proceedings 2002 - RECENT TRENDS 2016



proceedings 2002 - RECENT TRENDS 2016
Recent Trends
in Charged Particle Optics and
Surface Physics Instrumentation
o f th e 8 ,h In te rn a tio n a l S e m in a r,
held in S k a lsk y d v u r n e a r B rn o , C z e ch R ep u b lic,
fro m Ju ly 8 to 12, 2 0 0 2 ,
o rg a n ise d by
th e In s titu te o f S c ien tific In s tru m e n ts A S C R and
th e C z e c h o slo v a k M ic ro sc o p y S o ciety
Edited by Luděk Frank
P u b lish e d an d d is trib u te d by C ze c h o slo v a k M icro sc o p y S o ciety , p rin te d by G ra p h ic a l B rno.
C o p y rig h t © 2 0 0 2 by th e C ze c h o slo v ak M icro sc o p y S o c ie ty a n d in d iv id u a l c o n trib u to rs. A ll rig h ts
re se rv e d . N o p a rt o f th is p u b lic a tio n m ay b e re p ro d u c e d , s to re d in a re trie v a l sy stem o r tra n s m itte d in any
fo rm o r by an y m e a n s, e le c tro n ic , m e c h a n ic a l, p h o to c o p y in g , re c o rd in g o r o th e rw ise , w ith o u t th e w ritte n
p e rm is sio n o f th e p u b lish e r, e x ce p t a s s ta te d below . S in g le p h o to c o p ie s o r sin g le a rtic le s m ay b e m a d e fo r
p riv a te stu d y o r re sea rc h . Illu stra tio n s an d s h o rt e x tra c ts fro m th e te x t o f in d iv id u a l c o n trib u tio n s m ay be
c o p ie d p ro v id e d th a t th e so u rc e is a ck n o w le d g e d , th e p e rm is sio n o f th e a u th o rs is o b ta in e d a n d C S M S is
n o tified .
T h e p u b lis h e rs h ave m a d e e v ery e ffo rt to re p ro d u c e th e p a p e rs a t th e ir o p tim u m c o n tra st a n d w ith o p tim u m
re so lu tio n fro m m a te ria ls p ro v id e d by th e au th o rs. T h e p u b lish e rs a c c e p t n o lia b ility fo r th e loss o f im a g e
q u a lity re s u ltin g fro m su p p ly o f p o o r o rig in a l m a te ria l o r im p ro p erly p re p a re d d ig ita l reco rd s.
ISBN 80-238-8986-9
(xívíjaj ' 9
‘ja ip s g -jaj ‘asuaquoqos - 9 Áq pajiuapoiu)
AdoosojDiiu u o j p a p u o is s iiu a o jo q j
(S uopQ y ">| -y ‘BAoauaq g Áq pajR.iopoiuj
S3SU3J 3A|pafqo [i;uoi}iptíjj-uou puB sjijcIo puäBiu juauciujaj
OE t7!
(^ u e io )! 'A ‘3 uo[oq y Áq pojiuapoiu)
IM31 (aSBjjOA Avoi) aqi jo asn puB uSisap jo aanjnj
(asuaquoqas '9
J ‘aso y
0£ 11
suojjBj.iaqB jo UOIJ03JJ09
I I Mn[‘ Á i:p sjn i|x
( j a m q o s g ‘Á í[SA 0ireui0g
f Áq p a jB ja p o u i)
JM3S uí 33U33S3uiuitqopoq;B3 ‘jMas IBiuauiuojiAua
(S[3iUA\ H ‘Sbjusuuos g ‘qoBqpssBH d Áq pnuuopoiu)
sjBiuauiBpunj u o jp a ia
Ol Á jn f \B p s o u p a \ \
(a s u a q u o q a s
‘a s o g h ‘sæ [M BH g M p ajE isp o u i)
uoissas J3JS0J
q 'BA(uopn¡A¡ | ‘jq a d ç -g ‘so>[ire[AI ¡aj Áq pajejapoui)
u o i p n p o j d p u B s a p s o u S B ip jo p n p u o D ju ia s u í s iiiB s q - y
O ř tT
(çA oauaq g ‘sa^AÆH g Áq pajB japoiu)
II sapdo u o jp a p joj suoijBjndiuo3
o£’I [
(pr;ag g ‘ojun]Aí g Áq pan’.iapoiu)
I saijdo u o jp a p joj suoijBjndiuo^
0£ 60
6 Á[nf ÁBpsonj
S N O issn D sia a i a v i a N n c r a / s n o is s h s a v N m a s h h i j o
3 1 ÍK I3 H 3 S
It is not easy, but hopefully also not necessary, to adopt alw ays a fully new approach to
the preface w hen it opens new volum e o f a series o f proceedings from regularly organised
meetings. T his S em inar on R ecent Trends in C harged Particle O ptics and Surface Physics
Instrum entation is already the eighth in sequence. But there m ight be som ebody not having
had heard about the m eetin g so that the histo ry could be briefly su m m arized once again.
The germ w as the traditio n al su m m er sem in ar held in the Institute o f S cientific Instrum ents
in B rno on the occasion o f visits o f P rofessor Tom M ulvey from U niversity o f A ston in
B irm ingham . H is presence ju stified the sem in ar E nglish spoken and he always excellently
fulfilled his role o f m oderator and com m entator. In anticipation o f histo rical m ovem ents we
started to extend the participation already in sum m er 1989. T h at tim e five foreign guests
attended the m eeting still tak in g place in the Institu te’s library. In 1990, there w ere as m any
as 30 participants from 5 countries. T he th ird sem in ar in 1992 w as m oved to hotel Skalsky
dvur in B ohem ian-M oravian H ighlands w here it rep eats since th at tim e. T raditionally about
forty or fifty people w ere present and the m eetin g schedule has stabilised.
T he fifth sem inar w as the first one to w hich the p roceedings were published. T he p ro ­
ceedings o f the sixth sem inar ap p eared tech n ically a bit b etter but nothing at all was pub­
lished from the next m eeting in 2000. T h is took place im m ediately a fte r the 12th E uropean
Congress on E lectron M icroscopy in B rno and the organisers hesitated to force the p artic i­
pants into w riting one m ore text about (m ost probably) a sim ilar topic. So this brochure
should resem ble that from 1998 and we hope for ad ditional in fu tu re, alw ays b etter and
better but suggesting the trad itio n by th eir consistent design.
T he set o f up to ten topics, w hich fill th e sem in ar tim e o f five half-days, has been
variously com posed o f issues from electron optics and surface exam ination and often the
second p art o f the headin g w as less present. T he sam e is true th is tim e but still the surface
m ethods are not only presen t but even start to p enetrate new regions lik e the sem iconductor
processing and exam ination, progressively m ore surface sensitive at grow ing device in te­
O ne cannot overlook certain crisis developing itself in attendance at scientific con­
gresses and conferences. T h e m eetings seem to be too m any and, th an k s to pow erful tools
for electronic com m unication, people seem to b e less w illin g to travel. W e are repeatedly
a bit afraid and at least curious w hether th is tren d is going to im pact on us as well. A gain
not - when looking onto the list o f particip an ts one can only be happy in having o p portunity
to be there.
From its beginnings, th e sem in ar is claim ed to be a m eeting devoted m ore to putting
questions not answ ered yet than to rep o rtin g results. T his spirit usually does not really
dom inate in introductory p resentations and posters but used to be rem arkable in d iscu s­
sions. L et us looking forw ard to im pulses th at w ill arise this tim e from hours spent in wellknown hall the door o f w hich opens to the lakeside.
I w ould like to th an k in advance to all w ho co ntributed to the eighth R ecent T rends in
any respect.
L uděk Frank
T!ACU3[[n¡A¡ I pUB
q ‘TJBUIOQ-ia [\ ]/\[ ‘X^SAOUBUIO^ \
M m o D N O H X D a ia o n i n n v d s a o í i s n s m o t d i x v x s o h x o 3 3 3 a t i í m
V pire o_n;qsiiun>[ •§ ‘^jqno a
H a
a a z A a v N Y ycmaiiAi iv o r a a N ia A D a a n v n v d a H i
S3Hd a put: ysqtups s ‘ipsjirj 7, H ‘Joqosnny [\j
N o i i m o s a y Hod h o d a i a a v s x m o m o a a iN r a d h u m x iN n d n i n n v d s
S3T[d 3 PUR -13[>[U!A\ a ^sqosnRy w ‘qDSjnq x y ‘ysqtups S
peuiOQ-jg 'j/\i ¡/\[ put; BA0i3[p/q\ \ '>[cm.ij q
JO|pv XI pire so>[ut![Ai 'p\[
puO upJEpM
33I30Hd lAIVSa 3HX 30 NO IX V inO lV D
S N O ixvîiH aav X N S W N o n v s m
BAODuaq ^iiuinqog
BAoousq Epuinqog
SuiiMog f '\Tput; pray f [ j
oJuri[Aj oug
S 1 N 3 IN 0 3
I. M iillerová and L. F rank
M iroslav H oráček
F R E E E LE C T R O N S ( ‘E L E C T R O N A N T IB U N C H IN G ’)
H. K iesel, A. R enz and F. H asselbach
P. S onnentag and F. H asselbach
H. W ittel and F. H asselbach
R. A utrata and P. S chauer
R. A utrata, J. Jirák, V. R om anovský and J. Špinka
R. A utrata, J. Jirá k and J. Spinka
J. Jirák, R. A utrata and J. Špinka
L. S chneider and J. Jirák
R adek D rnovský
V. R om anovský, V. N eděla and R. A utrata
3 s u 3i [ u o ip s - g puB p>[J 3 i/\[ pv M aqosg
w aadx aoa
H M io a r a o a a o x x a i :3 iiv is ()> J i:ia a a
n v
a o N O iiv s iM id O
q o e d s u a ijy >1 p u e xjb¡^[ q ^
S H i a a a o N i a o H d a o N o s r a v d í v o o v : i D a a a a >i>J3>i
I V D I X d O - O ia N D V I M Q N V A dO D SO dD IIA I N 0 a i D 3 3 3 N 0 I S S M 3 0 1 0 H d
UU13UIJ9JS3^ f put! UURUIDULinj Q
‘ss u s q u o q o s 0 ‘p j u q u ja g y 'u n q i a r / i q
o n y
- s d x ‘S A Y d d v H H i i i i A j g - c n a u a o
d n í c í >j v i i >j
PM M aqosg jaj
S N o a iD a ia o iO H d -s d a
:>j 3 s a ' i v n v a i a i a
d n id v w i
v O N is n w a a a c i a y a n u A o a a N a
h iim
s n n s s a
i n i
uuB U iiiq o say -]/\[
put; ja rre g '[a] ‘j a p p u q o g
j ‘>(nÂSK.i>[ - y \ r a u s p o V ‘s s u a q u o q o s 'O
A dO D S O dD IP V N O H I D 3 1 3
N O IS IIA ia O lO H d D N I S n S 3 S S 3 D O d d a N 0 3 3 S 0 N V N - a f l S D N i o v m
p s s q s jn q ^ f u re fu y
IM3S 3 H I a o a S 1 N 3 1 M H D V L 1 V S N 3 3 1 3 N O VÍM lN 3 N V lV > J 3 d
BAOjBjncQ -g put; u p d a jg y
IM 3 1 3 0 V L 3 O A A \ 0 3 N I S3CIO W O N IN N V D S
>(Ut![0 >|
pue SuopQ y
S S d O D S O a D IIM .M 0 3 1 D 3 3 3 A 3 a N 3 I d d - d 3 S n 3 H O N
o u q n ^ cl p u ^ pt’3>| H a 'BAOUEJBg y
a d O D S o a o m N o a i o s i s o n i n n v d s 3 0 v l3 0 A -m 0 3 v a o a a o i o s a a o D
N o i i v i i a a a v 3 11 v i s o> j i d 3 3 3 n v a o n o j i v 3 í i n i s H a i n d w o D
put: ssu a q u o q o g
S a 3 3 I d S N 3 3 G N V N O I1 V H 3 3 3 3 D V ! N 3 a N 3 d 3 a 3 IM U O N i s n N O i i o a a í i O D N o u v a a a a v a v o i a s H d s
o n v
D iiv w o a H D
BIGJiny y put; _tont:qn^ fj
iN a iA ia a a s v a íM v a i o s d s a o N a o s a N iiA in a o a o H iv D
JBlttpi 0 putí t!|? p 3 N ^ ‘A^SAOUEUICry
w a s a v iN a iM N o a iA N a n i
SN 0>J1 3 3 3 3 3 V N 0 I S 3 0 N 0 U D 3 1 3 C I G N V N O IS 3 3 d S IC l
Eric Munro
M unro’s Electron Beam Software, Ltd., 14 Cornwall Gardens, London SW7 4AN, England
e-mail: m [email protected] com
1. Introduction - Concept and A dvantages o f Direct Ray-Tracing
For simulation o f electron and ion beam systems, direct ray-tracing offers an
attractive alternative to the more conventional method o f paraxial ray-tracing and aberration
integrals. Direct ray-tracing is similar in concept to methods used in conventional light optical
design, and it offers great conceptual simplicity, generality and robustness. In order to make
direct ray-tracing w ork successfully in electron optics, the key requirem ent is to obtain
sufficiently accurate and self-consistent values o f the electric and magnetic fields in the lenses
and deflectors, for the ray-tracing to be accurate and meaningful. This is achieved in the
present w ork by representing the field distributions analytically using series o f Hermite
Direct ray-tracing in electron optics has several key advantages. Entire bunches o f
particles can be ray-traced simultaneously, so Coulomb interaction effects can easily be
simulated. M ultipole lenses can be handled as easily as round lenses. H igher-rank aberrations
are handled automatically. Trajectories are com puted as functions o f time, w ithout restrictions
on ray slopes or directions, so cathode lenses, electron mirrors, ion traps, etc, are handled
straightforwardly, w ithout special treatments. Effects o f asym metry errors, which simply
introduce additional parasitic multipole field components, can easily be evaluated. In the
following sections, the method is sum marized and an illustrative example is presented.
2. Analytic Representation of the Electric and M agnetic Fields Using Hermite Series
The electric and magnetic fields o f each optical elem ent can in general fye expressed as
a sum o f round lens fields plus multipole components, each o f w hich are characterized by
a set o f axial fie ld functions (round lens, dipole, quadrupole, hexapole field function, etc.).
The axial field functions for each optical elem ent are first computed numerically at discrete
points along the optical axis, using any convenient numerical method. For example, Figure
1(a) and 1(b) shows an electrostatic octopole lens, excited as a quadrupole, and Figure 1(c)
shows the quadrupole field function, computed using a 3D finite difference method.
Figure 1: (a) A n octopole lens excited as a quadrupole; (b) Side view ; (c) C om puted quadrupole fie ld function.
A p s o p p s j j i j s q UB3 s u o i p u t y jb ix b XuB[^[ 'u o i p u t y p s j j i ; p u B u o i j s u n j p s j n d u i o a s q j u s s M j s q
JUSUISS j S b S q j JS S O p S q j ‘(/Y JO 3n]BA S q j JS3JB ] 3 q j ’S*l) JIJ S q j UI U35JBJ SUU3J SJOUI S q x
( 9)
:B |nouoj s q j Suisn p s j B p s p s sjb
‘s j u s p i j js o s s q x
(x )mHmDr¿ = (x ) f
:snqj ‘jas sisBq s s u s s s j i u u s j j sq j j o s u u s j / y Suisn
u o ijsu n j sq j jij u sq j
(z)Q ‘uoijjunf ppij /Dtxv pamduioo /(jjvouswnu y :f ajnSij
• q jp iM -Jp q SJI JO 3JnSB3UI 3AlJBJU3S3jdsJ B SI D pUB p p i j S qj JO >[B3(J [BJJU33 Sqj
:x ‘3|qBUBA SS3|U0ISU3UIip B 3UIJ3p JSJIJ 3M '(£ S j n S lj 33S) ‘(.z)Q ‘u o jjs u n j p[3IJ [BIXB UB JIJ OX
A jjsdojd s iq x 'sssuB jsip | bixb s S jbi jb ojsz oj o S | | b Asqj jBqj jd s s x s ‘s s u s s j s u n o j
su o ijsu ty sisBq sq j oj snoS opuB sjb suoijounj sisBq 3jiuu3]-[ s s s q j jBqj U33s s q ub3 j|
' ( ç p 9S u d
a jD O tu d A
‘0 1 + 3%u m
jv ju o z u o ff)
su o iio u n f s is v q
d iiw u d jq jifS jd
isai J
d ifj J o sifdüÀQ
b jo
:ç d jn S ij
01- 01+
~ \í
•z s j n S i j ui UA\oqs s jb ‘( x y i f - (x)oi¡ ‘s u o ijs u rg sisB q s j í u u s j j j q S p j s j i j s q j j o sqdB jQ
-V7{(* ) z~’
i i - fA - (*) V zf\= (y) ’
‘( z:ri - ) dX3Sío-li= (r) 0!/
:sBjnuuoj u o j s j n s s j 8 uimo [|oj s q j S u isn X[p;dBj p s jn d u io s s q ubs s u o ijs u n j J s p jo -q S iq s q x
(“ = <") ll
. .. .
( „ * * ) 0} = XP{X)
(2 )
^ j B u u o u o q j j o ,, p s ||B 3 -o s ) p s z i p u u o u puB p u o S o q j jo q jo q s jb ip i q w ‘s u o ijs u n j sisB q j o js s
3 js |d u io s B u ijo j s u o ij3 u iy s j í u i j s j j s q x u aauSap J o ¡v iiu o u X /o d d j i u u d j j s q j si ( x ) " // sjsqA \
( i ) ■■■ ‘
i ‘o = " j q j
‘( ^ T ~ f X3(y) “t f
= ( y)V
: sb p s u i j s p
sjb suoijsuty s jíu u s jj sq x
'suoiiounf sisvq djiuuaf{ j o ssu ss b Suisn ‘uo p o u n f oiiÁjnun
¡Doxa üb qjiM psjji) mou si ((a) [ sjn S ij u; jBqj sb qsns) uoijsunj p p ij [BIXB qsBg
M= 2
with relatively few terms o f the series. This is shown in Figure 4, where the axial function o f
Figure 1 is fitted with an increasing num ber o f term s o f the Hermite basis set. In this example,
a good fit is obtained with M - 8 .
M= 4
Y /
F igure 4: F ittin g the axial fu n c tio n o f Figure 1 using increasing num bers o f H erm ite basis functions.
In this way, each axial function is represented by an exact analytic function. Since
these functions are also analytically differentiable, the fields at any off-axis point can now be
com puted to arbitrary precision by power series expansions. The resulting fields are exact
solutions o f L aplace’s equation at every point, and are ideal for use in direct ray-tracing.
3. A ccurate Direct Ray-Tracing, Including the Effects o f Discrete Coulomb Interactions
D irect ray-tracing through these analytic fields is performed using a fifth-order RungeKutta method, with adaptive step size error control. A self-consistency o f order o f
1 picom etre over an axial range o f 1 metre can easily be achieved. This accuracy is ample for
use as an electron optical simulation and design tool. Electron trajectories are computed with
tim e as independent variable, so cathode lenses and electron mirrors can be handled as easily
as ordinary lenses, and aberrations o f all orders are autom atically included in the simulation.
Complete bunches o f particles can be ray-traced simultaneously. For complete bunches, by
adding the inter-particle fields (Figure 5) to the external fields, the combined effects o f
Coulomb interactions and lens aberrations can be sim ulated in a consistent and rigorous way.
The Coulom b forces can be computed, on each time step, using a hierarchical tree code
algorithm, that computes the forces betw een N particles in a time proportional to N log 2 N.
Spot diagrams o f the blur and distortion at the final image plane can be plotted out (Figure 6 ).
F igure 5: D irect ray-trace fo r a bunch o fp a rticles, including electric fie ld s
Ea due
to the Coulom b interactions.
Figure 6: Spot diagram s with (a) System atic initial conditions; (b) Random initial conditions.
dpouiDdp duo J o i u d m u 2 i j D s i m - x uni q q [ ifjiM. ‘jiouuno uivaq y u q q ç ¡ ü sijnsdu S uipuodsduuoj (f) p u v ( d )
lUdAAYio y u q q ç i d sjjnsdu S w p u o d sd ju o j (p) p u v (o) luduuno luvdq m o j j v ‘d u v jd dSviui jv suinuSojsiy
(q) p u D duv¡d dSvm i i d i u d j S d j p jo d g (v) sasudj djodrupvnb u n o f fo uiaisús v u o f sjjnsd}¡
d jn S ij
s u o iiD iA d Q
SÏ.: I S S T s S
Eis' i íiis s T S
■ssupuiuiXsE s p o j p s p puB suoipE JSiui q u io jn o j j o p s j j s psuiqiuoo sq i SMoqs snqj 3[duiEX3
jEuij siq j ís s p o a p s p aiodiijnm sq j j o su o j o juauiuSqB S íai-r rnrl 0 0 1 B
‘V u 00S J ° jus-uno
guies sqj JB sj¡ns3j Moqs (j)¿ puB (s)¡_ s i n S i j -junoDOB oju; suoipB jgjui q u io jn o ^ S u ;>[bj
‘y u 00S J ° lu s iin o uiEsq b joj sijn sa j g u ip u o d ssjjo o M oqs (p)/, puE (o)¿ a jn S ij -jus-uno uiEsq
(q)/. s j n S i j puB ‘3UB|d sSblui sq j ie sdEqs jods sq j SMoqs (b)/, s j n S i j 'pBJUi z j o 3|SuE-jiEq
uiBsq e qjiM jo d s sjBnbs arrl ç b SuiSeuii
o i j o X Sjsus uiEsq e qjiM ‘sssu s) sjodrupE nb
jn o j S uiuiejuoo ‘uisjsAs SuiSeuii 3]odpinui b jo j sjjnssa SApEijsrqi; SMoqs ¿ s j n S i j
3 |d u iB X a 3 A l j i u j s n m
F. H. R ead1* and N. J. Bowring 2
'D epartm ent o f Physics and Astronomy, U niversity o f M anchester, M anchester M l 3 9PL, UK
2Faculty o f Information and Engineering Systems, Leeds M etropolitan University,
Leeds LSI 3HE, UK
*e-mail address: [email protected]
The effects o f stochastic electron-electron interactions in beams have been the subject
o f many theoretical and computational studies (see for example the reviews [1-3]). M any
different approxim ations have been used. In the present study we have tried to avoid all
approximations, except the uncertainties that are inherent in a computational M onte-Carlo
calculation. As a first example we have applied our technique to the energy spreading in
a converging round beam, first studied carefully by Loeffler [4],
The strategy that we use is as follows:
(1) A converging beam with the required angle o f convergence and waist diam eter is
simulated using a large number N o f random ly selected trajectories. At a later stage o f the
strategy the num ber jV will be extrapolated to infinity (using a power series in 1IN). The beam
exists for a distance s on either side o f the waist. This distance will later be extrapolated to
(2) Each trajectory in the beam is integrated in small steps (the length o f which will later be
extrapolated to zero) using a specially developed computer program that automatically takes
account o f collective (i.e. non-stochastic) space-charge spreading.
(3) At each step o f each trajectory the average velocity is taken as the velocity o f an electron
e/ that might be coulomb scattered by the other electrons in the beam.
(4) The program searches for the nearest other trajectory and uses its velocity as
a representative velocity for the nearest electron
in the beam.
(5) The velocities o f e¡ and e¡ are used to establish their relative velocity urei in the centre o f
mass system and then the duration dt o f the step is used to define a path length / = ure¡dt.
( 6 ) Using the fact that the transverse spatial probability distribution o f the electrons is an
invertible distribution [5] it can be shown that the transverse distance r to the nearest electron
is given by
r = ( - ln(l - R ) / m l ) '12
w here R is a random num ber from 0 to 1 and n is the density o f electrons in the beam.
(7) The angle 6 by which electron e¡ is scattered is then given by the well-known expression
for the Rutherford scattering o f a non-relativistic particle
tan(# 12) = K /(mburel2)
where the effective im pact param eter b is r/2 and the effective force constant K is e / 4nso.
S17I (6 9 6 1 ) L I ’S ^ q d 'S u y 'J Z ‘» U í 3 0 l H " S
'0 6 6 1 'S ¿ q d 'u o j j o s jg - u o jjo s ia -A py o j 0 c j d d n s ‘(u o p u o q
'0 0 0 c ^3SJ3f M3J^ ‘I[BH 33IJU 3JJ 'p3 pU£ ‘SOISÁqj JOJ SpOqj3p\[ (ESU SU m ^ :BI3JBQ " J 'y
pUB >[JO^ M3Js[ ‘SS3ÍJ 3lU I3pB 0y) SUIB3Q S p iJJ E J UI SUOIJOBJ3JUJ qUIOjnO^ :U3ZUE]-[
•6861 ‘u o p u o g
MSfvi ‘SS3JJ DiuispBoy ‘so ijd o uojjoajg jo s s jd p u u j :jadsE^j g ‘S35jmbj-[ 'M 'j
s X q j 'u o j p s j g 'u o j p s j g 'A p y o j 3 £ j 'jd d n g ‘( u o p u o q p u s >[JO^ M3|\[ ‘s s s j j o iu is p s o y )
SLP 'd ‘O 'JOA ('p a ‘j s i jd s s -y ) s o ijd o a p i i r e j p 3 á re q 3 p s q d d y :jq s d s "a ‘s s o y h
o q j jB p o ju o so id o q
j [ im
[ i ] jq o d g puB s s o y Xq X pjB jnooB puB ^ i[|ty 3 JB 3
[i?] j s y j s o g Aq Xq p o u ijo p uiojsX s o qj ui SuipBOjds ASjouo o q j
jo j
s jo u i
psjBOjj puB
p su iB jq o s jjn s a j o q x
pOUOlJUOUI SB pSJBlodBJJXS Oq UBO SjpSSJ 3qj JBqj OS pSUBA 3JB SjapUIBJBd Jtldui SLJ^L ( l i )
•§mpB3Jds jBijEds pue XSwus j o SUIJ3J ui pszusjoBJBqa sjb
ss S bjs jn o SuiXjjbo - i s y y ( o i )
uiB aq [b u ij s q j j o s s ija s d o id s q j s a u o p s f B jj s q j ||B jo j (6 ) o j (£ )
'X3j3U3 puB u o ijis o d ‘UOIJOSJíp [Buy
SJI 3AI§ OJ Xl|BIJU3nbSS pappB OJE d s js qOBO JB ¡3 UOJJOOp JO X jp o p A Oqj UI SSSuEip aiJX (6)
•uiojsÁs ÁJOjBJoqBi o q j ui X S jsu s puB u o ijo a jip
i3Jn u o ijo o jip o qj jnoqB é 3[Sub Suuojjbos u io p u E j b puB q ojSub o q x (g)
p u ij o j p s s n u s q j sjb
Z00Z spuaux IU933X
Bohumila Lencova
Institute o f Scientific Instruments AS CR, Brno, Czech Republic, e-mail: [email protected]
The first order finite elem ent method (FOFEM ) is often used for electron optical
computations because o f its ability to handle local dependence o f material properties and
complicated geom etry o f the electron optical elements. We use a topologically regular mesh
based on irregular quadrilaterals. In order to avoid local computation error, we prefer to use
meshes with a sm oothly graded mesh step. It is possible to obtain accurate solution o f most
2D problems that arise in electron optics [1-3]. In order to give an idea o f computation speed
with the FOFEM , in a mesh with 200000 points we can obtain the potential distribution in
less than 1 minute on a 1 GHz PC. The previously used mesh sizes had mostly a few thousand
points; such a computation needs ju st a fraction o f a second. For meshes with n points in the
radial direction and m points in the z direction the computation time is proportional to n2m.
The computation needs below 32 MB RAM. An im portant practical aspect o f each
computation program is the user interface. O ur standard user interface is a DOS program used
since 1990 [4].
We have shown that the potential o f multipole fields (m agnetic deflectors and
multipoles made o f saddle and toroidal coils in the presence o f rotationally symmetric
magnetic materials and electrostatic m ultipoles defined on cut cylindrical surfaces) are
obtained with more than sufficient accuracy for all needs in electron optics [1]. It w ould be
nice to know also the accuracy o f the results o f the rotationally symmetric lenses. Because o f
the short com putation tim es and low mem ory requirem ents we have introduced a simple and
practical method o f the estim ate o f accuracy o f the computed potential. The advantage o f this
method is that it provides an estim ate o f the error at every point o f the mesh for given
electrostatic or magnetic lens. In the past the accuracy has been estimated only for points
lying on the optical axis [5] (except for [6 ]). The axial values do not help in identifying the
sources o f errors. For this we need the display o f error in two dimensions. These errors can be
quite small, and so it is mostly im possible to identify them from the display o f equipotentials.
If w e m anage to produce equipotentials o f the error o f the potential, we can identify sources
o f these errors or we can try to find a procedure how to get rid o f them. In order to obtain
a difference o f potential computed in tw o different fine m eshes^even in the same coarse
mesh, we w ould require an accurate 2D interpolation in the potential in a finer mesh to get
potential at all points o f the low er density mesh. Even if we have such a method, the ZRP
m ethod [6 ], at our disposal, such com putations w ould be tim e consuming.
We have m anaged to apply a method that is ju st a simple extension o f the existing
software package. It is based on comparing the results o f the computation with a given mesh
and with another fine mesh using tw ice the number o f mesh points in each direction in such
a w ay that the points existing in the low er density mesh are present also in the finer mesh - so
it is possible to compare the values o f potential obtained at the points o f the lower density
mesh. The num ber o f points is easy to double if we use constant mesh step but also if we use
a graded mesh. The difference o f the potential at each mesh point o f the lower density mesh
gives an estim ate o f the com putation error, and so we can easily visualize the sources of
errors, if we display this potential difference in the same w ay as w e display the equipotentials.
Even if we cannot display the equipotentials o f the FEM computations in meshes with more
'¿ se
(6661) LZW S3ÎI 's^Md 'RJ3W Jisui -pnjs[ ‘IBuiçiz T ‘pesy H 'J ‘?A03usq -g ‘ouqrQ -q
'££! (£661) P\OZ 3IdS ‘qWEg '3 [
6661 uojsog ‘JSMnj->[ ‘ssijdo spijjBd psSjEqs ui poqjsui jusuisp sjiuy sqx :pssqsjnq){ y
'9S (0661) 86ZV 's s y 'sXqd ui 'qjsjAI ijsui -p n ^ ‘eaosusx -g
'62£ (6661) LZPY 's s y sXqj ui -qjspM 'flsui -p n ^ ‘EAOOUsq -g
'061 (S66 I) £9£V 'ssy 's^qd ui MiaiAt 'flsui pn N ‘b ao su si -g
'60PP (Z.66 I) 89 'UJruisui -p s a s ^ ‘susq -jai ‘eaosusx 'g
'.10x 13 uoijE jnduios sq j jo j js s jjo s oj XqEnjusAS
puB uS issp s q j j o sjJEd [e sijijs szqEnsiA oj sn SMOjp XjjsuiuiXs u o ije jo j j o sixe sq j uo
sjuiod sq j je j o jj s sq j j o XBjdsip sq j ‘X [jeiiuiis -jo jjs sq j j3M 0| oj psSuEqs sq pjnoqs Xjisusp
jo X jjsuiosS qsarn sq j sjsqA \ q ssu i sq j ut sssE |d sq j oj sjuiod X[jBsp ji ‘s jn js id psjB sqduios e
sjin b sa iS UE3 [Bijusjod j o jj s sq j j o jsjsBJBqs sq j j i u s a jj -suoijEjnduios jBSijSBjd ui ss n jo j
juB jjoduii si p o q p u i s q x 'j o jj s uoijE jnduios sq j j o JsjsBJBqs sq j j o sje u u js s sq j jo j siqEjm s
3JE qSiqM ‘SSjdulEXS JS3J [BJSASS UO p3JBJJSU0Ul3p 3q [JIM p0qj3lU AV3U sq j JO 3Sn 3 q x
• p p i j s q j j o 3I1JBA JUEJSU03 e ’S ’3 ‘u o ijip u o s X jB punoq js j q s u i Q
X jp s d s 3AV s js q M j s p j o q s q j u o jo 3 p o jj3 3 [3 s q j u o o j s z o j [Bnbs si j o j j s u o ijE jn d u io s s q j
jE qj p u i u i s j sn j s q '( s u s [ s q j j o jjE d ju b s ijiu 3 is e ui d s js ju e js u o s 'S 's ) q s s u i s[d u iis b j o j puB
s s ijjs u io s S js s j 3 |d u iis j o j X juo p u g UB3 SM u j u s u o d x s s q j ‘X p jE u n jjo ju f i 'il7 '£ ^ q Zl\=u J0J
\=u j o j ‘£/t? j o j s e j b Xq p s q d iJ in u i
(z=u) s s u s p u s d s p sijB jpE nb j o j ’S 'i ‘ ( l - „£)/„£ o j |E n b s js q jE j si j o j j s |Enj3E
puE 0L'Z ^ q Z/Z=u J0J "Z J ° JOJ3EJ e Xq ( s s u s p u s d s p jE S u q )
s q p |n o q s ji
s q j u s q j ‘ai/ o j [E u o ijjo d o jd j o j j s s q j j o JsjsB JB qs s q j SEq z/H PUE H d s js qjiM u o ijE jn d u io s
s q j u s s M js q s s u s j s j j i p s q j j j '(/£ d s js qjiM q s s u i b s s n u e s s m ‘s S je j o o j s q p |n o M q s s u i
s q j j | ‘j o > /! / d s js qjiM u o ijB jn d u io s Jsq jo u E u io jj uiB jqo u e s s a \ q siq M ‘j o j j s s q j j o jsjS E JE qs
s q j u o s p u s d s p js m s u b s q x '[ ¿ ] s ju io d j o js q r a n u s j iu y u i u e qjiM jEqj u i o j j u o ijB jn d u io s
usA ië s q j UJOJj 33USJSJJIP s q j ’S 'i ‘i/ d s js qjiM u o ijB jn d u jo s u s a iS s q j j o
j o j j s s q j si
s 3 jb | M oq a\ou>[ o j jueav sm ‘J3A3MOH ' j / 1/ d s js s q j qjiM s u o jsq jo u E puB n d s js [BSidXj b qjiM
q s s u i s q j u i u o ijE jn d u io s s q j j o s s u s j s j j i p s q j j o j j s u e se js S s m ‘SAoqE p s q i j s s s p s y
■qSnous jjo q s s je sjuiod uoi[jim e oj s s o p j o sjuiod j o jsq u in u
uinuiixBui sq j Sum iEjuos ssq ssu i qjiM su iij uoijE jnduios sq j jEqj uibSe sssjjs sa \ qSnoqjjB
‘jss BjBp /ü s a s jo j p s u u o jjs d sq [|iav sisXjeue ue q sn s jBqj sso d d n s jou op 3
' ssj nui ut 9
punojB si q ssu i jssS jej sq j qjiA\ su iij uoijB jnduios s q x 'SA\opuiyV\ Jiq-J£ JOj pjBpuEjs Xjjiej
si ^ v y y eiAl 9S3 ‘JSASMOH 'sjuiod q ssu j 000008 qiiM uoijB jndtuos sq j jo j p \[y a gp\[ 8 ZI
jnoqE p s3u SAV ‘sju io d j o jsq u in u sq j oj (E uoijjodojd s je s ju s u is jin b s j X joujsuj sq j ssnE Ssg
sq j qjiM 3 |ij jndjno sq j 3J(eui osjb 3 ^ 's s e jjs ju i js s n suies sq j qjiM s s u s js jjip siqj ÀB[dsip
u sq j u e s S;\\ 's jo js q p sjn d u io s su o sq j puE q ssu i js s jb o s sq j j o sjuiod sq j je (Bijusjod m su
sq j j o 33U3JSJJ5P sq j jsn f suib juos 3 |ij jndjno [Buoijipps m su sq j p u s ‘su o p si u o ijs sjip qsES
ui q ssu i js s u s p S3U11J o a\j qjiA\ uoijB jnduios sq j u s q x 'p sjo js s je sixb sq j uo p js g puE [Eijusjod
puB sju io d q ssu i sq j je ¡Bijusjod sq j puE q ssu i u sa iS ui jsjij p s u u o jjs d si uoijE jnduios
pjBpuEjs ‘js s j XsBjnssB qjiM uoijdo sq j j s s p s sa \ j i [|euis X jsa XnBnjsB sje s§B>[SEd uiEjSojd
j o jJBd uoijE jnduioo sq j j o suoijBS¡j¡poui sq j jEqj si p o q jsu i siqj j o s§BjuBApB s q x
•qssui Xjisusp J3AVOJ sq j j o sjuiod sq j jb Xjuo ua\ ou >[ si ssnB ssq psÂE[dsip
j E iju s jo d
jo jjs ,,
sq j ‘uiEjSojd
sse jjsju i
ju s jjn s sq j qjiM sjuiod q ssu j Q0000Z UEqï
Bohumila Lencovä
Institute o f Scientific Instruments AS CR, Brno, Czech Republic, e-mail: [email protected] isibm
A new technique o f boundary elem ent method was proposed by Kasper in [1] under
the title “An advanced boundary elem ent m ethod for the calculation o f magnetic lenses”; see
also the recent volume o f AIEP [2]. W hat is new here? Obviously, for unsaturated lenses
some o f the formulas look a bit different than in [3] and the table in [1] is also different from
that in [3].
