This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 147.96.14.16

This content was downloaded on 14/10/2013 at 19:47

Please note that terms and conditions apply.

Electron beam induced compositional and structural changes in 

View the table of contents for this issue, or go to the journal homepage for more

1996 Supercond. Sci. Technol. 9 766

(http://iopscience.iop.org/0953-2048/9/9/010)

Home Search Collections Journals About Contact us My IOPscience

iopscience.iop.org/page/terms
http://iopscience.iop.org/0953-2048/9/9
http://iopscience.iop.org/0953-2048
http://iopscience.iop.org/
http://iopscience.iop.org/search
http://iopscience.iop.org/collections
http://iopscience.iop.org/journals
http://iopscience.iop.org/page/aboutioppublishing
http://iopscience.iop.org/contact
http://iopscience.iop.org/myiopscience


Supercond. Sci. Technol. 9 (1996) 766–774. Printed in the UK

Electron beam induced compositional
and structural changes in
Tl2Ba2CuO6+δ

C Dı́az-Guerra †, J Piqueras †, Yu Ya Tomashpolsky ‡,
N V Sadovskaya ‡ and C Opagiste §
† Departamento de Fı́sica de Materiales, Facultad de Ciencias Fı́sicas, Universidad
Complutense, E-28040 Madrid, Spain
‡ Karpov Institute of Physical Chemistry, Vorontsovo Pole 10, 103064 Moscow,
Russia
§ Laboratoire Structure de la Matière, Université de Savoie, BP 240, 9 Rue de
L’Arc-en-Ciel, F-74942 Annecy le Vieux Cédex, France

Received 9 May 1996

Abstract. Tl2Ba2CuO6+δ ceramics have been irradiated in a scanning electron
microscope. The resultant changes in the cationic stoichiometry and oxygen
content of the samples were investigated by cathodoluminescence (CL), secondary
electron emission (SEE) and EDX microanalysis. Tl depletion and Ba enrichment
are observed in the irradiated areas. CL spectroscopy shows that irradiation causes
the increase of a 540 nm emission band attributed to complex centres involving
oxygen vacancies. Other features of the CL spectra strongly suggest the presence
of radiation induced new phases. The SEE yield, 1, has been found to depend on
the oxygen content of the samples. This dependence can be explained on the
basis of two charge carrier subsystems and their mutual interaction. Measurements
of 1 allow us to detect an oxygen deficient region, sometimes extending up to a
distance of about 1000 µm, that surrounds the irradiated material.

1. Introduction

The effect of laser or electron irradiation on the properties
of high-temperature superconductors (HTSCs) is a subject
of technological interest related to the development
of lithographic and patterning processes of potential
application in superconducting electronics [1]. In addition,
the study of radiation effects has application in the
growth of superconducting thin films from target substrates.
Most of the previous reports on this subject refer to
the Y–Ba–Cu–O (YBCO) and Bi–Sr–Ca–Cu–O (BSCCO)
systems. Auger techniques [2, 3] have been used to study
compositional changes induced in YBCO and BSCCO by
laser or electron irradiation. In high-resolution patterning
studies of YBCO films by focused laser beams [4, 5],
it has been found that this material can be locally
modified by varying its oxygen content. Laser writing
induces semiconducting patterns on the superconducting
films by a reduction of oxygen content due to local
heating. Under certain laser irradiation conditions, changes
in the superconducting properties of YBCO are detected
which are interpreted on the basis of photoassisted oxygen
rearrangement [6, 7].

Radiation induced structural and compositional changes
in YBCO and BSCCO have been also investigated
by micro-Raman spectroscopy and cathodoluminescence

(CL) in the scanning electron microscope (SEM) [8–10].
CL microscopy enables us to detect the appearance of
radiation induced insulating phases as well as structural
changes influencing the intrinsic luminescence of the
HTSC samples. Luminescence effects related to electronic
processes in the oxygen sublattice in different HTSC single
crystals and ceramics have been often reported in the past
years (see e.g. [8–14]).

