Surface induced dissociation of HD2+ - A non-statistical behavior in the H+ and D+ fragmentation channels

Article abstract of DOI:10.1016/S0009-2614(03)00394-4

Surface induced dissociation (SID) of HD₂⁺ upon impact on a Tore Supra tokamak graphite tile was studied in the collision energy range from 20 to 100 eV. SID of the projectile ion HD₂⁺ on this graphite surface yields exclusively atomic fragment ions H⁺ and D⁺, whereas other possible fragment ions HD⁺ and D₂⁺ were completely absent. In contrast to the statistical ratio of 2:1 of D:H in the HD₂⁺ ion, the relative abundances of the measured H⁺ and D⁺ fragment ions were about of equal magnitude over the whole collision energy range studied. This non-statistical decay can be accounted for by vibrational predissociation through a minimal centrifugal barrier at a total angular momentum of about J～27-30.

Full text of DOI:10.1016/S0009-2614(03)00394-4

Chemical Physics Letters 372 (2003) 166–172
www.elsevier.com/locate/cplett
Surface induced dissociation of HDþ2 – a
non-statistical behavior in the Hþ and Dþ
fragmentation channels
1
A. Qayyum , C. Mair, W. Schustereder, P. Scheier, W. Hess , T.D. Mark
2
,3
*
€
Institut f€ur Ionenphysik, Leopold Franzens Universit€at, Technikerstr. 25, A-6020 Innsbruck, Austria
Received 9 January 2003; in final form 26 February 2003
Abstract
Surface induced dissociation (SID) of HDþ2 upon impact on a Tore Supra tokamak graphite tile was studied in the
collision energy range from 20 to 100 eV. SID of the projectile ion HDþ2 on this graphite surface yields exclusively
atomic fragment ions Hþ and Dþ, whereas other possible fragment ions HDþ and Dþ2 were completely absent. In
contrast to the statistical ratio of 2:1 of D:H in the HDþ2 ion, the relative abundances of the measured Hþ and Dþ
fragment ions were about of equal magnitude over the whole collision energy range studied. This non-statistical decay
can be accounted for by vibrational predissociation through a minimal centrifugal barrier at a total angular momentum
of about J ꢀ 27–30.
Ó 2003 Elsevier Science B.V. All rights reserved.
1. Introduction
explore and develop new methods for character-
izing the properties of gaseous ions, elucidating the
nature of the surface under study and to investi-
gate the basics of the ion/surface interaction. Be-
sides physical and chemical sputtering, interest has
been concentrated in particular on low energy (in
the range of tens of eV) collisions involving re-
flection, surface-induced dissociation (SID), charge
exchange reactions (CER) and surface-induced
reactions (SIR). In addition, of being of fundamen-
tal importance, polyatomic ion–surface reactions
are also relevant to a wide range of technologi-
cal applications including such different fields as
secondary mass spectrometry, surface modifica-
tions for the preparation of new materials (plasma
processing) and quite importantly plasma–wall
Collisions of slow ions with surfaces have been
increasingly used to characterize both gaseous ions
and the nature of surfaces [1–3]. Ion surface re-
active collisions is a research area which has been
rapidly growing the past 20 years in an effort to
* Corresponding author.
E-mail address: Tilmann.Maerk@uibk.ac.at (T.D. Mark).
€
Permanent address: Pakistan Institute of Nuclear Science
1
and Technology, P.O. Nilore, Islamabad, Pakistan.
2
Association Euratom-C.E.A., Cadarache, F-13108 St. Paul-
lez-Durance.
3
Also adjunct professor at Department of Plasma Physics,
Comenius University, SK-84248 Bratislava, Slovak Republic.
0009-2614/03/$ - see front matter Ó 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0009-2614(03)00394-4

A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
167
interactions in electrical discharges and fusion
plasmas [4,5]. Much of the modeling and diag-
nostics of fusion plasmas relies on the under-
standing of the atomic and molecular processes
which occur in the plasma and at the plasma wall.
In particular the role of molecules and molecular
ions is of utmost importance in the cold plasma
edge and during plasma–wall interactions.
appears that for HDþ2 only a few studies are
available in literature so far.
