PAPER • OPEN ACCESS You may also like
Single-ion induced surface modifications on - Contribution of terahertz waves to near-field radiative heat transfer between
graphene-based hyperbolic metamaterials
hydrogen-covered Si(001) surfaces—significant Qi-Mei Zhao,  , Tong-Biao Wang et al.
difference between slow highly charged and swift - Novel insight on the local surfaceproperties of ZnO nanowires
heavy ions Monika Kwoka, Anna Kulis-Kapuscinska,Dario Zappa et al.
- Low- and high-density InAs nanowires on
23 Si(0 0 1) and their Raman imagingTo cite this article: C Länger et al 2021 New J. Phys.  093037 Suparna Pal, S D Singh, V K Dixit et al.
 
View the article online for updates and enhancements.
This content was downloaded from IP address 134.176.66.200 on 05/09/2022 at 13:43
New J. Phys. 23 (2021) 093037 https://doi.org/10.1088/1367-2630/ac254d
PAPER
Single-ion induced surface modifications on hydrogen-covered
OPEN ACCESS Si(001) surfaces—significant difference between slow highly
RECEIVED
2 August 2021 charged and swift heavy ions
ACCEPTED FOR PUBLICATION
9 September 2021
C Länger1, P Ernst2, M Bender3, D Severin3, C Trautmann3,4 , M Schleberger2 and
PUBLISHED
27 September 2021 MDürr
1,∗
1 Institut für Angewandte Physik and Zentrum für Materialwissenschaften, Justus-Liebig-Universität Giessen, 35392 Giessen, Germany
2 Fakultät für Physik and CENIDE, Universität Duisburg-Essen, 47057 Duisburg, Germany
Original content from 3
this work may be used GSI Helmholtzzentrum für Schwerionenforschung, 64291 Darmstadt, Germany
4
under the terms of the Fachbereich Materialwissenschaften, Technische Universität Darmstadt, 64287 Darmstadt, Germany
Creative Commons ∗ Author to whom any correspondence should be addressed.
Attribution 4.0 licence.
E-mail: michael.duerr@ap.physik.uni-giessen.de
Any further distribution
of this work must
maintain attribution to Keywords: swift heavy ions, silicon, highly charged ions, scanning tunneling microscopy, STM
the author(s) and the
title of the work, journal Supplementary material for this article is available online
citation and DOI.
Abstract
Hydrogen-covered Si(001) surfaces were exposed to swift heavy ions (SHI) and slow highly
charged ions (HCI). Using scanning tunneling microscopy as analysis tool, the ion-induced
modifications on the surface were resolved on the atomic scale. SHI were found occasionally to
lead to changes which are restricted to one or two Si surface atoms. In comparison, HCI form pits
of several nanometers in diameter, depending on the potential energy of the HCI. These
observations are in contrast to many material systems for which similar effects of SHI and HCI
have been observed. The results suggest a high stopping power threshold for SHI-induced
modifications in crystalline silicon with major implications for the application in silicon-based
nanotechnology.
1. Introduction
The formation of nanostructures on the surface of solids by either swift heavy ions (SHI) or slow highly
charged ions (HCI) has been observed for many different materials, including oxides, halides, polymers,
and 2D materials [1–3]. Both ion types interact primarily with the electronic subsystem of the target
material; it has therefore been suggested that the subsequent relaxation mechanisms might be similar and,
as a consequence, should lead to similar modifications. The evident morphological similarity of SHI- and
HCI-induced nanostructures does indeed support this assumption. In addition, the two-temperature model
(TTM) [4], which has been applied to interpret SHI related data with great success [5–8], has recently also
been applied to explain data from HCI irradiations [9–12].
The TTM takes into account that the ion projectiles deposit their energy in the electronic subsystem
and, via electron–phonon coupling, produce a thermal spike in the atomic system. Depending on the
energy density and the material properties, the thermal spike can be sufficient to drive phase transitions
such as melting. The relevant phonons are generated on the timescale of picoseconds, while the initial
electronic excitation happens within femtoseconds and forms a highly localized non-equilibrium state. The
details of the coupling of these two subsystems are known to be strongly influenced by material properties
(e.g. conductive versus insulating, crystalline versus amorphous structure). Track formation in the bulk and
damage creation on the surface are thus both linked to a stopping power threshold which depends on the
materials investigated [1].