In the present work the effect of electron irradiation
on Tl2Ba2CuO6+δ samples has been investigated in the
SEM by using CL microscopy, energy dispersive x-
ray microanalysis (EDX) and secondary-electron emission
(SEE). SEE has been used in recent years in phase
analysis on the grain scale, in the determination of oxygen
content, and in the detection of phase transitions in YBCO
and BSCCO ceramics and single crystals [15–18]. The
exponential dependence of the SEE yield on the oxygen
concentration explains the high sensitivity of this technique
in the local determination of oxygen deficiency. Detection
of phase transitions is also possible due to significant
anomalies of the secondary-electron emission intensity near
such transitions. For these reasons SEE can be a suitable
complementary technique to CL in oxygen content and
distribution problems.

0953-2048/96/090766+09$19.50 c© 1996 IOP Publishing Ltd



Electron beam induced changes in Tl2Ba2CuO6+δ

Figure 1. CL spectra of the TH2-850b and TKO-930 samples before irradiation: (a) 80 K; (b) 300 K.

Table 1. Relevant physical parameters of the samples investigated (f represents the Meissner fraction).

Sample Batch Tc (K) f (%) Symmetry a (Å) b (Å) c (Å)

TM-850 B 91.1 43.9 I 4/mmm 3.8717 3.8717 23.225
TH2-850b B 52.4 49.1 Ccc2 5.4582 5.4811 23.213
TKO-930 A 0 — Ccc2 5.4467 5.4911 23.144

2. Experimental method

The synthesis procedure of the samples has been described
in detail in [19] and [20]. Tl2Ba2CuO5 and CuO powders
were used as starting components. Two different batches
were prepared from stoichiometric amounts of powders.
For batch A the powder was first kept for 3 days at
750◦C under 1 bar oxygen whereas batch B was not pre-
treated. The powder mixture was pressed into pellets
under 5 Kbar and then annealed under 100 bar of argon
or of oxygen with an appropiate thermal treatment. In
the first case tetragonal superconducting (TS) samples
are obtained while the treatment under oxygen yields
orthorhombic non-superconducting (ONS) samples. A
series of annealing treatments at low oxygen pressure has
shown that starting samples with different symmetry. i.e.
ONS and TS, transform into orthorhombic superconducting
(OS), indicating that the final state of the specimen does not
depend on the initial symmetry. In this work three samples
whose codes and relevant data are listed in table 1 have
been studied. One sample (TM-850) is TS, one (TKO-930)
is ONS and the last (TH2-850b) is OS.

Samples from batch A are x-ray pure. Some traces
of CuO and Ba2Cu3Ox impurities are detected; both are
estimated to be less than 0.1% vol. Optical micrographs
show only Tl-2201 grains. The samples from batch B
contain, besides the 2201 stoichiometric grains, less than
5% by volume of starting Tl2Ba2O5 impurity phase and
very small CuO grains (< 0.1 vol%). Further details on
these samples are described in [19] and [20].

Samples were polished with diamond paste and
observed in the emissive and CL modes in a Hitachi
S-2500 SEM at an accelerating voltage of 25 kV and
temperatures between 80 and 300 K. For CL measurements
an optical lens was used to concentrate the light on a
photomultiplier attached to a window of the microscope.
To record spectra an optical guide feeding the light to an
Oriel 78215 computer controlled monochromator was used.

Large areas of some samples, of 80×100µm2 in TKO-
930 and 50× 180µm2 in TH2-850b, were irradiated under
high-excitation conditions in the SEM. The irradiation was
performed at room temperature with an accelerating voltage
of 25 kV and beam current higher than 10−7 A. These
areas were irradiated until a final saturation condition,
as referred to morphological, total CL intensity and CL
spectral changes, was reached.

In order to detect possible stoichiometric changes in the
irradiated areas, EDX microanalyses were carried out in a
JEOL JSM-35 CF SEM.

