support of Chandra and XMM/Newton data evaluation.

Sample spectra taken from Lepson et al., Ap. J. 578, 648 (2002).

Grating

CCD Detector

EBIT

Focus

WavelengthR=44.3 m

Grazing incidence flat-field spectrograph at Livermore

Multichannel detection is a key element for precision spectroscopy.

This instrument provides the highest spectral resolution of any EUV equipment at any electron beam ion trap.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

0

5

10

15

Oxygen

O VII1-2

O VIII1-2

O VIII1-3

O VII

1-3

1-71-6

1-51-4

016 17 18 19 20 21 22 23

Xenon

Sig

nal (

arb.

uni

ts)

2

4

8

10

12

Wavelength (A)o

20.5 21 21.5 22S

igna

l (ar

b. u

nit)

250

200

150

100

50

0

O VIIE2

M3

E2(M1)

Wavelength (A)o

Xe XXVII

a

b

c

EUV spectra of oxygen and xenon recorded with a flat-field spectrograph at the LLNL EBIT

Many EUV spectra can be calibrated with the well known spectral line series of H- and He-like ions.

The decays of the lowest excited levels in �Ni-like Xe ions have been identified and �the wavelengths measured.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

XRS Microcalorimeter built at Goddard Space Flight Center for Astro-E / Astro-E2 spacecrafts

Covers X-ray energy range 300 eV to 20 keV with 6 eV line width at low E32 pixels of 0.6 mm x 0.6 mm eachWorking temperature about 60 mK

X-ray crystal spectrometers offer high spectral resolution, but suffer from low efficiency

Microcalorimeters feature a poorer resolution than crystal spectrometers, but are much superior to solid state diodes in low-energy access and in resolution.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

Xe E = 1450 eV

Pho

ton

ener

gy (

eV)

Time (s)

Electron beam on Electron beam off

XRS microcalorimeter data recorded at the LLNL EBIT with a 140 ms trap cycle. E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

35

30

25

20

15

10

5

0

C

ount

s (1

000)

a

3d - 4s3d - 4p

3d - 4f

Xe XXVII

Xe XXVIII

300

250

200

150

100

50

01400120010008006004002000

Cou

nts

Energy (eV)

b

O VII O VIII

Xe XXVII

XRS microcalorimeter spectra of Xe (E = 1450 eV)

a) electron beam on

b) electron beam off

Each signal pulse is time-stamped; the data can be sorted by X-ray energy or time within the trap cycle.

Time resolved spectra reflect level population dynamics.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

3d

3d9

10

4s

0

12

23

M1

M1

E2

E2

M3

311S DD

2.06E82.49E8

13689.2

2.26E82.72E8

53.771.3

1.86E41.88E4

Ni-like Xe XXVII; the transition probabilities shown do not take hyperfine mixing effects into account.

1

10

100

1000

10 000

0 20 40 60 80 100 120

b

Cou

nts

(log)

Time (ms)

1

10

100

1000

10 000

a

XRS microcalorimeter data of Xe

a) at the position of the M3 decay in Xe XXVII

b) at the position of O VIII Lyman alpha

When the electron beam is switched off, delayed photons arise from long-lived excited levels or from charge exchange (CX).

In (near) Ni-like Xe, there is only one level with a long (millisecond range) lifetime.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

1

10

100

0 10 20 30 40 50 60 70 80

Xe

Counts (log)

Time (ms)

Decay curve extracted from the XRS microcalorimeter data at the position of the Xe XXVII M3 decay

Apparent lifetime 11.0±0.5 ms

Microcalorimeter data of about one week total run time

This is the first atomic lifetime measurement using a microcalorimeter at an electron beam ion trap.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

0

5

10

15

20

25

30

35

40

52 53 54 55 56 57 58

Life

time

(ms)

Nuclear charge Z

Safronova 2005

Xe

BaCs

Results of LLNL EBIT lifetime measurements on M3 decays in Ni-like ions in comparison to theory (all neglecting any mixing due to hyperfine structure)

Very few calculations cover the magnetic octupole (M3) decays.

A shorter lifetime than predicted makes the ion less sensitive to density effects.

X

XX

Other calculations

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

3d

3d9

10

4s

0

12

23

M1

M1

E2

E2

M3

311S DD

2.06E82.49E82.41E8

13689.294.1

2.26E82.72E82.65E8

53.771.366.15

1.86E41.88E4

Xe XXVII

100

1000

0 50 100 150 200

Xe XXVIIC

ount

s (lo

g)

Time (ms)

natural Xe

isotopically pure Xe132

Soft-X-ray signal of Xe at SuperEBITThe even isotope Xe132 has no hyperfine structure. It features a single-component M3 radiative decay (and a tail from charge exchange (CX) processes). Natural Xe has about equal parts of odd and even isotopes and a more complex decay curve.

E TIAL B

20 Years of

LIVERMORE

since 1986

Spectroscopy

a) The spectroscopy of Fe (and whatever) will be superb from the corona, but the problem of elemental identification has to be taken care of on the ground.

There is a historic backlog of spectroscopic misidentifications, because not enough good laboratory data were available 40 years ago.

b) Most terrestrial light sources work well with short-lived levels, but those long-lived ones in the corona have to be confirmed. This takes a low density apparatus like EBIT.

c) The identification of long-lived levels in the lab takes some time resolution. Time-resolved spectra can harbour surprises. Consider the example of the metastable level in Ni-like Xe, that only recently has been explicitly left out of an extensive calculation. Microcalorimeter data on Xe demonstrate how the spectrum dynamics depend on a long lived level. In the soft part of the EUV, a different device will be needed, for example a microchannel plate-based detector (MCP).

Now that Hinode / EIS is orbiting, is there anything left to be done in the laboratory?

Working rangesBeam-foil spectroscopy : picosecond to hundred nanosecondsElectron beam ion trap : femtosecond and microsecond to hundred milliseconds Heavy-ion storage ring : millisecond to dozens of seconds

Atomic lifetimes (E1-forbidden decays) are of interest in - astrophysics (solar corona, planetary nebulae, AGN, etc.) - plasma physics (tokamak, spheromak, divertor)

EBIT experiments offer high spectroscopic accuracyBeam-foil data guarantee isotopic purity, reasonable charge state discrimination200 - 350 Å range: Beam-foil data on Fe are available, EBIT data may become available in a few years Heavy-ion storage ring lifetime experiments are in progress

Laboratory work on the EUV spectrum of Fe: Needs elemental purity, add time resolution