Complete analysis of the Na+3 fragmentation in collision with He atoms

Article abstract of DOI:10.1016/S0009-2614(99)00448-0

An experimental investigation of the fragmentation mechanisms of Na⁺₃ cluster ions in collision with He atoms at 263 eV centre-of-mass energy is presented. The relative populations of the three fragmentation pathways are determined. In particular, the kinematics of the three-body breakup is studied in detail. The analysis of the correlation between the velocity vectors of the fragments allows one to estimate the relative role of the electronic excitation or momentum transfer in the population of each pathway. The discussion of the fragmentation dynamics is based on a concomitant theoretical study.

Full text of DOI:10.1016/S0009-2614(99)00448-0

18 June 1999
Ž
.
Chemical Physics Letters 306 1999 233–238
Complete analysis of the Naq3 fragmentation in collision with He
atoms
1
)
M. Barat, J.C. Brenot, H. Dunet, J.A. Fayeton, Y.J. Picard, D. Babikov , M. Sizun
Laboratoire des Collisions Atomiques et Moleculaires Unite Mixte de Recherche C8625, Batiment 351, UniÕersite Paris-Sud XI, F-91405
´
´
ˆ
´
Orsay Cedex, France
Received 10 March 1999; in final form 22 April 1999
Abstract
An experimental investigation of the fragmentation mechanisms of Naq3 cluster ions in collision with He atoms at 263 eV
centre-of-mass energy is presented. The relative populations of the three fragmentation pathways are determined. In
particular, the kinematics of the three-body breakup is studied in detail. The analysis of the correlation between the velocity
vectors of the fragments allows one to estimate the relative role of the electronic excitation or momentum transfer in the
population of each pathway. The discussion of the fragmentation dynamics is based on a concomitant theoretical study.
q 1999 Elsevier Science B.V. All rights reserved.
w x
cess 2,3 . This study has then been extended to
1. Introduction
small Naqn ns3–9 clusters in an attempt to disen-
Ž
.
Ž
.
Collision-induced dissociation CID of metallic
clusters provides a means for investigating the dy-
namics of nuclei and unlocalised electrons after the
strong perturbation of a heavy particle collision. At a
few 100 eV centre-of-mass energy, the two fragmen-
tation mechanisms, namely electronic excitation and
momentum transfer in binary collisions between pro-
jectiles and target atomic cores begin to compete as
w x
tangle the various fragmentation mechanisms 4 .
However, only the dissociation in two fragments
could be analysed. In this Letter, we present a com-
plete investigation of the fragmentation pathways
including the kinematic analysis of the 3-body disso-
ciation of Naq3 . The technique of coincidence be-
tween 3 fragments has also been recently applied to
the study of the coulombic breakup of Hq3 5 and to
w x
w x
w x
predicted by the stopping power models 1 . Such a
prediction has been recently confirmed 2 by the
the photo-fragmentation of Arq4 6 . The present
w x
experimental investigation is concomitant with a the-
oretical investigation of the same subject given in the
analysis of the CID of Naq2 in an experiment based
on a multi-coincident detection of the fragments.
Such a simple system with a single active electron
has been considered as a prototype for testing vari-
ous theoretical approaches of the fragmentation pro-
w x
companion Letter 7 that allows for deeper analysis
of the fragmentation dynamics.
2. Experiment
)
Corresponding author. E-mail: sizun@lcam.u-psud.fr
1 Also Moscow Institute of Physics and Technology, Institutsky
per. 9, Dolgoproudny, Moscow region, 141700 Russia.
The experimental approach is based on the deter-
mination of the velocity vectors of the various frag-
0009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 99 00448-0

