Aq+ + B
R. Hoekstra, F.J. de Heer and R.
Morgenstern
~
A(q-l)+'(nl) + B+
c=: A(q-l)+(n'
l') + hu
q
In vords: a multiply charged ion A + collides
on a neutral target Band captures one of the
The importanee of transition
probabilities in atomie eollision
experiments
target electrons into an excited state
A(q-l)+(nl). This state viII decay byemission
of a photon, hu, and hence measurement of the
photon emission gives direct information on
the electron capture rate into A(q-l)+(nl).
The set-up used for this kind of measurements is shovn schematically in fig. 1. Inside the collision chamber the multiply char-
ABSTRACT
ged ions, produced by the Electron Cyclotron
Resonance ion source installed at the KVI,
cross a target beam. The photon emission is
The role of transition probabilities in Photon
Emission Spectroscopy of charge changing
collisions is discussed on the basis of
collisions , like He 2+ - and 0 3 + - H. For these
systems an accurate knovledge of transition
probabilities
is
essential
to deduce
observed vi th tvo monochroma tors, one being
sensitive in the "visible" spectral range
(200 - 700 nm) and the other in the vuv range
(10 - 80 nm).
The monochromators
can
be
state
positioned in such a vay that there is no
influence of polarization effects (Hoeks tra et
selective electron capture cross sections . In
addition
to
these
physical
aspects,
the
practical aspect of calibrating the sensi tivi ty of a vuv monochromator by the branching
al 1989b). The absolute target beam densi ty
ratio method is discussed.
induced radiation (Ciri~ et al, 1985). This
radiation is measured vith the spectrometer
profiles
are determined
by electron
impact
for "visible" light, vhich is equipped vith an
imaging lens system vhich enables measurements
INTRODUCTION
along the beam axis.
Electron capture
in
collisions
of
In this type of measurements there are
three
main
applications
of
transition
multiply
charged ions and neutral particles, especially
atomic hydrogen
is
a
subject
of
theoretical and experimental studies
and
Vinter,
1985).
The
interest
probabili ties and the related quantities:
branching ratio and lifetime, namely
extensive
in
(Janev
I - Sensitivity calibration
monochromator
these
charge transfer processes sterns not only from
11 - Identification
fundamental aspects but also from their importance in fusion plasma research and astro-
of
of
the
contributions
vuv
from
degenerate states in hydrogenic ions.
electron capture
Photon Emission
111 - Determination of capture cross sections
Spectroscopy -PES- (Hoeks tra et al, 1989a and
Dijkkamp et al 1985). The process for single
These
three aspects viII
be discussed
separately in the next sections on the basis
electron capture is schematically given by
of some collision systems of current interest.
physics. Ve study these
processes
by means of
R. Hoekstra et al.
from complex emission spectra.
221
I
Cl.)
atomie
hydrogen
souree
I
I
movable
electron
gun
ion beam axis
- - --- - -
%'
~'
___
diaphragms
. J:
lens
,-::~. ---~--()\
A
--
//
entranee slit
vuv-monoehromator
/'o,:~'~-~
-----{1-
mirror
,/00
/~
/1'/ ~~
~ faraday eup
~/ ","iili\ target be am
--- !in:
11111
II I II
11 111
11111
11 111
entranee slit
visible light
monoehromator
rotatable
mlrror
Q
I
Pigure 1. Schematical view of the experimental
set-up.
I - SENSITIVITY
CALIBRATION
OP
THE
impact of electrons on noble gases (van Raan,
VUV
HONOCHROHATOR
1973). In this range the sensitivity increases
smoothly
The
vuv
monochromator
used
is
a
grazing
wi th
decreasing
wavelength
(see
fig. 2) and therefore an interpolation between
incidence vacuum spectrometer equipped with a
the measured sensi tivi ty points seems to be
position sensitive microchannelplate detector,
justified.
enabling simultaneous detection of emission
wavelengths is tricky since the sensi ti vi ty
lines within ranges of about 20 nm. To de termine the absolute sensitivity of this vuv
on the reflection coefficient of the grating.
