HAL Id: jpa-00246354
https://hal.archives-ouvertes.fr/jpa-00246354
Submitted on 1 Jan 1991
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished 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.
Atomic metallic ion emission, field surface melting and scanning tunneling microscopy tips
Vu Thien Binh, N. Garcia
To cite this version:
Vu Thien Binh, N. Garcia. Atomic metallic ion emission, field surface melting and scanning tunneling microscopy tips. Journal de Physique I, EDP Sciences, 1991, 1 (5), pp.605-612. �10.1051/jp1:1991155�.
�jpa-00246354�
LPhys. I 1
(1991)
605-612 MM 1991, PAGE 605Classification
PhysicsAbsmw~
61.16F-68.45-79.70
ShonCou1u1unication
Atomic metallic ion emission, field surface melting and scanning tunneling microscopy tips
Vu Thien Binh
(I
and N. Garcia (1>2)(1)
Ddpartement
dePhysique
desMatdriaux(* ),
Universitd Claude BernardLyon
1, 69622 Vlleur- banne Cedex, France(2)
Departamento
de Rsica de la Materia Condenwda, Universidad Autonoma de Madrid, 28049 Madrid, Spain(Received4 March 1991,
accepted13
March 1991)R4sum6. L'obtention de faisceaux d'ions
(environ
10~ions/s) partir
d'dmetteurs ayant une di- mensionatomique
estpossible
et nouspr6sentons
id leur r6alisationexp6rimentale.
Ce travail reposesur le
principe
d'une fusion de surface sons champ unetemp6rature
d'environ un tiers de latempt-
rature de fusion de volume. Ces 6metteurs
pr6sentent
une structurepyramidale
de dimension nano-mdtrique
et terminde par un atome. Levoisinage
deplusieurs
dmetteurs, distants aussi dequelques
nanomltres, ddfinit alors des joints degrains
de surface. Nos rdsultats fournissent d'une partl'explica-
tion la
possibilit6
d'obtenir avec des pointes initialementmacroscopiques
une r6sclutionatomique
en microsccpie I eflet tunnel, et d'autre part ils montrent la
possibilitd
de fabriquer de manilre con- tr6lde des pcintes avec une double,triple,
etc... structures tdtines. Cette dtude a dtd rdalisde avec despointes
detungstIne
dont lescaractdristiques
sontanalys£es
parmicrosccpies
dlectronique et ioniquede champ.
Abstract. This work presents the
physical
realisation of metallic ion beams from atomic emitters vith currents ofapproximately
10~ ions per second. It also puts fonvard the idea of field surface melt-ing
atapproximately
one third of the bulkmelting
temperature. Undercooling
this melted surfbce,experiments
showpyramidal
structures of nanometer dimensionsending
in one atom alsoseparated
by nanometers, thenshaping
surface grain boundaries. Furthermore, this reveals whyit ispossible
to have atomic resolution in STMexperiments.
The formation of double,triple,
etc. atomic tetontips
is also possible. All this is shown ty field ion and field emission
microscopies
and atomic metallic ion emissionexperiments
presented here for tungsten tips.The
possibility
ofhaving
sources of metallic ions with atomic dimensions has beenlargely
in-vestigated iii
but their obtention has not beenpossible
until now. We show here thephysical
realhation. When a
large
electric field(F)
isapplied
to a metal surface(few
Volts perAngstrom)
(*
(UA CNRS).
6t6 JOURNAL DE PHYSIQUE I N°5
field
evaporation
has been observed intips ill by using
the field ionmicroscope (FIM).
At low tem-peratures (T), liquid nitrogen (LN),
fields of 5-6VIA
are needed to desorb the metallic ions from a
tungsten tip.
The activation effect isnegligible
at thistemperature
in thedesorption
processes and also surface diffusion is small so that the current of ions obtained isnegligibly
small(about
one ion per
minute).