For instance, Alvarez et al. [19] measured total
cross sections for the production of positive and
negative ions by collision of HDþ2 with a variety of
target gases in the energy range from 40 to 600
keV. In addition, total cross sections for dissocia-
tive recombination of HDþ2 were measured by
Mitchell et al. [20] reporting that these cross sec-
tions were quite different from those for Hþ3 and
Dþ3 . Moreover, in a collision-induced dissociation
study of HDþ2 in the gas phase, Alvarez et al.
[21,22] and Carrington et al. [23] observed the
occurrence of all possible fragment ions, i.e.,
Hþ; Dþ; HDþ and Dþ2 . In contrast, in a photo-
dissociation study of HDþ2 also carried out by
these authors [23,24], only two fragment ions, i.e.,
Dþ and Hþ were observed. They also noted that
the intensity of the fragment ion Dþ was smaller
(by a factor of 3) than that of Hþ and that the
spectra leading to Dþ and Hþ fragments via the
decay from long lived resonances were quite dif-
ferent. This immediately implies that a given res-
onance state populated in this photodissociation
study will decay either into the Dþ or the Hþ
channel.
Using a combination of an adiabatic (with re-
spect to vibration) and sudden (with respect to the
bending motion) transition state theory with
semiclassical corrections Schlier and co-workers
[25] were able to calculate corresponding branch-
ing ratios assuming that the trapping leading to
long lived resonances is due to total angular mo-
mentum barriers. They showed that because of the
narrow experimental window accessible in [23] low
total angular momentum values (J 6 27) will lead
almost exclusively to formation of Hþ while high J
values (J P 30) will lead almost exclusively to Dþ
formation. The range of J values leading to both
fragments is rather small and apparently not
sampled in the photodissociation experiment of
Carrington and Kennedy [23]. These conclusions,
i.e., that vibrational predissociation of Hþ3 and its
isotopomers occurs by tunneling through the
lowest accessible centrifugal barrier at the avail-
able total angular momentum, have been con-
firmed by Chambers and Child [26] taking into
account also parity considerations. They also
Energetic plasma particles strike solid surfaces
such as tiles, limiters and divertors of a tokamak
vessel, thereby eroding the surface material by
sputtering processes. Sputtered particles are ion-
ized in the plasma by electron impact and may hit
the surface again. In fusion devices carbon is
widely used as plasma facing material due to its
high thermal resistivity and low Z value, although
a disadvantage is the relatively large chemical
sputtering yield by hydrogenic species. The use of
low-Z materials such as carbon, boron and be-
ryllium has been beneficial in present large to-
kamaks [6]. Whereas a large body of experimental
data exists concerning physical sputtering, less is
known about chemical sputtering and reactive
collisions of ions at lower energies; the latter have
so far been limited to laboratory studies mainly
involving Hþ and Oþ projectiles and only recently
extended to systems such as CHþ3 and CH4þ but
with conflicting results [7]. It is interesting to note
in this context that in a recent workshop dedicated
to the atomic and molecular data situation for
tokamak fusion science [8] it transpired in a very
clear fashion that many more data (i.e., more ac-
curate and in a wider range than previously con-
sidered) are urgently needed in particular at the
eve of starting the most challenging period in fu-
sion research with the imminent construction of
ITER [9].
Because of the urgent data need in fusion
plasma technology (see also [10,11]) we started
recently first systematic studies on surface-induced
dissociation in the low collision energy range
(<100 eV) of hydrocarbon and hydrogen molecu-
lar ions including in the later case Hþ2 ; H3þ; D2þ;
Dþ3 and HDþ projectile ions [12,13]. Whereas for
most of these ions a number of gas phase studies
has been carried and thus their properties are
reasonably well understood (e.g., see the references
given for the triatomic species in [14–18]), it

168
A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
showed that the observed strong preference of the
Hþ þ D2 predissociation channel (over Dþ þ H2Þ
in the experiment of [23,24] is consistent with an
experimental preference for J values in the range
0 < J < 25.
After passing the mass spectrometer exit slit, the
ions were refocused by an Einzel lens and decel-
erated to the required collision energy before
interacting with the target surface. Shielding the
target with conical shield plates minimized field
penetration effects. The incident impact angle of
the projectile ions was kept at 45° and the scat-
tering angle (defined as a deflection from the in-
cident beam direction) was fixed at 91°.
The collision energy of ions impacting on the
surface is defined by the potential difference be-
tween the ion source and the surface and can be
varied from about zero to about 2 keV with a
typical projectile ion beam energy spread of about
1 eV in the case of the Colutron ion source [12,13].