A s an illustration o f the pow er o f the m ethod an open lens according to Shao and Lin
[4] is studied (this paper is actually not cited in [1] but it is given in [2], p. 228). Further, it is
clearly stated that such lenses cannot be computed with FEM because “it is extremely difficult
to construct a mesh system o f 500x500 in such a w ay that the far-reaching field is included
adequately” . This contradicts our discussion in [5] cited by Kasper in [2]! A nother K asper’s
extra requirem ent is that all the sharp corners o f magnetic circuit must be rounded (which
“does not cause an essential deviation from reality”). As the main reason it is given that at
these sharp com ers one must observe “saturation which is confined to a very narrow domain”,
an effect that is mostly not observed in the practical com putations o f unsaturated lenses.
W hat the author m eans by the difficulty to calculate magnetic lenses with the finite
elem ent m ethod in large meshes? He even cites our paper on the combination o f BEM and
FEM [6 ] (this program is unfortunately gone with the m ainframe com puter 10 years ago).
M oreover, w e have shown in [5] that it is possible to calculate open lenses successfully; then
we had maximum 32000 mesh points at our disposal. Fig. 1 (lens geometry and fluxlines) and
Fig. 2 (fine m esh, iron saturation and axial flux density in the polepiece region) show that the
problem can be analyzed already with some 80000 mesh-points. W e have put the border 1 m
away from the lens w ith 1.2 mm inner radius and 1.2 mm snout thickness; thep 98.74 % o f
excitation A -tum s are inside the border. If we apply Neumann boundary condition, we get
100.02 % o f excitation, and the maximum B(z) is higher only by 0.8 % (0.20278 T instead o f
0.20261 T). W hat is the accuracy o f the results? If we calculate with 30 points in the inner
radius, we get maximum B(z) o f 0.20427 T, and with 60 mesh points in the gap (the mesh has
then alm ost 800000 points) we get 0.20433 T (see more on our method o f accuracy estimate
and maxim um mesh size in these proceedings). It is thus shown that the results discussed in
[1] can easily be obtained with standard first order finite elem ent m ethod [7], which can easily
compute meshes with up to one million points.
The other example is a simple saturated asym metric objective lens studied by Munro
(see figs. 17-21 in [ 8 ]); it was used as a test lens in a diploma thesis by Zeh in 1986 in
Tübingen, see [2], The saturation phenom ena are in the “advanced method” taken into
account by including only the tangential component o f field H to estimate the saturation, and
the behavior o f the m agnetization inside the yoke is thus ignored. This is not physically
feasible and the results also show that the error maxim um B(z) at 15000 A t is 10 %. The non­
advanced BEM methods previously used w ere correct but rejected because o f the
complications in computed iron saturation. A citation from [1]: “The price to be paid for their
omission (o f nonlinear space currents) is that the field in ferromagnetic yoke itself cannot be
calculated correctly, but this drawback is not serious as long as the field o f the optically
relevant domain can be approxim ated fairly w ell.” In most objective lenses two or even more
•¿ 6 6 1 s s s j j 3 ^ 3
-(-pa ‘j j o p o
f) s o p d o
B}dt7 - ^duia^e pug - ť0V>l
E0+3Ï00000 E :U0T^eT.t3Xs le^ol
a p p JB j psSjBq 3
' 6Z£ (6661) LZPV
: u ' ‘o i u n p v ' 3 f e ]
's^ q d MPPM '-»sui - p n ^ ‘e a o o u s i g [/.]
‘BAOsusq -g [9 ]
'Li (t?8 6 l) 89 W ^ O ‘a u s i
'Ç81 (S 6 6 I) 6 Z.I -0SO J3M f ‘BA O susg g [s]
> £ ť £ (6861) 09 '-nsui P S
‘« H CTS'd ‘°®HS Z W
•£ (¿861 ) LL >Ipdo ‘l a d s r s -3 [£]
’1 0 0 3 S S 3 J J s i u i s p B s y ' 9 i x ' s A q j u o j p s j g 'S e i u j 'A p y : u i ‘j s d s B ^ j ' g [ j ]
'£¿1 (0003) OSťV ^
sA qj -qpjAj ujsui -pnM M sdsB^
[ l]
S33U 3J3J3>]
p o q j s u i s jB u x ix o j d d B A u b
33 s s E j d s i
I jn j e o i puB ÀOBjnaDB j u s p g j n s q jiA \ p s j n d u i o s s q u e s s u s ] u s d o X ju o i o n 'PM
o j s s u B q s o u SBq [ j ‘ i ] js d s B ^ i Aq p s s s n o s i p i/\[ g
s q j jB q j s p n p u o s s n q j u b d s m
• s s i p B jd j o j s s n o u j o s i p o q p u i psouB A pB s q j s j s q
o § - p s iB o q d u a o D s j i n b s i m s H E d
u o ijb jih e s
s q i puB p s s n
s jb
s ijs u
Seui ju s js jjip
D avid D vořák
Institute o f Scientific Instrum ents AS CR, Brno, Czech Republic,
e-mail [email protected] isibm
M isalignm ent aberrations arise from (small) imperfections o f lens geometry - either
because o f the inaccuracy in the m anufacturing o f individual elem ents (like electrodes and/or
pole-piece defects) or the misalignm ent o f the optical elem ents in the system (tilted or shifted
element). U nder such circum stances the ideal sym metry is disturbed and additional parasitic
fields are produced. In the effect they are producing additional image aberrations.
The first type o f errors prim arily occurs due to manufacturing faults if the surfaces are
not sufficiently planar and/or apertures are not sufficiently circular. For example, an
essentially rotationally sym metric lens showing slight ellipticity o f electrodes produces small
quadrupole field. The aim o f the com putation is to establish necessary manufacturing
The second type o f errors is caused by an incorrect assembly, when positioning the
elem ents o f a lens or m ore lenses w ith respect to one another. U nder this category come all
errors like shifting the lenses o ff axis, their tilt and in the case o f m ulti-electrode lenses
mism atching a single o f them. From analysis o f these errors we want, as a result of
computations, to improve the performance and introduce efficient correction elements.
The com putation o f aberrations starts with the analysis o f the geom etry o f the
problem. The unperturbed geom etry with basic sym metry is defined as well as the perturbed
geom etry o f the low er symmetry. The analysis proceeds in tw o steps. In the first step the
unperturbed field is found. In the second step we add the perturbation field due to the
sym metry breaking using perturbation method developed by P.A. Sturrock [3].»Having the
total field available, by solving the trajectory equations we acquire the paraxial trajectories
and the aberration coefficients. Such coefficients now include the aberrations o f the
unperturbed system as well as errors due to parasitic fields.
far we have concentrated on understanding the theory o f parasitic fields [3] as well
as on obtaining correct expression o f the aberration term s and understanding the derivation o f
aberration coefficients [1], N ow we are trying to compute the parisitic multipole fields in
misaligned electrostatic lenses using program packages ELD for the unperturbed field and
EMD for the perturbation fields. The influence o f these fields on the image is analysed by the
accurate ray tracing with the TRC package [2]. The results o f the case numerical simulation
for a simple thin lens are shown in Fig. 1.
'/.8 e ( I S 6 I ) ÍPZ V '« S ‘u o p u o i ‘-OOS "a -SUEJI
•sojiqj •ssujsuiuiXse oi snp sssus[ u o jp s [3 oijsuSbui jo suoTjBjjsqe 9 i]x v d 5p o x m is
ss§E>ped siEAvyos sq j j o o /.'£ UOISJ3A oj u oijejuauinooQ : q >[u ijsssi ¿\\ pue g
'ZOOZ i p d y ‘o iu g i s i ‘u o d a j jB iuajui 'SDBJsqy : a >l?JOAa
[j ]
[ l]
S33U 3J3J3> I
u iu iz o ¿iq sixv X u i pd]fii{s sud] 3i/j Adifo su ou jod jd f o uotinqijjsip (p) .‘uoijuojsip
udpjo puç SMOijS su d j Duidutiwis ájjvuojívjoj Bifl ¿iq d S n u ii dijj (o) .‘sixv jvjjjdo jnoqn jv ju jd u iu iá s si /; p u v
suoijoduip ¿i p u v X ifjoq u i u iv u iQ i/jpiM-fjvii svif jodfqo jv u tS u o d ifi (q) - J 0 N J 3 I o sijnsdu+ sudj (v)
d u n S ij
(3 )
Martin Oral
Institute o f Scientific Instrum ents ASCR, Brno, Czech Republic
The beam profile provides local current density distribution o f a particle beam. From
the known distribution we can optimize the axial position o f the specimen in a SEM,
understand image aberrations or find optimum function o f electron or ion analyzers.
Graphical plot o f the profile as a function o f the z coordinate shows what is happening with
the beam when it passes through the optical system. The first attempts o f graphical
presentation o f the beam aberrations and profiles based on aberration theory are discussed in
the recent diploma thesis [ 1 ].
In this paper a simple calculation o f current density j{r) o f rotationally symmetric
beam after an electrostatic lens is shown. We consider a point object from which
a m onochromatic beam o f particles is started. W e also do not consider diffraction effects;
such a com putation would be suitable for ion beams.
Let us suppose that we know the current density distribution in some entrance plane za
before the lens. We divide the radial interval [0 ,ram), w here ram is the maximum extent o f the
beam, into m small intervals [ 0 , r a i), [ 0 ,ra2),..., [ram-\,rm) that are not necessarily equally wide.
From the current density j{za) betw een the radii rak and rak+i the current is
Kk = ’T/CO cos ( 0 ,-W X d - d J .
where 0/r) is the angle between j(r) and the optical axis. This angle is typically very small, so
we can put cosf 6j(r))= 1 .
In the observation plane z=z* behind the lens we choose another set o f dividing points
rbk in n intervals. If n is sm aller than m by an order or two in magnitude, we can assume that
the particles in given interval [r«,rj/+ i) come from several, say S, intervals in the entrance
plane za; these intervals do not have to be one next to another. So the total currgnt density in
the /-th interval in the observation plane is
* ( rH+1
^ L j^ aki.s
,s = ~ T 2 ------ IT
n(rbM- r bl)
where Rbi is the center o f the /-th interval.
In order to find the values o f the index k/_s o f the point in the observation plane, we
have to calculate the paths o f particles. We need the start positions in za and end positions in
Zb. Each interval in the entrance plane is represented by a "test particle" started from the center
o f the interval with a given angle. (In some cases we may consider several test particles with
various directions and energies). Then we obtain its coordinate in z* and thus the index / o f the
interval in the observation plane. Moreover, if zj is outside any strong electrostatic and
magnetic field, i.e. the trajectories are straight enough, w e can get the radii in planes close to
given zb.
Any accurate ray tracing program can be used to find from the start points in z 0 the end
points in Zb for the trajectories o f test particles, e.g. TRASYS in multi-trace mode [2], The ray
tracing programs are more general, because they allow computations in any combination o f
electric and m agnetic fields. In some cases we can evaluate only the paraxial properties and
aberration coefficients with programs like EPROP, PROP or COM BI [3]. These results are
l!jd V ‘o u j g DOdS ‘isnuem J 3 s n o ¿'£ s X s b jx :b a o 3 U 3 1 g
[g ]
3 0 0 3 o i u g DOdS ‘[BMIBui j s s n o ¿ £ u o i s j 3 a
‘( i q i u o 3 ‘d o j j ‘d o j d g ‘Q i g ) u o ijB jn d u io o s s p d o s p i j j E d j o j s u i e j S o j j :BAOOUsg - g
s S u i p a s o o j d s s s q j ‘s u o ijB u s q B ju 3 u iu 3 ;|b s ija i ‘> j b j o a q p j a e q
1 0 0 3 o i u g i s i / o u j g n i 3 JM d 3 d l
[ ujtí] j
B u l ó l a :u E q jQ p A B j
s o o u o ia jo y
[lillil] z
ť0 " 9 0 -
8 0-
[tu rí] j
Z l~
ť l"
[ liir t ] j
•■'z=z 3UB(d s Seuii ubisstibS
sq j oj 3 so p ssuBjd z [BJ3A3S jb sp jo jd uiBsq aqj SMoqs aangij aq x 000 000 \=M sibajsjui jo
jsquinu aqj oj jEnbs si ssuojosfBjj j o jgqum u aqj puB umí z'0 oj dn juejsuoo si ap jo jd Xjisusp
juaxino jndui aq x AZ.61'0- 3 §ej[oa b SBq uiui 3 jsjsuiEip jsu u i qjiM >piqj uiu ij sp o jjo sp
JBJJU30 a q x '( M ass) sus[ aqj uio jj uiui 9 jb pasnooj sjsa \ Asqj puB suaj aqj jo juoxj ui uiui z
p sfq o JUtod B JB pOJJBJS 3J3M Á§J3U3 Jlun LJJIM SSpiJJBd 3LJX '[t7] u! UMOqS ‘UIUI l= Q sapojjoap
sjnjjsdB uiqj jsjn o jo jsjsuiBip puB uiui 3 qjSuaj qjiM sus; DijBjsojjosp [Bijuajodiun sujsuiuiXs
Z00Z XP U»J1 IU903U
M. M ankos and D. Adler
KLA-Tencor, 160 Rio Robles, San Jose, CA, U.S.A.
Cathode objective lenses were employed in the past in a variety o f electron
microscopes, including low-energy electron microscopes [ 1 ,2 ], photoem ission electron
m icroscopes [3,4] and low-voltage secondary electron microscopes [5], Increasingly, these
lenses are finding applications in scanning electron beam tools used in the semiconductor
industry, i.e. in electron beam inspection, review and m etrology [6,7] and are being developed
for pattern generation [8 ]. The continuing trend tow ard sm aller features however poses
a formidable problem for scanning electron beam tools due to their relatively low throughput:
the num ber o f pixels to be examined or written on a w afer results in inspection and writing
times that exceed practical limits. One o f the possible approaches to circumvent this problem
is to replace the serial acquisition process o f scanning electron microscopes with a parallel
scheme, w here a two-dim ensional image is detected on a scintillating screen and further
processed on a computer.
In electron lithography, a certain num ber o f electrons must be deposited in each pixel
at the substrate to expose the resist. Similarly, in w afer inspection, a certain number o f
electrons must be accum ulated in each sensor pixel to provide a useful signal with a sufficient
signal-to-noise ratio. The throughput is determ ined by the tim e required to deliver this
electron dose, so it is proportional to the maximum total electron current. However, the large
current required delivering the needed throughput results in increased electron-electron (e-e)
interactions, which blur the image and result in a loss o f resolution.
We have used the IMAGE software package from MEBS Ltd. [9] in our e-e
interactions calculations. The software computes electron-optical properties by propagating
bunches o f particles through realistic electrom agnetic fields using M onte Carlo simulation o f
the trajectories through the accelerating region and thick lens fields. The software can predict
the com bined effects o f the aberrations (geom etrical and chromatic) and the e-e interactions in
a cathode objective lens, in a rigorous unified way. In particular, this allows calculating
accurately the im pact o f electron-electron interactions in the accelerating region, where
electrons are starting at rest at large angles o f up to 90 degrees, an<ftn the following optically
thick lens, w hich so far have been neglected or treated by approximations only.
In a cathode objective lens, electrons are generated by illuminating the surface o f the
substrate with photons or low-energy electrons. Three comm only used cathode objectives
with constant accelerating field are considered: decelerating and accelerating electrostatic, and
com bined magnetic lenses. All three lenses have an accelerating field o f approximately
5 kV /m m at the substrate surface at maximum beam energy o f 20 kV. The geometry o f the
focusing electrodes and polepieces was chosen to give a practical lens design capable o f
forming a 4-5 tim es m agnified image o f reasonable size (few hundreds o f pm ) at a distance of
200 mm above the wafer. In the decelerating electrostatic lens, the electrons are slowed down
to energy o f approxim ately 4 keV, and in the accelerating electrostatic lens, the electrons are
accelerated up to energy o f approxim ately 6 6 keV. In the magnetic lens the electron energy
rem ains constant.
•suisjsXs uiEsq u o jjo s p jndqS nojqj qSiq joj uoijdo sjqisBsj b
sssusj 3Aposfqo spoqjEO sssqj jBqj spn p u oo o j sn a\0[[B s js q psjBjjsnji; ‘33p jn [q psonpui
o j psjspisu oo sq ubo uSissp sus| j o sjosdsE [BDijDEjd js q jo puB ‘pssn susj s q j j o sdXj sqj
qj|A\ X[S uojjs Xjba jou ssop 33p jrq q psjEpossB s q j puB suoijobjsjui s-s s q j ‘ p js u s S uj
•ssijisusp ju sjjn s
3-3 js3[|Buis sq j SBq su s; o ijE jso jjo sp SuijBJspooE sq j ‘sS ubj Xjisusp JU3XXT10 jsq S iq sq j
j o j 'sju sjjn o jsq S iq puB ss ijisu sp ju sjjn o js m o j je jsS jb[ s i SSBSJOSp j o s jb j s q j ‘X ]Suijssjsjui
p sjsp isu o o jou si ‘sjE jjsqns sq j jb azis p p i j sjqBjiBAB s q j jiuiq u eo qoiqM ‘suoijEjjsqE s ix b -jjo
u io jj u o ijnqujuoo sq j ís u o ijo e js ju i u o jjo s p -u o jjo s p j o jo o jjo sq j uo si snooj s q x 's u o ijo b js ju i
Xq p s u u o j sS buu pnjJiA sq j j o u o p n p s s j sq j sjsqM ‘six b p o ijd o sq j j b s u jn jq [b jo j sq j sjEnpAS
3A\ 'P3P 3P P 3JE s u o jjo s p sq j |[B jBqj os p ssn si sjnjjsdB jidnd ojsi 'A 3>1 01 1® 1X111 OH PUB
AS>I 0 Z 1® m u 0 L ¿pjB U iixojddB j o jn jq b SuiAsiqoB j o sjqBdBs sjb s s s u s | s s jq j jjb ‘sju sjjn o
lUBsq a\o [ X jsa j y ^op - _.'p = _.33p s i \3jnjBjpBtib UBissnB|) b Suisn >p ,m|q ¡b jo j sq j u io jj
sjusxino m o | X jsa jb op j r q q s q j SuijOBjjqns Xq psjE jn o p o si suoijobjsjui s-3 o j onp 33p jn jq
sq j ‘sz is p p i j puB ju sjjn o qoBS j o j 'as> [ OS
sju sjjn o uiBsq j o sS ubj b j o j Xjisusp ju s jjn s
uodn suB[d snooj UBissnBQ sq j jb jr q q uiBsq sq j j o so u sp u sd sp |BUO|jsunj s q j SA\oqs qs;qA \
‘ l 'S i j ui psjB duioo si ssdXj su sj SAijssfqo sp o q jss s s jq j sq j j o ssuB iiuojJsd sq x
(s u o jp s p o jo q d 'S 'i)
A 3 SZ'O ve ps>psd u oijn qu jsip uuBuizj|og-][SA\xB]/\[ b SBq XSjsus [bij¡ui s q j puB 'UBijjr>qiUB'|
si u o p n qu jsip Xjisusjui jBpSuB s q x
'u oijn qu jsip Xjisusp ju s jjn s u u ojiu n b qjiM ‘ sdBqs
sjBnbs b SBq bsjb uoissiuis s q x 'sju sjjn o uiBsq [b joj pus sszis p js y j o sSubj b j o j psjB|ns[B3
si jn jq uiBsq s q x 'Xjisusp ju s jjn s s q j SSSBSJSSP qoiqM ‘bsjb psjBuiuinni s q j Suisbsjoui Xq sus[
SAijssfqo spoqjB3 s q j ui ju sjjn s uiBsq usaiS b j o j pson psj sq ubs suoijobjsjui s-s s q x
'[Zl] uisjsX*^qj jo sssusj sqj Suisnoojsj
Xq Xpjsjduios psjBuiuiip sq sXbavjb jouubo snoojsp sqj ‘uoijnqujsip sSjBqo juspusdsp-suiij
puB uuojiun-uou sqj oj sriQ uoijbsijiuSbui ui sSuBqo b puB suB[d UBissnBQ sqj uioy sSbuii
sqj jo snoojsp b ui sjjnssj ‘uiBsq sqj ui Xjisusp sSjBqs ps Sbjsab sqj Xq psuiuusjsp ‘snsojsp
sSjBqs soBds sqx 'pssnpsj X[uo ‘psjosjjoo sq jouubs sjosjjs jusuissEpsip Xjojss Cbjj puB
qssjsog sqj qjog 'susj sqj jo Xjipnb SuiSbuii sqj ssjBJOusjsp Xpssjip ‘sstjpopA puB suoijisod
u ojjssp ui yiqs | ejojb[ sijsBqsojs B sb jp s ji sjssjiubui qoiqA\ ‘jssjjs (Jsy p o ']) jusuisoBidstp
Xjojosíbjj sqx 'uoijnqujuos uoijBjjsqB oijBuiojqo sqj Suisbsjoui Xq ssuBuuojjsd poijdo
-uojjosp sqj sjssjjb josjjs qssjsog sqx '[oí] snoojsp sSjEqo soBds puB (josjjs jusuisoB[dsip
Xjojosíbjj) jnjq Supnpojjui puB ‘(josjjs qosjsog) pBSjds XSjsus lUEsq sqj Suisbsjoui
Xq sus] SAijosfqo spoqjBO sqj jo ssijjsdojd SuiSbuii sqj josjjb suoijobjsjui s - s sqx ' suoijobjsjui
s-s pssEsioui ui sjjnssj qoiqA\ ‘jusjjns uiBsq jbjoj sSjbj b ssjmbsj jndqSnojqj PSSESJOUJ
Current d en sity [mA/cm 2]
Figure 1: E-e interactions induced blur as a fu n c tio n o f current density fo r varying total beam currents fo r
currents fr o m 1 /jA to 2 0 fjA at 20 k V in a decelerating a n d accelerating electrostatic, a n d com bined m agnetic
[1 ]
W. Telieps, E. Bauer, Ultramicroscopy 17 (1985) 57 .
R.M. Tromp, M.C. Reuter, Ultramicroscopy 50 (1993) 171.
G. F. Rempfer, W. P. Skoczylas, O. FI. Griffith, Ultramicroscopy 36 (1991) 196.
R.N. Watts et al., Rev. Sci. Instrum. 6 8 (1997) 3464.
I. M ullerova, Scanning 23 (2001) 379.
[6 ]
W.D. M eisburger, A.D. Brodie, A.A. Desai, J. Vac. Sci. Teclinol. B 10 (1992) 2804.
M. M iyoshi, Y. Yamazaki, T. Nagai, I. Nagahama, J. Vac. Sci. Technol. B 17 (1999)
[8 ]
M. Mankos et al., J. Vac. Sci. Technol. B 18 (2000) 3010.
E. Munro, these proceedings.
G.H. Jansen, J. Appl. Phys. 84 (1998) 4549.
M.M. M krtchyan, J.A. Liddle, S.D. Berger, L.R. Harriott, J. Appl. Phys. 78 (1995)
6888 .
M. M ankos, to be published in J. Vac. Sci. Technol.
L. Frank, I. M üllerová and M.M. El-Gomati *
Institute o f Scientific Instrum ents ASCR, Brno, Czech Republic, e-mail [email protected]
* D epartm ent o f Electronics, U niversity o f York, York, U nited Kingdom
D irect observation o f doped patterns in sem iconductor, usually on cleaved sections
through m ultilayers but recently also in plan view s o f patterned doping o f a technological
layer, is acquiring high interest because o f its straightforward application in the
sem iconductor technology. Plenty o f experimental data has been collected [1-4] from
conventional SEM observation and recently first results showed im proved contrasts attainable
with specimen immersed into electric field [5,6], M ain features are the sign o f contrast —the
p-type regions are always brighter than the n-type ones, and the contrast grows tow ard lower
Some particulars regarding the contrast interpretation and model o f the contrast
m echanism can be found in the literature and most complete summary was presented by Sealy
et al. [4], All these attempts relied upon the local differences in the ionization energy and only
recently at least the surface states were taken into account both via associated band bending
and subsurface electric field creation. Contrary to this, El-Gomati and W ells [5] were first
w ho interpreted the contrast as solely caused by the m etal-sem iconductor contacts between
the carbonaceous contam ination layer and the specimen. The carbon contam ination w as most
likely presented in all experiments announced so far.
This report aim s at sum marizing the main points indicating the im portant role o f the
surface status, the surface heterogeneity in particular, in the contrast formation. The data were
acquired solely for specimens dipped in HF, w hich is a step known to passivate the surface
against oxidation and also passivate the acceptor states. It has been found that:
A. if the p/n structure is coated with a Ni film o f about 10 nm thick, the contrast does not
change (i.e. the p-type is brighter), but
B. if the coating material is Cr, the contrast changes the sign and the n-type is brighter;
C. when the cathode lens is used and the specimen is immersed into electric field, the
contrast significantly grows but the increase is much higher under medium vacuum
conditions com pared to UHV;
D. after in-situ rem oving the surface layer by ions, one gets the brightness lowered for both
p- and n-type.
Hypothetical explanation o f all these phenom ena can be based on formation o f surface
potential barriers, including those connected with metal-sem iconductor contacts upon the
specimen surface. Let us suppose that after HF dipping the silicon surface does not oxidize
and the energy bands are flat. Then, a m etal-sem iconductor contact forms a barrier repelling
the electrons backw ard (the Schottky contact) on n-type silicon w henever the w ork function
o f the metal is higher than that o f silicon while for the opposite relation between the work
functions, the barrier appears on the p-type [7]. This arrangement enables the observations ad
A and B to be im mediately explained - the relevant w ork function relations are as needed.
Because also the w ork function o f carbon surpasses that o f silicon, the situation is like with
the N i coating, with the carbonaceous contam ination more intensive under w orse vacuum
■1 861 >[JOA MSN ‘^31!A\ 'ssoiAsp jojonpuoDituss jo soisXqj :dz§ ']A['S [/.]
-q ‘iJEUIOQ-jg 'IAl'IAI ‘?AOJ3nt!lAl 'I [9]
’\Z6Z(\00Z) 6 L SJ3JJ31 'sX q j [d d y ‘sipM . > I 3 I ‘u b u ioq-IH H
' II £ (000Z) 6 t -3 so jo i^ UOJJ0313 f ‘A\Eqs||Av >Jd ‘ipjsB 3 y p \ [ ‘Xjeds d '3 W
T 9£ (8661) 91 a ïouqDDX P S 3BA I' ‘su q [ 0 3 3 a ‘uiBf h ‘ssiqBUDA a [£]
'IZP (9661) n S ’lo u ip sx P S '313A Ï ‘- » q m w d ‘sajqEUSA Q [z\
>01 (S 6 6 l ) 8 S
XdoDsojDiurejjin ‘3 P 3 S T ‘3ÍP3! I ' I ‘s io a e x j ‘s im o h 'V ‘IPJSBD "HTM ‘d ia o jd j a a
•p3gp3[MOU>|3E X[[nj3jEa§ si 106S90IV ou JuriS jopun >13 § y J° XousSy lutVD
h im
osje sjbjocLiodui
juB jjoduq 'ssdX j ajBJjsqns puE Suidop sq j j o suoijEuiqiuoD S uiuibuidj
jo yoddng
ojuo jDEduii j p q j Xjijd a oj puB ‘uoijEAissBd DDBjjns jnoqjiM S^JSBUI
sp ix o sq j j o JBAOU13J jQj sp o q jsu i XjjBjnDpjEd ‘sDnbiuqosj uoijEJEdsjd jsq jo
sq pjnoqs sjuD iuusdxs
puB uoijBJoqBp
pDDU p p o u i siqj j o ssjn jE sj sm o g
SDEjjns Dqj qjiM psjDDuuoD Busuiousqd Xq pDjJDAUi ‘ duios jspun ‘jo PSSBSJDUI
Xq pssnBO si jsbjjuod qSiq jou sujos X[qBqojd jsoj/\[ 'ssipnjs snoiADjd uiojj sa\ oi[oj os|B siqj sjsijjBq DDEjjns oj snp si jsbjjuod uojjdd[d XjEpuoDDS xinj jou jnq Xjijo Ieui e X[uo jBqj snoiAqo
uiEDq uoi njis-ui sqx 'p3UiB|dxD os|E si j uisji sqj ‘snqx ' uoissiuid DdXj-u Dqj sjda\0[ JDqjjry
qoiqM ‘XjiJE|od ssjsasj sqj ui pssBiq 3J3q si jdbjuod DdXj-u - ¡ejdui sqj ‘jdaodjojai 'suoijipuoD
S. Schubert, R.Y. Lutsch, M. Rauscher, D. W inkler* and E. Plies
Institut für Angewandte Physik, Universität Tübingen,
A u f der M orgenstelle 10, D-72076 Tübingen
*ICT GmbH, A m m erthalstraße 20, D-85551 Heimstetten
[email protected]
Calculations for a m iniaturised low energy high current electron optical column
consisting o f an electron emitter, tw o electrostatic lenses, a scanning system and a secondary
electron (SE) detector are presented. The assembly o f the individual elem ents is based on
conventional components. A common Schottky em itter minimodul is used as electron source.
The electrostatic lenses are made o f simple platinum apertures used in conventional electron
m icroscopy as beam lim iting apertures. The deflection elem ents are made o f printed saddle
coils based on double-copper-coated Kapton foils. The SE are detected by means o f
a modified custom ary scintillator-photom ultiplier combination.
Column design
extraction electrode
Schottky emitter
ff 400pm
condenser lens
hal diaphragm
g. i 75Q|jm
SE detector
Fig. 2: Schem atic view o f the condenser lens
objective lens
Fig. I: C olum n overview
Fig. 3: Schem atic view o f the objective lens
The experimental system (fig. 1) currently being built is about 50 mm in height. In the
interm ediate region betw een condenser (fig. 2) and objective [1] lens (fig. 3) the primary
electron beam can be varied between converging and diverging mode. This leaves enough
space to place a double deflection system consisting o f m agnetic deflection elements [2 ] as
well as a detector between the lenses (fig. 1). The SE are accelerated through the objective
lens tow ards the in-column detector. Detection efficiency is additionally increased by
applying an extraction field at the sample. The detector region between the two lenses is set to
a potential o f 6 kV relative to the sample. This high potential is used to accelerate the SE
making them detectable by a scintillator-photom ultiplier combination.
The system is optim ised for final energies o f the primary beam EPB < 1 keV, a working
distance wd = 1 mm (between last lens diaphragm and sample) and an electric field at the
sample o f 500 V/mm. The goal was to achieve a resolution o f 50 nm with a probe current
greater than 50 nA.