A most interesting test case for the comparison of SHI and HCI is crystalline silicon, for which no
track formation was observed in the case of irradiation with monoatomic SHI up to a stopping power of
© 2021 The Author(s). Published by IOP Publishing Ltd on behalf of the Institute of Physics and Deutsche Physikalische Gesellschaft
New J. Phys. 23 (2021) 093037 C Länger et al
21 keV nm−1 [13, 14]. Only when using C60 cluster ions (total kinetic energy of 30–40 MeV) [15, 16] with
increased stopping power and effective energy density due to the lower ion velocity, ion tracks in the bulk of
crystalline silicon have been observed and a stopping power threshold of ≈30 keV nm−1 was deduced. On
the other hand, HCI irradiation of Si(111) 7 × 7 was shown to induce damage in areas of several
nanometers in diameter with a clear dependence on the potential energy of the employed ions [17–19].
The unusually high stopping power threshold was not reproduced in TTM calculations, which predict a
threshold between 5 and 10 keV nm−1 for ion track formation in Si [14, 20, 21]. Possible reasons for this
discrepancy might be found in the details of the energy transfer from the electronic to the atomic
subsystem, e.g. the temporal and spatial distribution of low energy electrons [22], resulting in a lower
electron–phonon-coupling in silicon than assumed in the calculations. It has also been speculated that this
seemingly high threshold is not caused by the details of the excitation process as such but that fast
recrystallization of the silicon lattice suppresses the formation of permanent damage or renders it at least
too small (or too short-lived) to be imaged correctly by the ex situ analysis tools applied. As an example, the
different response of various oxides to swift Xe ions have been discussed in this way [23].
For the further interpretation of ion-induced phenomena in silicon, it is thus of great importance to rule
out recrystallization as the origin for the high stopping power threshold. In our experiments, we made use
of hydrogen passivated Si(001) surfaces in combination with scanning tunneling microscopy (STM) in
order to explicitly address this point: on one hand, hydrogen desorption from these surfaces leads to distinct
signatures in STM measurements [24, 25] that can be resolved on the atomic scale. On the other hand,
desorption of hydrogen under ultra-high vacuum (UHV) conditions is an irreversible process. Thus, if a
thermal spike would lead to hydrogen desorption at the surface at any time after ion impact, this
modification would be observable even if the substrate itself recrystallizes during the further process. With a
resolution on the atomic scale, the experiments are sensitive to very small changes on the surface.
Furthermore, working under UHV conditions excludes the formation of silicon oxide on the surface, which
easily forms ion-induced volume tracks and surface hillocks [26].
For the irradiation, we used U ions of highest energy deposition available with monoatomic ions
(4.8 MeV/u, 1.14 GeV, dE/dx = 27 keV nm−1). Even for those extreme conditions, we observed, if at all,
only a very small effect of SHI on the H/Si(001) surface; the detected signatures were restricted to one or
two silicon surface atoms. We compare our results for SHI on H/Si(001) with the effect of HCI (Xe40+ and
Xe45+) carrying a maximum potential energy of Epot ≈ 60 keV (Xe45+) and a kinetic energy of 225 keV. For
HCI, we found substantial disorder and removal of material on/from the H/Si(001) surface caused by single
ion impacts, comparable to results on clean Si(111). We discuss our results in terms of a high stopping
power threshold for SHI-induced modifications in silicon.
2. Experimental
Surface microscopy investigations were performed at Justus Liebig University Giessen with a variable
temperature STM (Omicron VT-STM) mounted in a UHV chamber with a base pressure <1 ×10−10 mbar.
