Electronically induced modification of thin layers on surfaces
Interactions of thermally and electronically stimulated reactions in thin layers on surfaces are investigated. For self-assembled monolayers, thermal activation promotes many processes primarily induced by electronic excitations. We demonstrate that the film temperature is an important parameter f...
Збережено в:
| Дата: | 2007 |
|---|---|
| Автори: | , , , , , |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України
2007
|
| Назва видання: | Физика низких температур |
| Теми: | |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/121788 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Electronically induced modification of thin layers on surfaces / U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, M. Zharnikov // Физика низких температур. — 2007. — Т. 33, № 6-7. — С. 680-688. — Бібліогр.: 28 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-121788 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1217882017-06-17T03:02:58Z Electronically induced modification of thin layers on surfaces Bauer, U. Neppl, S. Menze, D. Feulner, P. Shaporenko, A. Zharnikov, M. Electronic Processes in Cryocrystals Interactions of thermally and electronically stimulated reactions in thin layers on surfaces are investigated. For self-assembled monolayers, thermal activation promotes many processes primarily induced by electronic excitations. We demonstrate that the film temperature is an important parameter for steering these reactions towards different final products. Using chemisorbed water on Ru(001) as an example, we investigate how the products of an irradiation induced reaction catalyze thermally stimulated dissociation of water molecules. 2007 Article Electronically induced modification of thin layers on surfaces / U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, M. Zharnikov // Физика низких температур. — 2007. — Т. 33, № 6-7. — С. 680-688. — Бібліогр.: 28 назв. — англ. 0132-6414 PACS: 61.82.–d; 79.20.La; 78.70.–g; 68.43.Vx http://dspace.nbuv.gov.ua/handle/123456789/121788 en Физика низких температур Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Electronic Processes in Cryocrystals Electronic Processes in Cryocrystals |
| spellingShingle |
Electronic Processes in Cryocrystals Electronic Processes in Cryocrystals Bauer, U. Neppl, S. Menze, D. Feulner, P. Shaporenko, A. Zharnikov, M. Electronically induced modification of thin layers on surfaces Физика низких температур |
| description |
Interactions of thermally and electronically stimulated reactions in thin layers on surfaces are investigated.
For self-assembled monolayers, thermal activation promotes many processes primarily induced by
electronic excitations. We demonstrate that the film temperature is an important parameter for steering these
reactions towards different final products. Using chemisorbed water on Ru(001) as an example, we investigate
how the products of an irradiation induced reaction catalyze thermally stimulated dissociation of water
molecules. |
| format |
Article |
| author |
Bauer, U. Neppl, S. Menze, D. Feulner, P. Shaporenko, A. Zharnikov, M. |
| author_facet |
Bauer, U. Neppl, S. Menze, D. Feulner, P. Shaporenko, A. Zharnikov, M. |
| author_sort |
Bauer, U. |
| title |
Electronically induced modification of thin layers on surfaces |
| title_short |
Electronically induced modification of thin layers on surfaces |
| title_full |
Electronically induced modification of thin layers on surfaces |
| title_fullStr |
Electronically induced modification of thin layers on surfaces |
| title_full_unstemmed |
Electronically induced modification of thin layers on surfaces |
| title_sort |
electronically induced modification of thin layers on surfaces |
| publisher |
Фізико-технічний інститут низьких температур ім. Б.І. Вєркіна НАН України |
| publishDate |
2007 |
| topic_facet |
Electronic Processes in Cryocrystals |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/121788 |
| citation_txt |
Electronically induced modification of thin layers
on surfaces / U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, M. Zharnikov // Физика низких температур. — 2007. — Т. 33, № 6-7. — С. 680-688. — Бібліогр.: 28 назв. — англ. |
| series |
Физика низких температур |
| work_keys_str_mv |
AT baueru electronicallyinducedmodificationofthinlayersonsurfaces AT neppls electronicallyinducedmodificationofthinlayersonsurfaces AT menzed electronicallyinducedmodificationofthinlayersonsurfaces AT feulnerp electronicallyinducedmodificationofthinlayersonsurfaces AT shaporenkoa electronicallyinducedmodificationofthinlayersonsurfaces AT zharnikovm electronicallyinducedmodificationofthinlayersonsurfaces |
| first_indexed |
2025-07-08T20:31:27Z |
| last_indexed |
2025-07-08T20:31:27Z |
| _version_ |
1837112165974147072 |
| fulltext |
Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7, p. 680–688
Electronically induced modification of thin layers
on surfaces
U. Bauer, S. Neppl, D. Menzel1, and P. Feulner
Physikdepartment E20, Technische Universität München, Germany
E-mail: feulner@tum.de
A. Shaporenko and M. Zharnikov
Angewandte Physikalische Chemie, Universität Heidelberg, Germany
Received November 7, 2006
Interactions of thermally and electronically stimulated reactions in thin layers on surfaces are investi-
gated. For self-assembled monolayers, thermal activation promotes many processes primarily induced by
electronic excitations. We demonstrate that the film temperature is an important parameter for steering these
reactions towards different final products. Using chemisorbed water on Ru(001) as an example, we investi-
gate how the products of an irradiation induced reaction catalyze thermally stimulated dissociation of water
molecules.