At 3t© K ion embsion is uncontrollable withspots
over the whole surface with very lowintensity [ii,
and cannot be used as metallic ion sources of atomic dimensions.Now let us think in the effect of an electric field
applied
to a metal surface. The mostimpor-
tant
physical
effect is that the metal screens veryeffectively
theapplied
field. In fact selfconsistent calculations [2] show that even forlarge
fields the secondlayer
of atoms from the surface does not notice the field. If the field ispositive,
thecharge
iscompressed
into the bulk(the image plane
moves into the
solid) denuding
the surface atoms fromcharge
andcreating
alarge
surfacedipole
as well as
polarbation
upon which the field acts,pulling
out the ions from theequilibrium position
do to a newequflibrium position
under fielddF;
and it acts because the field b not constant anymore at the surface.
Figure
la illustratesgraphically
this effect. Its consequences are that thebinding
energy of the atoms is weakened and ions can bedesorbed,
but also that the activation energy of the atoms for diffusion at the surface is also reduced. This was shownby early
exper-iments [3] and is also confirmed
by
selfconsistent calculations [2].Figure
16 also illustrates this process. Thebinding
energyT(F)
and the activation energyA(F)
under field can be written:T(F)
=To +E(F) (I)
A(F)
=Ao +Ea(F) (2)
where To and Ao are the
energies
without field andE(F)
andEa(F)
therespective
corrections for the field. On the other hand surfacemelting
occurs when the diffusion coefficient islarge enough
that the atoms can move
easily
at the surface but We bulk atoms cannot. Fbrmeranalyses
onsurface
melting
and surface diffusion [4] have shown that when the surface diffusion coefficient ism
10~~cm2 Is
the surface has melted. Fromequations (I)
and(2)
one can see thatby applying
afield to a surface we have two
parmeters
to melt it: the field that is localized on the surface and T tllat acts over the wholebody.
Thequestion
that we have put to ourselves is: would it bepossible
to melt a surface
by applying
a field at T much lower than the bulkmelting
on Tm, and at the same time have a localized ion source of metallic ions on We surface? The answer to thisquestion
isaffirmative and therefore we
present
theexperimental
evidence.The
experiments
areperformed
in anapparatus
in which a field emissionmicroscope (MM)
iscoupled
to a field ionmicroscope.
The base vacuum b better thant 10~~~ lbrr, and the
pumping speed
from the FIMimaging
pres-sure of rare gas of m 10~~ lbrr to ultra
high
vacuum pressure(UHV)
takes less than a few minutes.FI
imaging
areperformed
atLN,
and theheating
of thetips
to the desiredtemperatures
is ob-tained
through
a Jouleheating loop.
Thetemperature
at the end of thetip
b controlledby
anoptical micropyrometer
with aprecbion
of about 10 K The tungstentips
areprepared
from asingle crystal
wire with <ill>orientation,
and we use the thermalsharpening technique
near3000 K and in ultra
high
vacuum(< 10~~° lbrr)
to obtain the initial cleantip
[6,7j.
The vacuumduring
metallic ion emissions and FE h m 10~~~ lbrr. Metallic ion spots, FI and FE patterns arevisualized
through
achannel-plate coupled
with a fluorescent screen; and theimages
are recorded with ahigh
sensitive video camera,allowing
then furtherimage processings.
After
having prepared
thetip,
weproceed by sweeping
theparameters
described above thatregulate
the process: I-e- F and T. The values of Fapplied
arepositive
as to desorb ions from the surface. Infigure
2a wepresent
the observations of atomic metallic ion emission(AMIE)
from atomic dimension structures. Thespots
have anopening
of 3°approximately
and are well local- ized in somepoints
of the surface.They
also move in timediffusing
from onepoint
to another. AtN°5 ATOMICMETALLICIONEMISSION 607
(~) do ~~~~'~
~d
~
~
m#Moaf4pWlkdZWW
Vampod field
~
dF
""""m#lk an appllwd WWJd
~ , .«ddi
~~~ )~ °~~~~~~~~~"
~~~~~~~~
~ith an
applied
fief- c
m o
. ©
o .
et A<F) ted
, -
n o A~
, equi
z
Posit onj
(~ i
( Z
(ith an li.d
', z
,-
,"
E<F)
.ithout
ppli.d
Fig. I.