The STM measurements were performed in constant-current-mode using a tungsten tip and a sample bias
of Ut = −2 V; the tip thus images surfaces of constant density of states at the given energy, i.e. filled
electronic surface states in case of negative sample bias. Si(001) samples were prepared by degassing at
700 K and repeated direct current heating cycles to above 1450 K; cooling rates of about 1 K s−1 led to a
well-ordered 2 × 1 reconstruction [27, 28]. Hydrogen terminated surfaces were prepared by exposing a
heated Si(001) surface (Ts  650 K) to atomic hydrogen, leading to either a 2 × 1 or 3 × 1 H/Si(001)
reconstruction, depending on the exact surface temperature. Atomic hydrogen was generated by
dissociation of molecular hydrogen at a hot tungsten filament (TF  2000 K).
Figure 1 (top) shows filled state STM images of both, the H/Si(001) 2 × 1 and 3 × 1 reconstruction, as
well as a ball-and-stick model of the respective reconstruction. In the bottom, a schematic representation of
the performed irradiation experiments is shown. The samples which were prepared in the STM apparatus
were transported under UHV to the respective ion beam facility. Irradiation with SHI was realized at
UNILAC of GSI Helmholtzzentrum für Schwerionenforschung; beam parameters were: 238U28+ with
E = 1.14 GeV (4.8 MeV/u) at three different fluences of 0.2, 1.8, and 5× 1011kin cm−2. Irradiation with HCI
was performed at University Duisburg-Essen at the HICS beamline [29] equipped with a commercial
DREEBIT EBIS-A ion source; beam parameters were: 129Xe40+ with Epot = 38.5 keV at a fluence of
≈3× 1011 cm−2 as well as 129Xe45+ with Epot = 58.8 keV at a fluence of ≈1× 1011 cm−2. Kinetic energy
was Ekin = 225 keV in both cases. Both, SHI and HCI irradiation, were performed with the ion beam
perpendicular to the surface. After irradiation, the samples were transferred back to the STM chamber,
again without breaking the vacuum using a UHV transport system with mobile pumping unit (base
2
New J. Phys. 23 (2021) 093037 C Länger et al
Figure 1. Top: filled state STM images (5× 10 nm2; Ut =−2 V; It = 0.5 nA) of hydrogen-covered Si(001) surfaces with
2× 1 (left) and 3× 1 (right) reconstruction. Center: ball-and-stick models of the respective surface reconstructions (blue balls:
hydrogen atoms; yellow balls: silicon atoms). Bottom: schematic representation of the experiments: after preparation of the
H/Si(001) surfaces and sample transfer in UHV to the respective ion sources, the samples were irradiated with SHI or HCI (the
red arrow indicates the impact of an individual ion). After UHV transfer back to the analysis chamber, defects on the surface
(indicated by the white and red signatures) were investigated by means of STM.
pressure <1 ×10−10 mbar). In the load locks of the respective chambers, the samples were briefly exposed
to a residual gas pressure <1 ×10−9 mbar. Alle experiments were performed at room temperature.
3. Results
In figure 2, STM images (Ut = −2 V; It = 0.5 nA) of an H/Si(001) surface before and after irradiation with
238U28+ ions are shown. In all images, both before and after irradiation, the dimer row reconstruction of
Si(001) is clearly observed. The unirradiated samples exhibit a few additional bright and dark features
(figure 2(a)). The dark features are typically ascribed to missing dimer defects. The bright features are
assigned to silicon surface atoms which are not saturated by a hydrogen atom [24, 25]. The respective
dangling bond states show a higher density of states at the chosen sample bias leading to the increased
brightness in the images. In principle, additional adsorbates on the H/Si(001) surface can lead to such
bright features as well; however, in the case of the freshly prepared surfaces, we can largely exclude such
contaminations.
STM images after irradiation with the lowest fluence (0.2 ×1011 cm−2 which equals 0.4 single ion
impacts on the depicted area) and highest fluence (5 ×1011 cm−2 equaling nine single ion impacts on the
depicted area) applied in our experiments are shown in figures 2(b) and (c), respectively. After irradiation,
the well-defined surface reconstruction was still intact. However, more bright features were observed than
prior to irradiation whereas the number of dark features stays approximately the same. The number of the
additional bright features is increasing with increasing fluence, however, the observed density is significantly
higher than the applied fluence (approx. 5 to 10 times, compare figure S1 (https://stacks.iop.org//NJP/
23/093037/mmedia) in the supporting information). Reference samples kept in the irradiation chamber but
positioned outside the ion beam did not show these additional features. Therefore the results for the
irradiated samples indicate ion-induced effects which were either directly or indirectly induced by SHI.
