INTERACTION OF MULTIPLY CHARGED IONS
WITH TUNGSTEN SOLID SURFACES :
BACKSCATTERED IONS AND SECONDARY
ELECTRONS
M. Delaunay, C. Benazeth, N. Benazeth, F. Bourg, R. Geller, P. Ludwig, C.
Mayoral, G. Melin
To cite this version:
M. Delaunay, C. Benazeth, N. Benazeth, F. Bourg, R. Geller, et al.. INTERACTION OF
MULTIPLY CHARGED IONS WITH TUNGSTEN SOLID SURFACES : BACKSCATTERED
IONS AND SECONDARY ELECTRONS. Journal de Physique Colloques, 1989, 50 (C1),
pp.C1-277-C1-284. <10.1051/jphyscol:1989130>. <jpa-00229326>
HAL Id: jpa-00229326
https://hal.archives-ouvertes.fr/jpa-00229326
Submitted on 1 Jan 1989
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
JOURNAL DE PHYSIQUE
Colloque C1, suppl6ment au n o l , Tome 50, janvier 1989
INTERACTION OF MULTIPLY CHARGED IONS WITH TUNGSTEN SOLID SURFACES :
BACKSCATTERED IONS AND SECONDARY ELECTRONS
M. DELAUNAY, C. BENAZETH',
N. BENAZETH*, F. BOURG, R. GELLER, P.
LUDWIG, C. 'MAYORAL' and G. MELIN
C E A , CEN - D R F / P A D S I - 8 5 X , F - 3 8 0 4 1 G r e n o b l e C e d e x , F r a n c e
' . u n i v e r s i t B P a u l S a b a t i e r , LPS/CNRS U A - 7 4 , F - 3 1 0 6 2 T o u l o u s e C e d e x ,
France
Esu6'.- Nous pr6sentons des resultats concernant les ions r6flechis et les electrons
sec ndaires recueillis lors de l'impact A incidence normale d'ions argon multicharges
9) sur une surface de tungst8ne. Pour une charge donnee de la particule
ArqP (1 < q
incidente, ,le coefficient de r6flexion est sensiblement ind6pendant de 116nergie de
l'ion primaire dans le domaine 10-120 keV. Ceci peut Btre attribu6 h l'augmentation de
la charge moyenne des particules retrodiffusees lorsque 1'6nergie incidente augmente,
compens6e par la diminution du nombre de particules refldchies. D'autre part, pour une
Bnergie incidente donnee, la charge moyenne des particules reflechies augmente avec la
charge de la particule incidence. & coefficient d16mission electronique secondaire
total augmente lineairement avec le carre de la charge de l'ion incident. Ceci suggere
que le processus d'emission de champ pourrait jouer un r6le dans 1'6mission
electronique secondaire observee lors de l'interaction d'ions multicharges et de
surfaces metalliques.
<
Abstract.- Results are concerned with backscattered ion and secondary electron
~ <+ q < 9) ions on clean tungsten.
emissions from impact, at normal incidence of ~ r (1
For a given incident ion charge q, the reflection coefficient is roughly independent of
the incident energy in the 10-120 keV range. This can be due to the balance between
increase of the mean charge state of the backscattered particles when the energy
increases and the decrease of the reflected ion number. Moreover, for a given incident
energy, the mean charge of the backscattered particles increases with the charge of the
incident ion. It was found a linear dependence of the total electron emission yield 17
2
versus q . This result suggests that the field emission process could play a role in
the secondary electron emission due to the interaction of multicharged ions with
metallic surfaces.
1.
INTRODUCTION
Scattered ions and secondary electrons due to impact of highly charged ions on metallic surfaces play an important role in the performance of Electron yclotron Resonance Ion Sources
/I/. Electrons issuing from walls contribute to the electron density of the plasma, in such
a way that wall coating is a parameter as important as R.F. power, gas pressure or magnetic
field profile. The reflection with (partial) neutralization of multiply charged ions of the
plasma on the source walls must be taken into account for calculation of the plasma
equilibrium equations.
Another application of the knowledge of reflection and secondary electron emission yields is
the identification of particules with the same charge to mass ratio which cannot be sorted
10+ 4+
by a magnetic field (for example Ar
, 0 ).
We present here results concerning reflected ions and secondary electrons from the
interaction of multiply charged Argon ions ~
r (1
~ <+q
< 9) with
tungsten surfaces.
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1989130
JOURNAL DE PHYSIQUE
C1-2 78
2.
EXF'ERIMENT
The experimental set up has been described in details previously /2,3/. The multiply charged
ions are produced by an E.C.R.
ion source (ECRIS 10 GHz).
The extraction voltage can
be varied from 2 kV to 14 kV. Polycrystalline tungsten targets are cleaned in situ by
sputtering with a 3pA Ar2+ ion beam in a background pressure of 5.10-lo mbar.