monochromator,
absolute emission cross
However
extrapolating
to
lower
not only depends on the photon energy but also
sec-
To extend the sensitivity calibration down to
tions for electron and ion impact processes
lover wavelengths we make use of branching
are used. These cross sec t ions are gi ven in
table 1. The wavelengths cover a range from
ratios in Li-like ions (N 4 + , 0 5 + and p6 +). Por
these ions the 4t ~ 3t' transitions yield
radiation in the range of 35 - 65 nm where the
20 - 75 nm. If oscillator strengths are known,
i t is possible to get a few more calibration
points in the range of 50 - 80 nm by measuring
sensitivity is known, whereas the 4t ~ 2t'
transitions yield radiation in the range into
the
which
intensity of
resonance radiation af ter
Transition rates in collision experiments
222
we
want
to
extend
the
absolute
Table 1 Absolute emission cross sections (in
10-1Icm 2) used for the sensitivity ca1ibration
20
of the vuv monochromator (Oijkkamp et al, 1985
and references therein). Typical error: 20-25%
16
wavelength in nm
20.8
Ne 4+
system
energy (keV)
transition
ion
upper level
lower level
+
21. 3
+
He
30.4
e
He
+
0.2
60
NeIV
4
3s p
2p3 4SO
cross section
VI
iii 8
3s 20
2p3 20 0
VI
6
470
0 . 56
4
wavelength in nm
46.1
system
energy (keV)
GRINS calibration
18
e + Ne
71.8 - 73.1
e + Ar
0.20
0.20
2
wavelength
transition
ion
Nel!
2s2p6 2S
2s 22p 5 2po
upper level
lower level
cross section
4.80
À-
Pigure 2 . Sensitivity calibration curve of the
Arl!
3p44s 2p
3p5 2p o
vuv monochromator (GRINS). Branching ratio
4
method:O- p6+, . - 0 5+, Ä - N + and 0 - He.
• - direct processes (tabie 1)
1.71
sensitivity calibration, namely 9 - 20 nm.
As
an example of
the
branching
l----
ratio
2
method we will discuss the case of NV(1s 4p),
which is shown in fig. 3 . The branching ratios
lities
calculated
t
related to the known sensitivity at 62.9 nm,
c
K(62 . 9) by
=
~:~~ ~~~~:~~
3
f
-~
0·71
~
K(62.9)
' ·00
20_
.9_ _
2
with S(16.3) and S(62.9) the measured signals
at 16 . 3 and 62 . 9 nm. Unfortunately the signal
Pigure 3.
NV.
The
at 62.9 nm includes the third order of the
3p ~ 2s transition (3*20.93 = 62.8nm). Since
Schematic energy level diagram of
relevant
transitions have been
indicated together with their branching ratios
the excited N4 +' ions are produced in charge
R. Hoekstra et al.
d
by Lindgárd and Nielsen,
1977. The sensitivity at 16.3 nm, K(16.3) is
K(16.3)
s
4
have been deduced from the transition probabi-
p
and wavelengths (in nm) .
223
changing collisions of NS + the excited state
tions in H-like ions have to be compared with
population
the sum of the various theoretical contributions. Comparing in this way experiments for
C6 + and Oe+ - H collisions
with the most
can
be
steered
by
choosing
an
appropriate
target.
Charge
transfer
in
collisions on H populates 3l and 4l states,
2
whereas collisions on He populate only 3l
elaborate calculations it was found that there
states
is
(Dijkkamp
et
al,
1985) .
Hence
collisions on He can be used to deduce the
ratio between the first and third order of the
S
3p ~ 25 line. Charge exchange between N + and
4
2
H2 is used to populate N +"(15 4p). It should
of
calibration
procedures
visible
range
spectral
important
for
future
sections separately.
the
measurement of the 4
ratio
method
for
~
the spectrometer for visible light.
latter
15%)
is
by
lamps
and
known
to
the
ra tios
accurate
the
sensitivity
into
calibration
The
to n
He 2 +
of
4l
cross
sections
from
the
3 transition (Hoekstra
~
For
these
HeII(4l)
states
the
3 are given in table 2. Radiation from
=
trated on the target area, whereas radiation
the
longlived
states,
like
45
is
still
of
Table 2 Lifetimes, ~41' branching
decay to n = 3, 13(4l ~ 3).
vuv
range of
ratios
for
state
9 - 20 nm.
