Nuclear Instruments and Methods in Physics Research B 230 (2005) 438–442
www.elsevier.com/locate/nimb
Plasmon excitation and electron promotion in the
interaction of slow Na+ ions with Al surfaces
M. Commisso a, A. Bonanno a, A. Oliva a, M. Camarca a,
F. Xu a, P. Riccardi a,*, R.A. Baragiola b
a
Laboratorio IIS, Università della Calabria and INFM – Unità di Cosenza, 87036 Arcavacata di Rende, Cosenza, Italy
b
Laboratory for Atomic and Surface Physics, University of Virginia, Engineering Physics, Thornton Hall,
Charlottesville, VA 22904, USA
Abstract
We studied the energy distributions of electrons emitted in the interaction of slow Na+ ions with polycrystalline Al
surfaces. To study sub-threshold plasmon excitation we performed measurements as a function of emission angle which
showed the excitation of bulk plasmons, confirming the kinetic nature of the excitation process.
Electron spectra show narrow transition lines due to Auger decay in vacuum of sodium projectiles excited in the
2p-shell by electron promotion in collisions with Al target atoms. Several previously unidentified transition lines could
be attributed to the autoionization of doubly excited projectiles.
2004 Elsevier B.V. All rights reserved.
Keywords: Ion-surface impact; Ion scattering from surfaces; Secondary electron emission; Plasmons; Auger electron emission
1. Introduction
We present first results of experimental studies
of electron emission in the interaction of slow
Na+ ions with Al surfaces. Aluminum has been
widely used as a prototypical free-electron metal
surface to study electronic excitations induced by
the impact of slow ions. The choice of an alkali
*
Corresponding author.
E-mail address: riccardi@fis.unical.it (P. Riccardi).
projectile is motivated by the fact that, due to their
low ionization potential, sodium ions lack enough
potential energy to give rise to the so called potential electron emission (PEE) [1]. Alkali projectiles
are therefore well suited to study kinetic electron
emission (KEE) [2], which is due to the transfer
of the kinetic energy of incoming projectiles. Despite of the distinctive advantage of alkali ions to
study KEE from metal surfaces, such studies have
been undertaken only recently [3,4].
The purpose of our experiments was to study
plasmon excitation [5], and inner-shell electron
0168-583X/$ - see front matter 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.nimb.2004.12.080
M. Commisso et al. / Nucl. Instr. and Meth. in Phys. Res. B 230 (2005) 438–442
promotion [6], which has been extensively studied
for the analogous Ne+–Al system, allowing a
straightforward comparison of the experimental
results for two projectiles of vastly different potential energy.
The electron energy spectra reported here show
evidence for bulk plasmon excitation, agrees well
with previous studies, confirming that bulk plasmon excitation in slow ion-surface interactions
[5,7,8] is related to kinetic electron excitation induced by the bombardment.
Electron energy distributions show also several
narrow peaks, due to Auger decay in vacuum of
sodium projectiles excited in the 2p-level by electron promotion in collisions with Al target atoms.
The Auger spectrum of sodium reported in this
work is consistent with previously reported spectra
[9] and is dominated by the Auger decay of the
2p53s2 state. One of the purposes of this work
was to search for and identify the Na peaks with
energies above the 2p53snl limit. Promotion of
two electrons of the 2p-shell of the projectiles, well
established for Ne+ impact, helped us to attribute
all the atomic features in the Auger spectrum of
sodium.
439
(DE = 50 eV) and therefore a constant transmission over the measured energy range.
3. Results and discussion
In Fig. 1 we report energy distributions of electrons emitted from Al surfaces by 1 keV Na+ ions
2. Experiments
The measurements were performed in a UHV
chamber with a base pressure of 3 · 10 10 Torr.
Na+ ions were produced with a Kimball Physics
ion gun. The ion beam current was of the order
of 10 9 A and had a Gaussian spatial distribution
in both horizontal and vertical directions, as measured with a movable Faraday cup situated in the
target position.