We analyzed the spatial distribution of the bright features by evaluating the distribution of next
neighbor distances in the STM images; for comparison, Monte-Carlo simulations of a purely statistical
distribution were performed (figure S2 in the supporting information). The experimental results and the
simulations are almost identical (figure S2) pointing toward a purely stochastic distribution of bright
features after irradiation.
The experimental results can thus be summarized as follows: (i) we observe additional bright spots, but
with a higher density (5–10 times) than expected from the applied fluence when assuming that every ion
impact leads to one single bright feature. No additional dark features were observed. (ii) The bright features
do not show any spatial accumulation on the surface but follow a stochastic distribution. From the latter
observation, we can exclude that a single ion impact leads to several of the bright features, as they then
3
New J. Phys. 23 (2021) 093037 C Länger et al
Figure 2. Filled state STM images (Ut =−2 V; It = 0.5 nA) of hydrogen-terminated Si(001) surfaces before and after irradiation
with 238U28+ ions at different fluences. (a) Before irradiation, the H/Si(001) surface shows a mixture of 3× 1 and 2× 1
reconstruction with few dark and bright features. (b) H/Si(001) surface after irradiation with the lowest ion fluence used in the
experiments (0.2×1011 cm−2 equaling 0.4 single ion impacts on the depicted area). The surface reconstruction is preserved.
(c) For the highest fluence (5.0×1011 cm−2 equaling nine single ion impacts on the depicted area), a distinct increase of bright
features is observed after irradiation. (d) A blow up of (c) indicates the size of the bright features. (e) H/Si(001) after thermal
desorption of a small amount of hydrogen [24, 30]; the observed features are of similar size as in (d).
should be accumulated around the impact point: with the known mobility of hydrogen on silicon and its
temperature dependence [25], a complete equilibration of an initially inhomogeneous distribution is not
possible on the relevant time scales. The majority of the bright features may thus be attributed to contami-
nations adsorbed on the surface. They could result, e.g. from ion-induced sputtering in the vicinity of the
sample, e.g. from the ceramic sample holder plate. Indeed, when scanning the surface with the STM tip,
some of the bright features could be moved on the surface. This is not possible for covalently bound
hydrogen at the tunneling parameters employed.
Still, some of the bright features might directly origin from a single ion impact and ion-induced
hydrogen desorption. It is thus interesting to compare the size of the bright features with images of a
hydrogen terminated Si(001) surface on which isolated dangling bonds and dangling bond pairs were
created intentionally by means of thermal hydrogen desorption and diffusion as shown in figure 2(e) [24].
This comparison clearly indicates that the bright features on the irradiated surfaces, as shown in figures 2(c)
and (d), are restricted to one or two silicon surface atoms. We thus conclude that the single ion impacts do
not lead to surface modifications larger than one silicon dimer. Moreover, as the hydrogen-terminated
surface stays intact across the entire irradiated area, we exclude that the impact of a single ion leads to local
melting followed by perfect recrystallization. Hydrogen desorption occurs at a much lower surface
temperature than surface melting or ablation [30]; under UHV conditions, hydrogen desorption is
irreversible as the hydrogen partial pressure is low and hydrogen dissociation on silicon shows a very low
reactivity [24, 31]. Thus, even if the surface recrystallized after temporal melting (thermal spike), the
originally adsorbed hydrogen would be desorbed. In consequence, the molten/recrystallized areas would
4
New J. Phys. 23 (2021) 093037 C Länger et al
Figure 3. Filled state STM images (Ut =−2 V; It = 0.5 nA) of a hydrogen-terminated Si(001) surface before and after
irradiation with 129Xe45+ ions (Epot = 58.8 keV; Ekin = 225 keV). (a) Before irradiation, the H/Si(001) surface shows very few
small dark spots. After the irradiation, a slight increase of small dark and bright features is observed in areas outside (b) and
inside (c) the ion beam spot on the sample. On the irradiated area of the sample (c), well-defined holes of about 2.3± 0.3 nm in
diameter are observed. (d) Height profile through one impact area as indicated by the red line in (c).
show up in the STM images with higher brightness since the clean silicon dimers (without hydrogen
adsorbed) appear brighter due to a higher density of states probed by the STM at the given tunneling bias.