Typical Arq+ ion currents of about 0.1 pA are measured onto the target which is situated at
the center of an hemispherical collector. The target-collector voltage can be varied up to
+
-
500 V in order to collect all the emitted secondary electrons on the collector or to
prevent their emission from the target.
With an attractive field to the collector, the target current, I;,
current,^:, are
and the collector
:
I+
T
=
nI .q.e. - nR ?j e + n".e
I
I R
I+ = -n"e + nR q
C
I R . ~
(1)
(1')
and for a repulsive field :
with n1 : incident ion number per time unit, nR : backscattered ion number per time unit, n"
I
I
: secondary electrons number per time unit, q.e : initial ion charge, qR.e : mean charge of
backscattered ions, e : elementary charge.
1
So, the "true" incident ionic current II and the true secondary electron current I,,, can be
obtained :
It follows that the secondary electron yield y~n"/n:
rq
= :n
/ :n are equal to
and the reflection coefficient
:
y :(n" /):n
and
=
(Ie,, / 1;) .q
-
(
n
: /):n
(1; / 1;). (q / GR)
rq
It must be noted that the secondary (positive and negative) ions were neglected in the relaand
tions (1) and (2) . This assumption is supported by the fact that yields as low as 9.
+
5.10-~were observed for W+ ions under 4 keV and 40 keV Ar ions bombardment respectively
/4,5/.
3.
RESULTS AND COMMENTS
a. Reflection coefficients
The apparent reflection coefficient Rq
= [
(IR/l'
I). I q] (= rq.qR) is plotted versus ion kinetic
energy for a given initial charge state, in figure 1. Its value appears to be roughly
constant for a given ionic charge in the explored energy range. This could be due to a
balance between a decrease of the reflection coefficient itself rq and the increase of the
mean. charge of the scattered particule when the incident ionic energy increases. The
decrease of rq is a well established fact /6/ and the increase of
eR can be
understood in
the usual framework /7/ of charge exchange between a charged particle and a metallic
surface.
FIGURE 1.- Apparent ionic reflection coefficient Rq
kinetic energy (1 < q ,< 9).
-q
[
I
:
/ :1 ]
($)
versus Arq+ ion
From these results we have deduced the values of R~ versus incident ion charge state q for
15 keV Arq+ ions on clean tungsten (Fig. 2). For a constant incident kinetic energy it can
be assumed that the actual reflection coefficient rq is constant as well. Indeed, the atomic
collision parameters governing the trajectories of scattered atoms only weakly depend on the
charge of collision partners (through the screening of internuclear potential).
In assuming that rq is constant, the variation of Rq depicts actually the variation of the
mean charge ijR on q. For initial charge state from 1 to 8 , Rq roughly linearly increases
with q while for q equal to 9 a sudden rise is observed. Compaxing these results with those
of De Zwart et al. /8/,
+
the backscattering ion current can be splitted into a constant
contribution due to Ar , a linear increase of Ar2+ current from q
advent of Ar3+ ions for q
=
1 to q
=
8 and the
- 9. Assuming
that for incident Ar+ ions, the backscattered ion
+
flux is essentially composed of Ar
ionsl,the mean charge of the backscattered particles
can be assessed.
l ~ h i sis true within 2% error for 20 keV Arq+ ions with incidence and scattering angles
15" relative to the surface /8/.
R
JOURNAL DE PHYSIQUE
TABLE I
FIGURE 2.-Apparent ionic reflection coefficient R~ versus ion charge state q for 15 keV ~ r ~ +
ions on clean tungsten.
So, a value of rq
-
0.21 for 15 keV Ar ions on tungsten at normal incidence can be deduced
which is in good agreement with the theoretical value of Ref/6/.
These results confirm that the initial charge state plays an important role on the final
charge state of the backscattered particles, this is inconsistent with the Van der Weg model
/9/ of violent collisions.
b. Electron emission
Emission of electrons from a solid surface bombarded by highly charged ions is
dominated by potential emission as it was shown in references /2,3,10/. However it can be
noticed that in the energy range used in this work (Fig. 3)
energy ; from this it can be deduced
negligeable phenomenon.
increases with the kinetic
I
that the kinetic secondary electron emission is a non
/O
/O
0
I
, "
x x
X
l
"I .
X
O/O
/
=
o/
"
1
0'", .
O
o
x
x
I
,
/x
+
3
2
x
+
0
1
Vex.
+HO'
FIGURE 3. Electron yield y for impact of ~
extraction voltage.
4
For slow incident ions ( < 5.10 ins"'),
r ions
~ +on clean tungsten versus ion source
the secondary electron yield seems to be linearly
proportional to the potential energy avztlable from the approaching ion i.e :
4
y ;kWq wick Wq = i=l
C
'i-1.i
Ii-l,iis the ionization potential of the Z
(i-1)+ .
lon. Moreover, for upper incident veloci-
ties (v >, 5.10+~m s-I), and for impacts of ~ r on
~ W,
+ the yield for q >, 9 (case for which L
shell vacancies preexist in the incident ion) is below the linear dependence extrapolated
from the lower initial charge states /10,11,12/. We confirm this discrepancy in the
28-70 keV kinetic energy range (Fig. 4,5).