(
)
234
M. Barat et al.rChemical Physics Letters 306 1999 233–238
w
x
ments detected in coincidence. Briefly 2,4 , a mass-
neutral PSD has been replaced by a faster one based
selected chopped beam of Naq3 ions crosses at 908
on a delay line anode 8 , resulting in a 20 ns
Ž
.
w x
a cold He target beam produced by a supersonic
expansion. The ionic and neutral fragments are sepa-
rated by an electric field and detected on two posi-
response time. This was achieved at the expense of a
slight degradation of the position resolution, 0.3 mm
compared to 0.1 mm with the resistive anode. The
velocity vectors of all fragments of a given event are
obtained from the position and time information. The
most significant observables as, e.g., x the centre-
Ž
.
tion sensitive detectors PSDs . In all earlier experi-
ments, the neutral fragments arising from 3-body
fragmentation are received almost simultaneously on
the detector and cannot be identified by the PSD
Ž
.
of-mass CM scattering angle, F the angle between
q
Ž
Ž
based on the resistive anode technique 3 ms re-
the collision plane defined by the incident Na3
projectile and the recoiling He target and the disso-
.
sponse time during which the PSD is blind to a new
.
particle . Therefore for the present experiment, the
ciation plane, and e the sum of the kinetic energy of
Fig. 1. The Naq3 ™ Naq2 qNa pathway.
w x
a
I ´, x intensity contour map. The full and dashed lines are the predictions of the simple
Ž .
Ž
.
.
q
Ž
Ž
.
Ž .
Ž
.
impulsive model, see e.g. Ref. 2 , for cold full line and vibrationally excited dashed line Na3 ions. b I F, x intensity contour map.
Ž .
c
Ž . Ž . w x
‘ZZ correlation’ between the impact positions on the detectors of the neutral Zn and ionic fragments Zi , see e.g. Refs. 2,4 . The
arrows point out the backlash line, a signature of the electronic mechanism: Zi and Zn are in opposite directions, indicating a vanishing x
value. The main bulky structure signs the IM2 mechanism.

(
)
M. Barat et al.rChemical Physics Letters 306 1999 233–238
235
all fragments in the cluster centre of mass are deter-
spread F distribution corresponding to the grey
background in Fig. 1b for x-108.
Ž
.
mined Fig. 1 .
Ž .
Ž
.
The relative probabilities for the population of the
three pathways deduced from the counting rates of
the various fragments weighted by the detection
ii An electronic mechanism EM is responsible
for the backslash line in the ‘ZZ’ correlation of Fig.
1c. This mechanism also shows up as wiggles around
xf38, 0.5 eV-Erel -1 eV in Fig. 1a.
w x
efficiencies as discussed in Ref. 4 are given below
for a 263 eV centre-of-mass collision energy:
By an adequate filtering of the experimental data,
one finds that the relative contributions of the IM1,
IM2 and EM mechanisms to the Naq2 qNa pathways
can be estimated to 55%, 35 %, and 10 %, respec-
tively with an accuracy of ;5%. These values are
remarkably close to the theoretical predictions: 60–
68%, 23–31%, and 9%, respectively. The contour
maps of the two impulsive mechanisms are also well
reproduced by the theory, the A1 contribution in Ref.
Pathway A
Pathway B
Pathway C
Naq3 ™ Naq2 qNa
90% ,
4% ,
6% .
™ NaqqNa2
™ NaqqNaqNa
The relative Naq2 and Naq intensity is only de-
pendent on the relative ion detection efficiency and
is estimated with a 2% accuracy while the ratio
between B and C pathways that is dependent of the
absolute neutral detection efficiency is determined
with an accuracy not better than 20%.
It will be assumed in this Letter that all fragments
are in the electronic ground states, that is only the
less endothermal channel of each pathway is popu-
lated. The initial vibrational energy of the Naq3 ions,
unknown experimentally, was arbitrary set to 1 eV in
the theory as suggested in some previous work on
w x
Ž
.
7 . Notice in particular Fig. 1a the stretching of the
contours along the x axis for e-0.5 eV due to IM2.
The EM contribution, A2 in the theoretical analysis
w x
7 , is clearly peaked at xf08. The disagreement
with experiment cannot be considered as significant
due to inaccuracies arising in the data processing
around xf08.
w x
electron impact ionisation of sodium clusters 9 .
4. The NaH3 ™NaHHNaHNa and the NaH3 ™
NaHHNa2 pathways
3. The NaH3 ™NaH2 HNa pathway
Fig. 2a shows the correlation between x and e
the sum of the kinetic energies of the three fragments
in the cluster CM frame for the Naq3 ™ NaqqNaq
qNa pathway:
Fig. 1a shows the intensity contour diagrams as
functions of x and e. For large scattering angles, the
contour lines stretch along the curves given by a
model calculation assuming a binary He–Na elastic
Ž .
i
A first structure peaked at small x and e
Ž
.
collision considering a cold full curve or a hot,
values that accounts for ;40% of the population of
that channel is attributed to electronic transitions.
This mechanism should primarily correspond to the
C3 contribution of the theoretical analysis, a 3-body
breakup following an electronic transition towards
q
Ž
.
almost dissociated dashed curve Na3 ion, respec-
tively. Fig. 1b shows that for large x values, the
dissociation occurs around Ff08 as expected for a
Ž
.
direct impulsive mechanism IM1 in which a vio-
Ž
.
lent He–Na impact results in a direct ejection of a
the 2nd excited potential energy surface PES . How-
ever, the theory predicts too large an energy for e of
;1 eV. This is due to inaccuracy in the calculation
of the PES. Indeed a comparison with a full CI
calculation shows that the 3rd DIM PES is too
repulsive.
w x
fast Na fragment 2 .
A second structure shows up at x-108 and small
e values, the analysis of which has been given in
Ref. 4 . Actually, this structure corresponds to the
w x
sum of two mechanisms.
Ž .
Ž
.
Ž .
i A complex impulsive mechanism IM2 in-
volves an initial He–Na elastic collision followed by
momentum redistribution through Na–Na collisions.
In such a mechanism, the dissociation may occur out
ii The contour lines extend towards large e
values and roughly follow the curves of the binary
model indicating mechanisms that involve large mo-
mentum transfers. However, in contrast with the
previous case, the contours spread out over wider x
Ž
.
of the first He–Na collision plane giving a wide-