The SEE measurements were performed in a modified
BS-350 scanning electron microscope working in UHV
conditions (10−7 Pa) and operating with a field emission
W cathode. The SEE signal from the Everhart–Tornley
type detector was registrated by a digital voltmeter and
simultaneously recorded and treated by computer. The
accelerating voltage of the primary electron beam was
16 kV and the beam current was 10−10–10−11 A.

The SEE yield was measured from scanning areas of
23 × 16, 48 × 34 or 230× 160 µm2 in size. These
measurements were performed along several directions

767



C Dı́az-Guerra et al

Figure 2. Emissive (a) and CL (b) mode micrographs of
the untreated TKO-930 sample.

Figure 3. CL spectra recorded at 80 K from slightly
irradiated and non-irradiated areas of the TKO-930
non-superconducting ceramic.

Figure 4. Secondary-electron (a) and CL (b) micrographs
of an irradiated area of the TH2-850b sample.

across the irradiated areas in 16, 23, 34 and 48µm step
intervals.

3. Results

3.1. Cathodoluminescence

The CL properties before any treatment of the samples
used in this work have been previously reported [21].
An emission band at 540 nm (∼2.3 eV) whose intensity
increases on reducing oxygen content was proposed to
be due to a complex centre involving oxygen vacancies.
Another band centred at about 430 nm (∼2.9 eV) is
also observed in the samples. It has been previously
attributed to processes taking place in the oxygen sublattice,
but has a different origin than the 540 nm emission.
Figure 1 shows representative CL spectra of untreated

768



Electron beam induced changes in Tl2Ba2CuO6+δ

Figure 5. CL spectra recorded at 80 K (a) and 300 K (b) of the TH2-850b superconducting sample before and after strong
electron irradiation.

Figure 6. CL spectra recorded from the TKO-930 ceramic at 80 K (a) and 300 K (b) before and after strong irradiation.

TH2-850b and TKO-930 samples at 80 and 300 K. Both
bands are present at low temperatures while at room
temperature the 540 nm emission clearly dominates. The
appearance of CL images of all samples is similar showing
an inhomogeneous distribution of luminescence (figure 2).
Areas of enhanced intensity are related to pores. Inside
these pores well defined grain surfaces and no foreign
particles are observed, which indicates that the enhanced
CL is related to grain surfaces freshly exposed during
the polishing treatment [21]. Under normal observation
conditions in the SEM no changes are produced in the
topography or luminescence of the samples. When they
are deliberately irradiated, e.g. by using a beam current
higher than 10−7 A, surface changes accompanied by
increase of CL emission are observed. Some intermediate

excitation conditions can cause increase of CL intensity
without apparent topographic changes. CL spectra show
that this slight irradiation causes the increase of the 540 nm
luminescence band (figure 3) while after heavy irradiation
an additional band centred at about 590 nm (∼2.1 eV) is
observed in some samples.

For the reasons described in section 4, mainly the
results corresponding to heavy irradiation have been
considered. Figure 4 shows the secondary-electron and CL
images of an irradiated area of the TH2-850b sample (Tc =
52.4 K). The enhanced CL intensity is inhomogeneously
distributed inside the irradiated region. In figure 5 CL
spectra of the same sample before and after irradiation at
different temperatures are compared. The low-temperature
(80 K) spectra show that the main effect of irradiation is

769



C Dı́az-Guerra et al

Figure 7. A comparison between 80 K (a) and room-temperature (b) CL spectra recorded from irradiated areas of the
TH2-850b and TKO-930 samples.

the appearance of a broad band centred at 590 nm. A
component of this complex band is the original 540 nm
band, resolved in unirradiated samples. This is apparent in
the spectra recorded at 300 K (figure 5(b)) which shows a
band at 540 nm more intense in the irradiated regions.

The effect of irradiation on CL spectra of the non-
superconducting sample (TKO-930) is shown in figure 6.
Spectral changes are small compared with those of the
superconducting sample and the 590 nm peak is not
observed. The difference in the CL emission of both
irradiated samples is clearly observed in figure 7.