A fraction of the product ions formed at the sur-
face exited the shielded chamber through a 1 mm
diameter orifice. The ions were then subjected to a
pulsed extraction and acceleration field, which
initiated the time-of-flight analysis of the ions. The
second mass analyzer is a linear time-of-flight mass
selector with a flight tube about 80 cm long. The
mass selected ions were detected by a double-stage
multi-channel plate connected to a multi-channel
scaler (time resolution of 5 ns per channel) and a
laboratory computer.
In this Letter we present the first results on
surface-induced dissociations (SID) of HDþ2 via
impact on a graphite surface in the low collision
energy range from 20 to 100 eV. The main result of
this study is that surface-induced decomposition of
the excited parent ion leads to practically the same
amounts of Hþ and Dþ fragments over the whole
collision energy range. This is in contrast to the
statistical ratio expected of 1:2, as the parent mo-
lecular cation contains twice as much deuterium as
hydrogen atoms. The present results are discussed
in terms of the theoretical results obtained by
Schlier and co-workers [25] and Chambers and
Child [26]. Besides this non-statistical behavior in
the formation of Hþ and Dþ, another important
finding is the complete absence of the diatomic
HDþ and Dþ2 fragment ions which were observed
in experiments on collision induced dissociation
(CID) of HDþ2 in the gaseous phase [21–23].
2. Experimental
The surface used here was a graphite tile typi-
cally in use as a limiter surface in the fusion reactor
Tore Supra in Cadarache/France. It was main-
tained under ultra high vacuum conditions (about
10À10 Torr) in the bakeable turbo-pump evacuated
target collision chamber. However, even these
conditions did not exclude deposition of multi-
layers of hydrocarbon contaminants on the sur-
face, whenever the valve between the mass
spectrometer and the target collision chamber was
opened and the pressure in the target region in-
creased to the 10À8 Torr range. The surface was
thus a hydrocarbon-covered graphite surface as in
our earlier studies [12,13], thereby mimicking a
ÔrealÕ surface as occurring in many situations
dealing with plasma–wall interactions.
The present experiments were carried out with
the tandem mass spectrometer apparatus BESTOF
described in detail in our earlier papers [12,13,27].
It consists of a double focusing two-sector-field
mass spectrometer (reversed BE geometry), a sur-
face collision chamber connected to a linear time-
of-flight mass spectrometer. Projectile ions were
produced in a commercial Colutron dc plasma ion
source. For the gas discharge we used a 3:7 mix-
ture of H2 and D2 at a pressure of 0.2–0.5 Torr.
The main process is evidently ionization of hy-
drogen and deuterium molecules and subsequent
successive ion–molecule reactions leading to pure
and mixed diatomic and triatomic molecular
hydrogen isotopomer ions.
The produced ions were extracted by applying
an electrical field between the anode and the first
electrode of the focusing lens and accelerated to
3 keV for mass (and energy) analysis by the dou-
ble-focusing two-sector-field mass spectrometer.
3. Results and discussion
Fig. 1 shows secondary ion mass spectra after
surface interaction of HDþ2 primary ions at four

A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
169
Fig. 1. Product ion mass spectra generated by the surface impact of the hydrogen molecular ion HDþ2 at the collision energies of 40, 45,
50 and 80 eV interacting with graphite surface.
HDþ2 ! Dþ þ HD þ 4:45 eV
HDþ2 ! HDþ þ D þ 6:26 eV
HDþ2 ! Dþ2 þ H þ 6:23 eV
ð1bÞ
ð1cÞ
ð1dÞ
different collision energies. It can be clearly seen
from these spectra, that HDþ2 dissociates exclu-
sively into Hþ and Dþ, whereas HDþ and Dþ2 are
completely absent. This result is in contrast to gas
phase studies of collision-induced dissociation of
HDþ2 , where all possible fragment ions were de-
tected [21–23]. Nevertheless, photodissociation
studies on HDþ2 resulted in similar break up
channels as those reported here, with a complete
absence of diatomic fragment ions [23]. It is in-
teresting to note in this context that thermo-
chemistry of the different reaction channels [28]
(using for the enthalpy of formation of HDþ2 the
value of Hþ3 ) shows that the fragmentation chan-
nels leading to the diatomic product ions (reac-
tions 1c,1d having an endothermicity of about
6.3 eV) require about 1.9 eV more energy than the
channels leading to Hþ or Dþ formation (reactions
(1a) and (1b) having an endothermicity of about
4.4 eV)
It should be also mentioned in this context that
upon dissociation of HDþ2 into (i) HDþ + D or (ii)
H + Dþ2 , the product ions HDþ and D2þ, while still
in close contact with the surface, could capture an
electron from the surface and thus get neutralized.