(
)
236
M. Barat et al.rChemical Physics Letters 306 1999 233–238
q
Fig. 2. The Naq3 ™ Na qNaqNa pathway. a I ´, x intensity contour map. The full and dashed lines are the predictions of the simple
Ž .
Ž
.
q
Ž .
Ž
.
impulsive model as in Fig. 1. b I u, x intensity contour map. u gives the orientation of the Na3 plane with respect to the scattering axis.
Ž
.
and e ranges. A comparison with the theoretical
predictions shows that this wide structure is ac-
counted for by the sum of the C1 and C2 contribu-
contribution the tail at large x of the impulsive
mechanisms at large cos u.
More insight into the electronic 3-body breakup
can be found in the following sets of correlation
obtained after filtering the data for x-58. Fig. 3a
shows the corresponding data, displayed as a func-
Ž
w x.
tions Fig. 3C of Ref. 7 . Notice that the C1
contribution, which is very sensitive to the cluster
internal energy, is overestimated by the theory.
Actually more can be learnt about the 3-body
fragmentation dynamics by examining the shape of
ˆ
tion of the three angles n , n and i defining the
ˆ1 ˆ2
Naq3 triangle. They clearly exhibit a fragmentation
along isosceles geometries with a maximum corre-
sponding to the equilateral triangle, a preferential
geometry for fragmentation after electronic excita-
tion on the 3rd PES as predicted by the theory. Fig.
Ž
.
the dissociating trimer. The I u, x correlation be-
tween u, the angle that defines the orientation of the
trimer plane with respect to the scattered axis, and x
Ž
.
the scattering angle Fig. 2b shows a distribution
isotropic with respect to the trimer plane. Notice that
data for cos u)0.8 and small x are missing. They
correspond to a Naq3 ion plane parallel to the detec-
tor plane. In such cases, the two neutral fragments
3b displays the correlation between e and the n
ˆ1
angle obtained after selecting isosceles configura-
ˆ
tions with ifn illustrating a dissociation that
ˆ2
stretches symmetrically around the equilateral con-
Ž
Ž .
ously obtained for the I e, n correlation. On the
ˆ
Ž . Ž .
other hand, the I e, i correlation Fig. 3c presents
asymmetric contours with a tail extending for small i
angles towards larger e energies. We suggest that the
hit the detector simultaneously inside the 20 ns dead
figuration. The same result not shown here is obvi-
.
Ž
.
ˆ2
time and are not recorded. This is not the case at
large x, at which angle the trimer is skew with
respect to the PSD. One can tentatively invoke topo-
logical considerations to explain the increase of the
ˆ