CL spectra recorded from the region surrounding the
irradiated areas show a relative increase of the 540 nm
emission band, as compared with the unaffected material
(figure 8).

As described below, a high Ba content was measured
by EDX in the irradiated areas. BaCuO2 has been usually
found as a second phase in sintered HTSCs. In order
to detect the possible presence of small amounts of this
compound in these areas, the CL emission of a BaCuO2

ceramic was investigated. The spectrum recorded at 80 K
shows a wide CL emission in the blue–green spectral range.
A band peaked near 500 nm dominates the spectrum, while
a shoulder centred at about 410 nm can also be observed
(figure 9).

3.2. X-ray microanalysis

X-ray microanalysis was performed in the irradiated area
and at points at increasing distance from it. The
accelerating voltage was 20 kV. The L lines of Tl and
Ba and the K line of Cu were used. Measurements in
different unaffected regions show that the average cationic
composition is nearly stoichiometric (2:2:1).

In superconducting and non-superconducting samples
the irradiated areas have been found to be Tl poor and show
a relative increase of Ba content while only slight changes

Figure 8. CL spectra recorded at 80 K from the area
surrounding the irradiated region (a) and the unaffected
material (b) in the TKO-930 sample.

in Cu composition are detected. Variations of cationic
relative content extend to a distance of few micrometres
from the irradiated area as figure 10 shows.

3.3. Secondary electron emission

Figure 11 shows a plot of the SEE yield of the untreated
samples as a function of theirTc, normalized to the yield of
the non-superconducting sample. Since only three samples
are used in this work, the shape of the curve is unknown and
the line drawn might not correspond to the real situation.
However, the reproducibility of results demonstrates the
general trend, shown in the figure, of an increase of the SEE
yield followed by a decrease at highTc. Since a higherTc

770



Electron beam induced changes in Tl2Ba2CuO6+δ

Figure 9. The CL low-temperature (80 K) spectrum of
BaCuO2.

Figure 10. Cationic composition as a function of distance
to the border of an irradiated area of the TH2-850b
superconducting sample.

value corresponds to a lower oxygen content [19–22], this
result means that on decreasing oxygen content the SEE
yield increases in a first stage and then decreases.

In the ceramics investigated a higher SEE yield in
the irradiated zone has been observed. Since irradiated
regions have strong inhomogeneous surfaces this effect is
not directly measured in large areas including topographical
features. However, when areas of about 10× 10 mm2 are
investigated it is found that the SEE yield from irradiated
regions is very high as compared with non-irradiated ones.
The changes in SEE have a different spatial distribution
than the changes in the cationic composition, and extend to
a larger distance from the irradiated area. In the unaffected
region SEE is stable and lower than in the irradiated area.

Figure 12 shows the SEE yield in the TKO-930 sample
normalized to the value in the unaffected region, as a
function of the distance to the edge of the irradiated

Figure 11. Dependence of the SEE yield on the critical
temperature of the samples investigated. 1 values are
normalized to the TKO-930 (Tc = 0 K) yield.

area. The yield increases from the edge and decreases
after reaching a maximum value at a distance of about
600 µm. Comparison with figure 11 indicates that this
behaviour is compatible with a continuous variation of
oxygen content as a function of the distance to the
irradiated area. The curve of SEE yield shows that
changes in oxygen content extend up to a distance of about
1000µm from the edge. A qualitative evolution of oxygen
composition is also represented in figure 12. The two
ends of the curve correspond to oxygen content in the
borders of the irradiated area and in the unaffected region
respectively. Figure 12 shows that cationic composition
has no significative variations in this intermediate region
beside the irradiated area.

Figure 13 shows the results of similar measurements
in the TH2-850b sample. In this case1 reaches the
stable value at a distance of about 100µm from the
irradiated zone. A qualitative curve of oxygen content as
well as quantitative results of cationic composition are also
represented in figure 13.

4. Discussion

CL images of all untreated samples reveal the existence
of luminescence emission with an inhomogeneous spatial
distribution. This observation agrees with the previous
results referred to in section 1.