(a)
surfacelayers
of atoms without field. theequilibrium
distance isdo. the samelayers
but withapplied
field, where the atoms have beendisplaced
outwards to a distance dF. the illustation of thedamped
field is also sketched.
(b)
Activation barrier with and without field and thebinding
energy of atoms at the equilibrium positions and at the saddle points. The parameters are those of equations(I)
and(2).
6lB JOURNAL DE PHYSIQUE I N°5
lower
temperatures
thespots
remain fixed but have lower intensities. The three framefigures
from left tofight correspond
to different times. The values of F and T are m I-SVIA
and m 1500 K and thespot
currentintensity
is m 10~ ions/s. With theseparameters
andusing
a crudetheory
for iondesorption [I]
we found that the activation energy fordesorption
is m 1.8 eV and thebinding
atomic energy m 2.8
eY
much smaller than that of aplane
surface at zero field which h 8.7 eV The reduction of thedesorption
barrier b due on the one hand to the fact that the atom is on thetop
of aprotrusion
and has a smaller number ofneighbours
than those in aplane surface,
and onthe other hand to the
decreasing
in thebinding
energy due toF,
that now its effect is increasedby
the smeared out of the electronic
charge
at theprotrusion
[5j orSmouluchowsky
effect. We also found that under thegiven
field the surface diffusion activation barrier has decreased from 3 eV to 0.7 eVby scaling
with formerexperimental
resultsby
Bettler and Charbonnier [3] at smaller fields(0.4 VIA).
Thisyields
a surface diffusion coefficient of m 2 x10~~ cm2 Is
which is of the order of the estimate for surfacemelting
[4]. This is consistent with our observations and thepossibility
to obtain continuous flux of ions
desorbing
from one atomic site at low temperatures. However these values have to becompared
with theT~
m 3680 K needed to melt the W Therefore at values much lower than T~ it ispossible
to have aliquid
at the surfacelayer,
of the same atoms as thebulk, by applying
a convenient field. Thb also has a critical range over which theexperiment
ispossible
between 1.2 and 1.6V/fi~
The remarkable fact is that theliquid
isjust
at the surfacelayer
because the field has no effect on the second
layer
of atoms.By inserting
thetip
in LN under theapplied
field a very fastcooling
takesplace
that freezes thetip
structure and prevents diffusion.By
means ofMM,
I.e.reversing
fieldpolarity,
we observe thepicture
of the freezed structure~fig. 2b).
We find that the MM and AMIEspots correspond
each other very well in space,indicating
that both emissions takeplace
fromprotrusions
at the surface that have been createdby
the field F athigh
T due topolarization
anddipole
forces. These littlepyramidal protrusions,
we called them atomic tetontips
[6,7j. They always
emit the ions from the top atom of theprotrusions
because thebinding
energy is weaker(as
dbcussedabove)
and theeffective electric field
larger.
lb check that the
positive
field beams are W ions, we haveanalyzed
thespot
deviations due to theapplication
of amagnetic
field. Ourfindings
are that the electronspots
move with thb field, but thepositive
electric fieldspots ~fig. 2a)
do not. So the emittedparticles
areheavy positive
ions. because the current
intensity
is 10~ ions/s with the vacuum 10~~~ lbrr and T = 1500K,
itseems clear that ionization of residual gas or
impurities
is notpossible. Furthermore,
after theobtention of the
pyramidal protrusion
structure weapplied
to thetip, always
under UHV but atLN,
apositive
field of the same value or greater than that needed to see the AMIEspots,
and we see nospot
at all. Theseexperiments
have also beenreproduced
for different Wtips.
lb confirnl further the above observations we have the FIM as an additional
technique
that is well known to have anextremely good
atomic resolution. FlM results for the same freezedstructures than above are
presented
infigure
3a where the veryending
atomic tetontips
[6,7j
areobserved.