Thereby ionization energy is released and the ex-
cited neutralized products dissociate further into
(i) H + D or (ii) D + D, respectively. This process
has already been observed in glancing incidence of
Hþ2 on surfaces by Heiland and coworkers [29].
Since this electron capture is not likely to happen
in CID experiments, the product ions HDþ and
Dþ2 produced would not be destroyed. Obviously,
a similar neutralization reaction could also take
place for atomic ions in the outgoing channel, but
it is difficult to estimate whether in the atomic case
HDþ2 ! Hþ þ D2 þ 4:40 eV
ð1aÞ

170
A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
neutralization is less likely than in the diatomic
case.
due to any discrimination occurring in the mass
spectrometric detection system for ions with dif-
ferent kinetic energies. Because of momentum
conservation the two dissociation products of de-
caying HDþ should have quite different kinetic
energies (coming from kinetic energy release in the
decay reaction) and thus if discrimination plays a
role the abundances of Hþ and Dþ for decaying
HDþ should be different. Moreover, it is also in-
teresting to point out that the onset for SID occurs
for HDþ at a collision energy of about 30 eV,
whereas the onset for SID for HDþ2 sets in at about
40 eV. This difference can be attributed to the
slightly higher dissociation energy for HDþ2 (4.4 eV
for reaction channel (1a) and (1b), see above) as
compared to the dissociation energy for HDþ with
an energy of about 2.74 eV. Using the well estab-
lished [30,31] average value of translational-to-in-
ternal energy transfer on hydrocarbon-covered
metal or carbon surfaces of about 6–8% we find
that the difference of about 10 eV in collision en-
ergy amounts to about 0.7 eV in internal energy,
not enough to gap the difference in dissociation
energies of about 1.7 eV. It is well conceivable that
the missing 1 eV could already be present in the
diatomic ion coming from the Colutron source as
these ions are very likely produced via direct
electron impact ionization reactions, whereas the
triatomic ions are rather produced via ion mole-
cule reactions and thus will be in a thermalized
state (for more information on this question see a
discussion of internal energies and their influence
on surface induced dissociations in [7]).
The data from the secondary ion mass spectra
as shown in Fig. 1 can be re-plotted into so called
collision-energy-resolved-mass-spectra (CERMS).
In a CERMS plot the relative abundances of the
different ions (parent as well as product ions) are
shown as a function of the collision energy (see
Fig. 2 lower panel). Interestingly, the abundances
of Hþ and Dþ are nearly the same over the whole
collision energy range. This 1:1 ratio of the atomic
hydrogen isotopes is in contrast to the expected
statistical ratio of 1:2 and also different from the
ratio 3:1 observed in the photodissociation exper-
iment by Carrington and Kennedy [23]. It is in-
teresting to note that due to results also shown in
Fig. 2 (upper panel) for similar studies concerning
the surface induced dissociation of HDþ it can be
excluded that this result (the 1:1 ratio for the
abundances of Hþ and Dþ for decaying HDþ2 ) is
As mentioned in the introduction using a
combination of an adiabatic (with respect to vi-
bration) and sudden (with respect to the bending
motion) transition state theory with semiclassical
corrections Schlier and co-workers [25] were able
to calculate corresponding branching ratios as-
suming that the trapping leading to long lived
resonances is due to total angular momentum
barriers. The unimolecular dissociation rate then
depends on the structure of the total angular mo-
mentum barrier in each of these dissociation
channels. Classically, for a given J the centrifugal
term will be always lower for the Dþ channel than
the Hþ channel (due to the larger reduced mass of
the former channel) and this leads to a lower
Fig. 2. Collision energy resolved mass spectrum (CERMS) for
surface induced dissociation of HDþ (upper panel) and HDþ2
(lower panel) at a graphite surface in the collision energy range
from 20 to 100 eV. Lines drawn through the data points are
only given to guide the eye.