In a previous work [21] performed with the same
set of samples used here, it was concluded that the
CL intensity distribution is not due to impurity phases.
This conclusion was obtained from characterization studies
regarding kinds and amounts of impurity phases. A CL
band at 540 nm, whose intensity increases on reducing
oxygen content of the samples, was proposed to be due to a
complex centre involving oxygen vacancies. This band has
been also reported in YBa2Cu3O7−x and Bi2Sr2CaCu2O8+x

samples [8–10] and was related to processes in the oxygen

771



C Dı́az-Guerra et al

Figure 12. SEE yield in the TKO-930 sample normalized to
the value in the unaffected region, cationic composition and
oxygen content distribution, as functions of distance to the
irradiated area. Cationic composition has been normalized
to Cu content.

Figure 13. SEE yield in the TH2-850b superconducting
sample normalized to the value in the unaffected region,
cationic composition and oxygen content distribution, as
functions of distance to the irradiated area. Cationic
composition has been normalized to Cu content.

sublattice. In particular, a study of YBCO [9] has shown
that in this system the initial CL spectra do not depend on
the oxygen content, but during electron irradiation in the
SEM an intense 540 nm band appears in oxygen deficient
samples. Centres causing the 540 nm emission involve
oxygen vacancies but are formed only after irradiation
treatments, similar to the F centre formation in some ionic
crystals. In the samples used in this work the 540 nm
band is already present in the spectra of the untreated
samples, and the possible spectral changes caused by
electron irradiation would not be readily observable as in
the case of YBa2Cu3O7−x . In order to relate the electron
irradiation effects on the samples used in this work to their
structural features, such as oxygen content, a high beam

current—in the range of 10−7 A—was used. A large area
was irradiated in two of the samples and the CL spectral
changes were monitored until a final saturation condition
was reached. The final state of the samples regarding their
CL behaviour was then compared. In spectra of untreated
samples a band centred at about 430 nm (∼2.9 eV) is
also observed. A band at this energy has been previously
reported in ceramic or single crystals of YBa2Cu3O7−x

(YBCO) and Bi2Sr2CaCu2O8−x (BSCCO). It has been
attributed in the case of YBCO and BSCCO to an oxygen
quasimolecule [12] or to a complex involving an F centre
and a negative molecular oxygen ion in YBCO [13].

Electron irradiation causes an enhancement of CL
intensity as observed in CL spectra and images. After a
moderate irradiation, i.e. irradiation not causing topographic
changes, the same luminescence bands are present in the
spectrum. However, a clear change in the relative weights
of these bands is observed. Besides an increase of the total
CL intensity, the 540 nm emission band is enhanced after
irradiation and dominates the spectra.

In regions irradiated up to the saturation conditions,
topographic changes and a drastic increase of the CL
emission are observed (figure 4). CL spectra show that the
main mechanism responsible for the luminescence increase
depends on the sample considered. As figure 5 shows, the
CL increase in the TH2-850b sample is mainly due to the
growth of the 540 nm emission, while the 430 nm one
shows only a slight intensity increase. In the spectra at
80 K the different behaviours of both bands are shown.
At room temperature, spectra show only the 540 nm band
(figure 5(b)). Figure 6 shows the radiation induced spectral
changes in the non-superconducting sample (TKO-930). In
contrast to the case shown in figure 5, a moderate increase
of both bands is observed. For comparison the spectra of
both irradiated samples are shown in figure 7.

This result agrees with the proposal that the 540 nm
emission is related to oxygen vacancy defects and
with the observation in YBCO [9] that irradiation
can generate luminescence centres involving pre-existent
oxygen vacancies. It has been concluded from EDAX and
plasma emission spectroscopy measurements [19, 20, 22]
that the samples used in this work differ in oxygen content
as a result of mass losses during the synthesis process.
The TH2-850b superconducting sample has a lower oxygen
content than the non-superconducting one (TKO-930).