By
fielddesorption experiments
at LN andby
FIM we can see the differentlayers
of atoms thatconfigure
the atomic tetontips
formed athigh
T. Results of differentdecaped layers
are
presented
infigure
3b andthey
showclearly
thepyramidal
bash of the differentprotrusions meeting
in the surfacegrain boundary.
The existence of these boundaries and theshape
of thetip
indicate that the surface tension of the initial
tip
has beencompletely changed by
the presence of the field athigh
T.The observations desc&ed above are of considerable
importance
to understand STM exper- iments. Thequestion
that remains to be answered b: how b itpossible
that withtips
of100 nmradius or
larger
can be revealed atomic resolution swucture on surfaces? Afterperforming
theseexperiments
we believe that the answer lies in the formation of thepyramidal protrusion
end in g inN°5 ATOMIC METALLIC ION EMISSION &~9
zeW~
ifi~j
"OO©
GOB .fihB
fl9$
~~,e flf~W "~'i
©W~© '~
~~i
~e)
.iG~~l~
8 fl
)fl~
~-,£
uf~ufire
2~if
~d~ fl~f
((@
8#'~
~fi~~
a'5~
~~_3u ~~
iQ(if
°'~ullfifi
©~j>
(~li
~u~-m
~£%
Wjj>
Efi- i~8
" ~
)§(
~d>
11§ j~
'ny
_
fiflj~
#3u
%d"~
f.f~ iii
~S4z
~£
(ho eu~uEuUWQ
OO~.lo-OW~~
j~~'(
~O~
Off- flu-e
~f#j
J~j'5-~
".*u
j~E -m obp£
°4§~SE-E
610 JOURNAL DE PHYSIQUE I N°5
j-E bin
i~~
id
#I
u~E~mm
$Wo.-
n~'G~fl 6~
~'~~m
XC
~m
~g~-
©~
offel- y-
fl~
O©u~6 ijfl ES 4Z@m-
~'@ ~fi
~~J-
~~
~"o
fl~z
~i
~u EmOh
~t
~"]~
'r©
ECem
©m mm
iii
~~~'Z to OX
~j
E-R©~
l$3
~b
©~~o
©£i li~$
woif i©~©(
ff§ g~
.§~
~~
~.kj
=uH~tim
-~8~z
~
UWO_
'5~w -
"/~~
jo- ~~
"
"ii
i~~2
N°5 ATOMICMETALLICIONEMISSION 611
a
single
atom that we have described.Experimentalists working
in the STM field kJ1ow very well that beforereaching
atomic resolution different treatments have to begiven
to Wetips
[7j. Thesetreatments consbt
basically
inestablhhing
alarge voltage
betweentip
andsample
andpassing
a
large
currentduring
a shortperiod
of time.By giving
thesepulses,
atip
thatprovides
atomic resolution can be obtained. In ouropinion, during
thesepulses
thetip
surfacetemperature
rakesdue to electron bombardment of the surface because of the
large
currentpulse,
but under thepulse applied voltage
thatgive
rises to fields of m 1.5V/fi~
These arebasically
the same param-eters that we need to form our
pyramidal protrusions.
Ingeneral
we form manyprotrusions
overthe
tip surface,
but in a STMconfiguration
We mostprobable
ocurrence will be aprotrusion
at the apex of Wetip
because in this case the field is localizedjust
at the apex. In ourexperiments
the field h
practically
constant over the whole surface beforeforming
theprotrusions.
Thesestructures
ending
in one atomprovide
the atomic resolution and are the naturaltopographical
configuration
after thetip
has beenheating
under field and then fast freezed. Also Were have been STM observations that are believed to be due tomultiple tips
[7j. Thesemultiple tips
maybe the same as those
reported
and visualized in thin paper, but nowthey
can be built under well controlledparameters
and conditions. Theapplicability
of atip ending
in twopyramidal protru-
sions of m 1.5 nmheight separated by
m 3 nm in STM can be ofimportance
to measuredynamical
effects. For
example
coherencelength
inhigh
Tcsuperconductors
[8]. As aproof
that this can be reached with our field surfacemelting
method we present theexperimental
evidence infigure
4, in which atip ending
in twopyramides
ispresented
and even one can choose for thepyramides ending
in three atoms or one atom.Fig.