The polycrystalline Al samples (purity 99.999%)
was cleaned by sputtering with 6 keV Ar+. Sample
cleanness was assured by the absence of oxygen,
carbon and sodium signals in electron induced
Auger spectroscopy performed right before and
after the acquisition of each electron energy spectrum and by the constancy of the energy position
of sodium Auger lines during each spectral scan.
Emitted electrons were collected by a hemispherical energy analyzer lying in the incidence
plane and operated at a constant pass-energy
Fig. 1. Top: Energy spectra N(E) of electrons emitted from the
Al surface by 5 keV Ne+ impact versus electron emission angle.
The ion incidence angle is Hi = 30. Bottom: derivative dN(E)/
dE.
440
M. Commisso et al. / Nucl. Instr. and Meth. in Phys. Res. B 230 (2005) 438–442
as a function of observation angle He and for fixed
incidence angle Hi = 30 (angles are measured with
respect to the surface normal).
The broad structure visible in the 10–15 eV electron energy range identifies electron emission from
decay of plasmons [5]. Plasmon structures are usually visualized by the derivative of the experimental electron energy distribution [10], shown in
Fig. 1(b), where they result in minima at energies
Em = Epl / (Epl is the plasmon energy and /
= 4.3 eV is the work function for polycrystalline
Al). Consistent with previous studies [5,7,8], the
minimum around 11 eV in Fig. 1(b) is assigned
to electron emission from decay of zero-momentum bulk plasmons. This is in contrast to what
has been found for noble gas ion impact on Al
and Mg surfaces, where the bulk plasmon structure appeared overlapped to another structure assigned to decay of multipole surface plasmons
excited by the potential energy released when
incoming ions are neutralized by electron capture
from the surface [7,11,12].
This attribution is further confirmed by the variation of the intensity of electron emission from
bulk plasmon decay with the emission angle He,
shown in Fig. 2 normalized to the intensity for
Fig. 2. The intensity of electron emission from bulk plasmon
decay shows a cosine dependence on emission angle.
He = 0. These intensities have been evaluated by
subtracting a smooth polynomial background
from the derivatives of the spectra and integrating
the resulting negative peak [11–13]. Consistent
with previous results [8], the plasmon intensity varies like the cosine of emission angle, clearly indicating that bulk plasmon excitation occurs inside
the solid and that bulk plasmon decay act as an
isotropic internal source of electrons. Deviations
from the cosine dependence at large angles most
likely results from surface roughness.
The spectra of Fig. 1 show several narrow lines,
in the electron energy range 20–45 eV, due to
autoionization or Auger decay in vacuum of atomic excited states of reflected Na projectiles, created
by electron promotion [6] in binary collisions with
Al target atoms. To better illustrate these atomic
features, we show in Fig. 3(a) Na autoionization
spectrum for Ep = 800 eV, Hi = 70 and He = 0,
while in Fig. 3(b) we amplify the relevant portion
of the spectrum after subtracting a smooth secondary electron background. To correct approximately for the Doppler shifts due to the motion
of the Na (expected to be rather small in this
experimental geometry), the energy scale was adjusted so that peak labeled I assumes the value
of 25.7 eV to coincide with the kinetic energy of
2p53s2 autoionization line of neutral sodium [14].
The position of all the features indexed in Fig. 3
are listed in Table 1.
Besides the collisional excitation mechanisms,
the identification of the peaks requires consideration of the influence of possible surface neutralization or ionization mechanisms which are
peculiar to the ion surface interactions [1]. The
most intense peak I is generally attributed to electron emission from Na 2p53s2 decaying in vacuum
into the Na+ ground state 2p6 [14]. Within the
framework of the well known molecular orbital
(MO) promotion model [6], excitation of Na-2p
electrons is understood by the promotion of the
4fr MO, correlated to the Na-2p atomic orbital.