A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
171
centrifugal barrier height in the Dþ-forming
channel. On the other hand, tunneling through the
Hþ barrier is easier, but this is not enough to
compensate for the lower barrier height in the Dþ
case. Nevertheless, as argued by [25] and later by
Chambers and Child [26] using also parity con-
siderations it is also necessary to take into account
the zero-point energy, because the zero-point en-
ergy of D2 is by 341 cmÀ1 lower than that of HD.
As a result of this only the Hþ channel will be open
as long as the total energy is lower than the as-
ymptotic HD zero-point energy. When the total
angular moment barriers are plotted relative to the
zero-point energy of D2 (with the relevant zero-
point energies added for each channel) one notes
(see in Fig. 3 taken from [25]) that for low J the Hþ
barrier is always lower than the Dþ barrier. With
increasing J the Hþ barrier increases faster and
compensates for the difference in the zero-point
energies. This means that at lower J values HD2þ
will dissociate into Hþ þ D2, while at higher J
values the Dþ + HD channel opens up. In a more
quantitative manner using sudden transition state
theory with tunneling corrections [32] Schlier and
co-workers showed that low total angular mo-
mentum values (J 6 27) lead almost exclusively to
formation of Hþ while high J values (J P 30) lead
almost exclusively to Dþ formation. The range of
J values leading to both fragments is rather small
and apparently not sampled in the photodissocia-
tion experiment of Carrington and Kennedy [23].
Moreover, both models [25,26] predict kinetic
energy release values consistent with the experi-
mental findings [23].
Therefore, these models can account for the
deviation of measured Hþ=Dþ ratios in the previ-
ous photodissociation results and also in the
present SID study as compared to the statistical
prediction of a 1:2 ratio for the Hþ=Dþ intensity.
Nevertheless, the measured Hþ=Dþ ratios are dif-
ferent in the two experiments. One reason might be
a different internal energy population in the two
experiments. Moreover, it is quite interesting to
note that the Hþ=Dþ ratio is constant over the
whole collision energy range in the present surface-
dissociation experiments. According to the earlier
work of Kleyn et al. [33] about the rotational en-
ergy transfer in direct inelastic surface scattering of
NO on Ag (111) in the energy range below 1.7 eV
the rotational temperature scales with the normal
component of the incident energy. Thus one might
expect to see such a dependence also in the present
surface collision experiment, i.e., at higher energies
the Dþ channel should dominate according to the
discussion given above. However, the constancy of
the ratio Dþ=Hþ in the present experiment sug-
gests that the total J value does not change with
increasing collision energy which is not too sur-
prising taking into account the quite different ex-
perimental situations. The present conclusion
about conservation of the J number during the
surface collision is actually in line with a recent
observation concerning hydrocarbon ion surface
reactions [7]. In this recent study we could dem-
onstrate that different internal energy in a hydro-
carbon projectile ion is not equilibrated during the
surface collision but leads to different unimolecu-
lar dissociation rates in surface induced collisions.
In the present case this would suggest that ions
coming from the Colutron ion source have a cer-
tain amount of total angular momentum (ap-
proximately J ꢀ 27–30 due to the fact that we
observe about equal abundances for Hþ and Dþ
product ions) which is not changed significantly by
varying the collision energy, all what is changed is
the overall efficiency (see the decrease of the parent
Fig. 3. Quantal total angular momentum barrier for the HD2þ
system as given by Schlier and co-workers [25]. The energies are
plotted relative to the D2 zero-point energy and barriers are
given for values of J ¼ 20, 25, 30, 35 and 40.

172
A. Qayyum et al. / Chemical Physics Letters 372 (2003) 166–172
[10] P.Mertens, P.T.Greenland, (Eds.), Special issue on Hydro-
gen in the plasma edge, Contribution to Plasma Physics,
vol. 42, 2002 p. 573.
ion matched by an increase of the product ions in
Fig. 2) of the fragmentation due to the different
vibrational excitation induced by the ion surface
interaction at different collision energies. Com-
paring the present experiment with the earlier
photo-ionization experiment one would have to
conclude that the total J value is slightly higher in
the present one.
[11] U. Fantz, P.T. Greenland, Contr. Plasma Phys. 42 (2002)
694.
[12] A. Qayyum, W. Schustereder, C. Mair, W. Hess, P.
€
Scheier, T.D. Mark, Phys. Scr., in press.
[13] S. Cernusca, H.P. Winter, F. Aumayr, A. Qayyum, W.
€
Schustereder, C. Mair, P. Scheier, T.D. Mark, Int. J. Mass
Spectr. 223/224 (2003) 21.