4. FIM of a tip ending in two atomic tetontips
with m I-S nm height and separated by m 3 nm (this is obtained bycounting
rows of atoms in thepicture).
We can make thetip
finish in three atoms or one atom in a controlled way. thesetips
may be ofimportance
for sTMdynamical
experiments.In
conclusion,
it seems that we have shown that the obtention of atomic metallic ions'emhsion ispossible.
This isaccompanied by
the idea put forward of field surfacemelting
thatgives
rheto the formation of
pyramidal protrusions
or atomic tetontips
and meet in their basisforming
surface
grain
boundaries. These structures emit ions from asingle
atom source so that ions, as wellas coherent electron beams, are obtained. MM reveals the
emitting
structure and the exhtence of 2-Dmelting.
The obtainedtips
answer also thequestion
of thepossibility
ofobtaining
atomic612 JOURNAL DE PHYSIQUE I N°5
resolution in WM
experiments starting
with a 100 nm radiustips
as well as the exhtence ofdouble, uiple,
etc nanotips.
Thin may haveimportant physical
consequences and technicalapplications
innanotechnology
if one has clear ideas of how to use thb effecL Webelieve,
forexample,
that oursource of ions can be of
importance
forwriting
nano and atomicconducting
lines on surfaces. Alsothe coherent emission of our
tips
can makepossible
electronholography
with three dimensional atomic resolution [7j. Work in this field b now in progress.Finally
we would like to say that Weexperiments
arefully reproducible.
In fact the structures obtained are WeWermodynamical equilibrium configuration
of a Wtip
for We temperatures and fieldsreported
above. We believe that other metal surfaces should show similarconfigurations
and the same idea of field surface
melting
should beapplicable.
Acknowledgements.
We would like to thank Prof. R. Uzan for his interest in thb work and J.
Doglioni
for technical assbtance to set up thesystem.
This work has beensupported by
an EEC Scienceproject,
andby
the French and
Spanish
authorities. We also thank our partners of the EECproject
for support in the realization of theseexperiments.
Dbcussions with J.J.Saenz,
PA~ Serena, M.Pitaval,
D.Atlan and G. Gardet are
apprechted.
References
[1] Field Ion
Microscopy,
E.W Muller and II l§ong(Elsevier Publishing
Co. NY,1%9);
Held IonMicroscopy,
Held Ionization and FieldEvaporation, Pwg Su$
Sci. 1(1974)
1.[2] SERENA PA., GARCtA N. and Vu MIEN u1NH
(to
bepublished).
[3] BBnLER PC. and OMRmNNIER EM.,P%ys. Rev l19
(1%0)
85;Uwuot H. and COMER R., L Chem PhyK 37
(1962)
1706.Experiments
show that the activation barrierA(F)
cannot be zero(see
Ref. [1]). This is consistent with our observations, because below m 1000 K we do not see any AMIE even athigh voltage
due to the lack of diffusion.[4] DA£H J.G.,
Contemp.
P%ys. 3@(1989)
8%BIENFArr M. and GAY J.-M., Proc. of NATO ASI on Phase transitions in surface Films, June 19-30 1990, Erice,
Italy.
[5j smouLucHowsxi R~, P%ys. Rev 60
(1941)
Ml.[fl
Vu THtEN BINH,L Mcms. 152(1988)
355.[7j see for example
scanning Tunneling Microscopy
and Related Methods, NATO ASI Sefies E Vol 184, R.J. Behm, N. Garcia and H. Rohrer Eds.. For teton tips see page 409.[8] SCHRIEFFER J.R.