The crossings with other higher lying empty orbitals may result in the excitation of up to two electrons in the 4fr MO. The origin of peak I can thus
be explained by a one electron promotion process
during the projectile-target collision, accompanied
by electron capture from the surface in the
M. Commisso et al. / Nucl. Instr. and Meth. in Phys. Res. B 230 (2005) 438–442
Fig. 3. Top: Na Auger spectrum for 800 eV Na+ impact at Al
surfaces. Bottom: Amplified sharp lines after subtracting a
smooth secondary electron Background.
Table 1
Sodium Auger transition lines and proposed assignment
Label
Experimental
energy (eV)
Expected
energies (eV)
Initial state
Final
State
I
II
III
IV
V
25.7
28.5
32.9
37.0
41.3
25.7a
28.7a
32.6a
37.5b
41.4a
2p53s2
2p4(3P)3s2
2p4(1D)3s2
2p43s3p
2p4(1D)3s23p
2p6
2p5
2p5
2p5
2p53s
a
b
From [14].
Z+1 rule.
incoming or in the outgoing trajectory of the
projectiles.
441
The Auger spectrum of sodium reported in this
work is consistent with that reported in [9]. In that
work, peaks II and III were attributed to the decay
of the 2p53s(1P)3p and 2p53s(1P)3d configurations
of neutral Na. However, it is questionable that the
3d electron is stable against resonant ionization to
the metal conduction band. Here, we notice that
the energies of electrons from the decay of Na
2p4(3P)3s2 and Na 2p4(1D)3s2 into Na+2p5 [14]
are a better match with the positions of peaks II
and III, and in line with studies of the similar system, Ne–Al.
To our knowledge, assignment of peak IV, observed also in gas phase experiments [15], has not
been given. We assign it to the decay of
Na+2p4(1D)3s3p into Ne+2p5. An rough estimation of the energy of this transition can be done
by applying the approximate Z+1 rule [14], assuming that the difference in energy between the
Na+2p4(1D )3s3p and Na+2p4(1D)3s2 configurations is equal to the difference (4.6 eV) between
the energies of the states of aluminum ion:
Al+2p63s3p1P and Al+2p63s2. This yields an expected energy for peak IV equal to 37.5 eV, which
agrees well consistent with our measured value.
Finally, we notice that the energy of the weak
feature V is consistent with the energy of 41.4 eV
estimated in [14] for the decay of the state
Na+2p4(1D)3s23p.
Assignment of these peaks to the decay of 2pdoubly excited projectiles has significant implications that are consistent with those obtained in
the case of Ne+ impact [16–20]. Contribution of
2p4(3P) and 2p4(1D) states to peaks II and III, in
fact, implies the occurrence of a core-rearrangement process [16], converting the singlet into the
triplet core, which cannot be directly produced
by electron promotion.
The comparison with the case of neon projectiles can be further extended. In the case of Ne+
impact, experiments of electron emission [16,18]
and ion energy loss [20] showed that simultaneous
promotion of two 2p electrons of neutralized and
surviving projectiles leads to the formation of
Ne** and Ne+**, which, after autoionization
decay, are revealed respectively as backscattered
Ne+ and Ne++ with a characteristic energy loss.
In this contest, the production of Na projectiles
442
M. Commisso et al. / Nucl. Instr. and Meth. in Phys. Res. B 230 (2005) 438–442
with two 2p-vacancies can be understood considering that the ground state of incoming ions is the
same of neon atoms. Thus, simultaneous excitation of two electrons via the promotion of the
4fr MO leads to the production of the autoionizing states whose decay is observed in our
experiments.
4. Conclusions
In this study of plasmon excitation and projectile autoionization by electron spectroscopy we obtained two significant results. First, the spectra
show clear features due to electron emission from
decay of low momentum bulk plasmons by a projectile that cannot excite plasmons by the potential
mechanism. This result is consistent with the conclusions of previous studies, which have not been
as direct, that bulk plasmon excitation is indirect,
i.e. by fast electrons traveling inside the solid. This
implies that secondary effects, such as scattering
and electron cascade in the solid can give a significant contribution to electron excitation and
emission.