A new variable called the "total electron yield" can be defined :
=
Y + q. Indeed if we
assume that the neutralization entirely takes place out of the solid, whatever the electron
transfer processes, the surface provides a number of electrons per ion equal to y + q.
(Ysecondary electrons arriving at the collector and q electrons for the ion neutralization) .
For two values of the kinetic energy (28 keV and 70 keV), fig. 6 shows
that y
+
q is a
linear function of q2 for 1 ,< q ,< 9. This result confirms a previous one obtained for a
kinetic energy of 20 keV and for q included between 1 and 12 /12/.
This proportionality between the total flux of secondary electrons emitted from the surface
and q2 leads to explain the electronic emission in the framework of the "field emission".
JOURNAL DE PHYSIQUE
FIGURE 4.-Electron yield Y for impact of Arqf ions on clean tungsten versus ion potential
energy for 28 keV
incident kinetic energy.
FIGURE 5.- Electron yield for impact of Arq+ ions on clean tungsten versus ion potential
energy for 70 keV incident kinetic energy.
( e / ion)
201
FIGORE 6.-Total electron yield 17
-
y
+
q versus q- for 28 and 70 keV ArY'-ions.
Indeed following the approximate Fowler-Nordheim law for very high electric fields,&, the
2
flux of secondary electrons is proportional to E~ and so to q
.
This process could explained
the initial electron transfer from the conduction band of the metal to the potential well
created by the ion close the surface, whatever the subsequent relaxation processes of the
highly charged ion are, for example, Auger cascades.
This proportionality between rl and q2 is a new and confusing observation which requests
complementary results with other ion-target combinations.
Moreover we keep in mind the difficulties linked to the application of the field emission
theory to systems like multicharged ions-metallic surfaces. In particular the future theoretical model will have to justify the use of a constant ionic charge during the most part
of the incoming path of the particle towards the surface.
4.
SUMMARY AND CONCLUSION
Experimental results concerning reflected ions and secondary electron emission due to the
impact at normal incidence of ~
< < 9) on clean tungsten
r (1
~ +q
surfaces are reported. The
main results are :
-
The apparent ionic reflection coefficient is roughly constant when kinetic energy of
the incident particle Arq+ varies (in the 10-120 keV range). This can be explained by
an increase of mean charge of backscattered particles
-
For a given incident energy, the value of
qk
ifR while their number decreases.
increases with q so showing that the
final charge state of the reflected ion is dependent of the initial one, in discrepancy
with a previous model.
JOURNAL DE PHYSIQUE
Cl-284
.. The total electron yield 11 = y
+
q is linearly dependent on q2 that suggests that
field emission could play a role in the secondary electron emission due to the impact
of highly charged particles on metallic surfaces. However more experimental results
have to be obtained and there is need to think about the possibilities of adaptation of
the field electron emission theory to this type of experiments.
REFERENCES
/1/ GELLER, R et al., The Grenoble ECRIS status 1987, Int. Conf. on ECR ion source, Nov.
1987, Michigan State Univ. East Lansing (USA).
/2/
DELAUNAY, M., BENAZETH, C., BENAZETH, N., GELLER, R. and MAYORAL, C., Surf. Sci. 195
(1988) 455.
/3/
DELAUNAY, M., FEHRINGER, M., GELLER, R., VARGA, P. and WINTER, H., Europhys. Letters 4
(1987) 377.
/4/ BENNINGHOVEN, A., Surf. Sci. 53 (1975) 596.
/5/ JURELA, Z. Rad. Eff. 13 (1972) 167.
/6/ EiCKSTEIN, W. and BIERSACK, J.P., Z. Phys. B 63 (1986) 471.
/7/
ARIFOV, U.A.,
KISHINEVSKII, E.S.,
MLTKHAMADIEV, E.S
and
PARILIS, E.S.,
Soviet
Phys.-Tech. Phys. 18 (1973) 118.
/8/
De ZWART, S.T., FRIED, T., JELLEN, U., BOERS, A.L. and DRENTJG, A.G., J. Phys. B : A t
Mol. Phys. 18 (1985) L623.
/9/
VAN DER WEG, W.F. and BIERMAN, D.J., Physica 44 (1969) 177.
/lo/ DELAUNAY, M., FEHRINGER M., GELLER, R., HITZ, D., VARGA, P. and WINTER, H., Phys. Rev.
B 35 (1987) 4232.
/11/ De ZWART, S.T., Thesis, Groningen University (1987).
/12/ DELAUNAY, M., GELLER, R., DEBERNARDI, J., LUDWIG, P. and SORTAIS, P., Surf. Sci. 197
(1988) L273.