T I ‘1003 >ptuqsuui
‘o id o ^ s o j^ iu iu o u o jp p ig Jty Sun3E)jopuB|iojQ ‘ uiojsAs suo[ o p B jso jp o p posunjE iuiui
ui luouiojnseoui aoijnjosru joj poo ojppBS qjiM jiun SuiuuBog : jb p joqosriE>{ [¿]
'CI ‘ 1002 sptuqsuuj ‘3ido>[soj>[iuiu3uoj}>[3[g jnj SunSEjjopuBjiojQ ‘suopBoqddB
uoprqosoj qSiq 33e}|oa-mo) joj sosuoj u o jp o p opBjsojpop posiin}Biuip\ :[B p qosjnq [[]
uopoopp g§ oqi joj SB
•u opoopp uuirqoo-ui SuqqBUO sn q j ‘sopoiu q joq ui suo[
oqj q S n o jq i ssBd oj 0|q e ojb g § j o joqiunu sSjb[ A p u o p g jn s b jBqj UMOqs uooq SBq lj sopisog
•3|qBA3iq3B s; y u OS UBqj J3JB3j3 juojjno oqojd b qjiM uiu OS j o uoprqosoj b jEqj Moqs uuirqoo
p o p d o u o j p o p }uouno qSiq XSjouo ,\\o[ posunjB iuiui b jo j suoi|B[nD[BO paiuosojd o q x
■jodsB» -g J o jj jo (S0 3 ) suisjsXg p o p d o u o jp 3[g uiBjSojd
sqj qjiA\ jno poujBS ojow souopofBjj jo suopBjnuiis puB suopEinopo p p p oqx
A3 009=ad3
AS>| l- ad3
AS 009=ad3
A3 00t=ad3
A3 00^=ad3
0|oq ujB0q
ÄjBUJud u| }so| g s
SU8| 0A|P9iqO
sqj ßujssed 3S
:pouiB}qo ojb A3 01 °1 dn jo soiSjouo
Suijjbis qjiM g s Joj sopuopijjo uopoojjoo Suimojjoj oqi ‘a o Z Ie uinuiiXEiu oqj qjiA\ XSjouo
oqj joj uopnqujsip [pA\XBy\[ b sb ¡pM sb sojSub Suprejs oqi joj uopnqujsip ouïsoo b SuiuinssB
puB suo( OApoofqo oq} 3AoqB uiui oe (uiui I jopureip 3Joq) jo p o p p oqi SupBjj
•suoj OApoofqo o q j j o o p o u i
uooMpq uiBoq XjBuiud oqi jo suopipuoo Suisnooj joj papjnojBo ojom souopofBJi g§ jo p o p p
3qi SpjBAVOl A35! 9 JO ASjOUO UB Ol P3JBJ3J330B 3JB g§ 3qj SUOJ OApOOfqO 3qj SuiSSBJ o06 Ol
dn SOjSuB SuiJJBJS JE 3|duiES sqj 8uiAB3l g§ oqi SupOOJjOO sdp q ofdlUBS oqj JE p|OIJ UOpOBJJXO
o u p o p u y 'jo p o p p aqi spjEMOj suoj OApoofqo oqj qSnojqj poiBJopooB ojb g s oqj
SU 0JP3|3 \.llipuooos JO A3U3¡3IJJ3 U 0 p 3 3 ||0 3
•0|qBA3iq0B Xjisbo s; y u os UBqj jojeojS juojjno o q o jd b
qjiM uiu 0S jo uoprqosoj b uiB jqo O) jboS o q i uiui/a 00S = s|dlu,’s3 jo 0[duiBS o q j je p p y
uopoBjjxo o u p a p üb Suia\0[|b puB ‘\ s
[ = g \7 jo ([MHMJ) qipiA\ XSjouo ub ‘uiu qz = SAp
jo o z is oojnos [BnyiA b ‘pEjajs/yui s ‘o = ®I jo Xjisuojui juojjno jB|nSuB ub Suiuinssy 'suoj
- sopoui qjoq 3[iqM ‘opoui uopBJspssp jbujojui ui pojBJOdo oq o j pouinssB si sus[ josuopuoo
oqX 'A35l I PUB A35l p‘0 U00A\pq soiSjouo uiBoq joj p o ju o so jd oje suoiiB |no(E 3
SUO!4B|n3|l!3 UIB3q Xj b u i u j
M. Rauscher, R.Y. Lutsch, S. Schubert and E. Plies
Institut für A ngewandte Physik, Universität Tübingen, A u f der M orgenstelle 10, D-72076
Tübingen, Germany
email: [email protected]
In scanning electron microscopy resolution (probe size) measurements are most easily
realised by em ploying visual information, i.e. determining the smallest visible structure in
images o f a suitable test specimen. Furthermore, by use o f image information the alignment o f
the system ’s optical axis can readily be achieved.
Thus, a simple scanning and image acquisition system has been set up in order to
verify the calculated perform ance o f an experimental miniaturised electrostatic lens system
optim ised for high-resolution low-voltage applications [1], The latter consists o f a Schottky
em itter with extractor and an optim ised miniature immersion lens built o f conventional
platinum apertures (as used in electron m icroscopy) according to [2 ].
t cathode
platinum electrodes
CH Vitronit®
G titanium
I graphite
Scanning Unit
Scanning o f the beam is perform ed using a special saddle coil deflector based on
printed circuit technique [3] that is situated in the field free region between extractor and first
lens electrode (fig. 1). The saddle coil structure is transferred to a double-sided copper-coated
K apton foil by m eans o f photolithography and subsequently wound on a cylindrical former.
With this technique also more com plicated winding distributions can easily be realised.
Therefore an optim ised structure with respect to minim isation o f higher order harm onics [4]
has been investigated numerically.
Due to the limited available space within the system the overall height o f the scanning
module is kept at approxim ately 15 mm. Fig. 2 shows the individual components o f the
scanning module as well as the completed assembly. A commercial system (ADDA II , Soft
u o ijo u n f AOAAd di{i Suisn
n j p u v poifjdiu %SS~%ZI SuiM-Oifs A v q p u S sso a d d
d u ij
ïU d W d A ttS D d w
uoijnjosdtf :p
b a h S ij
p u S pd¡fiuSDp\¡ :ftf8i}f pjaS JAI3X
IS d l f o
M .3 1 A A 9 A Q 4 p J
:p u 2
uo udwpads
w fijU D n b . £ d A Y lS jJ
iuu / X
l!j s s n e o
•pspaxioo sq UE3 jsu b ] sq j souo SAOjduii j[;m pire lusijBiuSiiSB Suxuiboisj sq j Xq
p sjiu iij si u o p rq o sa j aqj £ 'S ij i u o j j siqisiA X jjB sp si sb ‘j s a s m o h -iuu g j o snjBA lEoipjosq}
sq i oj p a red u io s sS jbj j[ijs s i anjEA s iq x 'pajnsE aui u ss q SBq
[ j o X8tóua uiEsq XjEUiud e
jb iuu i £ j o snjEA [B jusuiijsdxs UE ju ssa id j y -(u o ip u n j j o j j s UEissnBQ sq j -3’i) u o ip u n j
[bdijXjbub sjB udoiddE üb Suisn jij 3JBnbs-jsBsi E Xq ao p o q p u i % 8 8 '% Z I 3H1
J s q ip
azis sqoad sq j sje iu ijss oj p ssn sq u sq j ubd p -Sij ui UMoqs sb a iy o jd Xjisusjui S u ijjn ssj s q x
•sjBq p u S sq j j o su o ssojob ubos3ui[ b SuijeinojEO X jiusnbssqns puE usuiioads sq i j o sS euii
ub Suum boB js jij Xq p s u u o jjs d s; juauiam sB siu u o ijn |o s3 j ‘S3U3J4 '3iqisE3j sq oj iou psAOjd
s s jn p r u js ijeius j o uoijBsqBnsiA sq j u is tjb u iS ii s e SuiuiEiuaj aqj 01 sn p ‘X p iB u n p o ju n
II) ( li .i in o.i ti sito [\j 11054n |<>so>| oj uopBojiddy
•lijqmassD pdiajduioj :ji{Si}[ siuauodmoo /vnpm puj :ifaq jiun Suniuoog
a m S jj
s||3qs ¡UJ3S
|ioj uo)dey pajnprujs
■ m tjO}B|nsu|
4J9A03 qjiM aqnj jaui|
_ P!JS uoqjEO psjEOD-pjoS
3UIJ B 'S s ) p s f q o uoissiuisuB jj siqBjins e j o [eu§is dno XEpEJEj sq j oj spuodsauoD qoiqM
‘Xjisusjui sSbuii sq j SuisqEnsiA puB sduiBJ ubos sq j S uijbjsusS jo j p ssn si (ui3jsX§ SuiSbuij
Conclusion and Outlook
Using a simple scanning unit with a saddle coil deflector module it was possible to
successfully perform alignment as well as resolution measurem ents in a miniaturised
electrostatic lens system.
The compact design o f the scanning module allows piecing together o f modules for
m ulti-stage deflection. Furthermore, several layers o f structured Kapton foil may be used to
additionally provide for astigm atism correction.
R.Y. Lutsch et al.: Initial resolution measurem ents o f miniaturized electrostatic lenses
for LVSEM , subm itted to Ultramicroscopy.
F. Burstert et al.: Novel high brightness miniature electron gun for high current ebeam applications, M icroelectronic Engineering 31 (1996) 95-100.
K..-H. Herrm ann et al.: The design o f compact deflection coils and their application to
minim um exposure system, Proceedings o f the 6 th European congress on electron
microscopy, Jerusalem (1976) 342.
F. M ills et al.: A flux theorem for the design o f magnetic coil ends, Particle
accelerators 5 (1973) 227-35.
/uioo'soijdouojjoop'M M M //:djjq ojis qo/\\ oqj uo 0|qE[iEAE ‘s u ie jS o jj O d 3
56 (£ ¿6 1 ) £ ť u in jjsu i - p s a o » ‘X3 (s;y ST [s]
IZ91 (0661) £S sX q j S o id d a y ‘X ejquiaix q ‘Xoy q [p]
6Z\l (Z00Z) £L -uiiuísui -p s A3» ‘pesy U j [£]
'ZSZ (6661) LLLÍ IoA 3 Id S
H '3 ‘u o j ji u j -jrç
'[Al ‘OJiqsEUin^ s ‘®>[3qpEg y ‘JOî[[EM y
H '0 '3 lz]
Z8ZZ (6661) 0¿ u jn jjsu i p s Aoy ‘u o j jn j j iN ‘ijb u io o l a w w ‘^ ji^ j ^ <B>pBf jm [(]
‘ije u io q 13
•uoijnjosoj XSjouo joqS iq qjiM uirujoods oqj j o jjEd p o jo o p s e
OAJOsqo oj posn jojoojop oqj j o jjBd jo u u i oqj puE p 3[B0S oq p|noo soSbjjoa poqddE oqj
qoiqM jo jjb Bjjoods uoijism boB pjjBJBd Suijoojjoo jo j posn oq p|noo j o p o p p oqj j o jJEd jojno
oqj o jo jo jo q x 'p u n o j si S uisnooj jo p jo jsjij qoiqM je jiu iq oqj A \opq jsn f si sañoso Suisnooj
jopjo-puooos qoiqM jb XSjouo o q x 'S uisnooj jopjo-puooos si ojoqj qoiqM jo j pun o j uooq osjb
SAEq su o ijip u o ^ 'ss ip popunojS oqj je posnooj oq ueo ( \ s Qpg oj Qp S 's ) soiSjouo j o oSubj e
qoiqM jo j p u n o j uooq OAEq suoijipuoQ 'qS iq oje sojSue ooubjjuo oqj uoqM >[Sip popunojS
oqj uo sjojoojop oqj spjBMOj souo jo o íejj oqj sjooyoj sjopuqXo oqj uooMjoq p p i j o q j
•[9] uiB jgojd Q cO d D 3MJ Sm sn psjE jnujis si uiojsXs o q x 'uoijoojip
jBipBj sjí Suo|E d o jp [Bijuojod u u o jiu n -u o u B SBq jEqj >[sip e Xq puo jo q jo oqj je p o s o p
si puB ‘pojunoui o je sjojoojop o joui jo ouo qoiqM uo puE jBijuojod punojS je si qoiqM )[sip e Xq
puo ouo je p o s o p si siopuqX o om j oqj uoowjoq oocds o q j jB ijuojod punojS je oq oj sonuijuoo
jopuqXo jouui o q x '[Etjuojod u u o jiu n b jb si qoiqM j o qoBO sjopuqXo m o jjb u j o soijos b Suisn
Xq poqddE si JspuijXo J3jno sq j uo u o ijb u b a jBijusjod s q x 'uo;jE jn§ijuoo ^ síp -o j-six E , sq j
‘jssX jeue M3U sq j j o suoijB jnSijuoo sq j j o ouo j o jusiudopA op oqj oquosop sm sjoj-i
•jopuqXo jojno oqj oj poqddB si u o ije u e a [Bijuojod (jEouq XpjBuiixoaddB
JO) JEOUq B jEqj jdooxo [ç > ] JOsXjBUE JOJJIUI |E0fjpUI|X 0 UAVOIDJ |[OA\ oqj OJ JB[IUIIS si josXjbub
o q x opoui S uisnooj jo pjo-puooos b SEq osje josXjbub mou e q x 'suoijoojip jBqjnuiizB j o o3uej
ojojdiuoo oqj JOAO jdoooB ueo ji jBqj os ‘XjjouiuiXs |e ix e SEq jBqj puE XSjouo j o oSubj opiM e
jdoooB oj 0(qB si jEqj [f] poquosop ueoq SBq josXjbub oijejsojjoojo j o odXj m ou y
opoui 3uisnooj jopjo-puooos e OAEq jou
soop y jH 3L() JBqi suioos j[ 'uoijtq o so j joqS iq b qjiM posXjBUB oq ubo uoiSoj X íjouo m o jjb u b
jBqj os opoui S uisnooj jopjo-puooos b oj qojíMS oj 0|qB oq oj oSbjueape ub oq osje pjnoM
ji X[[Eu;j 'XjjouiooS jBUBjd SEq ji osnBooq siqj op oj ojqE jou si y j H 3ll l ‘(ixoisjsdsip XSjouo
oje jBqj sopijJB d osXjeue puB jdoooB oj ojqB oq o j oSbjueape joqjjrij e oq pjnoM j[ -siqj op
oj 0|qB si V JH 3q j \u o ijism b o E p[|E JE d, se poqijosop oq ueo s iq x 'X[snoouBj|nuiis uinjjoods
XSjouo ojo|duioo b osXjeue oj ojqB oq oj oSejubapb ub oq p[noM-.q ■[ Q uEqj ssoj X][BoidXj
S u p q XSjouo ubouj oqj oj XSjouo j o oSuej oqj j o o ije j oqj ‘([Buis si ouiij ouo Xub jb pojoAoo
oq ubo jBqj soiSjouo jo oSubj oqj ‘[z‘\] (V JH ) JOsX[Bue p p i j oqoqjodXq oqj j o uoijdooxo
oqj qjiM ‘sjosAjEUB XSjouo o p ijjE d poSjBqo oijbjsojjoojo umou>| oqj jje ui jEqj suioos ji
SjO’[email protected] ISSOjppE [IBIU-O^
> in ‘d D I ¿ItM JOjsoqouEjM
‘pK°>j |JBq/W pjojjBJX ‘opjS[JEq,\\ ‘(o d o jn g ) XjojBJoqBq qojB 0so>| nzpEUiiqSj
X i] ‘l d 6 niA I JojsoqouBiAl ‘JOjsoqouE|/\| j o X jisjo aiu q ‘soisX qj j o juouijjB doa,
-J3>IIEM v PUE .o jB q siu in » s ‘^ H q n o a ‘*pRo>[ H d,
M3ZA1VNIV H O « y i i\i i v D i a a M n A 3 l a n v a v d a H i
V. Rom anovský*, M.M. El-Gomati*, L. Frank and I. M üllerová
Institute o f Scientific Instrum ents AS CR, Brno, Czech Republic
*Departm ent o f Electronics, U niversity o f York, Y ork YOlO 5DD, e-mail: [email protected]
The exploitation o f low -energy electrons for examination o f the material surfaces has
several advantages. One o f them is a small depth penetration o f the primary electrons (PE) to
the observed material. A t low energies the charging effects on non-conductive or slightly
conductive specimens are suppressed and also radiation damage o f the observed specimen is
decreased. Since the secondary electron (SE) yield is high, the signal to noise ratio (SNR) is
about one order o f magnitude larger than those in m icroscopes operated at 20 kV. M oreover,
exploitation o f low-energy electrons enables observation o f differently doped semiconductor
regions [1,2]. However, the use o f the prim ary electrons with low energy brings some
problem s e.g. low source brightness, increased aberrations and increased sensitivity to stray
fields (e.g. defocusing o f the probe by SE collecting field).
F igure 1: Schem atic o f the new design o f the fu lly electrostatic low energy scanning electron colum n
The first miniature electrostatic m icroscope with the field emission cathode was
constructed at the U niversity o f Y ork in cooperation with the ISI AS CR in Bmo several years
ago [3,4], The m icroscopy was fully functional, although during its usage several
constructional changes were necessary. Figure 1 shows the schematic configuration o f the
new microscope. For a better efficiency o f detection it was necessary to reduce the size o f the
m irror opening to 0.2 mm. This is mainly im portant from the viewpoint o f SE detection with
the energy up to 5 eV. N evertheless, so small bore restricted the viewfield too much.
Therefore, it was necessary to move the stigm ator-deflector (ST+DEF) unit below the mirror.
Furthermore, one o f the most challenging problems was the mechanical alignment o f the
w hole microscope. For this scope, it was chosen a construction consisting o f two independent
subassemblies. The alignment o f the individual parts is carried out by steel rods produced
w ith high exactness. Thereafter, the aligned parts were screwed together in a jig. The
alignm ent is finished electronically by using two independent deflector units (D efl and Def2).
These electrodes are positioned behind the mirror as shown in Fig. 1. In order to enhance the
resolution, the colum n was added by one more electrostatic (objective) lens (L2).
'66 (0002) 18 XdoosojoiuiEjJin ‘-pM‘ijEiuoryig ‘ i ‘EAOJ3[|nw ‘q ‘íjubjj
'9 1 2 'd ‘8661 uopuoq
‘suoijB3|uniuiuo3 j/\[OI ‘uoijosjq sqx :uj ‘q ‘^ u b ij q ‘BAOJ3||njAj ‘'WJAJ ‘;jBiuog-ig
•[£62 (1002) 6L '«31 's ^ d 'iddy ‘q ‘sipA \. “lAI'IM ‘p E iu o o -jg
>01 (£661) 8S XdOOSOJOIUIEJJjn ‘[E p ‘' a ' a ‘3TAOJ3d
[ 1]
oiug ISI jo pafoad [uiuoiui Aq puB (>fn) Áiopo^ [cao>| oqi Aq psjaoddns sem 100feud sii[x
•su o jjo sp
mojs j o
jEqj qjiM s g y j o Xjiaijisuss soE jjns sq j suiquioo
oj ja p jo ui s u o jjo s p j s S n y j o jazX puB y > \[3 e ojui uoijbxibjsui jo j p sp u sju i si uuirqoo
sqX 'SJBMyos uoijisinbDE s S euii j s j p q j o ssBqojnd s p n p u ; sdsjs ju su id o p A sp a jn j n j
■ psipxs ÁjSuipjOOOB
s3jB|d a o ip s |js p sqj jo j puB suoijb31jiu8euj stiouba jo j uoijEnjis sqj dBut 01 japjo ui spEui sq
p[noqs suoijBjniuis jsq jjn j puB spuuBqa xjs sqj jo qoB3 ui uoijBSíjqduiB jbuSis sqj ssousrqjui
jiun UBDS sqj jo uoijBJsdo jBqj subsuj siqx 'JsSSiq qonui s; B3je papEdui; aqj jBqj soijou
UB3 s/^ ’A0S- °1 pssBiq si jiun ubos sqj jo sp o jjo sp suo jBqj jdsaxs uoijBnjis aaiES sqj SMoqs
£ s jn S ij 'soBjjns u su ip sd s sqj uo p p ijM s p sqj ssjBDipuj sjBtibs sq x 'Jiun Suiuubos aqj uo sjb
ssSejjoa ou uaqAV jo jjiu j sqj pBduii jEqj o0 l+ siSub Sutjjejs puE a 3 01
Rl}M s u o jp s p
|buSis sq j ío j uMoqs si uoijBpm is sqj jo jp s s j sqj 1 s jn S ij uj 'pssn sem ‘suoijB[tuuis Q£ jo j
§uimo[|e ‘o '¿ 's js a uiBjSojd NOIIAIIS ‘uiejSeip Kej sqj jo SuipuEjsjspun jsjjsq j o j ’suojjoap
|bu8is sqj jo ssuojosTejj os|e ssousniju; jiun jopsysp-jojB U jiis sqj ‘d n -p s siqj u¡
'(>INS) ° P BJ 3SIOU oj [buS ts jsq S iq b sjsjjo jbuS is psiuuins s q x 'p p i j u o ijE Jsp o sp
sq j j o jusuiuSqE
jo j osje ssajss
sjo jo sjsp ajjsoddo
uiojj sjeu S is jo
u o su E d u io 3 'Xjiuiijejs Xjd
p u s XqdEjSodoj soBjjns puB uiEsq u o jjo s p sq j j o s o u s p p u i j o 3[Sue sq j j o sp u sd sp
suojjd 3[3 [eu S is
j o u o ijn q u jsjp jEinSuE s q x '[ euS is p su iu in s aq j
ju sp u sd sp u i xis ÄB|dsip
puuB qo qoB3
u iojj
jsjn d u io o [ojjuoo s q x 'SjaijqdujE sjEJBdss oj pajosuuoo sjjEd xis
OJUI pSpiAip SI JOJ33||OD 3qX 'p3J33J3p pUE p3JEJ3[33DB UIe Se 3JE 3S3qX 'SUOJJ33p /ÜEpU033S
jBugis s q x 'XSjsu s m o ] j p q j qjiM soBjjns sq j qjiM jsbjsjui puE u s u jp s d s sq j oj p sjE Jsp o ap
3JB ( g j ) SUOJJ3SJ3 ÄJEuiud 'pssBiq XpAijBSsu si u s m p s d s sq j sp o u i su3|-u; sq j qjj^v
liun unos diji /o s p o jio s p
duo uo
y í0 í ~
M O l 1-1I S
^ jn S ij
šiji uo
¡noyiiM uoiivjnuiis
N o m iS
:Z s u n S tj
I. Miillerova and L. Frank
Institute o f Scientific Instrum ents ASCR, Brno, Czech Republic
e-mail [email protected] isibm
In the course o f recent six years, a complex device for examination o f clean and welldefined surfaces with low energy electrons under ultrahigh vacuum conditions has been
developed and put into operation. The apparatus is intended for exploration o f novel image
contrasts, available at landing energies o f the scanning prim ary beam below 100 eV, under
a residual pressure in the specimen vicinity in the order o f 1 0 ‘ 10 mbar.
The instrum ent consists o f three vacuum chambers: the observation chamber,
a chamber for in-situ preparations and the loading chamber o f the air lock.
The observation chamber is equipped with a two-lens field emission electrostatic
electron column with electronically variable aperture (FEI Corp.). The energy range o f
prim ary electrons spans from 1 to 25 kV with the ultim ate resolution limit 12 nm at 25 kV and
0.1 nA beam current; maximum beam current is declared to be 100 nA. Main signal detector
is that for low and very low energy SEM, based on the cathode lens principle [1,2].
N egatively biased specimen forms the cathode o f the lens while the detector scintillator,
consisting o f the bored single crystal YAG disc, finely movable in all three axes, serves as the
anode at ground potential. Signal electrons are collim ated onto the detector by the cathode
lens field and, therefore, high detection efficiency is available down to lowest energies.
The observation chamber further contains a quadrupole mass spectrometer type QMS
200 (Balzers), operating within the mass range 1 to 200 amu and the pressure scale
1O'4- 10 ' 11 m bar with the Faraday cup or 10'5-10 "14 mbar with the channeltron. The device is
available for studies o f desorption under electron impact and also serves as ths leak detector
o f top sensitivity. Additional detectors under construction include a conventional EverhartThom ley detector o f secondary electrons and the hyperbolic field analyser [3] for parallel
acquisition o f electron spectra.
The preparation chamber contains a scanning ion beam gun for surface cleaning (IQE
12/38, Specs) with differential pumping. This is planned for in-situ specimen cleaning and
also for depth profiling and cratering o f multilayers at low ion energies, large tilt and
continuous specimen rotation. The chamber equipm ent will be completed by a set o f cells for
directed evaporation o f metals.
The specimen loading and manipulation system is in the final step o f manufacture.
This incorporates two UHV specimen stages and a loading trail for insertion o f the specimen
holder into the air-lock chamber. Specimen transport from the air-lock chamber into the
preparation cham ber and further to the observation chamber is made by means o f
a manipulator driven by magnets from outside (Delong Instruments, Ltd., Bmo). The
specimen stage for the observation chamber features the x and y movem ents ±5 mm, z-axis
m ovem ent ±10.5 mm, rotation ± 8 ° and two m utually perpendicular tilts up to ±5°. The
specimen manipulator for the preparation chamber has the x- and y-axis shifts up to ± 6 mm,
z-axis m ovem ent ±15 mm, continuous rotation and one tilt within ±60°. Both manipulators
accept a specimen holder, w hich bears five separate H V-insulated electric connections (for
SI I (1003) ÎZ Suxuireos ‘^uBjg q ‘BAOiaqnw \
'3833 (6661) 0 Z. 'uiaijsui p § A3>I ‘u o jjm j
‘pE uiog-ig
‘t o i pv ‘<»pEf -jm
'S03 (6661)
-osojoiim u ojpsig f ‘BAOJS[¡nw \ ‘^uEjg -q
' 6 SI (3661) St> XdoosojaiiuBjJin ‘BAOjaijnjAj g ‘oU3q
y o d d n g uio[>[ [ o a r j
siuajjBd Suidop j o
■psSpajMoiDpE Á[[nj3jBj§ si T06S90IY 'ou iUBJ§ -rapun >13 SV J ° <íou3§v ju b jq sqj jo
Xq psuSissp A piajduioo ojoa\ uisiireipaui §mpEO| sip put? sji.uE[ndiuinu u su ip ad s o q j
sisbjjuoo oj
[t7] A 3 J ° sjiu n oj UMop 3JB3S X gjsua sq j inoqSnoaq} uodijis ui
p s ju s u o puB ‘sjusuisaoui u s u ip s d s sq j j o sdoos p sjiuiij b X[uo
SuuniE sj ‘sjojB]ndiuBui u s u ip s d s yiqs3>|Biu qjiM p s u iio jœ d 3J3M sju a u iu a d x s jsjig
01 tô jsu a w m q lüm uud
3 l{i tfJ M a p o w SUd] SpOljlDO J lfl Ut p 3 jn S V 3 W SV UOlW¡OS3 J a S v U ll 3 ljl f o a n u d p u a d a p <<ga3 U3 3 l j [ .7 a u n S i j
/\a 'Â6 J8 U8 6 u|pue|
•Suipiinq sjn jp su i u ib u i sq j sp isjn o jo o p punojS sq j oj psAoui sq ||i m sd o a so jsiu i
a qj - j o p b j S u p iu iq aqj aq oj lusas suopEjqiA jBDiuEqDsui aqi |[ps jnq Suidurep uopEjqiA
SAISSBd JOJ U13JSXS E JO U0pEJU3UI3[duiI Xq pSAOadmi SEÁV U OpnjOSSJ sq j ‘jllSUldopASp
aoiAsp aqj j o s s jn o s sq j u j -sdojs je u S is s q j j o s p A q o/0Qg p u s y 0Q \ u ss M p q p s u iu u s p p
sbm qjpiM sS p s s q x p u su iu sd x a sq j j o j p ss n sem (jb § y ) I$ 5 1 S u s m p a d s jsa i uoqjBO
-uo-pjoS u o p rq o ssj-q S iq s q i '[ a jn g ig a ss - A3 01
puB a 35! I V¡ mu çz ‘As:!10I
je uiu £( j o u o p n [o s3 j aqj psuiBjqo pus s u o j p a p j o s s i S j s u s 3u;puB¡ snouEA j o j u o p rq o saj
ÇZ U^U Zl s! u n ^ 3qi JO UOpn[OS3J pSSJUBJEnQ
3§BUII JB3J sq i P31S91 3ABq
• j s p u i o j p s d s s sE u i s jo d r u p E n b s q i o s p s | o jju o o
s u ie s
s q q '( o u i g ‘- p jq ‘u s a s a x ‘s j q p i B S u ia js X s ) S u iA ïq o jB pu B S u i s s a o o j d ‘u o p is in b D E BjBp s 8 e u ii
pu B [ o jj u o o S u iu u e o s j o j u is j s Xs p a s B q
p s p o i p s p e Xq p s n o j j u o o s i s p o u i ] ^ g s s q x
•psiBuioiriB si 3 o081 °1 d n ino-3>[Bq j p q j puB
s d u in d 3 3 jj-]io j o s u b s u i
Xq paiBnoBAS s je sjsq u iB q o uopB JB d sjd puB u o p E A jssq o a q x
• jn o q
1 U Eqj a j o u i p u
js b [ o j p s j s s j u s s q
SBq ‘uinno B A 3 u ijB J 3 d o s q ; j o u o p B J O is s j q jiM a s q p S o j ‘aS uE qoxa u a u i p s d s ' 3 o0 0 t? I °1
p a u u B ] d s i S u p B s q u s u i p s d s s q x '( s i d n o o o u u s q i p UB ju a u iB iy S u p s g q s q i j o j ‘j p s j i u a u i p a d s
M iroslav Horáček
Institute o f Scientific Instrum ents AS CR, Brno, Czech Republic
e-mail: m [email protected] isibm
The possibilities o f acquisition o f the energy spectrum o f signal electrons or angular
distribution for every point o f the scanning m atrix on the specimen in SEM were studied and
suitable fast position sensitive detector was designed.
The basic task for the scanning electron microscope (SEM ) is to acquire twodim ensional image o f the surface o f the observed specimen in topographical or material
contrast mode, depending on the type o f detector used. The detectors collect the sum o f signal
electrons, i.e. w ork as integral detectors and can not acquire angular or energy distributions o f
electrons. To obtain more information from the specimen we can use the spherical detector for
detection o f the angular distribution o f signal electrons, or some energy sensitive detector for
acquisition o f the energy spectrum. Usual procedure is to acquire two-dimensional image o f
the specimen, next to select a limited num ber o f points from w hich we would like to obtain
energy spectrum information, then to direct the prim ary electron beam on these points and
measure energy spectrum point by point.
Thanks to advance in com puter techniques and detectors technology we can
autom atize and extend acquisition o f additional information from whole specimen. At present
we can obtain additional 2D information when detecting angular distribution o f the signal
electrons using area selective detector, or additional ID information when sensing the electron
energy spectrum using energy selective line detector. The main problem in realisation o f such
m icroscope is not in tim e consum ption or mem ory consum ption in the control computer but in
time consum ption o f the electron detection. Let us consider the scanning matrix o f 512x512
pixels. We w ould like to acquire the multidim ensional image in max. 10 min. Thus, the
m aximum time for acquisition o f the angular distribution or energy spectrum from one
specimen point is t = 600s / (512 x 512) = 2.3 ms. For detection o f the angular distribution o f
electrons in such short time we decided to design a detector based on the areal thinned back­
side illum inated CCD sensor working in the direct electron-bombarded mode (EBCCD)
(Fig. 1). The time shortage is the main reason w hy we cannot use any comm ercially available
camera. Such cameras are usually optimized for very long integration and process time
(slow -scan CCD cameras) and are equiped with high resolution sensor and cooler.
From the point o f speed the key param eter o f EBCCD (Fig. 2) is the gain G o f electron
bom barded sem iconductor (EBS) (a num ber o f signal electrons in the potential well generated
by one incident electron), w hich is related to the incident particle energy E. For a real CCD
sensor w e must calculate the EBS gain G as G = E s(E) I (3.65 eV) where s(E) is the
detection efficiency o f the CCD. In practice, the front-side illuminated CCD shows
a reasonable efficiency ( s> 0.1) at energy from 8 - 1 2 keV upwards, depending on the sensor
type. The available thinned back-side illum inated CCDs have an efficiency higher than 0.1 at
energies below 5 keV. To meet our resolution and timing requirem ents we chose the back
illum inated high perform ance CCD sensor CCD39-02 (80x80 pixels) from Marconi. For the
well capacity o f Nc = 300x103 electrons and uniform distribution o f the signal beam, the total
num ber o f electrons to be generated in all potential w ells is N„ = 3 0 0 x l0 3x80x 80 = 1 .9 x l0 9
'LLH ‘III 'IoA ‘OOOZ ouag ‘3 1 W 3W 13 30Jd :uI “ W ‘^ - “ H fc]
'ZZ ( 8 6 6 1) 01 Xdossojsij^i psisissv--i3 Jtiduio 3 jo iBiunof
‘>[sse.ioh [ l ]
ODD paivwuinijt dpis-yovq pauuiifj .7 dunSij
( iu r i£ - £ ) u o iS a i
’(pdAOiudu Suipjdiifs di{¡ tfjiM.) udqwvi/o ddoDSOAOiiu di{i ui Aojodjdp difi f o uoiiDunSifuoo ¡vjudiuuddxg : j d jn S ij
o¡[qnd3>¡ qo3Z3 sip j o X ousSy lUEJQ 3M1 ^<1 pspiAOid I OOd/OO/ZO I ou
3lT>Xq psjjoddns^si ju su iu sd x s 3i|x
•jno p su re s S u p q sje jusuipaequioq u o jp s p aspun sjusuisjnsBstu puB psdopA sp uosq
■siqEjins si sd o s so jsiiu sq j j o S3i}do u o j p s p sq j j o
jusuisS uexie [ j ‘ [] p sq sq q n d XpE3j|E ‘sso d in d siqj j o j 'A s u s p i jp u o i p s p p sq j j o K suspusdsp
KSasus s q i u io ij sm o[joj qsiqM pAS[ e uo s u o a p s p SuipjB quioq stp j o ÁSasus sq j jss jsnui 3a\
uib§ a 3 0 Q 3 sq i j o [3A3[ u m u u jd o uo Axres o x ’p sz iu iiu iu i si s u ip u o ijE iS sp i sq i oO£ = D I0J
s[iqM ss io u jnopBSj sq i Aq p sjiu ii| X[uo si spAS[ AsjS sjqBAJSsqo j o asquinu sq i puB p ssq iin
/C[[nj si onUEj oiuiEuXp oqi 00£ > £> J0J '00£ = D J0J sS u b j siureuAp oqi qjiM siqBJBduios
si puB f j / ’jv = Q SB psiEinojBS sq ubs (spA Sj AsaS j o asquinu) uoiin[oss.i ss su jq S u q s q x
g p 0 9~ si (SuqooD inoqiiM 'S'l) sjniE Jsduisi iu o o j jb 0 3 3 9 3 piBpuEjs j o sS u ej siuiEuXp sq i
puB JUEUlUIOp SI SSIOU JTlOpESJ Sqi (SUl 0 1 >) pSSn SUIll UOlJBiSajUl M0[ Sqi JOJ ’X iiuuojiunuou
p u S is >)JEp sn[d ss io u inopESJ oqi Áq p siiu iij si 0 3 3 9 3 aqj j o sS u ej DiuiBuXp s q x
•siusuisimbsj ano [ij[nj p|noo 0 3 3 poiBuiuin[[i opis-qot;q psuuiqi sqi
‘Á[iuonbosuo3 •[['() = e0 I xS/S9'£ x li\ = (A35I Ç)3 shjea popoou oqi p g sm SAoqn uopsnba
uioaj -3SI = e. 0 i x 3 x 6.0 ix i/61.0ix9 l x 60 ix 6 'l = ^ / / ^ 'HAř = 0
Josuas b pssu sm
suojps[s jo junouiB siq; sjbjsusS ox "V 6_0l = qI lusxino uiEsq ibuS;s suinssB s ^ -suojpsjs
H. Kiesel, A. Renz and F. Hasselbach
Institut fur A ngewandte Physik der U niversität Tübingen, A u f der M orgenstelle 10,
D-72076 Tübingen, Germany, email: [email protected]
The astronomers H anbury Brown and Twiss (HBT) were the first to observe in 1956
that the fluctuations in the counting rate o f photons originating from uncorrelated point
sources becom e, w ithin the coherently illum inated area, slightly enhanced com pared to
a random sequence o f classical particles [1]. This at a first glance mysterious formation o f
correlations in the process o f propagation turns out to be a consequence o f quantum
interference betw een two indistinguishable photons and o f Bose-Einstein statistics [2]. The
latter requires that the composite wave function is a sym m etrized superposition o f the two
possible paths. For fermions, by virtue o f the Pauli principle, no tw o particles are allowed to
be in the same state. The corresponding antisym m etrized tw o-particle wave function excludes
overlapping wave trains, i.e., simultaneous arrivals o f two fermions at contiguous, coherently
illum inated detectors are forbidden. W e succeeded, in spite o f the low mean number
(degeneracy) o f ~ 1 0 '4 electrons per cell in phase space, to observe these anticorrelations for
a beam o f free electrons for the first time.