Second, we have observed a number of narrow
Auger features in the electron spectra and have
proposed identification for all of them, including
previously unidentified peaks that can be assigned
to decay of 2p doubly excited projectiles.
It is interesting to note that understanding this
particular type of experiments requires the consideration of both delocalized solid-state concepts
(bulk plasmons) and a local description of twobody collisions (excitation of excited states), thereby stressing the wealth of information on atomic
collisions in solids that can be provided by electron
spectroscopy.
References
[1] H.D. Hagstrum, in: N.H. Tolk, J.C. Tully, W. Heiland,
C.W. White (Eds.), Inelastic Ion-Surface Collisions, Academic Press, New York, 1977.
[2] R.A. Baragiola, in: J.W. Rabalais (Ed.), Low Energy IonSurface Interaction, Wiley, New York, 1994, Chapter 4.
[3] J.A. Yarmoff, H.T. Than, Z. Sroubek, Phys. Rev. B 65
(2002) 205412.
[4] J.A. Yarmoff, T.D. Liu, S.R. Qiu, Z. Sroubek, Phys. Rev.
Lett. 80 (1998) 2469.
[5] R.A. Baragiola, C.A. Dukes, P. Riccardi, Nucl. Instr. and
Meth. B 182 (2001) 73.
[6] M. Barat, W. Lichten, Phys. Rev. A 6 (1972) 211.
[7] P. Riccardi, P. Barone, A. Bonanno, A. Oliva, R.A.
Baragiola, Phys. Rev. Lett. 84 (2000) 378.
[8] P. Riccardi, P. Barone, M. Camarca, N. Mandarino, F.
Xu, A. Oliva, R.A. Baragiola, Nucl. Instr. and Meth. B
182 (2001) 84.
[9] Q.B. Lu, R. Souda, D.J. OÕConnor, B.V. King, R.J.
MacDonald, Phys. Rev. B 54 (1996) R8389.
[10] M.S. Chung, T.E. Everhart, Phys. Rev. B 15 (1977) 4699.
[11] P. Barone, R.A. Baragoiola, A. Bonanno, M. Comarca,
A. Oliva, P. Riccardi, F. Xu, Surf. Sci. Lett. 480 (2001)
L420.
[12] P. Riccardi, A. Sindona, P. Barone, A. Bonanno, A. Oliva,
R.A. Baragiola, Nucl. Instr. and Meth. B 212 (2003) 339.
[13] N. Stolterfoht, D. Niemann, V. Hoffmann, M. Rosler,
R.A. Baragiola, Phys. Rev. A 61 (2000) 052902.
[14] P. Dahl, M. Rodbro, G. Hermann, B. Fastrupp, M.E.
Rudd, J. Phys. B 9 (1976) 1581.
[15] J. Østgaard et al., Phys. Rev. A 19 (1979) 1457;
G.E. Zampieri, R.A. Baragiola, Surf. Sci. 114 (1982) 15.
[16] G. Zampieri, F. Meier, R. Baragiola, Phys. Rev. A 29
(1984) 116.
[17] F. Xu, N. Mandarino, A. Oliva, P. Zoccali, M. Camarca,
A. Bonanno, R.A. Baragiola, Phys. Rev. A 50 (1994) 4040.
[18] F. Xu, R.A. Baragiola, A. Bonanno, P. Zoccali, M.
Camarca, A. Oliva, Phys. Rev. Lett. 72 (1994) 4041.
[19] L. Guillemot, S. Lacombe, V.N. Tuan, V.A. Esaulov, E.
Sanchez, Y.A. Bandurin, A.I. Daschenko, V.G. Dobrnich,
Surf. Sci 365 (1996) 353.
[20] F. Ascione, G. Manicò, P. Alfano, A. Bonanno, N.
Mandarino, A. Oliva, F. Xu, Nucl. Instr. and Meth. B
135 (1998) 401.