Figures 5–7 show that a lower oxygen content is
related to a higher intensity increase of the 540 nm band
during electron irradiation. Such correlation was observed
in [9] for YBCO. We suggest that, as in YBCO, electron
irradiation generates in Tl2Ba2CuO6+δ centres involving
oxygen vacancies which cause the 540 nm emission. This
band is clearly revealed in spectra recorded at 300 K as in
figure 5(b). However, at 80 K the radiation induced band
in the TH2-850b sample (figure 5(a)) appears centred at
about 590 nm. This band is complex and includes the
540 nm band, revealed in figure 5(b), and at least one
more component at lower energy. The latter would be a
consequence of structural changes which take place during
the severe irradiation treatment producing changes in the
cationic composition of the irradiated area. Due to the

772



Electron beam induced changes in Tl2Ba2CuO6+δ

strong deficiency of thallium and excess of barium detected
in this area, the possibility that the material decomposes into
BaCuO2 and BaO phases has been considered.

Barium cuprate (BaCuO2+x) has been often detected in
sintered 123 HTSC and it was also found after strong laser
irradiation of Hg based superconductors [24]. Comparison
of the CL spectrum recorded from the irradiated area in
TH2-850b with the CL spectrum of BaCuO2 enables us
to rule out the presence of the latter compound in the
mentioned area. The 590 nm luminescence band that
appears in the spectrum of the irradiated region (figure 5(a))
is not observed in the spectrum of BaCuO2 (figure 9). In
addition, the 500 and 410 nm CL bands, present in the
latter spectrum, were not found in the one recorded in the
Tl based superconductor after electron irradiation. Thus,
the 590 nm band is related to the formation of new phases,
different from BaCuO2, or to other irradiation induced
structural changes.

In the electron irradiated area an increase of the SEE
yield has been observed. This effect was also found in
irradiated YBCO [23], and can be explained as a result
of the electron irradiation induced changes in the cationic
composition and oxygen losses [17]. According to the
results summarized in figures 11–13, the increase of the
SEE yield could be partially due to the reduction of oxygen
content during irradiation. This would agree with the
enhancement of the 540 nm CL band, also related to oxygen
content, during irradiation.

In an extended region around the irradiated area, in
which cationic composition and topography have not been
modified, SEE measurements indicate a decrease of oxygen
content. In this region the reduced oxygen content is
accompanied by an increase of CL intensity, as compared
with CL intensity in unaffected areas. CL images do not
show clear changes in the luminescence distribution around
the irradiated region, but spectra show that the 540 nm
CL emission band is stronger in the region surrounding
the irradiated material than far from it (figure 8). Our
observations support the relation between the 540 nm
luminescence band and oxygen depletion, as confirmed by
the SEE yield measurements of oxygen deficiency around
the irradiated area. These results agree with previous
investigations in electron irradiated YBCO ceramics [23].
In [23] thermally driven diffusion processes were proposed
as the mechanism leading to the creation of the oxygen
depleted area surrounding the irradiated region.

The present observations show that both SEM based
techniques, SEE and CL, can be used to study processes
related to the oxygen sublattice of TBCO. SEE enables
us to detect variations of oxygen content with high spatial
resolution while a CL emission band is related to complex
defects involving oxygen vacancies. SEE and CL can
provide complementary information on the defect structure
in the oxygen sublattice of TBCO. The CL properties
of TBCO have been previously discussed [14, 21] and
compared with similar results concerning other high-Tc

superconductors. On the other hand SEE results on TBCO
samples have not been, to our knowledge, previously
reported. For this reason the present results, in particular
the qualitative dependence of the yield on oxygen content
shown in figure 11, will be discussed in the following.

In order to explain the SEE properties of Tl based
ceramics, the co-existence of two electronic subsystems
should be considered: free electrons related to oxygen
deficiency (I) and holes (II). The latter are known to be the
charge carriers in the Cu–O planes. This assumption does
not contradict models of electronic processes in HTSCs
with oxygen excess, such as BSCCO and TBCO [25–30].
In addition, an interaction between the two subsystems
is present when a change in free electron concentration
influences the number of holes [28].