The coincidence m ethod chosen in the present ‘antibunching’ experiment is an
alternative to H BT's correlation procedure for gathering information about correlations in
a stream o f particles [3]: Two detectors are coherently illum inated by an electron field emitter.
A ccording to the Pauli principle no tw o electrons o f a certain spin direction are allowed to be
in the same quantum state, i.e. to arrive at both detectors simultaneously. In other words, if
our detectors had a time resolution corresponding to the coherence time Tc , which is in the
order o f 10‘14 s, no coincidences would be observed. (In our unpolarized beam, suppression o f
fluctuations is only 50% com pared to a polarized one. Fluctuations in orthogonal
polarizations are independent).
The tim e resolution o f our fast coincidence counter o f Tr = 26 ps was about three
orders o f m agnitude less than the coherence time. For incoherent illumination o f the detectors
the electrons behave like classical particles. Due to the insufficient tim e resolution a certain
random coincidence rate is observed. By changing the illumination from incoherent to
coherent we expect a reduction o f the random coincidences (by a factor TJTr o f about 10‘3 in
our experim ent) due to the fact that w ithin the first 10*14 s after arrival o f an electron no
second one is allowed to arrive.
Our experimental set-up corresponds to HBT's stellar interferometer: The tiny
effective virtual source o f an electron field em itter illuminates via magnifying quadrupoles
and im age intensifying channel plates two small collectors. W ith increasing angular
magnification the effective lateral distance o f the collectors decreases and their illumination
changes from incoherent to totally coherent. In turn, a continuous increase o f anticorrelations
o f arrival times o f the electrons at the collectors is expected. Features o f our cold <100>
oriented tungsten field em itter are: Extraction voltage 900 V, total current 1.5 [tA, energy
width A E f w h m o f 0.3 eV w hich corresponds to a standard deviation o f AE o f ~0.13 eV, virtual
source diam eter ~36nm, brightness 4 .4 x l0 7 A/(cm 2 sr), coherence time Tc = 3 .2 5 x l0 '14 s,
coherent particle current 4.7x10 9 1/s, degeneracy 1.6x10’4.
'£££ (0861) 9S >l!ldo ‘« z j a q a s ’O
T M ‘Ill 'I°A ‘0003 o i u a ‘w a y n a m z i 3 °Jd ;U! ‘qoeqpssB H J
'ÍLP (£861) 3 £ l id d n s sXqg u o jja a jg # sa iu o jja a jg A py Mqads y ‘a s o y h
'9SZZ ( 8 /.6 I) 81 H Agy sX qj ‘ [B ja UBqBj^ h 'V ‘w nB a O ‘-"»FB I 3
[9 ]
'LL ( 6 S6 I) Z\ sXqd T UBqBjjsny ‘a j j j n O V ‘SS!MI Ö"H [£]
-6 frW (9S61) 8¿ l sJnjBN ‘ip a -m j - j ^ g [z]
'LZ (9 S 6 I) LL\ sjnjB N ‘ssim x f r y ‘uM oaa XjnquBH "H [ I ]
sa a u a ja ja y
•p3§p3[MOlD]3E X[|I\)9JEj S SI ÿBipSUISlUSSsSunipSJOJ 3ipSJtl3Q Oljl UIOJJ uocldns 3L| L
•paanpaj si s Sb
-LUBp uoijEipEj ‘snoauaS ouioq a jo u i si u a u ip a d s aqj j o uoijBuium jii aqj uaqM ‘[/J ja z ja q a s 'O
o) SuipjoooB ‘X [[Buoijippy -sju ajjn a q3iq je os[b mo[ u ie u isj pjnoqs puB \ s [ '() j o jo p jo
a q j uo si sajn jeaad u iaj m oj je so jn o s b qons j o peruds ÁSjouo a q x -pajaadxa si [9 ‘ç] SuiuapBOjq
X8ja u a snojBuiouB a q j j o u o ip n p ru aijsBjp b ‘X puanbasucQ 'a jd p u u d q tiB j oqi oj anp qjSua[
aouajaq o a b UBqj ss3[ j o saauBjstp uiqjiM oj jaq jo qoB3 SuiqoBOiddB u io jj pajuaA ajd ajB su o jj
- a a p p 3ZUB[oj :§ u isiu io jd Xj3a si [f?] auo p a z u B p d b Xq js jjiu is p p i j |bu o iju 3 a u o 3 b Suijnjijs
-qns Xq ‘‘S'3 ‘sa d o ss o js iu i u o j js s p a jn jn j ui SuiqaunqijuB u o jja a p j o uoijEaqddB a q x
qovjq ui pxfjoui suv s)uvdpD/vmiuni/t \\iudj3i\0j 3i]j svajv
U 0 )3 3 ]]0 3
3 1 ]) S 3 J 3 M 3 3 1 ])
‘S V 3 J V p 3 ) V U l W I Í ] ] l l< l)U 3 J3 t]O D 31]) ¡ U 3 S
-dudau s3sdi]]3 aiji swr?uSo)Did 3i¡) u j (ixai 33s) su0)03]]00 3i¡¡ fa
uounintun¡i¡f a 33U3J3t]03 f a u o ip u n f o sv Suii]3unqi)uy : j s u n S ij
-jiuia p p i j uajsS unj p[oa jn o j o sajnjE aj aijsuajaBJEqa
oqi lu o jj pait![n 0[B0 anjBA p o p o d x a X|[Boipjooq) aqj
qjiM sooj3b j] -(oippiiu ‘j S ig) £ j o
/\ [ / g
oijbj ssio u oj
|bu8is b qjiA\ £.01><9Z'1 = '“S' °1 sjunouiB uoi}Buiuin|[i
ju a jsq o o Xjjnj jo j sso u sp p u io o ui u o ip n p a j SApEpj
pSjnSBSUI 3 q x 'USAIS SI Moora^-Mooy
- g jj j o jSSU,
sqj uj -sd £ i =f j o A\opuiA\ uoqnjossj suiij aqj uiqjiM (
•gig ui >[Bod sq i j o Suiu3jjB|j b sb sjqisiA soiuoooq a itu j
-oads ju sjsq o a u i s q j snsjSA luojoqoo X|]BijiBd puB ju s
-jsq o o aqi ui ,y v sao u o p p u io o S uissiui aqj a 'i ‘ç [buS|S
SuiqaunqijuB a q x 'B jp ad s ju ajaq o a X)|BiUBd 'd sa j jua
-jaq o a aqj qjiM ju aja q o au i aq; S u iso d u iu a d n s Xj[Bns¡A
Xq paM oijoj B jp ad s aqj S u izijb u u o u puB Suiqjoouis
j o p ajsisu o a B jp a d s a u n i ,\\bí iB juatuuadxa aqj j o
a jn p a o o jd uoijBn|BAa a q x •([ -Sig ui sa u q ||ti[) uoijbu
-lu in n i ju a ja q o a X jpijjB d jo j uibSb jsbj a q j puB Xubjoj
‘XjjBjjjBd jo j § u ia\o [|o j a q j ‘( j 'Sig ui sau q p aq ssp )
sjo jaa jap aqj j o uoijBuiuinfji ju aja q o au i jo j js jij aqj
:qaBa s jq o£ jnoqB jo j pajBjnuinasB 3J3a\ B jjssds jn o j j o jbjoj b SuiqsunqijuB aAOjd oj ja p jo uj
P. Sonnentag and F. Hasselbach
Institut für A ngewandte Physik der Universität Tübingen, A u f der Morgenstelle 10,
D-72076 Tübingen, Germany, email: [email protected]
The W ien filter (i.e., a hom ogeneous electric and a hom ogeneous magnetic field both
perpendicular to the optical axis and to each other) is usually used as an energy analyzer or
a m onochrom ator for electron and ion beams. In an electron interferometer, it acts as a wavepacket shifting device for the laterally separated coherent electron wave packets [1,2]. The
W ien filter is said to be in its m atched (or com pensated) state if the electric and m agnetic
force for particles travelling on the optical axis and having the m ain energy component o f the
electron beam cancel each other, i.e. E = v B w here E is the electric field strength, B the
m agnetic induction, and v the m ain velocity com ponent o f the beam. If this condition is
fulfilled the phase difference betw een the two wave packets is left unchanged by the Wien
filter. This is due to equal but opposite phase shifts caused by the electric scalar and the
m agnetic vector potential (A haronov-Bohm phase shifts). The longitudinal w ave-packet shift
is caused by different electric potentials on the two laterally separated paths in the W ien filter
leading to different group velocities. The acceleration and deceleration o f the wave packets
actually takes place in the fringing fields o f the W ien filter capacitor. If the two wave packets
have the sam e position in the direction o f the optical axis before entering the W ien filter the
longitudinal shift caused by the filter leads to a decreasing overlap o f the tw o wave packets
when they are superim posed again (see Fig. 1).
W ie n f ilt e r
Figure 1: a) Electron biprism interferom eter with Wien filte r sw itched o f f (left) and Wien filte r in its excited state
(right): The w ave p a c k e t sh ift exceeding the coherence length leads to the disappearance o f the interference
frin g e s, b) R esto ra tio n o f c o n tra st b y a Wien filte r : The long itu d in a l sh ift o f the w ave p ackets caused by
electrostatic deflection elem ents (top) can be com pensated (m iddle) a n d overcom pensated (bottom) with the
Wien filte r [2J. The 11 m icrographs correspond to 11 different excitations o f the Wien filter, increasing fr o m top
to bottom.
T he decreasing overlap w ith increasing excitation o f the filter m anifests itse lf in
a decreasing fringe contrast. The value o f the longitudinal shift for vanishing fringe contrast is
identical to the coherence length o f the electron beam ; by Fourier analyzing the decrease o f
jou soop ouo jBqj
s jb o
jo b j
• j u o i u o jn s B o iii e
jo j
j o souioojno ju o jo jjip j o u oijisodjodns v, oAJOsqo
u b o ji s e
soisXqd m njuBnb j o u io jq o jd juouiojnseoui
oj dojs JSJTJ b si oouojoqoooQ 'S 3 A q XBpXioAO jn o ui ( o a j j b p u B pBop q jo q S u p q jb o b )
jo S u ip o jq o s j o (X jsnoouBjjnuns souioojno ju o jo jjip SuijBOipui juo u in jjsu i SuunsBOui b
o a jo s
j o jo ju io d b j o ‘o u j i j o u i b s oqj jB ojoqj j o a o p u B o j o q S u p q o jq B j b ‘' 8 'o ) sojBjsuoSp uo ijiso d j o
su o ijiso d io d n s OAjosqo jou op o m X q M ‘o [ d u iE x o j o j ‘u iB j d x o u b o oouojoqoooQ j u o u i u o j i a u o
s jí p u B u io js X s
(ju o u io [ S u b ju o )
u o a M j o q u o i j o b j o j u i o jq B p iO A B u n u b Á q
‘j u o u i u o j i a u o
su o ijB p jjo o uinjuBnb j o u oijB uuoj oiqisjoAOjji oqj q S n o jq j suiojsXs
o j 3np S u p q oouoioqoo j o
oqj qjiM
u in ju B n b
SI [ 9 ] 3 0 U 3 A 3 l¡ 0 3 3 p ‘JSBJJU 03 U] ’JUOUJOISUBJUO OjqiSJOAOJ
sso[ o jq is jo A S J q jiM X p A isn p x o j p o p OABq sm j b j o s
('[ 3 ] u ! p oquosop X p p q d x o si ssooojd s i q x ) 'Jojjij u o ;m oqj
poJQjsoj si sju o u io p uoijoegop
o q j j o s p p i j ojjB jso jjo ap oqj oj o n p jsoj s b m q o iq M o o u o j o j j o j u i M o q SMOqs q j 'S i j 'J O jp j
js jíj oqj JOjjB uoppiq Xjuo j n q jso[ jou si oouojoqoo jBqj s o jB J js u o u io p s iq j ‘o j o u u o q j j n g
qenbo sojbjs 3ui[joqt:| om j oqj o>pui pBojsui jnq ‘( s j u o u o d u i o o
p n b o OABq sjoipB d q j o q q o i q M jo ) ojbjs poB ds S u ijp q B j, b - q j i M u o i j B p j j o o ui joojop jsboi
jb jo - o juo joofojd jou op 8M o s n B o o q si siq x 'ssooojd siqj Xq Xjisuojui osoo[ jou op om jBqj
oouojojjip oqj qjiM jnq ‘josbjo uinjuBnb b sb uodn po>poi oq ubo js j jg a 3 i,\\ puo33s oq j uibSb
[B n b o
J3>[3B d OABM
sq j Xq jo u o jjb u u o ju i
3qj JO
q jB d -q o iq M
j o sso[ oqj Xq j o q jp ‘sXbm ju o jo jjip ui pojojcLioju; 3q UB3
siqj ‘u p S y -ouopun oq UB3 osoqj j o jsjjj oqj j o J33JJ3 oqj ‘uoijB jpxo ojisoddo qjiM s js jjij uoi.w
oa\j Suiuiquioo Xg :0[qisj 0A0j X[[bjoj si jo jjij U 3 i^
Xq pssnBS jsbjjuoo j o ssoj s q x
•so B ds j s j í j s q j X[uo uo S u ijo b s jq B A ja s q o Xub o j j o s d s s j qjiM s j s j j s j u i jo u u b d
s o B d s 3JBJS J s q jo u B ui so jB js jB u o S o q jjo qjiM pa|SuE ju3 OJB qniqw s s j b j s s s n B o a q js b j j u o o
j o sso[ b oj spBOj ju o u io jS u b ju o q o n ç Z u ! s o jb js JBuoSoqjjo Xj3 u is b o jo u i q jiA \ ju o u io jS u b ju o
SABq X ][b oj OM ‘posBOJOui si j o j j i j uoiyW oqj j o u o ijB jío x o o q j u o q M ‘0SBO s i q j iq z ® X u ! o jb js
jo n p o jd b jn q o jb js pojSuBjuo u b j o u j j o s j i s i j o j j i j u o i/^ oqj S uuojuo o jojoq j0í[0Bd o a b m jq d s
oqj j o o jb js oqj j i p o g ijs n f X n n j Xjuo si siqj j o ^ ¿ ui osoqj qji,« x u f s o jb js oqj j o ju o u io [ 3 u b ju o
OABq OM sj05[0Bd o a b m o q j j o BUJixBUi oqj j o su o p iso d oqj j o z y oouBjsip iB uipnjiáuo] o q j j o
o s n B o o q uoqj ‘( z s ix b |B o p d o oqj 8 u o jb ‘-o-i) u o o jo s jB q j oj jB u o S o q jjo s u o i j i s o d j o z ooBds o jb js
oqj j o puB u o o jo s uoijoojop o q j uo su o ijiso d j o x ooB ds o jb js oqj j o jo n p o jd j o s u o j b sb o p ijjB d
o q j j o oo B d s o jb js XjBuipjo oqj o j i j m o m j j :ju o u io iS u b ju o j o puiJ[ b sb s jo íjo B d o a b m o m j
o q j uaoM joq XBpp o q j jop isu o o o j s i j s b j j u o o u i OSBOJOOP o q j j o j d j o j u i o j X bm j o q j o u y
•pesBOJOUi si oojnos o q j oj oouBjsip oqj u o q M XjijiqBqsinSuijsip oqj
SuijnsBOUi j o j ojqBAjosqo ub sb ojqBjínsun o jo u i puB o jo u i souioooq [baijjb j o oujij oqj ‘rnoqj
uooM joq dB]JOAO Suisbojoui ub oj spBOj sja>[OBd oabm oqj j o 8uipB"ojds oqj osnBooq j n g
sqjBd oqj j o XjijiqBqsm Suijsip q puB XjijiqisiA sojouop 4 o j o q M [ > ZQ + ZA B jnuuoj oqj Xq
XuBoijeuioqjBui podsBjg oq osjb ubo u o ijB p j s iq x joS om (oouojojjojui) jnoiABqoq oabm sso]
oqj - X jjodojd opijJB d b - u o jjo o p oqj j o qjBd oqj jnoqB mou>[ om ojoui o q j - X jip n p ojoijjBd
jB q j jo B j o q x 'p o jjiu io s b m u o j j o o j o oqj u o q M o u j i j j o j u b j s u i oqj j o oSpo]Mou>i j o >[OB]
oqj j o osnBooq u o i j b u j j o j u i siqj jo3 Xj|boj jo u u b o o m sjuouiijodxo juosojd ui qSnoqj - sj 0>[0Bd
j 0q 0[0M„)
o a b m o m j oqj j o jq S q j j o soraij ju o jo jjip oqj u i o j j S u i s i j b u o i j b u i j o j u i (« 80^
q jB d -q o iq M (SuijjoS j o Á ji] iq ts s o ď ) S u is b o jo u i oqj oj onp S u p q sb pojojdiojui oq u b o JOjpj uoi .\\
oqj j o uoijB jp x o S u is b o jo u i qjiM s o S u i j j o o u o j o j j o j u i oqj j o j s b j j u o o u i OSBOJOOp o q x
J0J Xdoosojjoods XjqiqisiA s,uospqoiiAj qjiM XSojbub u i p o jnsB O ui
oq u b o so p ijjB d oqj j o u injjoods XSjouo oqj ‘jjiq s p u ip n jiS u o ] oqj j o u o p o u n j b sb j s b jju o o
In case o f the W ien filter the experim ents show that there is no decoherence emerging
from the - doubtlessly, abso lu tely reversible - interaction o f the electron w ith the
electrom agnetic fields or with the environm ent (capacitor plates o f the W ien filter). The latter
is negligible due to the large distance to the capacitor plates as we will see in the following.
An experim ent on decoherence in w hich a highly controllable environm ent is realized
by a conducting plate has been proposed by Anglin and Zurek [7] and is being in progress in
our laboratory. A charged particle beam (in our case an electron beam) is split and both wave
packets travel over a conducting plate with finite resistivity before they form an interference
p attern (see Fig. 2). C ontrast, being a m easure for decoherence, is influenced by two
param eters: O ne is the height y o f the trajectories over the conducting plate, the other is the
spatial separation Ax o f the coherent beams. The decoherence time is proportional to y /( Ax)2.
The reason for decoherence is the following: An electron above a conducting plate causes an
induced charge in the plate, and as the electron moves, so does the induced charge. Therefore
w e get an electric current in the plate w hich encounters ohmic resistance, thus a disturbance
o f the electron and phonon gas in the plate results. This disturbance is different for the two
paths o f the electron. So w e get entanglement o f the beam electron w ith the electron and
phonon gas inside the plate. A s dissipation is an irreversible process, a record o f the
electron’s path remains w hen the electron has crossed the plate and the two parts o f the beam
are brought together again. So in this m odel experim ent the object w hich gets ‘classical’
properties is the beam electron and the role o f the environm ent is played by the electron and
phonon gas in the conducting plate.
F ig u r e 2: S k e tc h o f d e c o h e r e n c e
e x p e r im e n t [7 ], The tw o d iffe r e n t
tra je c to rie s o f the ele ctro n o v e r the
c o n d u c tin g p la te le a d to d iffe r e n t
d istu rb a n c e s (sh a d e d reg io n s) in the
electron an d phonon gas inside the plate.
B y varying the param eters y and Ax (as well as the resistivity o f the plate), we can
adjust the strength o f decoherence ranging from negligible to strong decoherence and so
reproduce the transition from quantum to classical behaviour.
For our experim ent we use a com pact rigid electron interferom eter [8 ], The electron
beam em erging from the field em ission diode electron gun is aligne(fto the optical axis by an
electrostatic double deflection elem ent. Then the beam is split by a negatively charged
M ollenstedt biprism . The lateral separation o f the tw o rays is enlarged by an electrostatic
quadrupole (positively charged in the plane o f the beams). A fter passing the conducting plate,
the rays are directed tow ards each oth er by an o ppositely charged quadrupole. The
superposition angle is dim ished by another quadrupole in order to get larger fringe spacing.
As usual, the prim ary interference field is m agnified by two quadrupoles. To com pensate for
slight m isalignm ent, coils to rotate the image and deflection elem ents are incorporated. The
height o f the beam s over the plate can be adjusted as well m echanically by m oving the plate
as electron-optically by double deflection elements before and after the plate (Fig. 3)
In general, the electrostatic deflection elem ents w ill cause longitudinal shifts o f the
two wave packets relative to each other. H ence there will be loss o f contrast also due to this
effect w hich is absolutely reversible. Therefore, to discrim inate betw een this and the loss o f
A vau
'6PP~£PP ( 8 8 6 1) IL a sX q j Z 'suoijBsqddE pue s ju su iu sd x s iBjusuiBpunj
jo j ja p u i o j a j j a j u i u is u d iq u o j p a p A H fl sjnjB iuim p a zip ag g ru y iqoB qpssB jj \ j
[g ]
'0LZ-Í9Z d
‘(96 6 1 ) puouojAi ap s a iju o a u s y JSJXXX 'OOJd ‘('sp a ) u?A q u B q i u b j i f put; ‘p n B p H
-p n e jir) X ‘u E s u y y :uj 's s u s js q o o s p j o js s j u o i s p s i d y ip iw z 'H'AV 'u q 3 u y y f [¿]
'9661 u HJ3a ‘J s S u u d s 'X jo s q i
uinjuB nf) ut ppo^w ib o is sb |3 e j o aouEJB addy a q j puB a o u s js q o s a Q : jb p lu q n iQ q [ 9]
■ ¿ S I Z - K lZ ( 9 6 6 l) ¿ ¿ H3!
'A3» 'sX q j 'Xji[Bnb 3u¡ u y iu o ijb lu jo ju j X E ^ -q s iq M puB X jqiqisjA a g u u j 'jja iS u g 'Q - 'g [ç]
'69£ 'd :s-«3q •Z LÍ-9ÍÍ ( 9 8 6 1 ) 08ť s s s u a p s j o XuiapBsy
> (jo^ Majs[ a q j j o sjB u u y jb ^ s,ia8utpojqD S puB ‘sjo jB n u a jjy ‘s j s i j q d u iy :jaqnB [9 -j-g [p]
(S 6 6 I) £9C V S3>1 -sX qj ’qp]A[ pire ' íjsu j ' p n ^ '(X d o o s o jp a d s -ja u n o j) B jp a d s a p ijiE d
p a â ie q o j o juauiainsB aui o u p u io ja jja ju i uajJopuaqoByW H ‘JSJBqos ' y ‘qsE q p ssE H J [£]
'0 9 1 - K l
(£661) 8t* V 'A3îi ' S/iq d s ia p u io ja jja ju i BAEM-jaijBui-psgjEqs u; aouajaqoo [BuipnjiSuoj
S u u o js a j j o j aoiAap Suiyiqs-p:>pEd-3ABM y iJa q tj ua(A\. iqsE qpssE H ' j ‘s u b ^ o i ^ -jaj [ j]
'6 2 -8 3 'd ‘I Ï ° A ‘0861 s h S b h a q i ‘XdoosojaijAj u o j p s i g -jguo 3 'd o jn g q j¿ -o o jj ‘( 'sP3)
i u ° ° a 'O PUB o o ja p a jg <j :u i 'J a q i j u a i ^ b guisf} P î p E j s a b ^ u o i p a j g ub j o q jg u a q
a a u a j a q o j a q j j o ju 3iuajnsEaj/\[ o u p u i o ja j ja j u j p a j i Q ip u E jq o ^ 'O ‘ipaisu ajjo jrç '9 [ j]
p3§p3[A0U5[0E X]jnj3JBjS SI
0Z/Í901 VH °N JOBfluoo iopun yEipsuiauragsSunqosjoj aipsjnaQ Âq oouoioqooop uo juauiusdxa oqi jo yoddng
SuiMDjp siifi ui Utwoifs dJü judiuisnfpv jDdpdd
uoj tùvssdDdu Á¡qissod dun ifjjifM sjuduodwoj jvDijdo-uojjDdjd jjy judiuuddxd ai/j Jo dn-jds JijnuidijJS :ç dunSij
(ejaujea û03 01
P8|dnoo Aiieoüdo-Jdqy}
«j s â
»= ΙV
J0109||8p |B)U0Z|J<H|
P i
t).. fi
'33uajaqo3ap j o qjS uaijs
a q i Xq p a u iu u a p p uaqj si qaiqA\ 3U|ba uinuiixBui
s ji
|iju n
'g i j 33s)
js b jju o s
OS puB Sp>pBd 3ABM guiJ3JJ3JUI 3qi U33M pq 33UBlSip IBUipnjlguO] 3qi 3gUBq3 UEO 3AV J3J[IJ
u3fA\ p s q s p u i sq j j o u o i p j p x s sq j guiXjBA X g 'jjiq s jB uipnjiguoi siqj
uiBsq a q j ui J a q ij usiyw
s s n p o jp i 01 3[qBsuadsipui si
jo j
‘a o u a jsq o s sp
sjB susduioo oj q p d
u io jj
g u in su s
js b jju o o
H. Wittel and F. H asselbach
Institut fur A ngewandte Physik, A u f der M orgenstelle 10, 72076 Tübingen, Germany
e-m ail: hilm [email protected]
A fter the successful realization o f a spherically corrected objective lens for electron
m icroscopes by Rose [1] it is now necessary to reduce the chromatic aberration to reach subÄngström resolution. The energy spread o f electron beams is caused by the energy
distribution o f the field emission process and the Boersch broadening in regions o f high
current density. Two ways to reduce the aberration disk caused by chromatic aberration are
conceivable: Firstly, using o f a monochromator to diminish the energy width o f the electron
beam to about 0.2eV putting up with an unavoidable loss o f brightness and intensity.
Secondly, using an electron gun with an intrinsic energy spread low also at high current
densities. The reason for the Boersch effect are Coulomb interactions o f the electrons. In 1983
Rose and Spehr [2] predicted that the anomalous energy broadening in crossovers could be
reduced drastically by using a polarized electron gun. Because o f the Pauli principle in
a totally polarized electron beam every phase space cell is occupied by at most one electron.
Therefore the distance o f electrons also in high current density regions can’t be smaller than
about a coherence length. Consequently their Coulomb interaction and the anomalous
Boersch broadening should be drastically reduced. In the present work we develop such
a polarized electron gun, investigate in a first step the energy broadening at low currents and
in a second step we w ant to explore w hether the energy spread rem ains small at currents in
jiA-region necessary for electron microscopy.
For the developm ent o f a polarized electron gun we picked up an idea from a group in
Bielefeld [3, 4] who investigated this source for low currents. Onto a normal single crystalline
<100>-oriented tungsten tip a thin layer o f the ferromagnetic europium suJfide (EuS) is
evaporated and cooled below its Curie-tem perature o f 16.5K. The band structure o f EuS and
the emission mechanism (see text) are depicted in Fig. 1.
F igure
i . D u n u s i r u c i u r e o j c ,u o u r iu e m i s s i o n
pro cess
o j p o iu r
F igure 2: E xperim ental set-up
•u sS u iq n i jEjis.i3Aiun ‘(¿ 6 6 1) s is s q x a q d ‘-»JIROS 'V
' 9SZZ ( 8 Z.6 I) 81 9 'A sy sXqd ‘iq is y g ‘HJFH 'AV ‘UBqBpM h ’V ‘u m e g -9 M ^ s r ^ g
‘IS 9 I (ZL6\) 6Z 1131 -A3y 'sXqd ‘u u ;z 'A\ ‘p u s ip H 'AV ‘urajs>p 3
‘-» [[IM 'N
'£¿t? (£861) 3 £ I 'Id d n S ' s ^ d u o j p s i g puE s 3iu o jp s [ g ui -Apy ‘jq s d s 'y ‘s s o y h
' 111? (86 6 1) t S »W Sta aqosqE ^sX qj ‘uB qjn ">I ‘-»preH 'Vi ‘ss o y h
[I ]
AH 0 0 1 iv ssauiifSuq p3iüino¡vo atjj p u n a S v t f O A u o í p v a j x s w ssduiqSuq :g ‘d S ü i j O A uoipvujxa :ar¡ ‘aduSsp
¡ o j : j ‘sm p o j sounos p m j j i A :svu :sunS u o -ip a p p u d A a s f o sapuadojd si¡i /o u o s u v d iu o j :¡ a/qvj,
u o iid z u d
[ US- zlUD/v
[ US- z m o / v
S*l 0*1
Z’O ’ VO
A\-0 JZ
•y u m oj b j o s ju su n o jb dnoiS p p j 3 |3 ig sq j j o sjuoui3.insB3iu
oqj j o j[ns3j 3jb s s jn o s AV-Stig sq i j o uopBZiiBjod j o s s iS sp sq j puB pB 3jds XSjsuo s q x 'y rÍ9 '£
j o jusjano [bjoj b puB y\>] [ jsoui[B j o 3§ej[oa uopoEjjxo u b jb js ■ju o /y i'6 = 3 °1 ps-insBsui
s b m un3 sq j j o sso ujqS uq oqj [g] dnojS ano ut >)joav js iije s b uj ’UMoqs si X doosopiui
u o jjo o p ui posn sun8 u o jjo s p [E jsass j o s s iy s d o jd oqj j o uosiiBduioo b ajqEj SuiM oqoj sq j u¡
'sssaS oid ui si jsjouiojjoods XSjouo
siqj j o js s j s q x -japuioajoads p p i j SuipjBjoa dno X b p e jej b qj|A\ pornsEom sq [jim u o p n q u jsip
XSíouo o q x 'sasuaj opouSEui puB o u jo s p Xq posnooj puB s jo p s y a p ojqnop Xq p s p s p s p oq ueo
uiBsq u o jjo o p s q x 'ssou>piqj u u o jiu n qjiM pX bi b puB s n g j o sa n p ru js 3U||[bjsXjo 3[Suis b
p § oj sssooad SuijBsuuB jboijud X jsa b ui .»X B i-sng sq j joduioj oj juE jjodun si ji d p oqj ojuo
S n g j o uopBJodEAS io y y 'uopEJodEAS S u u n p siyb d p sq j ojuo iqS nojq sq ubo qoiqA\ ‘u sa o
-§ n g ue Xq apotre aqj j o ajo q aqj qSnoaqj s n g qjiM pajBJodEAS sq ueo p u s 3jejsoXjo u in ip q b
ojuo p aju n o u i si d p s q x 'Z '§ !J u í uA\oqs si sa jn o s ano j o dn -jss [E jusuiusdxa s q x
•(i 3|qE x s s s ) y u j o suaj
M3j e Xjuo j o s ju s u n s uiBsq je psqoBaj 3J3M a 33 '0 °J 1'O SB M0I SE pBSJds X Sjsus ub puB o/0g8
j o ¿ ss iS a p uopE zuE iod y -Suisiuioid X jsa s js m dnojS p p j s i s i g sq j j o sjjn sa j s q x
•J3j[ij uids b s e s p B jsX b )
5^3 sq j jEqj os
uids p s jja js jd e qjiM s s jb js uids Xdnoso s u o j p a p sq j s n g uj -uinnoEA ojui u sq j puB s n 3
sq j ojui JSiiiEq [suunj [ e i u s j u i u e q S nojqj p u u n j s u o jjo s p jEqj os usjsS unj j o ]3A9] lu u s g sq j
qiE3U3q UMop j[p
a x
sp u sq uoponpuoo sq j ‘d p sq j oj p p y uopoBajxs [E iusjxs qSiq e SuiXjddB
j a y y 'u o p e i u s u o uids oj jo sd ss j qjiM spuEq o m j o ju i puBq uoponpuoo sq j j o S u p jq d s e
o j spEO| uopoBJSjui sSuBqaxs opsuSEjM 'u o p E ju su o uids oures sq j 3AEq puBq- J p
ui s u o jjo s p |[E jEqj sn s j p j o jn j s^ p u n jj -puEq uoponpuoo puB p u sq 30U3[ea uso/wjoq puBq - ¿/
pszqEOOj p o jm J]Bq b SA\oqs s o je js opsuSEUioxiaj jsi ui s n g j o ain jo m js puBq oqx
R. A utrata and P. Schauer
Institute o f Scientific Instrum ents ASCR, Brno, Czech Republic,
e-mail: [email protected] isibm
Properties o f the scintillation - PMT detection system in SEM depend on the
properties o f the scintillator and the light guide, for all. Efficient electron-photon energy
transfer, very short decay time o f luminescence and an efficient transfer o f photons along the
light guide to the photom ultiplier are the decisive properties for the efficient detection system.
A variety o f the scintillation m aterials have been proposed for the detection o f signal
electrons in SEM. N evertheless, the range o f the suitable scintillators is restricted to powder
phosphor o f yttrium silicate (phosphor P47), plastic scintillators (NE 102 A) and single crystal
scintillator based on yttrium alum inium garnet (YAG) and perovskite (YAP) [1]
YAP single crystal has not been used for the detection o f signal electrons
comm ercially until nowadays. Relatively short wavelength o f maximum light emission at
370 nm w as a very serious limitation for commercial using o f YAP because the light
transm ission o f 370 nm w avelength is very low in light guide made from an organic glass.
Quartz glass is more suitable material but relatively expensive [2].
Further problem was very high self-absorption o f the generated light emission itself
(20% in 5 mm thick single crystal).
One more problem w as found in the polishing process o f the YAP surface. The surface
is very easily scratched and these cracks contain certain amounts o f im purities from
m echanically polishing material.
First o f all, a special light guide material was developed for transmission o f light o f
370 nm w avelength. This organic glass contains special organic dopants increasing the light
transm ission in a short wavelength region o f spectra (Fig. 1). The light from YAP is
transm itted to 95 % in the light guide rod 150 mm in length and 20 mm in diameter.
Secondly, owing to the additional treatm ent o f the YAP single crystal discs in oxygen
and hydrogen atm osphere at very high tem perature, the colour centres in the YAP or YAG
crystal lattice have been suppressed. Self-absorption o f generated light was halved to approx.