(
)
M. Barat et al.rChemical Physics Letters 306 1999 233–238
237
Fig. 3. Shape of the Naq3 triangle after dissociation for the forward 3-body breakup x-58 . n , n and i are the angles defining the
ˆ
Ž
.
ˆ1 ˆ2
ˆ
ˆ
.
Ž .
Ž
.
Ž .
Ž .
Ž
.
Ž
breakup geometry. a I i, n angular correlation contour map. b and c I ´, n and I ´, i contour maps, respectively.
ˆ1
ˆ1
Naq3 ™ NaqqNa2 pathway, a channel that cannot
ˆ
.
Ž
I e, i correlation corresponds to the sum of two
contributions: a first one symmetric around 608 simi-
be directly analysed with our technique.
Ž
.
lar to the I e, n correlation superimposed on a
ˆ1
ˆ
second one extending at small i angles and large e.
5. Conclusions
This latter configuration Naq3 ™ Na q NaqNa
q
Ž
.
corresponds to a quasi 2-body dissociation with very
small relative velocities between the two Na atoms.
This suggests that the mechanism responsible for this
fragmentation dynamics corresponds to the C2 con-
The present combined experimental and theoreti-
cal study was a first attempt to disentangle the
collision-induced fragmentation, including the 3-body
breakup, of a triatomic system. The relative roles of
electronic and impulsive mechanisms have been es-
tablished. On the experimental side, efforts will now
w x
tribution in Ref. 7 at small x. This basic EM might
also account, through the B2 contribution, for the

(
)
238
M. Barat et al.rChemical Physics Letters 306 1999 233–238
w x
2
J.A. Fayeton, M. Barat, J.C. Brenot, H. Dunet, Y.J. Picard, R.
be devoted to control better the internal energy of the
cluster ions and to extend the range of collision
Ž
.
Schmidt, U. Saalmann, Phys. Rev. A 57 1998 1058.
w x
3
D. Babikov, F. Aguillon, M. Sizun, V. Sidis, Phys. Rev. A 59
w x
energy. Experimental results 4 are available for the
Ž
.
1999 330.
two-fragment dissociation of larger Naqn clusters
w x
4
M. Barat, J.C. Brenot, H. Dunet, J.A. Fayeton, Y.J. Picard, J.
Ž
.
Ž
.
ns5–9 that will be completed by multi-fragmen-
Chem. Phys. 110 1999 10758.
w x
5
L.M. Wiese, O. Yenen, B. Thaden, D.H. Jaecks, Phys. Rev.
tation studies in the near future.
Ž
.
Lett. 79 1997 4982.
w x
6
P. Jukes, A. Buxey, A.B. Jones, A.J. Stace, J. Chem. Phys.
Ž
.
109 1998 5803.
Acknowledgements
w x
7
D. Babikov, M. Sizun, F. Aguillon, V. Sidis, Chem Phys. Lett.
Ž
.
Ž
.
306 1999 226, this issue .
w x
8
O. Jagutzki, V. Mergel, K. Ullmann-Pfleger, L. Spielberger,
The authors wish to thank F. Aguillon and V.
Sidis for stimulating discussions and continuous sup-
port during this work.
U. Meyer, R. Dorner, H. Schmidt-Bocking, SPIE Proc. 3438
¨
¨
Ž
. Ž
Ž
.
1998 also in: M.R. Descour, S.S. Shen Eds. , Imaging
Ž
.
.
Spectrometry IV, Proc. SPIE 3438 1998 , p. 322 .
w x
9
L. Bewig, U. Buck, Ch. Mehlmann, M. Winter, J. Chem.
Ž
.
Phys. 100 1994 2765.
References
w x
1
J.F. Ziegler, J.P. Biersack, The Stopping and Range of Ions in
Solids, Pergamon, New York, 1985.