We consider the basic formula describing the SEE
yield [16]

1 = βnλBe−λ/L

where β is the inelastic scattering cross-section of sec-
ondary electrons at primary electrons,n the concentration
of free carriers,λ the escape depth for secondary electrons
(SEs) andL its mean free path.B represents the probabil-
ity of the SEs overcoming the surface barrier, whileβnλ

describes the SE generation. The e−λ/L term describes the
SE scattering.

Changes in oxygen content (x) need to be considered as
a parameter. The influence of the exponential term is dom-
inant and different mean free paths should be considered
for charge carriers which belong to different subsystems
(I and II). For the free electrons related to oxygen defi-
ciencyLe = 1/neαe, while Lh = 1/nhαh for holes.αe and
αh represent the inelastic scattering cross-sections of sec-
ondary electrons at charge carriers of both subsystems. The
dependence of the exponential term exp[−λ(αene + αhnh)]
on x is related to the influence ofx on ne andnh. Whenx

(and oxygen concentration) decreases, the number of free
electrons grows whereas the number of holes drops. These
effects would respectively decrease and increase the value
of the SEE yield1. The net change in1 depends on
the interaction mechanism between the two subsystems (I
and II), initial hole concentration, difference betweenαe

and αh and permissible oxygen losses. For example, in
Bi2Sr2CaCu2O8−x a significant decrease of the hole con-
centration is observed for small oxygen losses [28], and1

increases. Further oxygen losses cause the increase ofne,
and decrease of1. As a consequence, a maximum in the
1(x) curve is observed. A qualitative analogy has been
found in our Tl based ceramics (figure 11). The model
described here is valid if the electronic processes related
to oxygen losses in this material lead to a decrease of the
hole concentration and to an increase of the number of free
electrons. The first seems to be experimentally proved, (see
e.g. [25, 27, 29, 30]). The existence of free electrons after
oxygen loss seems to be natural independent of the ori-
gin of such a deficiency (vacancy generation, removal of
interstitial oxygen atoms, etc). These considerations have
preliminary character, and further theoretical and experi-
mental studies should be performed to explain in detail the
complex behaviour of the1 function.

5. Conclusions

Electron irradiation of Tl2Ba2CuO6+δ (TBCO) in the SEM
causes changes in the cationic stoichiometry as well as vari-
ations in oxygen content. Irradiated areas have been found

773



C Dı́az-Guerra et al

to be Tl poor and show a relative increase of Ba content.
CL spectroscopy shows that irradiation causes the increase
of a 540 nm emission band attributed to complex defects in-
volving oxygen vacancies. Other features of the CL spectra
suggest the presence of radiation induced new phases.

The value of the SEE yield,1, has been found to
depend on the oxygen content of the samples. The
dependence can be explained on the basis of two electron
subsystems and their mutual interaction. Measurements
of 1 enable us to study variations of oxygen content
in the samples with high spatial resolution. In the
case of irradiated samples it was observed that variations
of the oxygen content extend in some cases up to a
distance of about 1000µm from the electron irradiated
area. This effect should be taken into account in the
development of lithographic and paterning processes in
high-Tc superconductors.

In the oxygen deficient region that surrounds the
irradiated material, an enhancement of the oxygen
deficiency related CL band has been observed. This
supports the relation between this 540 nm luminescence
band and oxygen depletion, as confirmed by the SEE yield
measurements.

Acknowledgments

The authors thank S G Prutchenko for EDX measurements.
CDG thanks the Spanish MEC for an FPI research
grant. This work has been supported by DGICYT (project
PB 93-1256), CICYT (project MAT 95-1184-E), NATO
(grant HTECH. LG 941034), the Russian Foundation
of Fundamental Research (project 94-03-0802) and the
Russian State Programme on HighTc Superconductivity
(project 93079).