10% in YAP.
Thirdly, owing to the m odified technological process o f growing the YAP single
crystals and the additional treatm ent o f the YAP discs, the decay time 1/e has been shortened
from the older value o f 30 ns to 17 ns now, for the incident electron energy o f 10 keV (Fig. 2)
Fourthly, the polishing process o f the YAP surface causes penetration o f polishing
m icroparticles into the surface microcracks. They can be rem oved by a washing process only,
in a special mixture o f acids at a suitable tem perature. The smoothness o f the YAP surface is
decreased by this treatm ent but the relative efficiency o f the electron-photon transfer is
increased. Schema o f the surface relief is shown in Fig. 3A,B.
All these technological steps enable to improve the scintillation properties o f single
crystal scintillators, namely o f YAP, which has the shortest decay tim e from all single crystal
scintillators. Its wavelength is spectrally well matched to all alkaline photocathodes, so that
there is no problem regarding the choice o f PMT. The short wavelength light transmission o f
'I (£661) 6 jusuiajddns /ídoDsoiDi[Ai SuiuuEog ‘"a E}Bj}ny [z]
'6 (966l) E sisájeue puB XdoDSOJDiui uEsdcung ‘ y EjBJíny [|]
saD uajajay
oiiqndsy qosz3 sip jo saoiraps jo Ámspesy sqj Aq ¿o I Í90Z S °N JubiS 3qi J3pun psjjoddns sem íjjom siqj_
•£ 'Sid
d V A g /i
oz -
( l u u ) 4 ) 6 u s | 9 a b /w
g judUijvdAj ivuoiiippv
d lfl A d lfv
puv y
SSdD O A d Suilfsijod dlf} A d } fv dOV/AUS
djSuiS 91/)f o f d j p A
(su o¿
D lJV U ldljD S : £ d A t l S l J
DVÁd /l ‘su¿i = Jy¿d/[)
DVA Puv d V Á ¡° soiiSUdiomüifD dutij^
d jn S jj
■(i) ssvjS d i u v S a o
p d í j i p o u i ¿ ¡ A id u
p u v ([) ssvjS
d iu v S a o
p A v p u v js
JVIDAdUlUlOD 9 l{ l J o V A JD dds UOISSIUISUVA[£ . [ d A l l S í J
•JO J3 3 J3 P g s g 3 i[ j u i 3 ]B p o i d n p s s n s[B js X a a
s q j j o s s d A } []B j o j s i q s j í n s \ i o i B u ; i u p s
j e j s /ü o
s j S u is q v a
3 3 B [d sj
s j S u is j u s i o i j j s
Ki s a
j ] 'jA H S u í s j o p a j s p
m o u si j v a
s p o q jB O O j o q d X JA Id s q i o j < j y A
s q i u i Ä 0 U 3 I3 IJJ3 q S i q j o ju a u iS A S iq o E s s j n s u s
s s b j S d iu b S j o i u o i j
a \3 |\ [
u io jj js js u b jj
jq S q
s p í n á j q S q p a d o p A s p Xj m s u
R. Autrata, J. Jirák, V. Romanovský and J. Špinka
Institute o f Scientific Instruments, Brno, Czech Republic,
e-mail: [email protected]
Specim en observation at a low accelerating voltage o f the electron beam (around lkV )
is new and attractive technique o f scanning electron microscopy. W hile detection o f the signal
o f secondary electrons (SE) with the help o f the scintillation-photom ultiplier system described
by E verhart-T hornley [1] does not depend on the prim ary beam energy, the detection o f
backscattered electrons (BSE), having their energy only a bit lower than energy o f the beam,
is lim ited by a low sensitivity o f sem iconductor or channel plate detectors to electrons o f
energy below 2 keV.
The threshold energy o f the scintillation-photom ultiplier BSE detector w ith a high
efficient scintillator, made o f the single crystal YAG or YAP, amounts to 1.3+1.2 keV. The
light output o f the detector, achieved for BSE below the threshold energy, is very low and
practically unusable. Insufficient num ber o f photons, generated due to incidence o f one low
energy electron on the scintillator, is the m ain reason for this problem . The reasons also
include relatively low quantum efficiency o f all known scintillators (not more than 7%) and
com plicated transfer o f photons from the scintillator to the photom ultiplier [2 ],
Solution to this problem is to accelerate the BSE before they im pinge upon the
scintillator. If a positive voltage is applied to the conductive coating o f the scintillator, an
electrostatic field is produced in w hich all signal electrons (SE and BSE) em itted from the
specim en are gradually accelerated in the direction o f the field and they im pinge upon the
scintillator with an energy corresponding to the voltage on the scintillator.
There are several ideas how to form the geom etry o f this field [3], The easiest way is
to cover the scintillator w ith a metal conductive layer, positively biased to e.g. 4 kV. If the
scintillator is exposed to im pact o f BSE, i.e. located close beneath the pole piece, all fast BSE
are accelerated in the field direction and im pinge upon the scintillator. But in this case not
only BSE but also SE are accelerated tow ard the scintillator. The resulting image is formed by
a m ixture o f SE and BSE and represents both the topographical and material contrast o f the
SE detection can be suppressed by a retarding grid placed tightly under the scintillator
electrode and supplied by a low negative voltage o f approx. - 5 0 to - 8 0 V. The negatively
biased retarding grid shields entirely SE leaving the specimen. The detector detects only BSE
and the image o f the specimen contains the material contrast.
If an image with higher content o f SE is to be recorded, the detector must be retracted
into a position outside the im pact o f BSE. The SE detector m ust be located in a suitable
distance from the specim en. SE should be extracted tow ard the scintillator by electrostatic
field so that BSE w ith a relatively high energy reach the scintillator in a very low num ber
only. The scintillator is supplied w ith a higher voltage o f + 6 kV and the grid below the
scintillator is supplied w ith 100 V positive voltage. Som e sm all portion o f BSE can be
attracted and im pact the scintillator but the image character is dominated by the topographical
contrast o f SE.
'I L Z l /l O / Z O l
‘° N
'£¿ ‘I TOA ‘26 w g y n a '^ojj :ui “ o ÍBjnn " y BjBJiny [£]
'6 (9661) € srs/CjBuy puB Adoosojoij^ UEadojng ‘’y BjBjjny [¿]
' 9PZ (0961) L£ lunjjsui p s f ‘ PM'JH ^[UJOqi “g-j, UBqjSAg [i]
Iu b jS j s p u n o ijq n d s ^
\p s z j
aq j jo
Á o u s S y iir e jr ) s q i Ä q p s y o d d n s sbav p a [o íd s iq x
'jo}ü]]ijups difj uo sdSvjjOA snouvA aoJ dSvjjOA p u S SuipuvidA di{i 'sa sjvuSis H S 9+ 3S Puv 3 S 8 9HI :P ^A nSij
PJaS SuipAV}9A dlfí UO SdSüJJOA SnOlAVA AOf dSüJJOA AOJDJJIJUpS dlfj 'SA SJVuSlS ^ S S Puv 3 S dm '£ ZAnSlJ
uivdq difj f o dSvjjOA SuijDAdpoov di/j f o su o p o u nf so sjnuSis g g g p u v
d ijj
AojDdidp 3 S 9 ~ D V A I° uoijVAnSifuoo aSvjjOA m o j
£ '§!3
tal ße<n
[a] PHßn
Z 's !3
A001" =pu6o—X—
A08" =pußfl -**—
AOS’ =P!J6n -»■AO=pußfl —
\ s
3S9 PUB 3S J° sjnjxiui b jo g§ jo uoipspp
JOj SB [pM SB ‘(/\3>| 0£ puB £ usawpq) g § g Á§J3U3 qSiq ‘(a3>[ £ oj C'() uiojj SutSubj)
3S0 XSjsua mo[ jo uoipapp joj pasn oq ubo jo p sp p psquossp OAoqt; sqj ‘Xba\ siqj u[
R. A utrata, J. Jirák and J. Špinka
Institute o f Scientific Instrum ents ASCR, Brno, Czech Republic,
e-mail: [email protected]
The scattering o f prim ary electrons to so-called skirt, appearing in the range o f
pressures used in regim es o f LV SEM (low vacuum) as well as in ESEM , has no substantial
influence on the spatial resolution for comm only used types o f imaging in the scanning
electron microscope. It demonstrates itself ju st as a higher share o f the noise signal. W hat is
more, it brings known substantial advantages, as that no preparation or modification o f nonconductive samples is needed and observation o f liquid phase containing specimens is made
As regards the X-ray analysis, the situation is rather more complicated.
In the ESEM the presence o f gas in the specimen chamber influences the results
obtained by X-ray microanalysis, which is caused by two effects.
interaction o f electrons with the gaseous environment leading to generation o f
X -ray signals
scattering o f the prim ary beam electrons to so-called skirt, which causes generation
o f the X -ray signals also in the vicinity o f the observed area on the specimen.
The first effect can be partly suppressed by reducing the distance, w hich primary
electrons pass w ithin the gaseous environment. G. Gilpin and D.C. Sigee [1] carried out
experim ents with a sample o f carbon and cyanobacterium Anabaena cylindrica at accelerating
voltage o f the beam 10 kV and for 100 s lifetime. The specimens were cooled by a Peltier
stage to the tem perature o f 6-8 °C. The probe spot diam eter o f 0.7 pm was used. The typical
count rate was taken to be l- 2 x l0 3 cps. Peak/background (P/B) ratios were determined with
respect to the background in vicinity o f the characteristic peak. For the carbon specimen in the
high vacuum (where the influence o f gas does not demonstrate itself) a dominant carbon peak
was detected but also a small oxygen peak appeared because o f im purities in the material.
Insertion o f w ater vapor into the specimen chamber led to increase in the P/B ratio for oxygen
because o f contribution o f the w ater molecules. It is interesting that the P/B ratio for carbon
did not change substantially for pressures up to approx. 1300 Pa. For a frozen biological
specimen comparison was made with situation in vacuum as well ás with the same specimen
not frozen at higher pressures. A scope o f elem ents contained in the specimen was found up
to the pressure o f 684 Pa identical to that in vacuum. Despite the P/B ratio was reduced, o f
course. The measurem ents were carried out also in the argon atmosphere. An interesting
effect was demonstrated: for w ater at higher pressures a strongly nonlinear increase o f the
oxygen peak occurred, while for A r a linear dependency was found out. Further interesting
results concern the influence o f the specimen distance, etc. The authors stated that the X-ray
microanalysis in ESEM provides with useful information, which, nevertheless, needs careful
Robert A. Carlton [2] focused him self on optimisation o f working conditions for X-ray
analysis in ESEM. He changed the accelerating voltage o f the beam from 10 to 20 kV and
w orked at pressures from 66 to 665 Pa. He states that in most spectra the elements o f the
substrate were detected as well. But X-ray counts can be expected also from the construction
•| e u S is [ n j a s n
a q j s i s i q j jB q j o s a j n s s a i d j o a s B a a a a p a q j q ji M s a s E a i a u i p u o a a s l a d s j u n o a ijsj; j o a s q u i n u a q x
puč i
' s g i g u i p a j u a s a i d a iB
33BJJ3JUI ¡ n -
n ^ aqj
io j
s a ia u a p u a d a p a q j
S S [d u iB X g
■ sjun o o a s o q j j o j o o í s jB n b s a q j s jB n b a s ju n o a jb jo j j o u o ijB iA a p p iB p u B js a q j a i o j a i a q j
p u s ‘s a ijs ijE js u o s s i o j j o BU3JUD a q j j a a i u s j u a u i a in s E a u i [B ija a d s a s a q x ’X a u a p u a d a p USAlS
3 q j JOJ UOISSOlSaj JBSUq p u i j UBO SM ‘d pUB J-0 SSn[BA pajnSB SU I (OA\J JSEB] JE) [BJ3A3S JOJ sO pUB
( sa - na ) j o u o i jB u iu u a ja p pu B b + x >[ = X X s u s p u s d a p jB 3 u q a q j o j u o i j b u u o j s u b j j X g
■ sO + ( UI d - ) d x 3 ( SD - no ) = J-D
l u j o j a q j o ju ; u o ijB n b s s q j
I ^
s j u a \3 . i o j
s jq is s o d si
1 O = UI
PUB ‘33U O P 3 J3JJB 3S 3 J 3 M S U O aj33]3 []B J I p a U IE jq O 3 q p [ttO M jB q j 3JBJ JUnOO - s3
‘S U 0 JJ0 3 J3 X j B u i u d a q j j o S u u a j j B a s X u e jn o q j i A \ p a u i E j q o a q p |n o A \ jB q j a j E j j u n o a - na
í[ ( u i d - ) d x a -
i ] so + ( u i d - ) d x a no = J-o
SI 3U I[ Á B J-X U 3A lS B JO J 3 JE I J U n 0 3 |EJOJ 3 q X
'3 i n j B J 3 d u i 3 J 3 jn [ 0 S q B - X p u B JUEJSU 03 U U B U IZ J|O g - >( ‘(S U O JJO ap JO q j E d ) 33U E JSIP - X
‘p a s n
S u is q
sq j jo
u o ijs s s
sso jd
S u u s jjB a s
jb jo j a q j s i o
‘s i n s s s j d
sq j
s ia q M
SE U 3A lS SI S U Q JJ33J3 p S S I S d S ip
jo u j o
o ije j
J S q S iq
a q j a u iiS a j S u u s jjb s s
JB U35[EJ S JU 3 U I3 inS B 3U l
s in s s s id
aqj jo
3 ] 3 u i s p s[[B O o s
[ B J ja a d s J O
p o q ja u i e p s u S is s p
jo j
-a jn ss a id
U O IJE JodE JJX a S 3 [q E U 3
o j d n B i j a a d s Xb i - x
aqj jo
a a in o s
„ o is z ,, s q j o j s a i n s s s i d
p o q js u i
S iq x
’p a a u a n i j u i X jS u o jjs a q u b o X i j a u i o i j a a d s Xe i - x
X aqx
j o j u o i j n j o s s j [E ijB d s a q j ‘s n q x ' j o d s u r e a q a q j u i o j j u n í j o
a q j p u a jx a
s p a ip u n q je js a s s j o
a a u B js ip b
ubo ju a u iu o iiA u a s n o a s E S a q j u i p a ia jjB a s
S U O JJ3 3 |3 JB q j ‘J3BJ a q j UO p S S B q SI 3 3 U 3 n |J U I JJI>[S 3 q j JO U 0 IJ3 3 J 1 0 3 JO B S p i UIBUI §U IJS3J3JU I
q j|A \
[g ]
la d d y
3 jjO [jB q 3
u a s u a jo s -s p jig
u a S jo f j o
u o ijn q u ju o o
'UOIJBUIUIBX9 j a p u n SJU 3U I3J3 3 q j j o s j u s u i a j i n b s j s S b j j o a j s a o u i n u i i j d o s q j q jiA \ j u s j s i s u o a
‘3 § e j ) o a
S u ijE ja ja o o E
sq j jo
u o ijb s iu iiu iu i
u in j
o s jB
s p u a u iu io o s j
p u s s b S s q j u i q j i M s u o j j o 3| s X j B u i u d j o X jo jo s C e j j u i n u i i u i u i b jb p s u u o j i s d s q p j n o q s j A i g s a
u i s i s X |b u b o j 3 i u i X e i - x u i n u i i j d o jB q j B s p i s q j j j o d d n s s j j n s s i s ; h - a o u s D s s j o n j j s b S b p a A j a s q o
a q a i a q d s o u i j B u o S j b a q j u ¡ 'u i u i j j o
u a u i i a a d s a q j j E q j ‘p a j E u i i j s a a y
i s j a u i B i p u i n u i i u i u i a A E q p | n o q s p iB p u B js b s e p a s n
'o j aA B q j o u o p s a p i j j E d j a q j o a [ iq A \ ‘u i m j a a d s S u i j j n s a j a q j
a a u a n g u i U Ea u i a q j j o a u i o § ’j a g j B j p u E j o j a a j a p a q j u a a A \ ja q u o i j i s o d i i a q j o j S u i p j o a a B ‘s j s i x a
s a j a i j j B d q a n s j o j a a j j a u o i j B j u a u o u e jB q j a j q i s s o d s u i a a s j ] ' u i t u j o a d s X e j - x a q j o j a j n q u j u o a
uB a ja S iB j a q j u i o j j
a jja u iq [iu i
a q j jB q j p a M o q s s j u a u i u a d x g
auo jo
a o u E js ip
s a p iy B d
a jq B Jisa p u n j o
a a u a s a jd
'X j i n j a j E a u a s o q a a q p [ n o q s s iB u a j B u i a s a q j jE q j o s s j E u a j E u i
pressure (Pa)
Figure 1: Counts p er second fo r Cu Ka as a function ofpressure, acquired from a point on pure Ni, 10 pm away
o f a straight interface to pure Cu [3].
It is obvious from the figure that fo r the zero pressure the number o f counts decreases to zero, so that the signal
fo rm ed by electrons scattered in so called skirt is in question.
pressure (Pa)
Figure 2: Counts per second fo r N i Ka as a function o f pressure, measured in a point on pure Ni, 10 pm o ff
a straight interface to pure Cu [3].
'U lH C M 3 U IS M 3 U /S M 3 U /U y B /'> [p ,3 0 S lJ 'M M M //:d jJ ll ‘S J 3 3 J J 3 JJl>[S U lB S g JO J S U 0 J P 3 X IO 3
;1AI3S A I P UE JAÍ3S3 u! A jjauiojjoads Xbj-x " 3 0
‘p d d y
‘g f
‘u a s u s jo s - a p u g
’ 16-S8 (Z.661)
61 8uiuueos odoosojoi[/\[ u ojjoo |3 Suiuubd§ jBjuauiuojiA ug aqj ui s a p ijjB j jo sisÁ]buboj 3 ija¡
: y >[ ‘u o jjjb q [z ]
X bj-x oqt uo su o ijip u o 3 S uijB jad o ju a u in jjsu j a u io s j o sjo a jj 3 o q j
'%Z~ZZ (£66 [ ) 6 L\ Ádoosojoi[/\] j o jB uinoj -sdoosojoiui u o jjo a p Suiuubds [bjusuiuojiaus
‘uidjiQ [j]
aqj ui su a u ip a d s p o iS o jo iq jaM jo sisXiBUBOjaiui A bj-x " 3 'Q
‘ssS ig ‘ 9
s a a u a ja p y
’l/.ZI/I0/£01:oN IubjS ‘oi|qnda>i vpazj sqjjo XousSy jueíq sqj Áq psyoddns sem >|íom siqx
■sajnssajd ju a ja jjip
jo j
sanjBA ç
t? jo juoiuojnsBsui
XjisBa uoijisoduioo u a u ip a d s aru j aqj
ssassB oj auo sajqßua puB 1AJ3S3 u ! sis Xjbubojoiui Xbj - x aqj j o saijq iq isso d j o jiioaiOAOjdiui
snoaSBjuBApB Xjba si poqjaui pauoijuaiu aAoqe aqj jBqj a p n p u o a ubd s u q
jo j
J. Jirák, R. A utrata and J. Spinka
Institute o f Scientific Instrum ents AS CR, Brno, Czech Republic
e-mail: [email protected]
The advantages o f the scanning electron microscopy working at higher pressures in the
specimen chamber are connected with the possibility o f observation o f specimens structures,
w hich are difficultly observable w ithout previous preparation for m icroscopes working with
pressures in the specimen chamber under 10"2 Pa. The pressure in the specimen chamber up to
approx. 2000 Pa brings the possibility o f observation o f specimens, w hich release gases,
specimens containing liquid phase, including w et biological preparations, reactions on the
phase interfaces, etc. A t higher pressures in the specimen chamber it is also not necessary due to neutralisation o f the surface negative charge by gas ions - to coat electrically nonconductive specimens by a conductive layer.
For detection o f signal electrons at higher pressures scintillation and ionisation
detectors are usually exploited. In our case the scintillation detector is constructed as
a double-deck one and serves for detection o f backscattered electrons. YAG:Ce has been used
as the scintillation material. The scintillator with a central hole in the bottom detector ensures
the integrated function o f the BSE signal detector and the pressure-lim iting aperture,
necessary for limitation o f the gas flow at microscopes working with higher pressures in the
specimen chamber. The single crystals in both detection stages are sym metrically divided into
the right and left halves. The interface betw een them is provided with an optical reflecting
layer. Both detection stages thus provide two paired signals, separately from the left and right
half. The possibility for further processing o f signals from both halves o f the individual stages
and their mutual combination create preconditions for the achievem ent o f a higher material or
topographic contrast from the studied specimen [1]. The bottom scintillation detector is used
for detection o f backscattered electrons at working specim en/detector distances usually within
1 - 3mm. A t larger working distance and also pressure in the specimen chamber, the level o f
the detected signal decreases. In order to w ork at high pressures, when the signal detected at
the w orking distance o f mere 1 mm is o f bad quality, suppression o f the electron scattering in
the high pressure environm ent requires to reduce the working distance even below this level.
In the case given a significant part o f backscattered electrons escapes through the central hole
o f the bottom scintillator to the upper space. Hence, it is advantageous to detect these
backscattered electrons by the upper stage o f the scintillation detector.
At higher pressures the ionisation detector detects the products o f collisions o f
electrons with the gaseous environm ent in the specimen chamber. The gas ionisation is
contributed by the secondary electrons (SE) as well as the backscattered electrons (BSE) and
prim ary electrons (PE). The electric field formed betw een the electrode o f the ionisation
detector and the specimen, w hich is on the earth potential, provides the necessary energy for
ionisation to the secondary electrons and to the products o f ionisation. The computations o f
shares o f the individual sources as regards the total detected signal for specific conditions o f
the specimen observation [2] show, that the SE share is biggest, PE share is lesser, while the
sm allest part o f the signal is form ed by BSE. Typical dependency o f the signal level on the
pressure for different w orking distances o f the specimen is depicted in Figure 1.
'ÍLV'L9V (9661 ) 81 Suiuuess ‘jo js s js q snossBQ ssdossojsiy\[ u o jjssjg Suiuuess
[BjusuiuojiAug sqj ui uoijsejsjui se 9 -u o jjssjg : g ‘p iq x ‘ JATV ‘pi,:uoG ‘ d ‘Mi;psJfIAI [z\
•d iz p ‘S8 6 I ‘U!P39 '3b|.i3a JsS u u d s ‘Adossojsij/\[ u o jjssjg Suiuubss " g M suipy [[]
T ¿Z l/l0/Z0r°N JuejS ‘oiiqnday qosz3 aqjjo XsusSy IUEJ9 aqj Xq payoddns semíjjom siqj_
sqj jo
s o i o q s s q j Áq p s s u s r q j u i s q UE3 sttjba s iq j_ '[B S id A s i s j n s s s j d j o
je j u s u iu o j ia u s
srqB A u i e u s s b
s n o s s B S u i u o ijb s iu o i js B d u ii s q j X q u o ijE s ijijd u iB je u S is s q j j o u in u iix B u i s q j
j o j s s j s p u o ijb s iu o ; j o j ‘js q u i E q s u s u i p s d s s q j u ; s j n s s s j d S u i m o j S s q j qjiA \ SSSBSJ3SP j o j s s j s p
u o ijB ||[ju i3 S s q j Xq p s j s s j s p jb u S is s q j s p q M
's s u o p s jsjjB s s:> p E q s q j j o u o i j n q u j u o s e q jiM
s u o j j s s p X iB p u o s s s jo /p u B s u o j j s s [s p s j s j j B s s j p E q j o u o i j s s j s p s jq B u s js q u i E q s u s u i p s d s s q j
u ; s s j n s s s j d j s q S i q je s jb u S is j o u o i j s s j s p j o j p s j i o [ d x s s j o j s s j s p u o ijb s iu o i puB u o i j B i j i j u p s
uoisnpuo 3
•su su ip sd s jsa \ j o dn SuiÆip j o uoijusASjd j o ju;odA\siA sq j u io jj
snosSEjuEApB si j; ssiA usqjo puB ‘s s jn s s s jd jsq 3 iq pjBA\oj uinuiixBui sq j j o jjiq s ‘uoijESijqduiE
jeuHis jsq S iq ssu iij je js a s s s 3 u u q jib p n p is s j b j o pBSjsui sjnodBA jsjb m psjBinjBS sq j j o
s s f] usquiB qs u s u ip s d s sq j ui p ssn ju su iu o jia u s sijosseS sq j uo ju sp u sd s p si uinuiixBui sq j j o
u o p iso d sq j SB |[SM sb uoijbsiuoi jSBduii sq j X q uoijBSíjqduiB p u S is sq j j o ju sjx s s q x 'SSUBJSIP
8 ui>[joa\ S u ise sjsu ; qjiM SSSBSJSSP s jn s s s jd j o siïjba s iq g 'uiruuijdo s je sbS ui uoijbsiuoi
jsBdiui jo j su o ijip u o s sq j qsiqA\ jb ‘s jn s s s jd j o srqBA u jb jjss b je ssuBjsip 8 ui>[joa\ u saiS b
jo j uinuiixBui SMOqs pAS[ [buSis sq j jo js s js p uoijB |[ijups Xq u o ijs sjsp sq j s>Ji[ug
O Z f° suouiDdjd
t (uviuud fo
t&udud ‘y \ q ç £ fo d p o A }D d ¡d A O iD d id p di{i uo d S v j j o A d i f j ‘s u o d v A A djvM . Jo l u d u i u o u m u d
uoijvsiuoj sdouvisip S w y j O M ludjdffip uof dunssdud diji uo pjoS uáoa/ \vuSis difi f o úoudpudddQ
‘u o jD d id p
d jn S ij
L. Schneider and J. Jirák
D epartm ent o f Electrical and Electronic Technology, Faculty o f Electrical Engineering and
Comm unication, Brno U niversity o f Technology
Údolní 53, 602 00 Brno, Czech Republic
e-mail: [email protected]
The Environmental Scanning Electron M icroscope (ESEM ) is a new form o f Scanning
Electron M icroscope (SEM), which supports a gaseous atm osphere and the presence o f liquid
w ater in the specimen chamber. This opens up new possibilities for the use o f ESEM to the
study o f many biological and wet based systems, w hich previously could not be observed.
609 Pa is the minim um pressure, at which w e can observe w et samples without any
degradation by evaporation. Nowadays scanning electrons microscopes are called
environm ental if they operate with pressure from 300 to 2000 Pa.
Theoretical background
Ionization detector is a device, which uses the atm osphere in the specimen chamber
for the detection and amplification o f the signal electrons.
The theoretical principles o f gas ionisation were described in detail in [5]. The total
amplified ion current can be found by summing all cascade contributions. These contributions
are given by
prim ary electrons:
1 PEG ~
(1 )
PE i
secondary electrons:
I SE G = I P E Ô k a
(2 )
a nd backscattered electrons:
_ T
1 BSEG ~
1 P E '/ ^ B S e F ' 1
W here Ipe is primary beam current; p is pressure; d is working distance, d BSE path
length; Ô and q are SE and BSE coefficients and k is:
exp {ad) -1
a {1 - / [e x p (W ) - 1]}
w here a and y are Towsend first and second ionisation coefficients, respectively.
With know ledge o f the ionisation properties o f a given gas, we can calculate the
amplification o f electrons in the gas as a function o f pressure, working distance and electron
beam energy.
UlUl Pl pUD ¿
'£ £ = D snipvj Adino ‘uim j j = q smpvu uduuj ÁSojouifDdj q j j Áq pdADddud dpouiodp uojoajjoj :£ dunSij
0p O jp 0|0
■ m ui ç p u B p ‘£ ‘
z ‘\ J o j o a j s p p u e u s u i p s d s u s s M j s q s o u B j s i p S u i> [ jo m p u B \
0 0 £ s s S e jjo a
sp o x p a p q ji M p s j n s B s u i s b m souspusdap s q x 'J s q u i E q o u a u i p s d s a q j u; e j OZ s j n s s s j d
sqj j e p a m s B S í u s b ‘y d o t 71 ‘sjuauiuadxa ||B j o j s u i b s s q j s b m j u a u n o uiEsq X i B u i u d sq x
E J 000 l °j 00 I UlOJJ |EA J3JU ¡ u i JIE j o ainssajd S ui> [JO M sqj u o ‘U O qjEO p u B p | o S j o s u s u i p s d s
u i o j j p s u i B j q o ‘(B u S is aqj j o souapuadap aqj p a m s B s m s * s s p o j j o s p ssjq j sssq j j o j
sas - uiiu p i puB ¿ ‘£'£ sbm puE psSuEqo smpEj jsjno sq x '(uiui [ x ) suies
aqj s a p o jp a p [¡b jo j sem snipEj sp o jjo sp jsuui sq x 'JsquiBqo u su ip ad s aqj ui sjnssaid sqj uo
souspusdsp sjí pus [euSis p a p a p p sqj sp sjjE s p o j p s p sqj jo azis sqj Moq au iu u a p p oj jsp jo
Ul JOJOSJSp UOIJESIUOI sq j JO SUISJSXS s p o jjo s p JBJtJUUB JU3J3JJIP 33jqj pajE dsjd 3yW
fj] p3)J3¡3p 3JV 7¡¡¡qJo SsdÁJ SHOUDA (o) pUD (q) S3lUll¡OA Uf ‘p3)33/3p (jUlVlU 3JD (d) 3WHJOIÍ U[
p p if DUID3J3 uuofiun qjm s3pojj33]3 ua/nuun 3UJU3DUOJ í(q ssutnioň. uoipdidp fo uorioupdds :i dunSfj
Á JB iu u d
'S U O JJD S p p3J3JJE0S>[D B q J O U O lJtiqiJJU O O S u iM O jS J 0 3 J [0 0 S3UO
S j e j ‘S U O JJD S p
Á JB pU O aSS X jU lEU I S J3 3 J I0 3 S n ip E J JS9[|E U IS 3 q j q jj M S p O J J O S p s q x • I ’S l J UO UM OqS SJB JO J3 S J3 P
u o ijE s iu o i
sq j jo
s a p o jjo s p
J E jn u u E
o u ju a o u o D
q ji M
p s jB O o q B
s siu rq o A
u o ijo s js q
Results o f measurem ent
The measured dependences o f the relative signal level from the gold specimen on
pressure for all three electrodes are shown in Fig. 3.
Sim ilar dependences o f the material contrast are shown in Fig. 4. M aterial contrast, in
our case, is taken as the difference o f relative signals from gold and carbon specimens.
Both dependences are plotted for the working distance 4 mm. The signal level was the
highest for this working distance for all three electrodes.
■ a = 14 mm
• a = 3.3 mm
A a = 7 mm
P [Pa]
Figure 3: Dependences o f the relative signal level from the gold specimen on working pressure fo r electrodes
with outer diameter 14, 7, 3,3 mm. Working distance 4 mm, electrode voltage 300 V, primary beam current
140 pA.
■ a = 14 m m
a = 7 mm
• a = 3 .3 m m
Figure 4: Dependences o f the material contrast on pressure fo r electrodes with outer diameter 14, 7, 3,3 mm.
Working distance 4 mm and voltage 300 V. primary beam current 140 pA.
The experim ents proved that the relative signal level from the gold specimen, as well
as the material contrast, are the highest when the largest annular electrode is used. This
[3 ]
’d £01 ‘0661 SSSJJ OlUISpBOy ad03S0J3ip\[
S uiuubds [EjusuiuojiA ug aqj ui s o ia s q j o p s p Q snossBQ j o X jo sq x " Q 'O ‘so jb iiu b q
■d OZZ ‘S661 PJOJ x O ‘SS3J J uopu3JBi3 -ss se o pgsiuoj :- y ‘u o a p§ug
•(IJ03Z3 ui) -d \,L9 ‘£961 EW d ‘U N S 'Xjouiq B>[izXj BAO[n^3[0j^
-y ‘uo¿ pddjH
'ÍLV-L9V (9661) 81 Suiuub 3§ -jopapQ snossBQ ssdoosojDipv uojpsjg Suiuueos
[BjuauiuojiAug sqi ui su o ijo ejsju j SBQ-uojpsig : g ‘piqx
‘PIB uoq “j ‘qjipsaspv
'd ZZ\ ‘8861 SS3Jd oiuiapeoy
u o jp a jg
Suiuub 3 §
lEjusuiuojiA ug j o
:'Q '£)
‘so jb iiu b q
'oiiqndsx i)D3Z3 aq)jo ÁouaSy luEJQ 3m Xq papiAOid U£[/I0/Z0I 011 lUBJ§ 3lP Xq papoddns si jusiuusdxa aqx
's p o j p s p sq j
JO UOISUSlUip 8UISB3J3UI qjIM SMOiS JBqj [BuSlS p s p s p p JEJOJ 3qj OJ SUOJJ33p p3J3JJE0S>[3Eq JO
u o ijn q u ju o o sq j
Xq pasnBO
sp u s jj q jo g -sam sssjd
jsm o j oj ssaoui
a p o j p s p aqj j o snipBj
p souEjsip Sui^íom
Xq uoijEAissqo ajduiBS aqj jo j suoijipuoo jsa q s q x umuiixBUi si uoijesiuoi
•jsquiBqo u s u ip a d s sq j ui b j o o ^ - 00£ s jn s s s id jib puB uiui
are sp o jjD sp siq j
pB diui j o ssso o jd sq j u; g § g j o u o ijn q u ju o o jBqj os ‘g s g [jb jsouijb puB g s s p a q o o s p o jjo s p
Radek Dm ovský
Departm ent o f Electrotechnology, Faculty o f Electrical Engineering and Communication,
Brno U niversity o f Technology, Údolní 53, 602 00 Brno, Czech Republic
e-mail: dm [email protected]
Summ ary
This w ork is focused on the influence o f the primary electron energy on signal detection with
ionisation and scintillation detectors in the environmental scanning electron microscope.
1. Introduction
Ionisation and scintillation detectors are used for detection o f the signal electrons in the
environmental scanning electron m icroscope (ESEM). The principle o f the ionisation detector
is based on amplifying effect o f the impact ionisation. Signal o f electrons is amplified in
gaseous environm ent in the specimen chamber o f the ESEM [1], Scintillation detector detects
only backscattered electrons (BSEs) at environmental conditions. Energy o f these electrons is
sufficient to excite scintillation in the scintillator o f this detector.
For both detectors the level o f the detected signal depends on the gas pressure in the
specimen chamber. Experim ent is focused on influence o f the prim ary electron energy on the
signal detected with these detectors.
2. Experiment
Construction o f the used detector combines ionisation and scintillation detector in one system
(Fig. 1.). The single crystal (yttrium alum inium garnet doped with Ce) with the central hole o f
about 500 p.m in diam eter is located above the sample and is used as the pressure-lim iting
aperture [2 ] separating spaces with different
pressures o f gases. This design provides the
detection o f BSE with large collection angle under
environmental conditions. Thin ITO layer circle
electrode on the bottom o f the single crystal with
a potential up to 500 V affects the impact ionisation
in the gas environment and is used as a collection
electrode o f the ionisation detector.