[14] T. Oka, Phys. Rev. Lett. 45 (1980) 531.
[15] T. Oka, Philos. Trans. R. Soc. London 358 (2000) 2363.
[16] J.T. Shy, J.W. Farley, W.E. Lamb Jr., H.W. Wing, Phys.
Rev. Lett. 45 (1980) 535.
Acknowledgements
[17] A. Carrington, J. Buttenshaw, R.A. Kennedey, Mol. Phys.
45 (1982) 45.
This work has been carried out within the As-
€
sociation EURATOM-OAW. The content of the
publication is the sole responsibility of the authors
and does not necessarily represent the views of the
EU Commission or its services. Work was also
[18] I.F. Schneider, A. Suzor-Weiner, Contrib. Plasma Phys. 42
(2002) 578.
[19] I. Alvarez, C. Cisneros, A.C.F. Barnett, J.A. Ray, Phys.
Rev. A 14 (1976) 602.
[20] J.B. Mitchell, C.T. Ng, L. Forand, R. Janssen, W.
McGowan, J. Phys. B 17 (1984) L909.
€
supported by the FWF and the OAW, Wien,
Austria and the European Commission, Brussels.
It is a pleasure to acknowledge discussions with
Prof. Zdenek Herman, Praha.
[21] I. Alvarez, H. Martinez, C. Cisneros, J. de Urquijo, Phys.
Rev. A 45 (1992) 285.
[22] I. Alvarez, C. Cisneros, J. de Urquijo, H. Martinez, Nucl.
Instr. Meth. B 53 (1991) 438.
[23] A. Carrington, R.A. Kennedy, J. Chem. Phys. 81 (1984)
91.
References
[24] A. Carrington, J. Chem. Soc. Faraday Trans. 2 (1986)
1089.
[1] R.G. Cooks, T. Ast, M.A. Mabud, Int. J. Mass Spectrom.
100 (1990) 209.
[25] M. Berblinder, J.M. Gomez-Llorente, E. Pollak, Ch.
Schlier, Chem. Phys. Lett. 146 (1988) 353.
[26] A.V. Chambers, M.S. Child, Mol. Phys. 65 (1988) 1337.
[2] A. Amirav, Comment. At. Mol. Phys. 24 (1990) 187.
[3] L. Hanley, (Ed.), Special issue on polyatomic ion–surface
interactions, Int. J. Mass Spectrom., 174 (1998).
[4] W.O. Hofer, J. Roth, Physical Processes of the interaction
of fusion plasmas with solids, Academic Press, San Diego,
1996.
€
[27] C. Mair, T. Fiegele, F. Biasioli, R. Worgotter, V. Grill,
M. Lezius, T.D. Mark, Plasma Source Sci. Technol. 8
€
€
(1999) 191.
[28] J.L. Franklin, National bureau of standards USA,
NSRDS-NBS 26, 1969.
€ €
[29] K. Schmidt, H. Franke, T. Schlatholter, C. Hofner, A.
[5] R.K. Janev, Contrib. Plasma Phys. 38 (1998) 307.
[6] P.C. Stangeby, G.M. McCracken, Nucl. Fusion 30 (1990)
1225.
€
Narmann, W. Heiland, Surf. Sci. 301 (1994) 326.
ꢁ
ꢀ
ꢀ
€
[30] J. Zabka, Z. Dolejsek, J. Roithova, V. Grill, T.D. Mark, Z.
Herman, Int. J. Mass Spectrom. 213 (2001) 145.
[31] F. Biasioli, T. Fiegele, C. Mair, Z. Herman, O. Echt, F.
[7] A. Qayyum, Z. Herman, T. Tepnual, C. Mair, S. Matt-
€
Leubner, P. Scheier, T.D. Mark, unpublished.
[8] IAEA Technical Meeting on Atomic and Plasma-Material
Interaction Data for Fusion Science and Technology,
€
Aumayr, H.P. Winter, T.D. Mark, J. Chem. Phys. 113
(2000) 5053.
€
[32] E. Pollack, J. Chem. Phys. 82 (1985) 106.
[33] A.W. Kleyn, A.C. Luntz, D.J. Auerbach, Phys. Rev. Lett.
47 (1981) 1169.
Julich, 28–31 October, 2002.
[9] ITER physics basis editors, et al., Nucl. Fusion, 39 (1999)
2137.