In the following, we investigate how the results obtained with SHI compare with HCI irradiation of the
H/Si(001) surface. Figure 3 shows filled state STM images (Ut = −2 V; It = 0.5 nA) of a hydrogen-
terminated Si(001) 2× 1 surface before and after irradiation with 129Xe45+ ions (Epot = 58.8 keV;
Ekin = 225 keV). Before irradiation, the H/Si(001) surface exhibits a few small dark spots, which correspond
to defects originating from sample preparation (figure 3(a)). Some bright features as present in figure 2(a)
were also observed on larger scale images. After irradiation with 129Xe45+ ions, well-defined pits of in
average 2.3 ± 0.3 nm in diameter appear in the irradiated area of the sample (figure 3(c) and figure S3 in
the supporting information). Within the error bars of the applied fluence, the number density of the pits
(0.5 × 1011 cm−2) agrees well with the number of ion impacts per surface area. The red line in figure 3(c)
indicates a line scan through the impact area, the corresponding height profile is shown in figure 3(d). The
apparent depth of this pit is ≈ 0.5 nm, i.e. several atomic layers in depth. Due to the limitations of
measuring depth profiles with high aspect ratio by means of STM, this value can be seen as a lower limit.
After the irradiation, a slight increase of small dark and bright features is also observed; they are assigned to
the UHV transfer and sample handling procedure before and after the irradiation (compare figure 3(b)).
Irradiation of H/Si(001) with 129Xe40+ ions (Epot = 38.5 keV) also leads to clear signatures of
ion-induced damage on the surface (figures S3 and S4). However, in contrast to the effect of the 129Xe45+
ions, the observed pits are less well defined in shape and less deep (≈ 0.2 nm). Some accumulation of
disordered material is observed at the rim of the pits and the pits are smaller than after irradiation with
Xe45+ ions (figure S5). Within the error bars, the number density of observed pits (2 × 1011 cm−2) also
agrees well with the applied fluence.
4. Discussion
The presented results for SHI experiments on crystalline silicon substrates differ qualitatively from the
results of SHI on various insulator materials [1, 32] or compound semiconductors [13, 33]. In both
material classes, SHI irradiation leads to the formation of permanent nanostructures. Except for the
STM experiments by Osmani et al on the impact of xenon ions on Si(111) (Ekin = 0.12 GeV;
dE/dx = 12 keV nm−1) [14], only very few experiments were reported for crystalline silicon so far. No
5
New J. Phys. 23 (2021) 093037 C Länger et al
structural change induced by the impact of SHI was resolved in this former study, even when irradiating
under glancing angle of incidence (< 2◦ with respect to the surface) resulting in a higher energy deposition
close to the surface.
From the observation of the intact hydrogen-covered Si(001) surface after irradiation with SHI, track
formation followed by complete reconstruction/recrystallization is largely ruled out; from our STM analysis
we can exclude surface modifications larger than a few surface atoms in size. The stopping power of the
uranium ions applied in this study was 27 keV nm−1 (figure S6), the results thus suggest a threshold above
27 keV nm−1, higher than the value predicted so far from the thermal spike model [14].
Furthermore, the results obtained with SHI on hydrogen-covered silicon clearly differ from the results of
our experiments with HCI on H/Si(001). The latter show permanent crater-like damage structures of
1–3 nm in diameter; the observed pit size is larger for ions with higher energy. These results are comparable
to the results of experiments with highly charged iodine ions (q between 30 and 50) on Si(111) 7× 7
[17–19] which produced surface damage with radii between 1 and 3 nm, as well.
At this point, we would like to note that with Ekin = 225 keV, the Xe ions employed in our experiment
are close to the maximum of the contribution of nuclear stopping to the total stopping power (figure S7).