The dependencies o f the signal level on the
pressure o f w ater vapours in the range from 100 to
1000 Pa were measured for different values o f
prim ary electron energies. The measurem ents were
carried out on the standard specimen o f gold with a
constant value o f the primary beam current o f
Figure 1: Configuration o f the detection
140 pA.
The measured dependencies are plotted for the scintillation detector in Fig. 2. Similar
dependencies for the ionisation detector are shown in Fig. 3.
‘81-ci'd‘(C66l) 9Z oaaa
[BjuauiuojiAug joj jo p a p Q u oxp ajg pajajjBosîpBg
ssdoosojoipv u o jp a ia S uiuubos
; >| ‘BjBJjny \z\
: f ‘B^iiidg ‘"f
SumuBOS |B iu3iuuoiiA ug sqj u;
so ia sq
'd jol
s s a i j oiuiapB oy s d o o s o jo i^ u o jp a jg
j o p a p Q snossBQ j o A joaqx : ' a ' 0 ‘sojb |; u b q [[]
S 3 .1 U 0 .1 0 p N
A OOt' spojiaa/a jo p a p p uoiiostuoi uo aSoj/OA
‘w u i
p aauoistp S u i ^ u o m
ý ja y
Z l PUD SI ‘OZ tâusua uoupaja
lù v iu u j .lopatap uoiivsiuot uof smodvA jatVM fo ajnssajd uo p¡o 8 uiouf pAd] p u S is f o sapuapuadaQ :ç a jn S tj
(ed) ajn ssa jd
wui [ aouDjsip SutyjoM A ^y Z l Puo SI ‘OZ A ÏÏJ J U J u o jp a p
jopa/ap u o fw fíltu p s j o f sjno dm ja/DM f o ajnssajd uo pjoS uiojf
/a ta /
¡ouSis f o sapuapuadaQ
iù d u iu j
:z a jn S ij
(Ed) a jn sssjd
00 U
•X S w U S U O j p S p
XjBUIljd S lllS B S JO S p
qjlM S3SB 3J3U I Ä jyos p A 3 [ JB uS lS 31]^ 'lU SU IU OJIA US SE§ UI S U O j p S p XjBUIljd X § J3 U 3 J 3 M 0 [
3qj j o 8uu3J}E3S jaq S iq 3 t |j S3}BSU3diU03 X[qBqojd j b u S i s p s p a p p aqi j o XSjaus J3 A \o [ qjiAV
3 S 3 J° uotjBstuoi p B d u q aqi j o u o ijn q u ju o o ñuisB.uoin sq j j o p a p p
u o ijb s iu o i
jo j
pA3[ jbuSis puB ooj s u o j p a p paj3HB3s>[0Bq puB XjBuiud j o SuuaijBos sasnBO sjnodBA j3jbm
sq j j o a jn ssa jd Suiseajoui s q x 'sa iS jsu a jsm o j jb s 3 ¡n 33 |o u i sbS qjiM s u o j p a p p 3J 3UB3s>pBq
puB XjBuiud j o SuuauB os j3 § jb | sq i si u o sb sj s q x 'XSjaus u o j p s p XjBuiud SUISBSJSSP qjiA\
s s in p a p pA 3 | [buSis aqi jBqj Moqs j o p a p p uoijB ](pups aqi jo j sa p u a p u s d s p p am scau i s q x
u o |s n p u o 3 •£
V. Romanovský, V. Neděla and R. Autrata
Institute o f the Scientific Instruments AS CR, Brno, Czech Republic, e-mail: [email protected]
Commercially available Environmental and Low Vacuum scanning electron microscopy
permits a low pressure o f gas in the specimen chamber [1], These instruments are possible to be
used for observing uncoated nonconductive or pure conductive or water containing specimens
without the necessary drying and using other special preparation techniques. Contemporarily,
these microscopes come in use mainly in biology, medicine, as well as in the semiconductor
industry. As the principle o f these microscopes allows the presence o f gas in the specimen
chamber, the area o f their application can be much wider. It is the ability to stabilize specimens
thermodynamically for static experiments or effect o f chemical or kinetic changes for dynamical
experiments that is relevant. However, the situation becomes more complicated because the gas
surrounding the specimen plays its role at magnification and consequent detection o f the SE
signal and takes part at the creation o f the signal background. The aim o f this paper is to provide
a background into the behaviour o f various gases in the environmental SEM as an aid to
researchers designing experiments.
amplification o f the ionisation detectors in
environmental SEM microscopes have
been a very complicated matter. For partial
simplification the parallel electrode system
has been considered, which is presented in
Fig 1. The size o f the resulting
amplification o f the detector is dependent
on the gas pressure in» the specimen
chamber, working distance, energy of the
primary beam electrons, type and
properties o f the working gas and the
sp e cim e n
voltage on the electrodes o f the detector.
Figure 1: Configuration o f ionisation detector fo r
The computation o f the amplification has
environmental SEM
been analyzed in [2 ] by using the
Paschen's law. Then, we can calculate the
amplification by applying the product o f pressure and distance. The voltage on the electrodes can
be increased only up to a certain limit, from which the sustained discharge starts changing to the
self-sustained discharge. Our aim was to find this limit, because from the viewpoint o f the
amplification it represents the best working conditions for the given configuration o f the system.
Fig. 2. shows the realized measurements o f three parameters obtained from the specimen by line
scan. The first parameter (T) is the signal magnitude from a scratch on gold, which characterizes
the significance o f the topographical contrast. The second parameter (M) is the difference
•ÇC61 pJOjx'0
‘s s s j j
XjisjsAiuf] p jo jx o 'P3 Z ‘s s s b q psziuoj
'0 6 6 1
u o j j s s [ g S u iu u b o s p j u s u i u o j i A u g s q j u t
s s ia s q jo js s js q
s s s jj
‘p S u g [2 ]
s i u i s p B s y ‘s d o s s o j s i p \ ¡
sn o ssB Q j o A jo s q x :
' 0 ‘s o j b ]i u b q [ [ ]
o iu g
I S I J ° Ju b j S [ t îu jo iu i Xq p r o ‘ U J I / 1 0 / J 0 I ° N 'd o > [ i p a z g s i p j o A o u s S y j i r e i g s ip \ q p s y o d d n s
p s fo a d s iq x
A00£ I o sdpouiodjd di¡i uo s S d jj o a pun utut£
9DUVJSJP SuiyuoM difi id SdSüS lUdAdfßp Aof y¡ (q pun j (v dzis /vuSis di¡i Jo sdpudpudddp pdunsvdui difj :ç dunSi
ooe f
00s i
■jBoipBJ j o sdAj 9j s s q j j o s s in s s p u i sqj uo
sjnjdBS 3i(i Aq s u o jjs s p jbuSis j o sso] 3i(i Aq pssriBO [buSis u; SSESJSSP B sbav jjn ssj sq x 'uo ijssjsp
3qj JO 3Sn
sqj qjiM sju siu u sd x g jsquiBqs |B iju sjsjj|p sqj ui pssnBS si ssb sjsu i sjn ss sjd sijBUiBjp b ‘sjn jjs d s
sqj qSnojqj uoiSBAJsd Asbs j p q j puB ssjnss[oui sji j o szis [p u is sq j j o ssnESsq ‘jsasa\oj-[
B j 0001 SAoqB sjn ss sjd b jb 5|Bod b psjiq iq x s qDiqw
‘u in q sq si s b S S u i j s s j s j u i js q jo u y -ip sqj ui u s 8 o j j i u
SSUSSSjd % 81 JO SStlBSSq ‘JBJIUIIS /ÜSA SJE UsSojJIU pssn sift f o soopns st{¡ Jo ¡W13Q .7 ajnSy
puB JIB JO J UOlJBSIjqdlUB Sqj JO S3Ajn3 3qX '1|3M SB
u o 8 j b q jiM psuiBjqo s j s m sjjn ssj pooS /CiSy\ -su su ip sd s
S u i u i b j u o s -j s j b m
sqj j o
jo SUISBSJSSP 3 q j si qsiqM ‘sSBjuBApB j b s j S jsqjouE
SI 3J3qX SjqBJIBAB /([ISBS SI puB SSIJIJiqB SuiiijqduiB
ju s jp s x s 3ABq qsiqM ‘jnodBA j s j b a \ qjiA\ psuiBjqo u ssq
3ABq s)|ns3j }ssq s q x '9J S
S A ijeS su o jp sp sqj puB
‘u in ip q ‘u s S o j j i u ‘jnodBA jo p w ‘j i b j o j jb u S is [aj
u o 3 jb
puB X jo ssnjBA pajnsBOUi 9JB aaaqj q
‘b £ 3 ; g
•|bu 8 is ui JSBJJU03 sSpa sqj j o sjBqs sqi ssu iu u sjsp
j j o azis s q x 'qsjBJSS sijj j o sS ps sq j u io j; psuiBjqo jbuSts sq j si ( j ) jsjsuiBJBd p jiq j s q x 'jsBjjuos
¡bijsjbuj sq j j o sjBqs sq j SMoqs jaj j o szis sq j ‘s jo js js q x uoqjBS puB p¡o 8 u io jj S]Bu8is ussM jsq
V. Romanovský*, V. Neděla, O. Hutař**
Institute o f the Scientific Instruments AS CR, Brno, Czech Republic
* Department o f Electronics, University o f York, York, Great Britain, e-mail: [email protected]
** Cypress Semiconductors, San Jose, California, USA
The environmental scanning microscope is based on convenient adaptation o f classical
SEM employing the focused primary electron beam. For the proper function o f the electron gun
the pressure in this part o f the microscope must be less than 10"3 Pa. The working pressure in the
specimen chamber in environmental microscope reaches the order o f hundreds o f Pa [1].
Therefore, the effective vacuum separation o f these two parts is necessary. This can be obtained
by the use o f two or more apertures limiting the gas flow. It is also necessary to evacuate the
space between the apertures by a vacuum pump o f a high pumping speed [2], The increasing of
pressure in the specimen chamber brings some difficulties, especially the dispersion o f signal
electrons by the influence o f interaction with the gaseous environment.
Figure 1: Simulation o f dispersion o f PE in the specimen chamber: at the pressure o f 500 Pa, 63.7% PE is not
dispersed while at 2000 Pa only 16% PE is not dispersed.
The simulations were carried out by using the program “Electron flight simulator, ver.
3.1” intended especially for calculations o f dispersion o f electron beam in gases and solids. The
program uses the Monte Carlo method for calculation o f the electron trajectory. In Figure 1 there
are the results o f the simulation for air and energy o f primary electrons o f 20kV. Owing to
dispersion o f primary electrons, caused by the collisions with gaseous environment, the primary
electron beam is surrounded by a skirt. Thus, the primary beam spot on the specimen broadens.
The skirt o f scattered electrons is distributed over such a large area, as shown in Fig. 1,
that the average over the specimen structures produces a nearly constant background, which can
be subtracted by adjustment o f the black level o f the signal. A highly resolved image can thus be
obtained from the signal generated by unscattered electrons. Under certain observation conditions
a resolution comparable to conventional SEM can be obtained also in the environmental SEM.
•s i s A | e u b o j s i j /\[ p u B
u o ije u u o j s 8 e u i[ j o
soisÁqj ‘Xdoosojo;p\|
u o jjo s jg
(8661) U0 R!P3 puC
Suiuireos :-g ‘j s i u p y [3 ]
•8661 ssaid
D iu is p E s y
( B ju s iu u o jiA u g 8 u iu u e s § u o j j s s [ g
JBLU9JUI Aq pire ¿0ïÇ902‘S ssousps jo Xuiapeoy ipazD
s u o ijE p u n o j : 'Q '0
‘s o j e j i u e q [ [ ]
•oiug isijo pa (bid
jo XousSy jubjq aqj Xq psjjoddns sem pafojd siijx
Ájuo (eajsjui psjfuiq B uiqjiM psijijsnf si iAjgsg jo sijsuojsBJEqo b sb p pue
jo jsnpojd sqj jo ssn sqj ‘sousg 'Sjnsssjd jo ssusn|jui sqj o; joijsju; st Xsuspusdsp sqj puB
jsjsEJEqs JBSuq jo si ssuBjsip SupiJOM Suisesjsui sqj qjiM jusjjns ^reunid sqj jo dojp sqx
•szis ppijMsiA sqj joj sjusuisjínbsj sqj jssui ÁjjjEd puB suoij|puos uinnsBA sqj 8uiqsBSj
joj sjqEjms sjb qoiqM ‘ussoqs sjsm sjnjjsdE uiojjoq sqj joj uiiu ç'o puB sjnjjsdE jsddn sqj joj
uhhSI’OJ° S3Z!S 3M1 JusuisjnsEsiu puB suoijBjnsjES sqj jo j ’ujui p oj dn ssuBjsip SuppoM sqj ssn
oj pire BdOOOS OJ dn jsqureqs usuipsds sqj ui ssjnsssjd qjiM >|jom oj s|qBU0SESJ si ji suojjssp
XjBiuud jo juiodMsiA sqj uiojj puB psjspisuos uoijBjnSijuos sqj joj jBqj spnpuos ues suo
•(qjSus) Ajojosíejj sqj 7 puB ss[nss|oui jo
~ b<
= {X )J
i p s j i o j d x s s e m u o i j B p j S u iM o n o j s q j K s u s p u s d s p s iq j j o u o ijE |n o |B S s q j j o j ’u i i u j j o
s s u B js ip S ui> [jo m s q j je jiE j o j p s j B j n s p s s i U 0 isi[[0 3 3U0 q jiM puB u o i s q j o s Kub jn o q jiM u s u i p s d s
s q j u o S u ip u B ] u o j j o s p s q j j o Á jq iq E q o jd s q j
' S y u j - s j n s s s j d S u i s e s j s u i s q j q jiM SSBSJOSP
Á[[E3ijBUiEjp h im j o j s s j s p g g g u o i j E q i j u p s s q j j o X s u s p y j s s q j jB qj o s s d E q s J E |iiu is b SABq [jim
u o i s j s d s i p g § g s q j j o Á o u s p u s d s p s q x 'ju n o u iB j s j b s j S e o j s s A p s u i s q j S u ij js s s e jje js s b § puB
(g < j) s u o j j o s p /Ü B u iu d s q j u s s A u s q s s s s s o o j d u o i s i q o s s q j s js q M ‘b ,j q q z
UES s u o
p jo q s s j q j b s s i j o u
'S i J u ] •(di ) j u s j j n o Á jB iu u d s q j j o d o jp s q j o j s s u s n |j u i js S S iq SEq js q iu B q s u s u i p s d s
s q j u i s j n s s s j d s q j jB qj p s M o q s ju su isjn sE S iA i 'u i u i 8 o j o m o jf s o u B js ip S u i> [jo M ^ q j je b j OOSJ °1
0 1 w o y s jn s s s jd s q j jo j p s u iu u s js p sem Á su s p u s d sp s q x 'y d
00I7 o j
j s s s e m lunnoB A u i j u s j j n s
s q X 'j u s j j n o u iB sq X jB iu u d s q j j o SuiSESJO sp u o (p ) so u B js ip S u i^ j o m s q j puB (’id) js q u iE q s
u s u i i s s d s s q j u i s j n s s s j d s q j j o s o u s n g u i s q j X jiju E n b o j p s u u o j j s d s e m ju s u is jn s B S j^
(p) d o u D i s t p d i f j u o p u v
d ifl Jo Á o u d p u d d d p d ifj
udjß? puv uoisijjoo v tnoifjm udiupdds di/j uo Suipuvj
mvdq uoujodjd Jo Á.'/ijiqvqoud pdjvjnojvj .7 dunSij
p j ) u d q iu v ifD u d i u p d d s u i d u n s s d u d
u i ¡udA uno u d iu p d d s d ifx
:ç d u n S ij
< --------
[D d ld
0000 L 000 L 001
tvd] d|
P. Schauer and R. Autrata
Institute o f Scientific Instrum ents AS CR, Brno, Czech Republic, [email protected] isibm
1. Introduction
In our laboratory the equipm ent for the investigation o f cathodoluminescent (CL) properties
o f solid specimens has been built [1], The equipm ent is based on the completely rebuilt Tesla
BS242 electron m icroscope, supplied by the light collection UV transmitting system, and by
a fast and efficient photom ultiplier tube (PM T) as the detection unit. The CL properties o f
specimens (slim disks) can be measured in continuous or pulse modes. So, in addition to the
efficiency and decay times the emission spectra can be measured for 10 keV electron beams.
2. M easurem ent system
The layout o f the m easurem ent system is shown in Fig. 1. Contrary to the direct measurem ent
o f the efficiency or decay characteristics, at the spectral m easurem ent the CL emission is
guided to the entrance slit o f the m irror m onochromator (Carl Zeiss Jena SPM 2), and the
PMT (EMI 9558B) is positioned at the output slit o f the monochromator. As the signal to
Wehnelt control
Beam current
Specimen temperature
S /-M e te r
IEEE-488 bus
Light guide
chrom ator
HV Supply
CL Signal'
Lock in
S /-M e te r
Figure 1: IEEE-488 bus layout fo r the control o f the CL emission spectra measurement.
dSuvi/DXd v jv p
u o fpuv
S u jijo j ju o j
Suissaoov u o f a o o f m u i SSP~33313lH Suuvij puD O g jj o [ j j V0S£Z8 fuans V 3H1 :Z b m S i j
•S3SuBl[D 3|JJ|[ qjIM jaqjOUB OJ UiajsXs 3U0 UIOiJ
|3A3] apoa aaanos sqj jb a[qBjiod si (jua[| 8 y puB ¿ j j Xq p ad o p A ap - 3 3 IS ) XjBiqix lOJJuo^
luaiurujsiq pjBpuBjg ju ajiS y aAijBiuajiE aqj ‘ySIA °l Xjbjjuo ;} 'UiajsXs auiBS aqj uo uru oj
sjuauinjjsui juajajjip smojie jsqj prepuBjs (aauB qiy uiajsXg) XB|d5pSn|d [XA 3qi °1 Suipjoaas
pauSisap si ji asnBaaq ‘juauiajnsBaui aq) joj paxiajajd sbm XiEjqq o/I VSIA 3Ml '86 SM opui^
jo j sjaA up o/I (VSIA) s jn p a jiq a jy aaEMyos juauirujsuj p n jiiA ju ajiS y Suisn pa|[ 0jju 0 D
sea\ pjEa g u o aq x 'sajqEa q i j o b!a sjuauirujsui paddinba g id Q p npiA ipui oj papauuoa
pjBa siqj Xq p auu o j si snq g u o aqx -prepuEjs XEjd-puE-Snid aqj qjiM ajqijEduioa puB
ajqE jnSijuoa ajBM yos si pjBa aq x s/g^i OS/, oj dn (yj/\[Q) ssaaaB Xiouiaiu jaa.qp oj jo u a d n s si
jEqj aauBouojaad UiajsXs puB o/I sapiAcud S uuajjnq aq x 'saajsuEjj snq 1 3 J aqj lu o y sjajsuBjj
9 Id O sajdnoa-ap S uuajjnq ui-j[inq qjiM pjBa siqx ' Z '^!3 ui UMoqs pjBa aoBjaajui g iJ Q I3 d
VOS £28 saiSojouqaax ju ajig y aqj Suisn uaAup si snq g iJ Q aqx 'BJBp pauiBjqo aqj sassaaojd
osjb j E q j ‘j j aqj Xq pa[joajuoD si snjBJBddB Suunssaui aqj pus ‘(piEpuEjs 881^3331 SE
umou)( ‘g id O ) snq aaEjjajui asodind [EiauaS aqj oj pajaauuoa ajB sjuauirujsui [BnpiAipui aqx
snq ju » u i 3 jn s e 3 ui p u c io jju o 3 •£
qojjuoa j p u q a ^ pijuanbasuoa
aqj oj puB juauiajnsBaui j u a j j n a uisaq aqj oj pasn s b m a§Ba XBpBJBj aqx 'pajnsBaui aq uBa
amjBjaduiaj aqj puB ‘pajBaq aq ues u au ip ad s aqx 'uauipads aqj aAoqB paaBjd si uiSBjqdEip
Su;>[UB[q aq x 'apmS jqSq aqj SupBj [ j ] aaquisqa u auipads aqj ui pauoijisod si (XjBssaaau j i
ajEjjsqns juajEdsuEjj b qjiM) uauipads j u a i s d s u E j j pajESijsaAui u y uosuas u 0 ijn| 0Aaa e Suisn
paqqsaS si yiqs qjSuapAEM aq x '3 d puE jajjaAUoa y /Q aqj Suisn 3[qB|iojjuoa jo jo u io jjaap
aqj qjiM jiun [ j b u i s b Xq paaB]daj s b m qjSuapAEM jojBuiojqaouoiu aqj jo jjiqs jEnuBUi
aqX 'JOjEJauaS SuijBjnpoui uiEaq aqj jo jndjno SuuaSSuj aqj u i o j j ua>(Bj si a s p d aauaaajaj aqx
•aajauij|OAOUBU ui-^ao] b pus uisaq pajEinpoui aqj Suisn apoui snouojqauXs b u i uoissiuia x 3
aqj ajnssaui oj Xjsssaaau si j i ‘h e u i s Xjba b j b sauq [Bjjaads jo j s o u i jo s o i j b j punojS>(aBq/asiou
M 3
10 C o nfig - 10 L ib ra rie s C o nfig u ra tio n
O ptions
H elp
T h is utility configures 10 interfaces. It m ust b e run w hen e ver a new 10 interface is installed in the
com puter or when ch an ge s need to be m a d e to an existing 10 interface.
T o configure a new interface, se le ct it in the A v a ila b le Interface T y p e s list a nd c lic k on
Configure. T o e d it a p re vio u sly configured interface, se le ct it in the Configured Interfaces list
and d ic k on Edit.
A v a ila b le Interface T y p e s
Configured Interfaces
S IC LN am e
n /a
n /a
Auto A dd
H elp
Figure 3:
V IS A N am e
82335 GPIB
82340/82341 GPIB
82350 PCI GPIB
E2075 GPIO
E8491 IEEE1394 to VXI
LAN Client
LAN S erver
V IS A LAN Client
VXI Com m and M odule
B e m o ve
The main window o f the IO Config. This utility assists fo r the setup o f the GPIB
lö Ä - V IS A A ssistant
V iew
GPIBO 9: 0 IN ST R l
riDIRft;•17"IMCTD -—I
Instrument Driver | Formatted I/O | Memory I/O | Attributes |
No Instrument Driver Configured
Address String
|g p ib O::9::0::INSTR
Configure: _ |
Prim ary Address: |g~
Secondary Address: |q
For Help, press F1
Figure 4: Typical window o f the VISA Assistant utility showing the system instrument view.
•juud ui ‘
(ZOOZ d3S) ‘bouJV qi'ios ‘uBqjnQ ‘
ç\ IAI33I jo oojj : » ‘BjBijny put: ¿ ‘ja r a n p s [z\
'Zl~\ ‘6 (9661) 'lddnS ’asojoiiM guiuuBag r j \i3nBqas pue y ‘E jB jjny [i]
qoszQ sq j j o X ausSy
ju b jq
aqj jo 8 1SO/I 0/303 '° N JUBj S aqj Aq pajjoddns
ay q n d sy
s;q ]
sea\ jjjom
•sauB jqiq o i JuojiS y sq j 3u isn + + 3 p r a p o g a q jjii u sjju m sbm
(su o jp a x io o pire uoijBjqi|B3 S u ip n p u i) sjBM yos Suissaaoad puB SuunsB siu BjBp a q x 'pauiBjqo
3J3M B jp ad s UOISSIIU3 ÄJIAIJISU3S 3p0qjB30J0qd 3qj JOJ SUOlJ33JJOa [BuSlS pUB UOlJBjqi]B3
qjSu3[3ABM aqj j o
uoijE aqdde Ja y y
'3[qB[iBAB ajaM su o ijejad o ajo u ia j put; ‘>(joA\jau
3i|l oi P3J33UU03 sea\ 3 j o q x 'sju su iru jsu i p sd d m b s 8 8 ^ 3 3 3 1 ja q jo p u s ‘sja ja u iy p u u y I ()t?H
ju a y S y ‘jiu n qajiMS u o ijism b a s BjBp V0Z.617£ Ju s ijS y sifl u io jj/o j snq aqj qS nojqj A\oy
BjBp aqj |ojju o o o j 3|q ;sso d sbm ji ‘sa u E jq ig 0 /1 lu s ijS y a q i puB 86-SMopuiyw SIM 3lP S u isy ijfl
i u a u i a j n s n a | \ •§
sju au iru jsu i aqj qjiM p a jsp o ssE are jsq j sajnqujjE aq u o sap oj puB ‘se sje Kjouiaui
ajuM p u s pB3j oj sn pawo|[B osje ju E jsissy y s i A '0 /1 pajJBUUOj jjo d d n s sjuauinjjsui
oj sS u u js a A p aa j puB puas oj p a sn sbm y -snq g u o aqj j o jsa j aqj jo j pasn sbm (p 'S i j ) Kjyijn
ju E jsissy V S IA 3ll l 'lUBiSojd aqj u; payBa uoijau n j aqj j o S u u js jajauiBJBd aqj ui passBd si
uoijB uuojui s iq x 'aaB jjajui a q j jo j sarqEA u oijE jngyuoa XjBssaaau jaq jo sb [[a.w se ‘pjBa g |< |0
aqj oj ja q iu n u jiu n jboiSoj puB auiBU aoBjjajui aqj pauSissB (£ '3 | j ) S y u o 3 o i 'sauB jqy o / I s'il
qjiw p a sn si ji a jo ja q pa.inSyuoo aq jsnui aaB jjaju; a q x 'sa ijy ijn ju E jsissy y s i A PUE S y u° 3
01 ju a p S y Sujsn juauiajnsB aui aqj jo j p a jsaj puB pajsnfpE ojoa\ snq puB pjBa g [ j Q a q x
uoHE4n§ijuo3 o /I
G. Schonhense 1 and H. Spiecker 2
’institut fur Physik, Johannes Gutenberg-Universitat, 55099 Mainz, Germany
2LaVision Biotee GmbH, 33619 Bielefeld, Germany
e-mail: [email protected]
Scherzer’s famous w ork on the impossibility to correct the chrom atic and spherical
aberrations in electron-optical systems com posed o f round lenses [ 1] has initiated many
attempts to find a w ay circumventing his theorem. Despite o f extensive work, it was only
very recently that the spherical aberration o f a TEM w as successfully reduced by multipole
elem ents [2 ] and mirror correctors are ju st ready to prove their capability o f correcting
chrom atic and spherical aberrations in cathode-lens microscopes [3], Technically, both ways
are very demanding.
Besides m ultipole- or m irror-elements there is another way o f circumventing the
preconditions o f Scherzer’s theorem, i.e. the application o f tim e-dependent lens fields [4],
Early attempts, the so-called “high-frequency lenses” (for an overwiev, see [5]) failed because
the crucial phase condition betw een the microw ave-excited lens and the necessary bunching
o f the electron beam could not be fulfilled. W e present a novel theoretical ansatz for both cc
and cs correction in photoem ission electron m icroscopy (PEEM ) and low-energy electron
m icroscopy (LEEM ) making use o f the highly precise time structure o f pulsed photon sources
like electron storage rings or pulsed lasers, e.g. driving the photocathode o f a LEEM. The
new approach is based on rapid switching o f acceleration fields or lens fields and / or timeresolved im age detection. A s the electrons’ tim e o f flight is an im portant parameter, the
method is especially useful in low-energy microscopy. We do not aim at the ultimate
resolution like m ost o f the previous attempts on aberration correction,, but rather at
a substantial enhancem ent o f the imaging performance at m oderate resolution in the range o f
a few nanometers.
C, - correction
The contribution o f the chrom atic aberration is most important for Synchrotron-radiation
excited PEEM (X-PEEM ). Due to the high photon energies in the-soft x-ray range, the energy
distribution o f the low-energy secondary electrons being imaged is several eV wide thus
giving rise to a dominant chrom atic term 8 C= cc a AE / E. The cc-corrector exploits the timeof-flight o f electrons with different energies in a drift space [6 ], The spatial and temporal
dispersion at the end o f the drift space facilitates three ways o f correction:
1. The straight-forward way is to actively vary the excitation o f p a rt o f the lens system
synchronized with the pulsed electron beam such that for all electron energies (arriving
one after the other) proper focusing is achieved.
2. An alternative way is to introduce a switchable (parabolic) acceleration field behind the
drift space w hich will invert the energy distribution o f the electrons. As all electronoptical elem ents behind the accelerator act on the inverted energy distribution, the
'SJ3S[nd o iu o jp s p jjB-sqi-jo-siEis juassjd j o siseq sq j uo u o p o sjjo a
so put: 3o j o suopBuipsa Q\pRiput;nb a \ i 8 [|im 0/\\ sosbojoui uopB.uoqE opBUiojqo ‘m rp ui ‘put;
pguspBOjq aq him uopnqujsip XSaaus aqj asiMjaqjo asnBosq qojiMS aqj j o su ip -ssu jsbj Xjsa b
s s jin b s j siqj ‘X[[E;u3 uiu a d x g '( p p p sus[ sq j j o y e d SuiSjsauoo aqj 3'i) jusuoduioo aojoj
[BipBJ 3AUB7Ï3U qjlM UOl8sj 3qj JOJUO SUOflD3(3 Oqj SB UOOS SB p[3y SU3( sq i JJO qDJIMS X(pidBJ
oj si Xjqiqissod s u o '[ , ] sSbuii [BnjjiA s 4}i ui (°d pus) so psiWAUi ub qjiM susj SuiSjSAip b
8uisi[B3j j o sXbm usdo sp p ij juspusdsp-auip jBqj p b j sq i j o asn ss>(Biu jj 'jBuiojqoB [Bopdo ub
s q q u issa i j o p 3jjon-so aqj ‘sopdo uojoqd ui 8o[bub ou SABq uopoajjoo-’o j o sXbm sq j S B a iaq ^
U0J133JJ03 - J
(d jn u s d o ¡sv.nuoo s S joj ‘t u r f 0 ( = A ° j ) saSowi jv ijuvd
p3/D3S-3J Jo Uins 3 1 ‘dS d IUI p3¡33JJOO (o) ‘UOflHJOSSU SUllt JHOljllM aSoUII MDJ (q) ‘S3UI1) JV.XU.W ¡U3jjffip 01
S u ip u odsd jjoo SdSoiUl ¡ v p jv d Jo S3U3S (o) '(zuw ffl 'MUß ‘A3USJ3Q {/ ÚS3}Atl03 ÜJDp Ml)A) U01J33M03 pdlSlSSD
-ajDMlfoS p u v U0lt0313p dSvUll lUdpUdddp-dWtl SutZtJIltl UOllüJJdqö JjlDUlOAljJ /VU3]OJ Jo U01J33M0J .'[ 3jn8t¿]
u o p rq o ssj u¡
uib 8 |BuoijippB ub p p iX him aSBjpA S u istp o j s q i j o u o ijb u b a snouojqouX s 'sVtooj ui si b j
S i j ui sSbuii su o aqi X|uo jBqj os oijbjs jds>[ sba\ sopdo adoosojoiui aqj ‘s j s h 'JuapiAS si
o puB q u o o w p q jsbjjuoo puB u o p rq o sa i ui uibS jb 3 [d o q j o j '3 i [ ‘jso[ si Xjisusjui ou jBqj
os uibSb dn p su iu in s sjb ssSbuii psjBos-sa jjb ‘X jjeu ij "jo p b j ajBudoaddB aqj qjiM sSbuii
qoB3 SuqB os-sj Xq p s p s x io o m ou si ssSbuii [BpjBd sq j ui uopBjjsqB opBUiojqo jBjajB] a q x
q [ ’S i j ‘uopB jpqB opBuio.iqo S uojjs b SA\oqs sq t;|s-au ip |[b j o o§Buii-uins o q j u o p rq o ssj
3UII1 jnoqjij\^ 'Bi 'S i j sss ‘siBAiajui XSasua jjbuis S u p u a sa jd a j (ltsqBis-3uip„) ssSbuii
|BpjBd ojui [s] j o p a p p suijX E pp S uia[os 3j - j ‘X‘x sq) 3uisn psujBiqo sSbuii ub S u ip u ip Xq
U 0 1 J D 9 U U 0 D s a i l J O 3 [d lU B X 3 JSJIJ B SM O qS [ O jn S l.J - [ ¿ ] P M g g j - X UI U ()p tl[0 S 3 J J O 3SB 3J3U I
(B p U B JS qn S B S p |3 lX
( iS J lJ J X 8 j3 U 3 S u iS b U I I 3 A IS J 3 d S |p B UI 3 > iq
X S J3 U 3 *3'l )
J 3 J I P j q S i [ j - j o - 3u i p b X pB 3J[B jB q j p s j B J j s u o u i s p 3A B q 3A \ ‘X p u 3 3 3 y ’y i q s j o u o p B i o j s S b u i i
‘(U 0 P B 3 I J I u 8 e U I s S b U II p U 3 p U 3 d 3 p - X 8 j3 U 3 = ) U O p B H S q B 3 p B U I 0 jq 3 IB J3JB] 3 q j P 3 J J 0 3 UB3
S u i s s s s o j d s S b u i i i B D u a u i n u q ii M u o p B u i q u i o o u i
uotioaiap aS m ut pdA¡os3u-3uiii
‘X h b u i j
‘J0 1 B J3 [3 3 3 B 3 q i A 3 l j v pU B d j o f o q S 3 p d o
§ U ip U 0 d S 3 J J 0 3 S 3 tl|B A 33 3 A p B S s U pUB
S A p iS O d 3 q i J O U O pB []33U B 3 3 p j d u i 0 3 B 3 A 3 |q 3 B S n q j UB3 S JS P U IB JB d S U 3| pUB J0 1 B J3 [3 3 3 B
j o u S issp j s d o j j
'o o i
‘p3}j3Aui si ui3]sXs SU3[ ju sn b ssq n s sq j j o uopE jjaqt:
o p B U io jq o
[1] O. Scherzer, Z. f. Physik 101 (1936) 593
M. Haider, S. Uhlemann, E. Schwan, H .Rose, B. Kabius, K. Urban, Nature 392 (1998)
H. Rose, D. Preiksas, Nucl. Instr. Meth. A363 (1995) 201;
G.F. R em pferet al., Microsc. Microanal. 3 (1997) 14.
[4] O. Scherzer, Optik 2 (1947) 114.