45
11 - IDENTIFICATION OF CONTRIBUTIONS FROM
DEGENERATE STATES IN HYDROGENIC IONS
~41
(nsec)
13(4l
In hydrogenic ions the l levels are quasidegenerate and therefore light emission from
~
3)
14.2
0.42
4d
4/
0.77
2.26
4.53
0.042
0.25
1.00
4p
emitted dovnstream the ion beam axis (fig. 1).
Therefore the measurement of the emission
different
l levels can not be resolved
spectroscopically. So measurements of transi-
Transition rates in collision experiments
case
electron
from
of
the
shortlived states, e.g. 4p is mainly concen-
knowledge
the wavelength
For
lifetimes and branching ratios for transitions
(ratio
of
transition
probabilities) it has been possible to extend
monochromator
separate
different.
impact emission cross sections.
Due
branching
diagnostics
(error
calibrated on sensitivity
standard
plasma
the fact that the lifetimes of the states are
35) and one in
the vuv (53.7 nm, 3p ~ 15), which links the
sensitivity of the vuv monochromator to that
of
particularly
et al, 1989c). To that end we have exploited
HeI(153p) which has a branch in the visible
spectral range (501.6 nm, 3p
are
colliding on atomic hydrogen we have deduced
result
branching
and
(Fonck et al, 1984) and hence it is essential
to know all the l state electron capture cross
for
table 1. Furthermore the figure inc1udes the
the
theory
(Boileau et al, 1989) • Due to the strong
fields in a tokamak the l states are mixed
Li-like ions are shown in fig. 2 together with
the direct results of the processes given in
of
between
experiment for the dominant capture channels,
of the 3p ~ 25 line to the signal at 62.9 nm
is about 75 and 50%, respectively. Therefore
the measurements have to be performed with
optimal counting statistics. The results of
type
agreement
whereas for the non-dominant high-n states
there were considerable differences (Hoekstra
et al, 1989a) . However these transitions
between high-n states, yielding light in the
be noted that at typical impact energies of 50
and 75 keV the contribution by the third order
this
good
profiles along the ion beam gives information
224
on the ! state population. This method, of ten
used in collision experiments on static
emission profiles P
targets and in beam foil experiments, has only
also is due to the fact that the transi tion
once before been used in a beam experiment,
namely by Aumayr et al 1984 for H+ - Li
collisions.
are accurately known. As a typical example of
least squares fit . This is feasible since the
41
the experimental results obtained in this way
fig. 4 shows
P4l (z) ~ _1_
v~
f
4l
T(z') e
z)/v~
4l
dz'
experimental
0
K
I) 13(4! ---7 3) P (z) a
Ir
4l
with K an absolute
13(4! ---7
4f
the
state,
results
increase
smoothly,
whereas the theory shows a structure around
4 keV/amu.
with v the velocity of the ions and ~41 the
lifetime of state 4!, z the position along the
ion be am axis and T(z') the target density
profile. The measured signaIs, S(z) are equal
to
S(z)
the results for
together wi th the theoretical predictions by
Fritsch, 1988. It can be seen that the
41
-(z' -
are weIl known, which
probabilities, lifetimes and branching ratios
Neglecting cascades the emission profiles
along the ion beam axis, P are described by
•
(Z)
Vith knowledge of lifetimes and branching
ratios of
the HeII(4!)
states
it has
been
possible to deduce for the first time high-n
(n ~
4 is a high-n level since the capture
goes predominantly to the n ~ 2 level) state
selective cross sections in He 2 +_H collisions.