However, the absolute value is low (2 keV nm−1 in equilibrium charge state) when compared to the total
potential energy available from the 129Xe40+ and 129Xe45+ ions. This goes along with the clear increase of the
induced damage with the potential energy of the impacting ions when keeping the kinetic energy constant,
which also points toward a major influence of Epot for the formation of the observed damage.
So how can we explain the apparent difference between the impact of the SHI and HCI on crystalline
silicon? In earlier experiments with 30 and 40 MeV C60 ions of ≈ 50 keV nm−1 electronic stopping power,
track formation in Si was evidenced. From extrapolation of these data, a threshold of ≈ 30 keV nm−1 was
predicted [15, 16]. With 27 keV nm−1, we are just below this predicted threshold. In principle, the surface
should exhibit a somewhat lower threshold for mainly two reasons: (i) the broken bonds at the surface
reduce the effective binding energy (ii) it will reflect and thus increase the effective heat flux (although this
might be of minor importance for perpendicular incidence). Regarding C60 projectiles compared to our U
ions, one has to consider the following difference: the much lower velocity of 30 or 40 MeV C60 clusters
leads to a higher energy density than in the case of the 1.14 GeV U ions. At high velocities, the energy is
spread into a larger volume and thus the required critical stopping power is expected to increase [34].
Regarding the effective energy deposition of SHI and HCI, it has to be considered that the electrons
generated by SHI are typically high in energy and leave the impact area rather fast [22, 35], thus lowering
the effective energy density. Furthermore, the strength of electron–phonon coupling depends on the energy
distribution of the excited electrons and thus on their thermalization. In contrast, the impact of an HCI is
associated with the creation of many electrons of rather low energies [36] confining the energy closer to the
impact point. It is these electrons which effectively transfer the energy to the lattice by creating phonons
[20, 35].
The energy effectively deposited is usually estimated to be between 50% and 80% [10, 37, 38]. Taking
additionally into account that the energy of the HCI is deposited within a depth of 1 nm (as a conservative
estimate [9]), an energy per track length of 30–46 keV nm−1 is derived for Xe45+ ions. For Xe40+ ions with
an energy per track length of about 20–30 keV nm−1, the HCI-induced damage features are substantially
smaller. With a less efficient energy deposition in the case of SHI, it seems reasonable to expect a stopping
power threshold  30 keV nm−1. Our data thus rule out the occurrence of a thermal spike in Si which is
high enough to induce track formation and surface damage structures with monatomic SHI. We attribute
this to a high stopping power threshold due to inefficient energy deposition in crystalline silicon [22].
5. Conclusion
In conclusion, we observed significant differences in the interaction of SHI and HCI with a crystalline
hydrogen-covered Si(001) surface. For SHI with a stopping power of 27 keV nm−1, the upper limit for
single-ion induced surface damage was determined to be in order of 1–2 silicon atoms in size. In contrast,
the impact of single HCI with a potential energy between 40 and 60 keV results in surface pits of 1–3 nm in
diameter. As we could exclude recrystallization of silicon after SHI impacts, the striking difference between
the effect of HCI and SHI was attributed to a high stopping power threshold operative in the case of SHI.
This high threshold could result from fast spatial energy redistribution in combination with low electron
phonon coupling for the electrons excited in crystalline silicon by swift heavy ion impact. The U ions we
employed in our experiments exhibit the highest stopping power which can be realized in silicon with
monoatomic SHI. Our results thus suggest that nanoscale materials’ processing based on the impact of
single monoatomic ions cannot be applied to crystalline silicon.
6
New J. Phys. 23 (2021) 093037 C Länger et al
Acknowledgment
We acknowledge financial support by HICforFAIR, HGSHire (CL), DFG through CRC 1242 (Project ID
278162697), and BMBF (Projects 05K19RG1 and 05K16PG2). The results presented here are based on a
UMAT experiment, which was performed at the M-branch of the UNILAC at the GSI Helmholtzzentrum
für Schwerionenforschung, Darmstadt (Germany) in the framework of FAIR Phase-0.
Data availability statement
The data that support the findings of this study are available upon reasonable request from the authors.
