Secondary defect dissolution by voids in silicon

Article abstract of DOI:10.1063/1.117485

Clustering of point defects into dislocations as a consequence of ion implantation and annealing in silicon has been inhibited through the presence of cavities produced by helium ion implantation and located at depths up to 1 μm. We will show that annealing at 1200°C will result in the complete absence of extended defects for the case of either partial damage produced by boron ion implantation or continuous amorphous layers produced by germanium implantation. The suppression of dislocation formation does not depend on the depth at which cavities are located, on their density and on the purity of the adopted silicon wafer. The results are interpreted on the basis of the measured efficiency of voids in the capture of interstitial silicon atoms.

Full text of DOI:10.1063/1.117485

Secondary defect dissolution by voids in silicon
V. Raineri and S. U. Campisano
Citation: Applied Physics Letters 69, 1783 (1996); doi: 10.1063/1.117485
View online: http://dx.doi.org/10.1063/1.117485
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/69/12?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Gettering of Fe to below 1010 cm−3 in MeV selfimplanted Czochralski and float zone silicon
Appl. Phys. Lett. 69, 4203 (1996); 10.1063/1.116986
A transmission electron microscopy study of microstructural defects in proton implanted silicon
J. Appl. Phys. 80, 4767 (1996); 10.1063/1.363414
Role of silicon surface in the removal of point defects in ultrashallow junctions
Appl. Phys. Lett. 69, 2228 (1996); 10.1063/1.117136
Influence of dislocation loops created by amorphizing implants on point defect and boron diffusion in silicon
Appl. Phys. Lett. 68, 3570 (1996); 10.1063/1.116640
Lifetime control in silicon devices by voids induced by He ion implantation
J. Appl. Phys. 79, 9012 (1996); 10.1063/1.362633
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.113.111.210 On: Sun, 21 Dec 2014 02:18:25

Secondary defect dissolution by voids in silicon
V. Raineri
CNR-IMETEM, Stradale Primosole, 50-I95121 Catania, Italy
S. U. Campisano
Dipartimento di Fisica, Corso Italia, 57-I95129 Catania, Italy
͑Received 8 April 1996; accepted for publication 10 July 1996͒
Clustering of point defects into dislocations as a consequence of ion implantation and annealing in
silicon has been inhibited through the presence of cavities produced by helium ion implantation and
located at depths up to 1 ␮m. We will show that annealing at 1200 °C will result in the complete
absence of extended defects for the case of either partial damage produced by boron ion
implantation or continuous amorphous layers produced by germanium implantation. The
suppression of dislocation formation does not depend on the depth at which cavities are located, on
their density and on the purity of the adopted silicon wafer. The results are interpreted on the basis
of the measured efficiency of voids in the capture of interstitial silicon atoms. © 1996 American
Institute of Physics. ͓S0003-6951͑96͒04838-3͔
High dose helium ion implantation in crystalline solids
give rise to small ‘‘precipitated’’ bubbles which evolve into
empty cavities by gas desorption after annealing at interme-
diate temperatures. Recent work has shown that the internal
surface of these cavities is quite efficient in trapping metallic
impurities and therefore can be successfully used as a local-
ized gettering center.1,2 The surface of these cavities may, in
addition, act as a generation-recombination site supplying a
quite efficient means for localized control of the minority
carriers lifetime in silicon devices.3 All these results indicate
that such an internal surface is highly reactive; it may there-
fore be efficient as a generation recombination site for point
defects as well.
duce a reference sample. The energy deposited by the boron
ions is not enough to form amorphous layers while the ger-
manium implant produced a 0.3 ␮m thick amorphous layer.
Since the presence of carbon and oxygen noticeably influ-
Point defects in excess of the equilibrium concentration
are generated during ion implantation; their clustering as a
consequence of the annealing process gives rise to well-
known effects such as end of range dislocation loops for
heavy damaging implants or to threading dislocations for
low damage ion implantation. These extended defects may
be electrically active or may trap dopants thus worsening
device performances.4–6 In this letter we report experimental
results on the interaction among ion implantation point de-
fects and voids internal surface.
Czochralski grown ͑100͒ silicon wafers were implanted
with 40, 100, or 300 keV He ions at doses of 3ϫ1016 or 6
ϫ1016 ions/cm2. Annealing at 1100 °C for 5 min allowed
helium to evaporate from the wafer and the formation of a
stable void layer. Systematic transmission electron micros-
copy observations on cross-sectioned samples revealed that
the width of such a layer is 0.3 ␮m while, depending on the
implantation energy, it stands at a depth of 0.2, 0.5, or 1 ␮m,
respectively. Moreover, the density of the cavities changed
according to the helium ion dose thus producing a set of six
groups of wafers. Wafers of each group were then implanted
by boron ions at 80 keV energy and 2ϫ1015/cm2 dose or by
germanium ions at 70 keV energy and 5ϫ1015/cm2. The
same implants were done on virgin material in order to pro-
FIG. 1. TEM plan views of samples implanted with 80 keV
2
ϫ1015 B/cm2 with no voids ͑a͒ or with voids formed at a depth of 0.5 ␮m
͑b͒ or of 1.0 ␮m ͑c͒.
a͒
Electronic mail: raineri@ct.infn.it
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
Appl. Phys. Lett. 69 (12), 16 September 1996 0003-6951/96/69(12)/1783/3/$10.00 © 1996 American Institute of Physics 1783
130.113.111.210 On: Sun, 21 Dec 2014 02:18:25