41
calibration constant,
the branching ratio for decay to
n ~ 3 and a
the electron capture cross
4l
sections. By ~easuring the signals at some
3)
111 -
DETERHINATION OF CAPTURE CROSS SECTIONS
FROM COMPLEX EMISSION SPECTRA
25-30 positions along the ion beam axis, it is
possible to deduce the cross sections from a
Photon emission following electron capture in
3
+_ H (03+(1/2/2p) ---702+ (1s22s22pn!) plays
0
3
an
ES
important
role
in astrophysical
studies
(Heil, 1987). Therefore fully quantal calculations have been performed (Heil et al, 1983
and Bienstock et al, 1983), which predicted
that
the
3p SL
states
are
dominantly
populated.
However
translational
energy
spectroscopy
o
(Vilson
et
al,
1988)
and
PES
(Hoeks tra et al, 1989d) showed that the 3s SL
are dominantly populated.
A useful evaluation of the PES measurements is only possible wi th a fair knowledge
of transition probabilities, since a conside-
capture
into 4 f
v_
06
rable number of emission lines falls outside
the
wavelength
ranges
covered
by
the
spectrometers.
Since a complete set of
O·ra.u.
Figure 4. Electron
He + (4f)
capture
into
deduced from emission profile measurements.
Solid curve - Atomic Orbital calculation by
Fritsch, 1988
R. Hoekstra et al.
transition probabilities was not
available,
- especially two electron transi tions of the
2
type 2s 2p3p ---7 2s2/ were lacking - cal cu la-
225
tions have been performed (Hansen, 1989). The
(Hansen, 1989). More details will be given in
importance of these two electron transitions
can be shown from measurements on the 3p lp
a forthcoming article. Knowledge about these
and 3p 3 D states.
determining level populations and interpreting
the measured emission spectra.
two-electron
Fig. 5 shows the emission
profile for the 3p lp state. The solid curve
transitions
is
essential
for
is a least squares fit with a lifetime of
10 nsec, the dashed curve is a profile with a
lifetime of 27 nsec,
which is
CONCLUSION
the lifetime
deduced from the transition probability gi ven
calculated (Hansen, 1989) is 11.8 nsec, which
It has been shown that an accurate knowledge
of transition probabilities is important for
photon
emission
spectroscopy
of
charge
is in fair agreement with the measurement. His
transferring collisions . Vi th this knowledge
by IHese
transition
et al,
1966 for one-elec t ron
2
to
2s 2p3s lpO.
The
lifetime
it has been possible to extend the sensitivity
calibration of a vuv monochromator. Furthermore it has been possible to resolve
contributions
from
degenerate
states
in
hydrogenic ions and to determine capture cross
sections from complex emission spectra.
ACKNOVLEDGEMENTS
This work is part of the research program of
the Stichting '/ oor Fundamenteel Onderzoek der
Materie (FOM) with financial support from the
Nederlandse Organisatie voor Vetenschappelijk
Onderzoek (NVO) .
REFERENCES
Figure 5. Emission profile along the ion beam
axis of the 2s 2 2p3p 1 P ~ 2/2p3s 1 ~ transition. Solid curve least squares fit and dashed
curve profile calculated with the lifetime
deduced from Viese et al, 1966 see text.
Aumayr, F., M. Fehringer and H. Vinter, 1984 Inelastic H+-Li(2s) collisions (2-20keV):
11. Electron capture into H(2p) and H(3l).
In: J. Phys. B: At. Mol. Phys.,
calculations
show
that
2
for
2s 2p3p states some 25-45% of
decays
via
all
the 6
the population
two-electron
transitions
to
measured the ratio
between one- and two- electron transitions for
3p 3 D and found it to be 1.8 ± 0.4 which is in
agreement wi th the calculated value of 2.1
2s2p3 states.
Ve
with atomic hydrogen.
have
Transition rates in collision experiments
17, 4201-
4211.
Bienstock, S., T.G. Heil and A. Dalgarno, 1983
Charge transfer of 0 3 + ions in collisions
In:
Phys. Rev. A.,
27, 2741-2743.