ORCID iDs
C Trautmann https://orcid.org/0000-0001-7058-6340
M Schleberger https://orcid.org/0000-0002-5785-1186
M Dürr https://orcid.org/0000-0002-4676-8715
References
[1] Aumayr F, Facsko S, El-Said A S, Trautmann C and Schleberger M 2011 Single ion induced surface nanostructures: a comparison
between slow highly charged and swift heavy ions J. Phys.: Condens. Matter. 23 393001
[2] Li Z and Chen F 2017 Ion beam modification of two-dimensional materials: characterization, properties, and applications Appl.
Phys. Rev. 4 011103
[3] Schleberger M and Kotakoski J 2018 2D material science: defect engineering by particle irradiationMaterials 11 1885
[4] Toulemonde M, Dufour C and Paumier E 1992 Transient thermal process after a high-energy heavy-ion irradiation of amorphous
metals and semiconductors Phys. Rev. B 46 14362
[5] Toulemonde M, Paumier E and Dufour C 1993 Thermal spike model in the electronic stopping power regime Radiat. Eff. Defects
Solids 126 201
[6] Toulemonde M, Dufour C, Meftah A and Paumier E 2000 Transient thermal processes in heavy ion irradiation of crystalline
inorganic insulatorsNucl. Instrum. Methods Phys. Res. B 166–167 903
[7] Daraszewicz S L and Duffy D M 2011 Extending the inelastic thermal spike model for semiconductors and insulatorsNucl.
Instrum. Methods Phys. Res. B 269 1646
[8] Lang M, Devanathan R, Toulemonde M and Trautmann C 2015 Advances in understanding of swift heavy-ion tracks in complex
ceramics Curr. Opin. Solid State Mater. Sci. 19 39
[9] El-Said A S et al 2008 Creation of nanohillocks on CaF2 surfaces by single slow highly charged ions Phys. Rev. Lett. 100 237601
[10] Karlusic M and Jaksic M 2012 Thermal spike analysis of highly charged ion tracksNucl. Instrum. Methods Phys. Res. B 280 103
[11] Dufour C, Khomrenkov V, Wang Y Y, Wang Z G, Aumayr F and Toulemonde M 2017 An attempt to apply the inelastic thermal
spike model to surface modifications of CaF2 induced by highly charged ions: comparison to swift heavy ions effects and
extension to some others material J. Phys.: Condens. Matter. 29 095001
[12] Kozubek R, Ernst P, Herbig C, Michely T and Schleberger M 2018 Fabrication of defective single layers of hexagonal boron nitride
on various supports for potential applications in catalysis and DNA sequencing ACS Appl. Nano Mater. 1 3765
[13] Kamarou A, Wesch W, Wendler E, Undisz A and Rettenmayr M 2008 Radiation damage formation in InP, InSb, GaAs, GaP, Ge,
and Si due to fast ions Phys. Rev. B 78 054111
[14] Osmani O et al 2012 Damage in crystalline silicon by swift heavy ion irradiation Nucl. Instrum. Methods Phys. Res. B 282 43
[15] Canut B, Bonardi N, Ramos S MM and Della-Negra S 1998 Latent tracks formation in silicon single crystals irradiated with
fullerenes in the electronic regime Nucl. Instrum. Methods Phys. Res. B 146 296
[16] Dunlop A, Jaskierowicz G and Della-Negra S 1998 Latent track formation in silicon irradiated by 30 MeV fullerenes Nucl.
Instrum. Methods Phys. Res. B 146 302
[17] Tona M et al 2007 Nano-crater formation on a Si(111)-(7× 7) surface by slow highly charged ion-impact Surf. Sci. 601 723
[18] Tona M et al 2007 Nanofabrication on a Si surface by slow highly charged ion impact Nucl. Instrum. Methods Phys. Res. B 256 543
[19] Tona M, Watanabe H, Takahashi S, Nakamura N, Yoshiyasu N, Sakurai M, Yamada C and Ohtani S 2007 Potential sputtering from
a Si surface by very highly charged ion impact Nucl. Instrum. Methods Phys. Res. B 258 163
[20] Murat M, Akkerman A and Barak J 2011 Can swift heavy ions create latent tracks in silicon? Nucl. Instrum. Methods Phys. Res. B
269 2649
[21] Khara G S, Murphy S T, Daraszewicz S L and Duffy D M 2016 The influence of the electronic specific heat on swift heavy ion
irradiation simulations of silicon J. Phys.: Condens. Matter. 28 395201
[22] Medvedev N A, Rymzhanov R A and Volkov A E 2015 Time-resolved electron kinetics in swift heavy ion irradiated solids J. Phys.