P.W. Hawkes, E. Kasper: Principles o f Electron Optics, A cad.Press 1996, II p857 ff
[6 ] H. Spiecker et al., Nucl. Instr. and Meth. A 406 (1998) 499.
[7] A. O elsner et al., BESSY Annual Report 2001, and to be published.
[8 ] A. Oelsner, O. Schmidt, M. Schicketanz, M.J. Klais, G. Schönhense, V. Mergel,
O.Jagutzki, H. Schmidt-Böcking, Rev. Sei. Instrum. 72 (2001) 3968.
’9 £" 6 l (6661) 681 p s d s ssbjaj f -juj ‘ss su s| s|o d n jso pue sjodrupE nb j o p sso d u io s
suisjsX s sus[ s p E js o jjs s p j o suopE jjsqE s q j j o uopE Siuiiuij^ ‘p E s y H 'J ‘BAOUBJBgyg
'S2£-9l£ (S661) 99C V s s y 'sXqj qisjAl pus Jisui -p n ^ ‘jo js s jjo s
a jo d iq n m b Xq ^\[gs sS ejjo a m o| b ui u o p s s jjo s u o p E jjs q y :jspiB]gj/v ‘qsE Z ’f
•psipnjs u s s q osje SAEq jo js s jjo s sq j j o s js s js p [ESiuBqssui oj sjiu iq ssu b js[o j s q x
sq j j o uopEJJsqE spEU iojqs sq j j o u o p s n p s j sq j
jojssjjo s
p ssn sq p p o q s sp o u i uopB JspssB -piui sq j
3]odnpo-3[odrupE nb
uisjs Xs
jssjjo s jouubs suojb
ssspqjjSA Sj\[ *su3| SA pssfqo sq j j o uopEJJsqE sp E iu o jq s puE p s u s q d s sq j
s § bj | oa
u e su i sq i jBqj p u n o j u s s q SBq j] -[f] psuiuiBXS u ss q SBq
spEU iojqs puB p s u s q d s sq j j o s s u s p u s d s p s q x 'sus) SApssfqo stqd
p su s q d s p jo j sq j
sq j
s jssjjo s
jsq jo sq j puB
jo jssjjo s
jo jssjjo s
sq j j o uopBjjsqB
sq j j o uopEJJsqE p s u s q d s sisu u ju i
qsiqM j o su o pun o j u s s q SAEq ss S ejjoa s p d n js o j o s p s o m x 'p sip n js u ss q SBq
s p d n j s o sq j SupEApsB j o subsuj Xq uopBjjsqB p s u s q d s sq j j o u o p s n p s j s q x
sjssjjo s
ss Sbjjoa
•ssnpA uinuuuiui jp q j SAEq sju sp p js o s
uopBjjsqB puE ssSejjoa spdrupE nb sqj jo ssnpA sqj sssus] jsjn o oavj sqj jo qjSusj
sqj se 3uo[ sb ssim j s; sssusj jsuui oavj sqj jo qjSusj p su jsu io sS sqj usqA\ jBqj punoj ussq
pi[BJBd S upS jno ub ssaiS jBqj spoui sid o sssp j e u u o j oj ssSejjoa spdm pB nb sqj âupsnfpE
jo j psdopA sp ussq SBq sjn p sso jd y 'sqj3us| [Enbsun jo sssusj p jo qjSus[ suies sqj jo sssus[
9 jo jsisuos jBqj sjo jssjjo s Xpnjs ubs suo uopE puiis siqj qjiAV 'SU3I sp d n jso -sp d ru p E n b
psuiquios b u u o j oj sjJEd p n b s jq 8 p ojui sjojs [BuipnjiSuo] Xq jns si qsiqM jo qsBS
ss p o jjs sp [BS|jpui[Xs psuSqB usass jo sjsisuos psjB[nuiis ussq SBq jBqj uisjs^s sq x
•SUip S[qBUOSBSJ E Ul puE XpjEjnsoE pszX|EUE
sq oj suisjsXs ju s js jjip j o jsq u in u sSje[ e sm o ||e uopB |nuiis jsjn d u io s b s[iqAV X qB jusuiusdxs
'[z] a £ 0 d 3 m EjSojd lE uoisusuiip-ssjqj
uopE ziuipdo puE sisXjeue sjb Xpnjs sq j j o suiib s q x
sspsusjSBJBqs [B sp d o -u o jjssp sq j j o
os op oj jjn s ijjip s q pjnoM jj 'jo js s jjo s sq j j o
sq j Suisn jo js s jjo s s[o d njso-sjodnjpE nb sq j j o u o p E p u iis jsjn d u io s b dn jss SABq s m
jo js s jjo s
siod n jso -sio d n jp B n b
UA\oqs u s s q SBq
uopBjjsqB sq j
jo j
'S u isiu io jd
jsq j [ j] X jp ju su iu sd x s
e ui
jiu iq u o p n io ssj sq j
•p ssn p sj sq
qjiA\ p sd d in b s V\[3S s S ej [oa- m o [
sq j su isss
js z js q s g Xq psjssS S ns
Àq p s js u js s j sq
pBSjds XSjsu s
jo jssjjo s
sXbm jo
u o jjss [s
sio d p [n u i
SuiXoiduis ‘u o p s s jjo s
jsquinu,_E S u o u iy •jojEuiojqsououi
sq j ss u is ssijSBjd u; juB jjoduii
sq j
XqBnsn si uopBjjsqB p s u s q d s s q x 'sus[ SApssfqo sq j j o suopE jjsqB spB uiojqs puE p s u s q d s
sq j Xq p sjiu iq si s s d o sso js iu i u o j js s p S uiuubss
s S bj [oa - m o | jo
u o p n io ssj s q x
3jO'[email protected]">[UEy ISSSjppE IIEUI-S*
^Ifl ‘d O I LIIA1 JSJSSqsuBj/\[ ‘s p is jjB q ^ ‘p jq (sd o jn g ) XjojBJoqBg qsjBSSS» nzpBUiiqsf
'SJ1 ‘l d 6 e iW JSJssqsuBj^i ‘JSjssqsuE ^[ j o X jisjsa iu q ‘ssisX q j j o jusuijjB dsQ 3
BISSII» ‘120^61 S j n q s j s js j -js ‘sjn jp su i [E sm q ssx -o sisX q j s jjo f g y ,
£s u q n 3 x j puE *..pBs>i u g ^ b a o u b jb q - y -g
3 d O D S O M 3 I l\l N O M 1 3 3 1 3 ONI1MMVDS a O Y l l O A M O l V H O J H 0 I 3 3 M Î I 0 3
N O IX V M H S a V 3 I I V 1 S O M 1 3 3 1 3 N V 3 0 N O I X V l f i m S M 3 1 I l d M 0 3
A. Delong and V. Kolařík
Delong Instrum ents s.r.o., Bulharská 48, 612 00 Brno, Czech Republic
1. Introduction
It is beyond any doubt that electron microscopes are capable o f contributing a great deal to the
achievem ent o f scientific knowledge, which cannot be achieved otherwise. Randomly
selected examples may include medical diagnostics (thin sections, viruses and others),
nanom eter resolution material diagnostics, environment diagnostics as well as a number o f
other areas. Their further dissemination, along with a generation change over will no doubt be
highly desirable. Current electron microscopes (particularly, TEM s) are being accepted with
certain embarrassm ent and reservations. One o f the main reasons consists in the fact that the
current electron microscope is a relatively expensive piece o f equipment. Some types are also
rather bulky and difficult to install. Therefore, the question arises w hether or not it might be
practicable to produce - provided that some compromise is resorted to - routine, desktopdesign, easy-to-transport microscopes, featuring low pow er demand and little place
requirem ents and, last but not least, more attractive prices than the current general-purpose
microscopes. Given that the price o f a fairly good optical microscope ranges from 100,000 to
200,000 USD, so this is ju st the sum, w hich should not be exceeded by most electron
2. Ways for dim ension decrease o f electron lenses
Having a look at a TEM design (the situation being similar with SEMs), it is apparent that
there are hardly any savings to implement on it. The microscope overall dimensions are
determined mainly by the electron lens size. Particularly, the design o f a good-quality
objective lens is usually well optimised. Therefore, it can hardly be reduced in size
considerably, unless we choose to change the only adjustm ent-accessible parameter, i.e., the
exciting coil current density. This is the approach chosen by T. M ulvey [1], His papers are
comm only known. From these papers it follows that a magnetic electron lens with
a rem arkably reduced magnetic circuit is contrivable provided that there is a w ay to dissipate
a considerable pow er from the heavily current-loaded exciting coil. Although this problem
may be solved in principle, any such solution features an essential drawback, namely, the high
input power, most o f w hich must be dissipated by means o f a-cooling medium, preferably
water. This is why this solution has not appeared in comm ercially available microscopes.
We made an attempt to replace the exciting coil with advanced-design permanent magnets,
although we were aware o f the crucial problem to overcome - the difficult control o f the
m agnetic field, in conjunction with no possibility to annul the magnetic field at need. The
m ost up-to-date perm anent magnets make it possible to produce magnetic lenses with
param eters considerably different from those o f AlNiCo magnets used formerly. Tentative
calculations based on the software provided by B. Lencová [2] have shown that - with regard
to the need to realize high diam eter-to-length ratio magnets - similar parameters as those o f
T. M ulvey’s lenses may be achieved [3]. Fig. 1 shows a simplified section o f a low-voltage
TEM objective lens in comparison with an exciting coil equipped objective lens. As 2 gaps
had to be created in the perm anent magnet m agnetic circuit, one gap was used to act as
a short-focus condenser lens, while the other as a second zone objective lens (T ab.l).
Z/8 I
(•joaqj) Z£-q
(SZ'O) S'0
8 's e
z - O is i
(ffZ psinsB aui) z í
(e o ) s ’o
£LV 0
(Dlnysm ny3diuB ) js p u ie jB d S m jp x 3
(pBj) ldo»
u oqrqos 3.i (uiu) g
(a s ) n v
(u n ) 33
( n u £/. 10'0="Y) A 3>1 s
( uiu ÇQO'O =Y) A3>I 09
(X) Xjisusp xrqj umuiixB]/\[
(g lX psjnsB sui) ç X' l
(uiui) q)Su 3 [ [Booj
9 1SO
6£ L 0
(LU l psjnsB aui) L L \
lauSvui luauvwusd i/iru
su s¡
itoafqo 3 i / i f a sjsisu ivw d
(p 3 u n s D 3 U t)
p jivjn j/n ') :¡ 3¡qv¿
'pSASiqDB 3q 0} SiBSddB [B0§ paSBSIAUS s q x
•sp b jsjjb 0} p s[ SAEq jq3[iu qoiqM ‘guijsBJjuoo [Epui-XABaq )noqjiA\ ‘s p s fq o [BoiSojoiq ‘sb
qans ‘s p s f q o jsbjjuod-a\oj j o isbjjuoo sSbuji aqj ssb sjo u i oj Suim o[|b sdoosojoiui b sb ‘jsqjB j
‘jnq ‘iM gx B J ° u o is js a p sijq d u iis sw u i b sq oj p su g isa p u ssq jou SBq M 3 1 sSbijoa-m oi s q x
'( 2 '3 y ) jsqjouB oj sp o u i guijBjado suo
u io jj s jn p s o o jd j 3 A0 -q 3jiA\s ajduiis Xjjibj b S uunjB sj ‘(¡Ajas ‘JM 3IS ) adoosojoiui Suiuubos b
7 loo S u ftp x a ijihM (q isu S diu luauviuasd ifiiM (a
attwog t(f,su3p xíq! w m uixvui 3WDS 31f! Ifll.w S3SU3I 3A1133ÍqQ . [ 3MSiJ
p u s ‘(1 A I3 X ) s d o o s o j o i u x s u s [ b q j o q s b a j q B o q d d B s q o j u i s j s X s [B 3p d o - u o j p s | S
oj -
s a in o s u o j p s p
s s a u j q S u q q 8 i q p u B J s p i u B i p b q ji M u o i p e u u o D u ; -
u b u S is s p
3 [ q iS B 3 j p s A O jd q
The contrast transfer function (CTF) (Fig.3) shows
that it is particularly the chromatic aberration to act
as the main resolution power limiting factor. Its
correction will be inevitable in the future if
a rem arkably higher resolution is to be achieved.
As the low-voltage microscope has its specific
application area, an attempt was made to produce
a TEM whose beam energy was to be higher.
An energy o f 60 keV has proved to be a good
compromise, as will be shown below. Fig. 4 shows
a simplified section o f an objective lens o f the
same design as in the case o f the 5 kV TEM, and,
as above, in comparison with an exciting-coil
equipped objective lens with the same value o f
Bmax for both lenses. In T ab.l the objective lens
param eters are listed.
Fig. 5 shows the contrast transfer functions (CTF).
may conclude, above all, that for a wellFigure 2: Ray diagrams o f TEM and
designed 60 keV objective lens the resolution
pow er limiting factor is the spherical aberration
rather than the chromatic
one. The initial resolution
power threshold, 0.33 nm,
may further be improved
to reach 0.12 nm provided
aberration is corrected by
means o f an existing
method. If, in addition,
the chrom atic aberration
is reduced,
too, the
power will
equal 0.1 nm for the
param eters
T ab.l.
A pplication o f perm anent
magnet based electron
objective lenses makes it
produce electron micro­
scopes featuring satis­
FigureS: Contrast transfer function (CTF) fo r 5 keV.
factory resolution even
when considerably re­
duced in size.
s jq y ‘(O V A auqiEjs/Oo
-ajSuis ‘Ä[[Bnsn jsoui)
Ady 09 AOf (dID )
u o ijo u n f A d fs u m i
jsd jju o j
:ç d jn S ij
-n |O p O l|J B 3 B J O
-U 0 3 ‘UISJSÄS S u i p j 0 3 3 j
s8 eU II
jB ijj
Ä B p -JU 3 S 3 jd
p sjo u sq oj si
j i •§>)
QZ\ se qonui se qâisM
ÄBui JsquiBqo S uim sia
3§BUII sq) ‘3|duiBX3 JOJ
‘0§ 'jqSl3A\ SdOSSOJOIUI
jssjbsj S
sq j
sq j ssjn q u ju o o
‘X ijE psdsg
3 uia\ sia s Sbuii | buij sqj
qjIM UI3JSÄS uoijo sfo jd
aqi si sp o u SuijBjado
lu a js A s u o ija a f o .i d
p a ij!|d u i!s e
7too Suijpxd ifjiM (q ‘jduSvui luduvuiudd ijjiM (v
‘XDtuiOg ¿jjsuBp x n jf wnwixDiu auivs 9i/j ijjiM sdsudj dAijodfqo :p d jn S jj
optics and a CCD camera, makes it possible to not only record the image but also display it in
a monitor screen. In this way, the current projection system and the image-viewing chamber
are becom ing useless.
In the low-voltage microscope, we have chosen to use a high-quality objective lens o f an
optical m icroscope to serve as a coupling element between the lum inescence screen and the
CCD camera. In this way, a portion o f the overall electron-optical magnification has been
taken over by this objective lens. In conjunction with the electronic magnification o f the CCD
camera, the m onitor screen image magnification may reach 1000 and 200 for the beam energy
o f 5 keV and 60 keV, respectively. The electron microscope projection system is thus reduced
in length considerably. At the same time, the large numerical aperture o f the objective lens
increases the transm issivity o f the whole detection system. The detection quantum efficiency
DQE reaches as much as 0.7, which means that a single electron in the im aging beam gives
rise to 0.6 electrons in the CCD camera.
Figure 6: Carbon replica o f pearlitic steel. Left - 100 keV. Right - 5 keV.
A very prom ising detection system is the CCD chip, on the back side o f which the electron
beam is incident. The chip is made thinner and its detection quantum efficiency is very high.
For an electron beam energy ranging from 6 to 8 keV, it reaches as many as 1000 signal
electrons per a single primary electron. Simultaneously, the service life o f such a CCD
detector (1050 x 1050 dpi), w hich is currently used as an image intensifier, turned to be
extended significantly.
‘III t o a ‘0003 o iu g ‘31 JAHW13 aoJd :uj ‘e ^ p a e j ¿ ‘^ iib io ^ a ‘ITP^IH "S ‘S u o p a y [£]
'631 ‘(£ 6 6 1) 6 ¿ l XdoDsojDijM f ‘EAOouoq -g ‘S u o p a 'V ‘ssuiB py j [3 ]
'O lf
- 6 ££ ‘3861
§ B [ tó A J o S u u d s
‘s o >[m
'AVd) sosuoj u ojjoop
o ij o u § b ]/\[
‘X o
a p ja
j_ [ l ]
S 3 0 U 0 J3 J3 > J
joqS iq opnjiuSEiu j o jo p jo ub Xq oq sXea\ je [jim ooud j p q j ‘joa o m o h ' sojjuoo juouidopA sp
pUE qOJEOSOJ p0Zl[Ep0dS UI OAIJOOJJO-JSOO X¡ui3uiud Oq p[nOO UOlJBZqijn j p q j puu JOA\Oj oq
jjim ‘s>[sbj
p ojuouo qojEosoj
X jp p o d so ‘sosnjEJBddE qons
poou o q x 'o ijp o d s
uojjB Jsdo-ouijnoj pojopisuoo oq jouubo ossqj joaom ojj '( a s ^I OOt7 oí 003) soiSjouo u o jjo o p
JoqSiq XpEJopisuoo ojm boj suoijBoqddB ieoijXieue 3ui3uo||Eqo jo u o ijn p s o j qSiq qoiqM
u; ‘y s |
ouios XpEjniEU
u o jjo o p j o uoijB3i[ddB oqj qoiqA\
‘juouidinbo ouijn o j
o jo q x '[B p ijsu sq XiJBpoijjBd aq oj sjEoddB sodoosojoiui
jo j
‘suio]qojd j o jo qiunu
jeoj S b oa[os
oj jo sn oqj Suimoije
pojopisuoo oq p p o o sodoosojoiui u o jjo o p pouoijuoui oqj ‘uoiuido jn o uj
u 0 ¡s n |3 u 0 3 -p
•qonui ooj j/\[3 i a 35[ £ 3Ifl u jo jj jnoXEj p u s azis ui JOjjip
jo u []ia\ j| 'OAoqE pojonb uooq OAEq ooiAop e qons j o sjopuiEJBd [BDjjojooqj o q ¿ ']Ai31 0§EjpA
-mo[ oqj j o jEqj sb jdoouoo jb [u u ;s oqj j o ‘uSisop doj>(sop ui ‘IA I3S /JA I3IS M 31 A 3T 09 ®
pouSisep OAEq 3M ‘pMgx u o jjo o p 3§Bj[0A-A\0[ oqj qjiM poureS oououodxa oqj uo posEg
• ( s u o i j o o s u i q j ) s jo o f q o j s j p i q j X [ jq 8 q s u i s u o j p o p j o S u ijo jjb o s o ij S B p u ; o q j o j o n p
s i q o iq M ‘u o i j B j j o q E o i j B u i o j q o j o p o j j o o q j S u i j B u i u u p j o o S B ju B A p E
(asa J 0
3 S 4 J !M J A ia S -10 I M 3 I S ) u io j s X s S u iu u b o s
b sb
o jB ja d o
o s jb
SB q S u i S e u i i J A Í 3 Í S
Xb u i o d o o s o j o i u i o q x
•joXb] Suijeoo [B pui uiqj b qons ui jsbjjuoo Xub
X[pjBq oj osij uoaiS SBq sSbuii u o j p o p A3T 001 3l[l ‘Jsejjuoo SuipuBjsjno ub s o jn p o j oSeuii
POJEJOUOS mEoq u o j p o p A 3^ 9 3R1 SE O jaq^ 'SiEpui XAEoq qjww po)SBjjuoo X pqgqs ‘po^s
oi)i|jB3d j o Boqdoj uoqjBO e SMoqs 9'S i j 'p o u u iju o o uooq SBq jsbjjuoo ui qjA\oj3 p o p o d x o
oqi ‘X ]jEpads3 ’(joqjo qoso u io jj uiui g j o oouBjsip e je pooBds sa jo q oooid ojod uiiu p qjiM
u o isjo a oqj oj saqddB siqj) u iu 3 inoqE j o u o i p p s o j e Su¡qoB3j si odoosojoiui sq j ‘Xpuojjn^)
■|tysssoons sq oj psAOJd OAEq uiojsXs u o ip o fo jd posBq suo| SAipofqo jq8i[ e
puB sosuo] u o j p o p p u S b iu jusuB iujsd j o uoiiEoqddB ‘'O'i ‘u o ip n p a j jq S p M puB ozis soiAop
oqj oj S0q0E0jddB qjoq jEqj SMoqs odoosojoiui uiEoq SuiSeuii u o j p o p a 3T S sS eijoa-m o] oqx
poonpaj oq []ias. ooud oqj )Bqj podoq oq oj si jj ' ( a s n 0 0 0 ‘9 I °J 0 0 0 ‘W u io jj sjsoo j p s j i
diqo oqj) jO| b ojinb si qoiqA\ ‘Q S fl 0 0 0 ‘0£ p u n o j sjsoo diqo b qons qjiA\ poddinbo bjoujeo y
Z00Z spuaux luaoay
P. Štěpán and E. Coufalová
Delong Instruments Ltd., Bulharská 48, 612 00 Brno, Czech Republic
e-mail: [email protected]
Low voltage transmission electron microscope (LVTEM) with 5 keV electron energy
[ 1,2] was enlarged o f scanning modes as scanning transmission mode (STEM) with 3.5 and 5 keV
electron energy, scanning mode with secondary electrons detection (SE) and scanning mode with
back scattered electrons detection (BSE). The main idea why a low voltage TEM was developed,
is an important gain o f contrast at low acceleration voltages. This improvement is interesting
especially for low atomic number materials which has to be artificially changed (stained,
shadow ed...) to obtain better contrast at standard 100 keV or higher electron energies.
The detector for STEM uses as scintillator the YAG-Ce screen, which forms a converter
o f electron image in TEM mode. The detector for SE SEM uses P47 scintillator and for BSE
SEM uses solid state detector. The field emission gun (FEG) with the Schottky cathode represents
very small virtual source, which needs a relatively small demagnification. This is done by the
magnetic field o f the first part o f objective working as second zone lens (Suzuki lens with
permanent magnets). The disadvantage o f this design is that the magnetic field can not be
controlled, and only the accelerating voltage and the Z-object position can be changed. It can be
used only electron energy around 3.5 and 5 keV.
LVTEM is able to image low atomic number samples without any staining or shadowing
by heavy metals [3]. There is only requirement for specimens - to be as thin as possible, which
means around 20 nm. The sample preparation o f such thin specimen could be for some materials
very difficult. STEM mode can be than powerful tool for thicker samples. The typical example o f
STEM image is on fig. 1, which is a small gold particles on carbon film. The contrast is very high
for an unstained sample and the resolution is also sufficient. We can clearly observe membranes
o f the nucleus and the other microstructures. From the same reason o f elimination o f inelastic
scattering the STEM can be successfully applied for observation o f stained samples (fig.2 and 3a).
The detection o f secondary (SE) and back scattered electrons (BSE) could be very useful
for very thick samples with a surface microstructure but as it is obvious on the figure 3b, the
material contrast is also observable and could be successfully used for thicker or stained
specimens. The last example o f LVEM imaging possibilities are two pictures from BSE mode on
figure 4a and b. The specimen is a bulk Mo tape, which was observed in topographic and material
[1] Delong A., Hladil K., Kolařík V.: A Low Voltage Transmission Electron Microscope,
Microscopy & Analysis (Europe) 27, January 1994.
[2] A. Delong, K. Hladil, V. Kolařík and P. Pavelka, Low voltage electron microscope I. Design. In: Proc. EUREM 12, Brno 2000, Vol. Ill, 1197.
[3] William P. Sharp, Eva Coufalova, A nnin Delong, Vladimir Kolarik, Robert W. Roberson,
I.S.T. Tsong: Use o f the low voltage TEM with biological specimens: a feasibility study. In:
Proc. MSA Meeting, Philadelphia 2000.
[4] A. Delong, E. Coufalova, P. Stepan: Low voltage STEM. In: Proc. ICEM 15, Durban 2002
(in print).
if j i m . dA D dvuoD ) d S D w i
fy d v u S o d o j
(Dp SlJ
d jn S ij
(ddDi opy) jsDJjuoD Sm puuD ifo 3 s g . Dp s u n S ij
•lujíf uoqjD j uo sdpjjJDd pjoQ : j d u n S ij
d jn S ij
(jdAtj s, }üu) uoijods p d u w js yf3ii¡i m u qç :ü £ d jn S ij
'(D£ d u n S ij
i/jiM. dUDdmoo) isdàiuoo jDUdjviu 3 S 3 :cl£ ^ u n S tj
(djKjoifdwà]) uoijods p d u w js y dii/j uiu ç§
Anjam Khursheed
National University o f Singapore, 4 Engineering Drive 3, Singapore 117576
e-mail: [email protected]
In this paper, a variety o f different permanent magnet lens designs will be presented in the
context o f improving the performance o f conventional Scanning Electron Microscopes (SEM).
Firstly, it will be shown how an add-on lens attachment can improve the image resolution
o f a conventional SEM and extend its mode o f operation down to low landing energies (< 1 keV).
The add-on lens attachment is placed on top o f the normal specimen stage, below the SEM
objective lens. The SEM is operated as normal, where the objective lens is used for fine focusing
o f the image. The specimen is placed within the add-on lens unit, which consists o f an iron circuit
and a NbFeB permanent magnet slab, as shown in Figure 1. A secondary electron image o f the
add-on lens top plate hole is first imaged, and centred via specimen stage movement. The
strength o f the SEM objective lens is then readjusted to focus the primary beam on to the
A primary beam is retarded to low landing voltages by biasing the specimen negatively.
Field distributions were calculated by some o f the KEOS programmes [1], which use the finite
element method. Other programmes in KEOS, using standard third-order perturbation o f the
paraxial equation, were then used to calculate the on-axis aberrations for the add-on lens. The
focal length, f the spherical aberration coefficient, Cs and the chromatic aberration coefficient,
Cc, were calculated for a 6 kV primary beam retarded down to have landing beam energies of
1 keV and 600 eV. The primary beam was assumed to be parallel at the entrance o f the lens (zero
magnification condition). The simulation results are summarised in Table 1. The calculated
aberration coefficients are less than 251 |im, which is typically more than a factor o f two smaller
than those predicted for the corresponding pure magnetic immersion add-on lens.
Landing energy
1 keV
600 eV
/ (nm)
Cs (nm)
Cc (nm)
Table 1: Simulation predictions fo r aberrations o f mixed field lens attachment
To predict the image resolution from the simulated aberration coefficients, the root-sum
formula reported by Barth and Kruit is used [2]. Using the root-sum formula for the beam energy
spread o f 2 eV (tungsten gun), the probe diameter is predicted to be 2.67 nm for a landing energy
o f 1 keV, and 3.26 nm for a landing energy o f 600 eV. The probe diameters here are aberration
limited ones, in that they do not take into account the gun brightness.
Figure 2a shows the image obtained with a conventional JEOL 5600 tungsten gun SEM
(without any lens attachment) for a primary beam voltage o f 1 kV for a tin-on-carbon test
specimen. Very little o f the specimen can be resolved. Figure 2b shows an image using the add­
on lens attachment for a landing energy o f 600 eV (primary beam voltage o f 4 kV and specimen
biased to -3.4 kV). There is obviously a dramatic improvement in resolution when the add-on
'W 3S [EU0 i}u3a u o 3 b Xq pauiBiqo aq oj jbab
u o ijn |o s 3J isaqSiq aqi aq jpM Xeui saSBiui asaqi ui uoiinjosaj |Bi}Bds a q x 'uoiiEjnuiis Xq p a p ip a jd
si ieqA\ oj a s o p ‘pauiBUE aq ued uiu £'o punojE j o uoiiti|osaj aSBiui ub p q i Moqs oj u a u ip a d s
IAI3X uoqjB 3-uo-p|oS u oiinjosaj qSiq e jo j p aiuasajd aq him sq n saj jEiuauiuadxa j a ip jn j
•uojiniosaj jo u a d n s puB ‘u o piuijap a8pa jaqSiq
3AEq oj sjBsddB luauiqaEUE suaj uo-ppB uoissiiusueji aqi qiiM u a jp i sSbuii p p i j >(JBp a q x 'pasn
si a>I 0£ J ° XSJ3U3 ureaq XjBuiud puB 000‘00l J ° uotjboijiuSbui sSbuii u y -suopBoqddB J^3X
lu a ja jjip j o XpiJEA b jo j ji ui sajoq sBq iBqi (uiu 001 >) uoqjEa j o u i|ij uiqi b j o sisisuoa u a u ip a d s
siq x 'luauiqoBUB suaj uo-ppB aqi inoqiiM puE qiiA\ ‘u a u ip a d s £-99 JS
uoqjB ^ XaoBq JESy
aqi jo j ¡Ajgs 0£3X scHl!M d B uo p a u p iq o saSBun p p i j }|JEp ajEduioa qp puB ep sa jn S ij
•suiaisXs 1AI3XS pai^aipap
oi jB|iuiis XiqiqEdBO uoiirqosaj b aAEq him sua] uoissiuisubji uo-ppB aqi p q i p ip a jd a jo ja ja q i
si]nsaj uoiiEjnuiis a sa q x 'uiu 磣'0 aq oi p a p ip a jd si aaBjjns u a u ip a d s aqi ib ja p u re ip aqojd
paiiuiq uoiiBiiaqB aqi ‘uiaisXs uoissiuia p p i j [Buuaqi b jo j paidX i ‘uiEaq XjBuiud aqi ui \ s p'Q
j o pBSjds XSjaua ub si ajaqi iBqi SuiuinssB puB ‘suopEjjaqE Suiuiquioa jo j Binuuoj uins-iooj linx>[
puB qiJBg aqi 8uisp| 'uiui 969'0 J ° iúD) lu a p ijja o a uoiiBJjaqB aiiEuiojqa b puB uiui ¿Z9'0 j o (SJ )
lu a p ijja o o uoiiBjjsqB p a u a q d s b ‘uiui 9£6'0 j o qi 8 ua[ jbooj b aABq jjim uiEaq XjBuiud Aa>[ 0£
Suiuioaui pijBJBd B iBqi s p ip a jd uoiiBinuiig 'suaj aqi j o dB8 jad d n aqi u; E |sax ç 'I punojB j o
qi8u3J!S PJ3IJ JBIXB >[E3d E 31EJ3U38 UE3 pUB qSiq IUUJ PZ SI SU3( UO-ppB 3qX *iU3UiqOBHE SU3] 3qi
j o dE§ J3ddn sq i j o sjppiui sq i ui psoE[d si u s u ip s d s s q x 'li 3AoqB Xposjíp psoBjd si lusuiqoEHE
sus) 3qi puB ‘sSeis u s u ip s d s sqi uo paDBjd si j o p a p p pajSHBOS^DBq y '(JAI3XS) 3 do 3sojoi[/\i
aiEjado oí PV3S ub 3|qBU3 oi p ssn si susj u oisjsuiuii oipuS eui pEdrnoo b 3J3H '£ S-inSij ui UMoqs
si siusuiqoBUB 1AI3S SE P3Sn 3C1 UB0 sssuaj p u S e u i lusuE uusd Moq j o sjduiBxs js q io u y
•uiu p si JM3S 3lP J0J psjonb uoiiniosaj isaq X jsa sq i ‘ssiS jsu s SuipuB]
q 8 iq ib u sa s puB ‘S38buii X8j3U3 SuipuEi mo[ qans apiAOjd 01 sjqB iou si siu su iu sd x s sssqi
jo j p ssn JAJ3 S un 8 usisS uni aqi ‘Xijbuijou iBqi p sio u aq p p o q s i[ 'uiu gz'£ J ° u o ip ip s jd psiiiuq
uopBjjsqB p3iB|tiuiis aqi qiiM X]qBjnoABj sajEduioo uoiirqossj siq x 'uiu g £ sq oi punoj sem ii puB
‘iq S p q sil j o o/o0 8 oi o/o0 £ XpiBuiixojddB u io jj sa su jeuSis aqi qoiqM js a o aouEisip aqi aq oi ua^jBi
si uourqosaj a q x 'aSBuu u o j p a p XjEpuoaas XSjaua SuipuB] \ 3 009 ® ssojbe ua>[Ei sba\ ubos aui] b
‘luauiqoBUB sua[ uo-ppB aqi Suisn Xq uourqosaj paAOjduii aqi aiBuipsa o x 'pasn si luauiqaBUE suaj
Primary Beam
25 mm
Figure 1: A Mixed Field Immersion Lens Attachment
1.2 nm
1 keV
JEOL 5600 tungsten gun SEM
x 100,000, 3 mm working distance
600 eV
(4 kV beam, -3.4 kV specimen)
Figure 2: Experimental results from the mixed field immersion lens attachment
(a) Without lens attachment
(b) With lens attachment
'101 (9661) lOI W d o
'3 T [c]
'03611 sio d íS u is ‘I U S D S S J 3 sSpiy J U 3 > [ 01 ‘saodiSujs
jo XjisjSAiun JBUOIJBN ‘SuussuiSug [Eoupaig jo juauiUEdsQ ‘(S66 l) SOS'S ‘pssiisjnq^f y [j]
S 3 0 U 3 J S |0 > |
JUdimfDDIJD SUd] UOlSSJ UISUDUJ uo-ppv IflJAi (q)
solido JA13S pUOlJUOAUOO SuiSfL ( d)
W 3 S uoissiuio p p i j 0 Î1 X sdlUHd D uo sdSviui p p i f yuDQ :p d jn S ij
uiu p p z \
m uppzl
G. Schönhense1, A. O elsner1, A. K rasyuk1, C. M. Schneider2, M. Bauer 3 and
M. Aesehlim ann 3
1 Institut fur Physik, Johannes Gutenberg- Universität, 55099 Mainz, Germany
2 Institut fur Festkörper- und W erkstoffforschung, 01171 Dresden, Germany
3 Fachbereich Physik, U niversität K aiserslautern, 67663 K aiserslautem , Germany
([email protected] m ail.uni-m
Owing to its parallel image acquisition, photoem ission electron microscopy (PEEM)
ist w ell-suited for real-time observation o f fast processes on surfaces. Pulsed excitation
sources like lasers or synchrotron radiation, fast electric pulsers for the study o f magnetic
switching, and/or tim e-resolved detection can be utilized. A standard approach being also
used in light-optical im aging is stroboscopic illumination o f a periodic (or quasi-periodic)
process. A sim ilar technique is pum p-probe excitation via pulsed laser radiation.
Alternatively, tim e-resolved image detection with sub-ns resolution can be employed. In this
case the excitation source does not need to have any time structure. W e report on two recent
experiments, one em ploying synchrotron radiation and the other one fem tosecond laser
D ynam ics o f fa s t m agnetization processes
is being studied using m agnetic X -ray circular
dichroism (M XCD) at BESSY II (Berlin).