References

[1] Tolpygo S, Nadgorny B, Shokhor S, Tafuri F, Lin Y,
Bourdillon A and Gurvitch M 1993PhysicaC 209 211

[2] Auciello O, Kraus A R, Santiago Avilés J, Schreiner A F
and Gruen D M 1988Appl. Phys. Lett.52 239

[3] Matsui S, Ichihashi T, Yoshitake T, Miura S, Satoh T and
Mito M 1990 J. Vac. Sci. Technol.B 8 1771

[4] Shen Y Q, Freltoft T and Vase P 1991Appl. Phys. Lett.59
1365

[5] Sobolewski R, Xiong W, Kula W and Gavaler J R 1994
Appl. Phys. Lett.64 643

[6] Kawamoto K and Hirabayashi I 1994Phys. Rev.B 49 3655
[7] Osquiguil E, Maenhoudt M, Wuyts B, Bruynseraede Y,

Lederman D and Schuller I K 1994 Phys. Rev.B 49
3675

[8] Gómez P, Jiḿenez J, Mart́ın P, Piqueras J and
Doḿınguez-Adame F 1993J. Appl. Phys.74 6289

[9] Gómez P, Piqueras J, Sayagués M J and Gonźalez-Calbet
J M 1995Solid State Commun.96 45

[10] Torres A, Jiḿenez J, Ǵomez P and Piqueras J 1995Mat.
Res. Soc. Symp. Proc.vol 373 (Pittsburgh, PA: Materials
Research Society) p 431

[11] Cooke D W, Rempp H, Fisk Z, Smith J L and Jahan M S
1987Phys. Rev.B 36 2287

[12] Stankevitch V Get al 1993Phys. Rev.B 47 1024
[13] Fugol I, Politis C, Ratner A, Samovarov V and Zhuravlev

V 1994 J. Lumin.62 291
[14] Gómez P, Piqueras J and Opagiste C 1994Solid State

Commun.91 747
[15] Tomashpolsky Yu Ya, Sevostyanov M A, Sadovskaya N V

and Kolganova N V 1989J. Microsc. Spectrosc.
Electron.14 213

[16] Tomashpolsky Yu Ya 1993Ferroelectrics144 231
[17] Tomashpolsky Yu Ya and Sadovskaya N V 1995

Ferroelectrics163 129
[18] Tomashpolsky Yu Ya and Sadovskaya N V 1995

Ferroelectrics170 97
[19] Opagiste C, Couach M, Khoder A F, Abraham R, Jondo

T K, Jorda J L, Cohen-Adad M Th, Junod A, Triscone
G and M̈uller J 1993J. Alloys Compounds.195 47

[20] Opagiste C, Triscone G, Couach M, Jondo T K, Jorda J L,
Junod A, Khoder A F and Müller J 1993PhysicaC 213
17

[21] Dı́az-Guerra C, Piqueras J and Opagiste C 1996PhysicaC
259 121

[22] Opagiste Cet al 1993PhysicaC 205 247
[23] Dı́az-Guerra C, Piqueras J, Tomashpolsky Yu Ya and

Sadovskaya N V 1996Phys. Status Solidia 155 525
[24] Ren Y Y, Chang H, Xiong Q, Xue Y Y and Chu C W

1994PhysicaC 226 209
[25] Shimakawa Y, Kubo Y, Manako T and Igarashi H 1989

Phys. Rev.B 40 11 400
[26] Martin C, Maignan A, Provost J, Michel C, Gerveu M,

Tournier B and Braveau B 1990PhysicaC 168 8
[27] Shimakawa Y, Kubo Y, Manako T, Igarashi H, Izumi F

and Asano H 1990Phys. Rev.B 42 10 165
[28] Nagoshi M, Suzuki T, Fukuda Y, Terashima K, Nakanishi

Y, Ogita M, Tokiwa A, Syono Y and Tachiki M 1991
Phys. Rev.B 43 10 445

[29] Kubo Y, Shimakawa Y, Manako T and Igarashi H 1991
Phys. Rev.B 43 7875

[30] Shimakawa Y 1993PhysicaC 204 247

774