FIG. 3. Void empty volume for volume unity ͑%͒ versus the annealing
temperature. In all cases the annealing time is 1 h.
layer and on the purity of the silicon substrate.
The formation of threading dislocations and end of range
defects is usually attributed to clustering of interstitials. In a
first approximation the number of interstitials gettered by
voids at different temperatures can be estimated by consid-
ering the evolution of the cavity empty volume for unit vol-
ume. If a great amount of interstitials is trapped at the void
surface an epitaxial growth occurs, then the empty volume
decreases. This behavior was observed for the implanted He
doses at all the considered energies but at temperatures
higher than 1000 °C as plotted in Fig. 3. While up to 1000 °C
the empty value remains almost constant versus the anneal-
ing temperature, at higher temperatures it decreases. Then,
the number of gettered silicon atoms can be calculated by
considering the filled volume at a fixed temperature. In the
four considered cases it is in the 2ϫ1016–1ϫ1017/cm2
range. This large number of Si atoms is gettered during void
evolution. Stable conditions are reached after thermal treat-
ments longer than 1 h.2 This means that in our case a large
part of Si atoms can be gettered at the void surface during the
thermal treatment. The number of interstitials introduced by
the implants, according to the plus one model,7 does not
exceed 2ϫ1015/cm2, i.e., less than the 10% of the void get-
tering capability.
In conclusion secondary defects dissolution was ob-
served at temperatures higher than 1000 °C if a void layer is
present deeper than 1 ␮m. The result is independent on the
void density, i.e. the internal surface for unit volume. More-
over, the effect is independent of the kind of primary damage
formed by ion implantation and it is related with the void
evolution. No influence of impurities ͑C,O͒ was also ob-
served. Voids act as traps for interstitials during high-
temperature annealings when the empty volume is reduced
due to the internal surface growth.
FIG. 2. TEM cross sections of samples with a void layer at a depth of 1.0
␮m implanted with 70 keV 5ϫ1015 Ge/cm2 and annealed at 1000 °C for 30
min ͑a͒ or 1200 °C for 3 min ͑b͒.
ences the diffusion of point defects in silicon, all the experi-
ments were repeated by using epitaxial silicon wafers, which
contain a much smaller amount of these impurities.
Transmission electron microscopy ͑TEM͒ at 200 kV has
been extensively used to reveal the nature and the amount of
residual damage. In Fig. 1 the plan view micrographs are
reported, taken under the two beam condition with the ͑220͒
spot excited, for boron implanted samples annealed for 3 min
at 1200 °C. Figure 1 shows the network of dislocations re-
sulting from the processing in the reference sample. Figures
1͑b͒ and 1͑c͒ show the results of the processing made on
samples containing a void layer located at depths of 0.5 and
1.0 ␮m, respectively. As it appears no dislocations are vis-
ible and this result is independent of the void density and on
the presence of carbon or oxygen in the sample. It is worth
noting that the implant and the diffusion caused by the ther-
mal processing produce a boron profile which extend up to a
depth of 0.9 ␮m, while the extended defects are centered at
the end of range of the boron implanted ions ͑0.3 ␮m͒. The
depth of the void layer can be deeper than the possible junc-
tion depth, as shown in the extreme case reported in Fig.
1͑c͒, where the thickness of analyzed material produces a
loss of resolution which does not allow a good observation
of the cavities.
Similar experiments have been performed on samples
containing a different kind of damage, i.e., amorphous layers
which result, upon annealing, in the formation of a well-
defined layer of dislocation loops at the end of range. Figures
2͑a͒ and 2͑b͒ show cross section TEM micrographs of
samples implanted by germanium ions at the above-
mentioned conditions and annealed for 30 min at 1000 and
1200 °C, respectively. Also in this case the effect of the
1200 °C annealing is evident with the disappearance of the
end of range dislocation layer while annealing at 1000 °C
produces only a small reduction in their density. Also in this
case, the result is independent on the depth of the cavities
We thank M. Furnari and S. Pannitteri for the skilled
technical assistance and we are grateful to C. Polizzotto for
the assistance in the experimental work.
1 S. M. Myers, D. M. Follstaedt, G. A. Petersen, C. H. Seager, H. J. Stein,
and W. R. Wampler, Nucl. Instrum. Methods B 106, 379 ͑1995͒.
2 V. Raineri, P. G. Fallica, G. Percolla, A. Battaglia, M. Barbagallo, and S.
U. Campisano, J. Appl. Phys. 78, 3727 ͑1995͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
1784 Appl. Phys. Lett., Vol. 69, No. 12, 16 September 1996 V. Raineri and S. U. Campisano
130.113.111.210 On: Sun, 21 Dec 2014 02:18:25

3 V. Raineri, G. Fallica, and S. Libertino, J. Appl. Phys. ͑1996͒
͑in press͒.
Saris, Mater. Sci. Rep. 6, 275 ͑1991͒.
6 N. Cuendet, T. Halicioglu, and W. A. Tiller, Appl. Phys. Lett. 68, 19
͑1996͒.
4 K. S. Jones, S. Prussin, and E. R. Weber, Appl. Phys. A 45, 1 ͑1988͒.
5 R. J. Schreutelkamp. J. S. Cluster, J. R. Liefting, W. X. Lu, and F. W.
7 M. D. Giles, J. Electrochem. Soc. 138, 1160 ͑1991͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
Appl. Phys. Lett., Vol. 69, No. 12, 16 September 1996 V. Raineri and S. U. Campisano 1785
130.113.111.210 On: Sun, 21 Dec 2014 02:18:25