D: Appl. Phys. 48 355303
[23] Rymzhanov R A, Medvedev N, O’Connell J H, Janse van Vuuren A, Skuratov V A and Volkov A E 2019 Recrystallization as the
governing mechanism of ion track formation Sci. Rep. 9 3837
[24] Dürr M and Höfer U 2006 Dissociative adsorption of molecular hydrogen on silicon surfaces Surf. Sci. Rep. 61 465
[25] Dürr M and Höfer U 2013 Hydrogen diffusion on silicon surfaces Prog. Surf. Sci. 88 61
[26] Khalfaoui N, Rotaru C C, Bouffard S, Jacquet E, Lebius H and Toulemonde M 2003 Study of swift heavy ion tracks on crystalline
quartz surfacesNucl. Instrum. Methods Phys. Res. B 209 165
[27] Schwalb C H, Lawrenz M, Dürr M and Höfer U 2007 Real-space investigation of fast diffusion of hydrogen on Si(001) by a
combination of nanosecond laser heating and STM Phys. Rev. B 75 085439
7
New J. Phys. 23 (2021) 093037 C Länger et al
[28] Mette G, Adamkiewicz A, Reutzel M, Koert U, Dürr M and Höfer U 2019 Controlling an SN2 reaction by electronic and
vibrational excitation: tip-induced ether cleavage on Si(001) Angew. Chem., Int. Ed. 58 3417
[29] Skopinski L et al 2021 Time-of-flight mass spectrometry of particle emission during irradiation with slow, highly charged ions
Rev. Sci. Instrum. 92 023909
[30] Dürr M, Biedermann A, Hu Z, Höfer U and Heinz T F 2002 Probing high-barrier pathways of surface reactions by scanning
tunneling microscopy Science 296 1838
[31] Dürr M and Höfer U 2004 Molecular beam investigation of hydrogen dissociation on Si(001) and Si(111) surfaces J. Chem. Phys.
121 8058
[32] Khalfaoui N, Rotaru C C, Bouffard S, Toulemonde M, Stoquert J P, Haas F, Trautmann C, Jensen J and Dunlop A 2005
Characterization of swift heavy ion tracks in CaF2 by scanning force and transmission electron microscopy Nucl. Instrum.
Methods Phys. Res. B 240 819
[33] Schnohr C S, Kluth P, Giulian R, Llewellyn D J, Byrne A P, Cookson D J and Ridgway M C 2010 Swift-heavy-ion-induced damage
formation in III–V binary and ternary semiconductors Phys. Rev. B 81 075201
[34] Meftah A, Costantini J M, Khalfaoui N, Boudjadar S, Stoquert J P, Studer F and Toulemonde M 2005 Experimental
determination of track cross-section in Gd3Ga5O12 and comparison to the inelastic thermal spike model applied to several
materialsNucl. Instrum. Methods Phys. Res. B 237 563
[35] Osmani O, Medvedev N, Schleberger M and Rethfeld B 2011 Energy dissipation in dielectrics after swift heavy-ion impact: a
hybrid model Phys. Rev. B 84 214105
[36] Wilhelm R A et al 2017 Interatomic coulombic decay: the mechanism for rapid deexcitation of hollow atoms Phys. Rev. Lett. 119
103401
[37] Karlusic M, Akcöltekin S, Osmani O, Monnet I, Lebius H, Jaksic M and Schleberger M 2010 Energy threshold for the creation of
nanodots on SrTiO3 by swift heavy ions New J. Phys. 12 043009
[38] Karlušíc M et al 2015 Response of GaN to energetic ion irradiation: conditions for ion track formation J. Phys. D: Appl. Phys. 48
325304
8