Boileau, A., M. von Hellermann, L. D. Horton,
J . Spence ar.d H. P. Summers, 1989 - The de-
226
duction of low-Z ion temperature and
densi ties in the JET tokamak using charge
exchange recombination spectroscopy. In:
Plasma Phys. Controlled Fusion, 31, 779
Ciri~, D., D. uijkkamp, E. Vlieg and F.J. de
Heer, 1985 - Selective electron capture into HeII(n,!) subshelIs in collisions of
2
He + wi th atomic and molecular hydrogen.
In: J. Phys. B: At. Mol. Phys., 18, 47454762
In: J. Physique ColI., 50, 387-392.
Hoekstra, R., F.J. de Heer and R. Morgenstern,
1989c - ! state selective charge exchange
2
cross sections for collisions of He + on
atomic and molecular hydrogen. In: Invited
Papers XVI ICPEAC, New York, in press.
Hoekstra, R., K. Boorsma, F.J. de Heer and
R. Morgenstern, 1989d - Photons shedding
light upon charge exchange processes in
collisions of 24 keV 0 3 + with atomic hydrogen. In: J. Physique ColI., 50, 349-352.
Janev, R.K. and H. Vinter, 1985 - State selective electron capture in atom - highly
charged ion collisions. In: Phys. Rep.,
Dij kkamp , D., J. Ciri~, E. Vlieg, A. de Boer
and F.J. de Heer, 1985 - Subshell-selective
electron capture in collisions of C4 +, NS+,
o6+ with H, H2 and He. In: J. Phys. B: At.
Mol. Phys., 18, 4763-4793.
Fonck, R.J.,
D.S. Darrow and K.P. Jähnig,
117, 265-387.
Lindgärd, A. and S.E. Nielsen, 1977 - Transition probabilities for the alkali isoelectronic sequences Lil, NaI, KI, RbI, CsI,
Fr!. In: At. Nucl. Data Tables, 19, 533-
1984 - Determination of plasma-ion velocity
distribution via charge-exchange recombination spectroscopy. In: Phys. Rev. A., 29,
3288-3309.
Fritsch, V., 1988 - Calculation of partial
633.
Raan, A.F.J. van, 1973 - An absolute intensity
electron-transfer cross sections in 1 - 84
keV/amu He 2++ H collisions. In: Phys. Rev.
calibration method for vacuum ultra-violet
spectrometry based on electron impact
A., 38, 2664-2667.
Hansen, J., 1989 - private communication.
Heil, T.G., 1987 - Multiply charged ions in
astrophysics. In: Nucl. Instrum. Methods.,
B23, 222-225.
excitation. In: Physica, 65, 566-578
Viese, V.L., M.V. Smith and B.M. Glennon, 1966
Atomic transi tion probabili ties NSDRS-NBS4
vol 1
(Vastington
DC:
US
Govt
Printing
Office)
Vilson, S., R.I!. McCullough and H.B. Gilbody,
Heil, T.G., S.E. Butler and A. Dalgarno, 1983
Charge transfer of doubly charged and
trebly charged ions with atomic hydrogen at
thermal energy. In: Phys. Rev. A., 27,
2365-2383.
1988 - State selective electron capture by
slow 0 3 + and Ne 3 + ions in H. In: J. Phys.
B: At. Mol. Opt., 21, 1027-1037.
Hoekstra, R.,
D. Ciriè,
F.J. de Heer
and
R. Morgenstern, 1989a - State selective
electron capture in collisions of C6 + and
AUTHORS'
0 8 + on atomic and molecular hydrogen
studied by Photon Emission Spectroscopy.
In: Phys. Scr., T28, 81-90.
Hoekstra, R., M.G. Suraud, F.J. de Heer and
R. Morgenstern, 1989b - Alignment of p- and
d- states of three electron carbon ions
produced in collisions of C4 +(ls2) with H .
2
R. Hoekstra et al.
ADDRE~. SES
K.V.I., Zernikelaan 25, 9747 AA Groningen,
The Netherlands.
F.O.M. Institute for Atomic and Molecular
Physics, P.O. Box 41883, 1009 DB Amsterdam,
The Netherlands.
227