Switching in Perm alloy dots was induced by a fast
m agnetic field pulse generated by passing a short
current pulse through a strip-line. Figure 1 shows a
PEEM -image: the magnetic dots on the strip-line
(grey bar) appear sharp and only those on the
insulator on both sides o f the strip-line evidently
suffer from charging. The image w as taken by
m eans o f a special tim e-resolving image detector,
the so called delayline-detector (single electron
counting with x,y,t-resolution). We originally
achieved a time-resolution o f about 500 ps [1], now
Figure I: PEEM image o f a Cu strip line with
150 ps (cooperation with the group o f Prof. H. ferromagnetic Permalloy micro-structures
on Si (FoV = 200 um)
Schmidt-Böcking, Univ. Frankfurt). In an early
approach we used a comm ercial ultrafast gated CCD
cam era with a minim um gate-interval below 500 ps [2], Owing to its parallel detection o f all
arrival times, the efficiency o f the delayline detector turned out to be much higher. Time
resolution in the present experiment, however, is mainly limited by the electric pulse
conditions at the stripline being characterized by a rise-tim e o f the order o f 1 ns.
The strip-line geom etry is illustrated in Figure 2 (left part). A short current pulse I
passing through the copper strip line prepared lithographically on SiO induces a magnetic
field H as indicated. The insulator material and its thickness are chosen to achieve a 50 Ohm
impedance. The integral along a closed line o f magnetic field equals the current passing
su p p si saSviui juDovfpv uaaMjaq ¡uaiuauoui aqj ‘a s/n d p p a tf aiiauSDiu jq i Kq pauifap si (dSvuii ifa¡ dot)
o ja z aiuii aqx aui¡ d u is aqj uo to p ¡(oi¡Duuaj tur! g v ui ¡sdjiuoj Q J X f\rf a aouapuadap a m ij : f a jn S tj
Á f
4 m
jEqj s.woqs £ o jn 8i .| j o jjn sa j oqi ju a u iu a d x s js jij jn o ui SuiqajEui sauB psduii j o u iajq o jd stqj
j o 3jids3Q 'joiABqoq paAjasqo aqj o j a s u ssaiS sjo jaa u u o a jEO upaja sq j ui as[nd ju a jjn a aqj
j o u o ija a |j3 J E jEqj X|a>|;| si ji ’jop XojjBiujad ®J ° am i] SuiqajiMS a isu u ju ; sq j p a g a j oj a3jB]
ooj je j si ( |3A3j /(ajg jq S u q ) aSEUii q jjn o j puE (jop sq j uiqjiM pA aj Aaj3 >(JBp) aSBun jsjij
aqj uaaM jaq su £ \ j o a a u a ja jjip auiij a q j, sjn ja ru js uoijEzijauSBiu aqj j o uoijbijba b sajEaipui
jsE jjuoa ui aSuEqa o q j •([£] aas ‘m3[A3j ju s a a j e jo j) jsE jjuoa-Q 3X IM 3lIJ Suijiojdxa ( u q ja g )
II ASSH 8 I® uoijBipEj uojjojqauX s pas|tid S uisn p a u u o jja d sbm juauiajnsE aui a q x (su p'p j o
sdajs) j o p a ja p auijX Bpp aqj Aq jas ltqB[S-3iuij„ puE as[nd p jaij u a a w p q XBjap-auiij aqj ’sa jop
uiri 9 e u; jse jju o s 3ij3U§EUi ui aSuBqa aqj sju a sa jd £ a jn S jj u; UMoqs sa u a s sSbuii aqj_
pa jjo jsip XpuEaijiuSis jou si sjop aqj j o aäEuii aqj ‘jaAaMoq
‘aui[ d u js aqj j o aaBjjns doj jB|j aqj u o 'uinuiixE ui si aaBjjns aqj oj [ajjEJBd juauoduioa
p jaij aqj j o juaipEjS [E00| aqj ajaqM ouq d u js aqj j o sapis q joq uo su o jjjo jsip S]B3 A3j sSbuii
a a u a ja jjip a q x ZHIAI I j o ajBj b je sasjnd p ja ij paqddB qjiM a§Buii aiuiBu/Cp b puB aijEjs e
j o aSEUii a a u a ja jjip e si (jjEd jq S u ) j a jn S jj ui UMoqs ajduiBxa a q x 'aSBiui aqj j o „Jrqq,,
ajqB jajoj b A|uo asnEa sasjnd p[aij aijauSBiu jjo q s aqj jEqj ajEaipui s jjn s a i js jij a q x
(lySu) z fjfy ¡ f a 3}vu v id paijddn
Suiaq sasjnd p p i f aiiauSDUi qiiM auo ja q /o aqi puD (pai/dda p p i j iudisuoo ¡p m s) asDO oiid/ s aqi ui pauw /qo
auo ‘saSDWi
omj f a aSDiui aouauaflip p u n (¡faj) ajdiuas aui\-dij)S oiibuS diu aqi f a tüjauioag :£ a m S tj
a p o j } 0 3 |g j e a y
J O ) e |n s u
e u ! |d u ; s
o ß e u ß e iA i
•sjop ÂojiBuijad aqj j o uoijEzijauSBUi aqj qajiMS oj ju aia ijjn s ‘x u i 01 j o
uinuiixEui e SuiqaEaj sa s|n d p p i j uiEjqo 3M (uju 0 0 l inoqB p ssau>p;qj ‘uirl OS - 01 = 9
qjpiM ) X jjsuiosS a u q -d u js puB j3S |nd uaAiS sq j j o j I = sp j j | '3'] ‘dooj p[3 ij sq j qgnojqj
H ot-electron lifetim e contrast on the femtosecond scale is observed using a pump-probe
technique. As compared with synchrotron radiation (photon pulses typically 50 ps) fs-lasers
provide pulses being about three orders o f magnitude shorter. In this regime it becomes
possible to directly observe “hot-electron dynam ics” at metal and sem iconductor surfaces [4].
In the first pum p-probe PEEM experim ent [5] illustrated in Figure 4 a pulsed Ti:Sapphire
laser (40 fs, second harm onic hv = 3.2 eV, pulse energy <1 nJ) was used as excitation source
for the PEEM . A variable optical delay betw een pump and probe pulse allows to study the
lifetime o f the electrons excited by the pum p pulse. Since the laser runs at a repetition rate o f
80 MHz only a few photoelectrons are produced per photon pulse. Space charge effects
cannot occur at these conditions.
Laser System
ijr J
Figure 4: Schematic set-up fo r the observation o f fem tosecond electron dynamics in a photoemission
microscope (PEEM).
As hv is sm aller than the w ork function o f the sample, one-photon transitions are
strictly ruled out at the given conditions. At least tw o photons are required to obtain
photoem ission. Figure 5 schem atically shows the energy-level diagram o f such a two-photon
transition. The first photon h v l (“pump photon”) excites an electron from the valence band
region into an unoccupied state betw een the Fermi energy E f and the vacuum level Evac- In
the diagram the time axis is pointing to the right. This excited state (several eV above EF) has
a short lifetime and decays via a transition into a state close to E f (arrow pointing
downwards). The second photon hv2 (“probe photon”) arrives at a later tim e with the time
delay being precisely selectable by the optical delay stage, cf. Figure 4. If the “hot electron” is
still in its excited state, the second photon can lift it to a state above the vacuum level and thus
create a free photoelectron. Finally, these photoelectrons generated by two photons form the
image in PEEM.
Obviously the local lifetim e o f the “hot electrons” determines the local image intensity
if the tim e delay betw een hv l and hv2 is varied in the region o f the lifetime. This facilitates
a quantitative determ ination o f the lifetime by recording the image intensity as function o f the
tim e delay. Furtherm ore, the lifetim e distribution on an inhomogeneous sample surface can be
made visible as “lifetim e contrast” .
'9SPZ (L 661) 89 u iru jsu i p s A a y ‘js u y g S u p M
-o[3oruB isoa o
ÍZZ (ZOOZ) PL a 'S /íq j '[ d d y ‘u u B u iq q o ssy w
‘uueuisi/w ’3 ‘Js n e g -p\[ ‘jpiunps q [ç]
‘asu aq u o q ag 'O ‘A sX sjpuy o ‘sjJE qos ’JA[ ‘qiEJOj
'SPP (S6 6 I) TOI 'sA qj uw q 3 [ ‘lp}UBj\[ V d P™ ‘°®0 Ä
‘n v j f MaqjIAl G T 'y ‘HV"P3^ BSi a 3 H ‘JS B U iuajpim ps ' y j ‘u u B u iq q asay -jrç 'S 's [p]
(ju u d ui) ’sÁ q j ’S o j j -d a y ‘a su sq u o q o s q p u s tó p p u q a s 'PM’3
'6 6P (8661) 90F V m 3JAl P«B 'flsu i -pnjs[ ‘asuaquoqog O P«® ‘uuE u izu p H 'f l
‘p^-iapM -JAJ ‘e f n q y t y y ‘3J3 qsup[> [ - 3 ‘3>[uaw a ‘u s q p i z 'R 3 ‘Jpim qos O ‘« » p a id s H [Z\
'896£ (1003) ZL u itu jsu i p s -a s » ‘g u p p o g -ip iiu q D S H ‘p p i n g B f o
‘pgjsjA i 'A ‘ss u a q u o q o s r> ‘sie[-s[ ■fprç ‘z u E p ^ p iq o s -j^ ‘jpiuiqog o ‘» u s p o 'V [ l]
s a a u a ja p y
•zuiep\ uinaju9zs§unqosjoj ssipiiijBipsusssiMjBuajBj/vi
/Cq SE 1PM Sí £98i N £[ puB ç/j
so sjireag j a w a ¿q psp“nJ om sjirarausdxa aqx
'3|BDS-Su ao -sd aqj uo su o ijisu b ji ssEqd jaq jo jo [9] g u p p u i [edo| sb |[9m sb suisjs/Cs
[Buoisuaiuip-M oj puE pajnprujsouE U ui sjaiiiEO agjBqa j o soisX qd-sj aqj Kprus oj avo]je ]jim
ji sjn jrq jb s u aqj u i uoijrq o saj [EiiEds q giq je saiureuXp u o a p s|3 -jo q j o soijsuspB JE qa sq j oj
ssaooE saAig puE i/v g g d io j p o q p u i isejju o d a \s u e s js jjo u o i p j p x s jasE j-puooasoiuiag
•S3AIXIE asjn d a q o jd aqj s jo ja q uo ijisu ejj
uoijEjLoxsap b auogaapun SABq S[BuajBiu q ioq ui s u o j p a p joq jsoui m ou aou;s pajBsddBSip
lsoui]E SEq 33U3J3JJIP siqj (puB d jq 8u ) ÄBpp sj 002 IV M opnpuooiuias sq ; ui a u u p jq
u o a p a p - jo q ja q g iq B j o aouapiAS guiAig snqj suogExaq 3 y aq i UBqj tó jq S u q JBsddE sjEq syB Q
pajsA oaun aqj ‘(puB d ajjuao) sj 09
ps^B jsp ajE 3S[nd a q o jd puB drnnd aqj j i '(p b jijjb ub
si B3JE jq g u q sq j) suogBxaq g y aqj j o s u iu sq j jb X[uo s s s u e jsbjjuoo e (p u E d y a [ ) X sp p auiij
inoqjiM a g su ii u o issiuiaojoqd uojoqd-OAVj aqj u¡ 'aoBjjns syB Q b uo um oj8 sem (suogExaq)
a jn p r u js o ja iu i- g y y '9 a j n g i j ui uM oqs si guigEuii a u iip jq j o ajduiExs js jy y
([g] w ojfi u irf09= A O J iCopp s f OOZ
ifjfM aSaun aqojd-dwnd :)t{Su ‘(pa)DDuiqns puno.iSi/ovq sspid-aßuis \ 3.Smut awiiafij,,) ¡ÍD¡sp s/ 0 9
ijliM aSpiui dqoud-diund :9Aiudd ‘X vpp 3 U4ii jnoipiM dSvtui uoissiiudofoqd uojoqd-ohij :iß rI 'dODpns
dpiudsuv luni/jvS v uo (suoSvxaq) dmioruisojoiui 8 y n ui suouiDDjd ioq f o isn.uuoj dimidfi'j :g a jn S ij
a u iji
a o u a iB A v
/¿Vi if ) dspid dqoud puD (¡a ijj ¿s/nd drnnd
U3dň\}3q Ánpp duiii sjqvuvA ü s m o j j v anbiuqoaj
aqoAd-dmnd Difi fX]3Aij03ds3J ‘p \d \ lunnovA puv
luujr] 3qi dioudp mA^f puv tyj uoissiutaojoqd uojoqd
-owl m psAuasqo soiiuvuXp uoAirrDp-ioif Suiquosop
awoqos dwii
snsusA pxaj-Wjauj
: a m [ J „U 0 J J 0 8 I3 IO H “
■ ■J L a ■ ■ ■ ■ ■ ■ ■ ■ ■ ■
M. E scher1, M. M erkel1, Ch. Ziethen2, P. Bernhard2, G. Schönhense2, D. Funnemann 3 and
J. W esterm ann 3
'Focus GmbH, Am Birkhecker Berg 20, 65510 Hünstetten-Görsroth, Germany
2Johannes G utenberg Universität Mainz, Institut für Physik
3Omicron NanoTechnology GmbH, Taunusstein, Germany
[email protected]
G rowing demand for spectroscopic information in microscopy triggered the develop­
ment o f energy filters for microspectroscopic or spectromicroscopic applications [1]. Here we
present recent results with the retarding field analyser IEF (Imaging Energy Filter) [2] at­
tached to the Photoelectron Emission
M icroscopy (FOCUS-IS-PEEM ) that
is able to provide high-pass filtered
images. Three examples will be pre­
The first example are results
with fie ld em ission sites o f a thin
cathode ray tube device structure [3].
Each pixel o f the device consists o f
a 15x84um 2 area where random ly
grown field emitters are distributed.
The field emitters are pyramidal
structures about 150nm w ide covered
by a aperture acting as an anode elec­
trode. As a voltage is applied to the
emitters (typically less than 30V) the
field emission starts. Fig. 1 compares the im aging o f photo-emitted and field-emitted elec­
trons. W hile with threshold photoem ission the emissions site cannot be resolved probably due
to field distortion in the vicinity o f the
emitters,, the strength o f emission
m icroscopy appears clear with im ag­
ing the field-em itted electrons. Be­
sides the spatial distribution o f the
em itting sites (only part o f all em it­
ters) also their brightness and tem po­
ral behaviour can be examined. By
m oving the contrast aperture the an­
gular distribution o f different emitters
Electron kinetic energy (retard voltage)
can be distinguished.
o f fie ld emission sites. The
Y et with a low voltage o f
spectrum shows both
about 10 V the photoem ission is
photoemission (PE) and field emission (FE) peaks
equalled by the field emission. This is
in the energy distribution.
'V S fl ‘TB3 ‘'d jo ^ ssiSojouqosx jusosspuBO jo Asajjnoo S[dlUBS [c]
'961 (1003) 0817 'P S J jn S “ P J3 ’IM ‘I ^ I A I [Z\
•(¿661) W ’u io u s q j jE p ^ -$ -osofloads uojj 3S[g f ‘(jojips) h ‘s p y [i]
S 3 3 U 3 J 3J 3 >(
'6S¿¿N£[ PUB W 8¿N ei °N Z>U dHWH ¿q papunf ¿lU«1
sjusuiuadxs s i u
UMoqs sq oj jsX SBq
ssS euii s s u s js jjip j o uoijBjjsqB sijE uiojqs jom o] oqi oj onp u o ijrq o ssj [RiiBds js jjs q p sja sd x s
u y 'u o ijB jp x s A 33 ’13 Pub aidures S y psjTvprujsun e qjiM psuiBjqo sbm jjn s s j s iq x 'usAOjd
sq pynos A 3iU 0SI> J ° u o ijrq o ssj X Sjsus ue puE sS uijjss [je jo j sjqisiA jo u sjb sp u S sq j jBqj
Moqs sjusiu sjn sB su i j s j i j 'u irlg '8 oj uMop azis p u g sq j S u p n p s j Xq s u io s js a o sjb s>pEqMBjp
s s s q x 'so q d o u o jjo s p S uipjB jsj sq j j o sS uijjss u iejjso jo j p su iss sip sq p in o s lusjjBd p u S
sq j o s ¡ y 'S3SB3 js s q ui a 3 £'0 ^ luo s! u o ijrq o ssj X Sjsus ssoqM sp u S q ssu i uirlg 9 qjiM psuiBjqo
s js m SAoqB UMoqs ssiduiB xs qjoq ui p ssn sp u S pjBpuBjg 'sp u S p ssn j o azis q ssu i sq j Xq p sji
-iui[ sba\ JOSÁjBiiB p[3ij S uipjB jsj b qjiM X doasojjosdsojoiui j o u o ijn jo ssj sq j MOU OJ df)
'3|q|SSod sq o j uw oqs sbm su iu 0 5 2 x 0£3 J ° B3JR ojduiBS e j o Xdo3sojj3sdsoj3ij/\[ 'psuiBjqo u ss q
SBq s u o j j 3 s p - § j x M1!M uju q^ [ j o u o ijn jo ss j [BijBds e u o ijE jp x s A 3 00 ¿ M1!M '(doj) ¡yoj-ej
SuiX[j 3 pun sq j p u s (uiojjoq) ssjE nbs S y sq j j o u o ijn q u jsjp |B|jBds sq j M oqs ssjnjESj oidoosojj
- 33ds sq j j o ssSbuii ssEd-puBg 'ssjnjB sj ju su iu io jd sjb s u o jjs s p j p e x sq j p u s sjn jsru js puEq
- 33 U3 |b a 3 y s q x 's i p s i u i j s j sq j JBSu u iru jssd s sq j SMoqs £ ’S i j 'ss sjn o s X bj-x XiojBJOqB]
S uijsixs oj sp u o d s a jjo s qsiqM ‘X Sjsus uojo q d A 3 0 0 £ l ° l
uoijBipBj qjiM psuiBjqo sjsm
sjjn ssj o id o o s o jp s d s n X sssg je suquiE sq jojB[npun JMQS-1/6W1 3lP 1B uoijBipBj u o jjo jq s
■)/v3 d puvq 30ua¡a\-Sy puo (do¡) fp v± ai/i o/ Sm puodssjjoo saSouii ssvdpuvq :¡ijSu
uoiim pxa uoioifdA^OO^l M>’M V1 uo sajDnbs S y a i/ / j o M iim uap puo uirupads ssvd-ijSii¡ :tfo¡ : f aunSij
(As) A6jsu3 Bujpuig
-uXs qjiM psSBUJi sbm [ioj uin|BiuEj b uo (p o u a d u ir l¿ '2 l) ss jE n b s-S y SuijBJodBAS Xq p sjE d sjd
3|dujBS y usAOjd si su o j/ja /a g j \ SutSntut j o XjijiqiSBSj sq j 0|diuBX0 p u o sas sq j s y
•gjnjanjjs ssiA sp sq j oj sn p s|B |ju3jod ssb j
-jn s SuiSusqo sq j u io jj s j|n s s j u o |jn q u js;p X Sjsus u o jssiu iso jo q d sq j j o SuiuspEOjq s q x 'p su i
-iubxs s je ssjis js m s j SE JsdjBqs ssu io o sq >[Bsd sq j j o qjpiM sq j ípsijddE s8 e j)o a sq j oj spuods
-sjjo 3 >(Bsd u o issiu is p p i j s q j j o uo ijiso d sq x 's jjn ssj o id o sso jjo sd s sq j Xq z 'S iJ ui UMoqs
G. K. L. M arx and R. Allenspach
IBM Research, Zurich Research Laboratory, CH-8803 Ruschlikon, Switzerland
e-mail: gem @
Photoem ission electron m icroscopy (PEEM ) has proven to be an invaluable analytical
tool for the investigation o f magnetism on a subm icron scale. PEEM experiments are
generally done at a synchrotron, because a tunable, high-flux x-ray source is needed to excite
core-level electrons in sufficient quantity. The decay o f these excited electrons produces
a cascade o f secondary electrons, w hich is detected. Because specific core levels are involved,
synchrotron-PEEM is able to collect elem ent-specific information. Recently it has been
shown that a dichroic effect exists also for excitations near the emission threshold, i.e., in the
conduction band [1]. Such a laboratory-type PEEM w orks with an UV light source (“UVPEEM ”) and is conceptually rather simple compared with PEEM at synchrotrons. In
particular, the emitted electrons are mainly photoelectrons that have not suffered additional
scattering processes in the sample.
Synchrotron-PEEM is able to map magnetic structures below nonmagnetic overlayers
[2], The probing depth is on the order o f several tens o f nanometers, as expected for the lowenergetic part o f the secondary electron cascade [3]. The situation is less clear for UV-PEEM:
The emitted electrons originate from both the magnetic film and the nonmagnetic overlayer,
w ithout a possibility to distinguish them by elem ental analysis. Here w e report on an experi­
mental study to determine the probing depth in UV-PEEM , and compare it with the probing
depth o f a related all-optical experiment, the magneto-optical Kerr effect. In the Ag/Fe model
system, it is found that the probing depth o f UV-PEEM is essentially given by the penetration
depth o f the photons.
The linear magnetic dichroism as observed by UV-PEEM requires the same geometry
as the transverse K err effect: The m agnetization direction is perpendicular to the plane o f
incidence o f the incoming light, whereas the light polarization lies within the scattering plane.
Experim entally, we have used unpolarized light from a Hg high-pressure arc lamp (100 W,
h v [ 5 eV). The electron optics o f the PEEM consist o f a triode objective lens, a transfer lens,
a projective lens, and a fluorescent screen read out by a 12-bit C € D camera. A contrast aper­
ture resides in the back focal plane o f the objective lens. The magneto-optical Kerr effect was
measured in a microscope w ith the Hg lamp, followed by a band filter for green light
(512 nm). The longitudinal K err effect was used, w ith an angle o f incidence o f 45°.
Polycrystalline Fe films o f 90 nm thickness were grown by molecular beam epitaxy at
room tem perature onto Si(100) wafers covered by native oxide. Subsequently, a Ag overlayer
o f varying thickness has been deposited through a shadow mask. The Ag thickness has been
determ ined during growth by a quartz microbalance, and calibrated ex-situ by profilometry.
To determine the attenuation o f the m agnetic signal by the overlayer, the number o f
electrons I em itted from oppositely m agnetized domains in the Fe film has been measured.
The experim ental asym m etry Ap hence is defined as A P = (I(M + )-I(M -))/(I(M + )-I(M -)), with
M+ (A/-) being the m agnetization parallel (antiparallel) to the scattering plane normal. The
coupling o f the incoming light to the num ber o f electrons emitted can be understood by the
•¿661 ™ squisj\v ‘H 3 A ‘^dossoisijM u o jjssjg :ui ‘isríEg 'g [ 5]
M3M ‘s s s j j oiuisp E sy ‘¡I spqos ui sjubjsuos [Boijdo jo >[ooqpuEH [^7]
'Z (6 £ 6 l) l 'IBUV sobjjsjui jin s ‘qousQ V M ‘qEss 'ďW [£]
'8S9 (£661) 6 SZ s o u s p s ‘ [E 13 iq o js T [ z \
'888S (OOOč) t>8 '«31
's^Md ‘ssu sq u o q ss
'D ‘sjsu q g f-H
T X O [l]
S 33U 3J3J3JJ
■ [(m i
W 3 3 d s m j o UEOI 3 Ml i o j
ssuaquoqD S
3 § p 3 [M O U 5 ( O B A [ | n j 3 j E i 8 o ;\ \
p sijijsn f jsS u o j ou si s s u s q ( \ s 0 l _ 0) ^S-isus mo| j o sS u e í sq j ui sn b iu q ss j SA ljisuss-ssEjins
X[qS;q b si PV 331 JEqj j s q s q uouiuioo sq x :(lA ig g i) X doosojoiiu u o jjs s js A 8jsus-m o| ‘Á p iU B U
‘sn b iu q ss j psjE |sa b siu ssu o d s jjn ssi s s sq j u io jj ss u sn b ss u o s is q jin j y •sjsAe) sijsuS bui
p s u n q j o u o ijeS ijssau i ub io j p a p a šu jou si á jia ijisu ss je ju su isis jBqj su esu i siqx 'uisijsuSbui
>[inq s q o id ‘s 3 u e j a 35I 3lP u ! ss iS isu s u o ijB jp x s qjiM ‘j^ g g j-u o jjo jq s u A s puB ‘ASisus
uojoqd \ s g Xjuo qjiM ‘iA iggd-A fl m °q ‘s s u s g '[£] SAino jESiSAiun,, sq j Kq p s js ip s id
qjBd 33JJ u e su i sijSE pui sS jbj sq j qjiA\ ju su issiS e ui ‘uojoqd Suiuiosui sq j j o qjdsp Suiqoid
sq j UEqj jsS je ] si íÍSjsus
iu u s j
sq j SAoqB \ s g je s u o jjs s p o jo q d sq j j o qjd sp sdBDSS sq x
•Xjiaijisuss q jd sp sq j io j jo js b j g u iu iu u s js p sq j si jqSij j o qjd sp uoijB ijsusd sq j jEqj suesui
qsiqA\ ‘j s s j j s a is-)! iE3ijdo-ojsu§Eui sq j j o q jd sp S uiqoid sq j oj sso[3 si jse iju o s uiEiuop sq j
j o uoijEnusjjE sq j u io jj p s u iu u s js p q jd sp S uiqojd sq x 'S y j o uiu 0£ u sq j sio u i AVO[sq p su n q
s j u; sujsjjB d uiBuiop sijsu S eu i SuiSeuii j o sjqEdBO si JA ig g ď A fl lElfl s p n p u o s s^\\
•uiu z [ g j o qjiu3|SABA\ b
je [>] S y j o y puB u sju ejsu o s pssijdo sq j u io jj psjBins|BO luu £' \ z J ° 3niBA sq j qjiA\ sssjS e
qsiqA\ ‘uiu S 'l£ = ¥Y
‘( ^ y j p - ) d x s ~ x y oj p sjjij puB p s u iu u s js p sq ubs j s s jjs x is ^ sq j ui
•V[/ XijsuiuiXse sq j ‘X[snoSo]BUB sjín f) 'uiu j ' 9 [ = Jy j o q jd sp S u iq o id b spjsiX ( J \ / p - ) d x s ~ d y
me( Xbssp (B ijusuodxs ue qjiA\ EjEp sq j oj jij y 'ssSbuii w g g d - A n J° SSUSS B UIOJJ
sssu>(Diqj isX b iisao ’s a ssujsuiuiX sb p s u iu u s js p X [[Bjusuiusdxs sq j SMOqs \ s in 3 ig
Z1url OPIXQOZ s? MajApppiJaiu
u/u p z fo ssauy.niji S y o uof aSvuii /Y , 7 7 c / (tyžu) a o j mu c n - -Yy f o puv
-ioj utu Z'9l= JY 1°
ijuhp Sutqoud v sppiú. <h»/ 6vo0p jviiuauodxa uv ijijw Dfvp 3 ift 01 ii f y ssduyjiijj S y SuiifsmaA oj pazipwuou
uddq z a v i j vjvp
'ssdu^oti/i u d X v p d A o S y 's a sauiauiuiKsD pdmwuoidp <(nviu3w uddxg (lfo\) : [ sunSt^j
[ l] soBjjns sq j oj iE ]n sip u sd jsd jo jo s a jusuissE [dsip sq j Suisq
:q qjiA\ ‘ *(j ~ / oj spB3[ puE ‘s u o jjs s p s s i; j o u o issiu is io j Á iosqj sq j puE suoijEnbs [jsavxei/\i
M. E scher1, M. M erkel 1 and G. Schönhense 2
’Focus GmbH, Am B irkhecker Berg 20, 65510 Hünstetten-Görsroth
2Johannes G utenberg U niversität Mainz, Institut fur Physik
e-mail: [email protected]
The ultim ate lateral resolution o f a typical Photoelectron Emission M icroscope
(PEEM ) crucially depends on an optimised contrast aperture to balance the diffraction term
with the spherical and chromatic aberrations. For PEEM applications w ithout energy filtering
the type o f the objective lens plays an im portant role as has been shown by Chmelik et al [1],
The authors found for all configurations examined a best resolution below 10 nm which is
sufficient for many practical applications.
Many PEEM users do not aim at the ultim ate resolution, but at higher intensity, that
can be realised with larger apertures. For example applications with X-ray excitation
(XPEEM ), w here bright laboratory sources are yet to come need higher intensity to get man­
ageable conditions. Synchrotron applications with limited beamtime, or spectro-microscopic
studies w here series o f images have to be taken need highest intensity to get short exposure
With larger apertures the contributions o f the objective’s spherical and chromatic aber­
rations to the resolution deterioration become essentially large, and thus an optimisation o f the
objective geom etry is mandatory.
The goal o f our study was to obtain an optimised geom etry for an electrostatic tetrode
objective. W e have chosen an electrostatic objective for two reasons. First the sample should
be at ground potential to allow easy in-situ preparation (cooling, heating, etc.) and the column
energy o f the electrons should be as low as possible to allow easy adaptation o f retarding [2 ]
or dispersive [3] energy filters or TOF-detectors [4], Furthermore the investigation o f mag­
netic sam ples [5,6] w ithout disturbance o f the magnetic properties is feasible.
The largest influence on the properties o f an electrostatic tetrode objective lens, as
shown in Fig. 1 gives the second electrode (Focus). We calculated the spherical and chromatic
aberration coefficients for geom etries with varying length L and diam eter D o f the focussing
electrode. The bore o f the extractor electrode and the microscope column were kept constant
as well as the sample distance that was chosen
to w ithstand an accelerating voltage o f 30kV.
For each geom etry the column potential was
varied from 0.6 o f the extractor potential down
to the point where no focussing with a fixed
image position was possible. The calculations
were made by direct ray tracing using
Simion 6.0.
Knowing the constant aberration due to
the accelerating field, the aberration coefficients
o f the lens (w ithout extractor field) can be cal­
Figure 1: Geometry o f the electrostatic tet­
culated. Fig. 2 shows the lens aberration coeffi­
rode objective studied with typical
cients for a typical electrode length with varied
'£ ¿ 6 £ (6661) OL 'uim jsui - p s 'Aay ‘ I® la 'S ‘sja p u y
'¿156 (6661) I I -»W W -suapuoo "sÁ qj f " o ‘asuoquoqos
'11721 (6661) S sjajjax pire AvaiAa^j aoEjjng ‘-je ja 'm ‘q o n ^
' 6 6 t7 (¿661) 9 0 t V 'm aw
'-»sui - p n ^ ‘-[B ja h ‘ja jp a id s
'6 0 1 1 (¿661) 16-88 'm ouaqd jE p ^ # -osojjoods uojjoaja 'f ‘'P Ja 'O ‘ipiuaqog
[9 ]
'961 (100Z) 08ť 'P S 'Jjn S ‘ P 13 ' n ‘P>RP\[ [ť\
s e i (6861) £8 T ildo ‘ T ^ O ‘x je w pue ‘ h x ‘uasEjjpuaA ‘t ‘T!PU*io [ 1]
sa a u a ja ja y
¿88ÍNÍI ZM d a n a ty p^punf ¿puDd sdm i/mm. sii/j
■aSEjjOA S u i s s n a o j a q j
J E a u [ B j j u a j o d u u i r q o o a q j q jiA \ p a u i E j q o a q
s j p s a j j s a q a q j jE q j M o q s s S u i p u i j a q x
•oj pauiiE suoijipuoo
’ SdJD U ipjO O D d j d u w s OJ
SAdfdA A\7
-D A A dqV d t p f o S m p V A d l j j
UOlJVJlDXd j \ f i X
A O f I jip iM A d p ' I If JIM U O Ijn q iA J S ip
u v u i s i jn s d A d A n j A d d v
bS avj
d A n jA d d v
JSDAJUOO uniQQ£ V AOf JÍAdUlf a dAftSlJ . £ dAUSlJ
S u i> [ jo a \
j o j [EOidÁj ‘(aSEJioA uuirqoo a T c '1 Je j -lajaure
-ip uirloO£) a jn jja d e aSjB] b j o j XjijuBnb siqj
SA\oqs £ - S i j '[ ¿ ] pEajds XSjaua aqj o s j b - jn o i
-ABqaq ssEd-A\o| sjí qS nojqj - puB pEajds ajSuB
aqj sa u ija p j| ‘ajnjjadB aAijoafqo aqj uo juapuad
-ap X ju ib u i si ju a u i j o a jnS ij siqj J/SiaadX J0á
■juaui j o ajriSij e se jrq q a§Eiui tsua[ aAijoalqo
aqj u; Suij[nsaj uoisnjuoo j o osip aqj >[ooj s/a
§ u i> [ jo m
uinuiijdo aqj puij o x
p o o u n o u o jd
s so ] s ; X jjo u io a S a q j u o a a u a p u a d a p a q j ís ja ja u r e
- i p [ | e u i s p u B sa 3 E j[O A u u i n |o o q 3 i q
u o ijB x ia q B
o ijE iu o jq o
'p u n o j
jo j
js m o j
u in u i
- i u i u i a j n j o s q E a q j o s [E a j a q A \ ‘s [ B [ j u a j o d u u i n
- jo a
m o
3a i§
jo j
jn o j j o
o q j si s iq x
s ja ja u re ip
jo jo b j b o j
d n u o ijo n p a j
p a o u n o u o jd
js o u i
'JO jiE U is s i S u i | [ i j s u a | a q j a j a q M
a p o jjo a p
ja § jE |
p a s iim u iu i
jo j
UEO u o i j E j j o q E [ E o u o q d s a q j p a j o a d x a s y
7 dAYlSlJ lil pdlltfdp BAD 7 pUD Q SdSvjJOA UlUUJOD pití,%
saiAidiuodS lusAdffip ao/ sjuapiffdoo uoiiüAAdqv (ji/S ia) jijvuioai/ d p u v (ip¡) jDJiAdijds p a jv jn jjv j
dA nSij
■3nr " ”n

Podobné dokumenty

proceedings 2006 - RECENT TRENDS 2016

proceedings 2006 - RECENT TRENDS 2016 Recent Trends in Charged Particle Optics and Surface Physics Instrumentation


proceedings 2004 - RECENT TRENDS 2016

proceedings 2004 - RECENT TRENDS 2016 Recent Trends in Charged Particle Optics and Surface Physics Instrumentation Proceedings o f th e 9 th In te rn a tio n a l S e m in a r, held in S k a lsk y d v u r n e a r B rn o , C z e c h R ep...