PACS: 61.82.–d Radiation effects on specific materials;
79.20.La Photon- and electron-stimulated desorption;
78.70.–g Interactions of particles and radiation with matter;
68.43.Vx Thermal desorption.
Keywords: thin layers, electronic excitations.
1. Introduction
Thermal excitations drive all kinds of reactions from
diffusion to bond formation and bond breaking. The spec-
trum of excitation energies available to surmount activa-
tion barriers is a continuum described by the Boltzmann
distribution. The energy range relevant for experimen-
tally accessible reaction rates at a distinct temperature T is
limited to less than ~ 12 kT. The attempt frequencies of
thermally stimulated reactions given by the ratios of parti-
tioning functions of the initial and the transition states are
large, typically 1013 s–1 and beyond; the initial and the
transition states are commonly assumed to be in thermal
equilibrium [1].
For reactions induced by electronic excitations, either
by energetic particles or photons, the situation is com-
pletely different. Particularly photons from modern laser
and synchrotron sources cover the entire valence and core
electron excitation range. Specific electronically excited
states at distinct sites in molecules and condensed matter
can be selectively prepared by narrow bandwidth excita-
tions, either such which are highly dissociative and break
bonds, or such which turn an inert particle into a reactive
one in order to make bonds. Apart from very few excep-
tions [2], electronically induced processes proceed far
from equilibrium.
In practice, however, both scenarios may well be re-
lated. A good example is cryomicroscopy [3]. Energetic
electrons in scanning or transmission electron micro-
scopes, or photons in x-ray microscopes break bonds, but
cryogenic conditions prevent structural changes mainly
by hindering diffusion (see Ref. 3 and below). On the
other hand, products of an electronically stimulated reac-
tion may act as catalyst for further thermally activated
process. In this contribution we report such interrelations
of both reaction types for thin layers on surfaces. In the
first section we focus on thermally stimulated processes
promoting irradiation induced modifications of thin orga-
nic layers on metal and semiconductor surfaces. Using
© U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, and M. Zharnikov, 2007
1
Also: Fritz-Haber-Institut der Max-Planck-Gesellschaft, Dept. CP, Berlin, Germany.
chemisorbed water on the Ru(001) surface as an example,
we show in the second section how the products of an
electronically stimulated process may catalyze a ther-
mally induced conversion, thereby dramatically amplify-
ing the effective conversion cross-section.
2. Examples and results
2.1. Thermal effects in beam induced modifications
of self-assembled monolayers
Self-assembled monolayers (SAMs, Fig. 1) are well
ordered arrangements of molecules consisting of
headgroups with a specific affinity to the substrate, an
aliphatic or aromatic backbone and an endgroup consti-
tuting the outer surface of the film (see Ref. 4 for further
details).
The enormous potential of SAMs in many fields of
technology is due to the flexibility in combining different
structure elements (headgroup/backbone/endgroup) in
order to suit the requirements of even very special appli-
cations (see [5] for an overview). SAMs serve as transfer
agents in micro and nano printing techniques [5], as func-
tional interfaces in bio sensors [5,6], as lubrication in
micromechanics [5], corrosion protection [5], as ultrathin
resists for lithography with ultimate resolution [5,7], but
also as well defined model systems for organic layers on
surfaces in fundamental research. Here we focus on the
last point, using them for investigations of basic tempera-
ture effects on beam induced modifications. Many studies
of irradiation effects in SAMs by electrons and to a lesser
extent also by photons are reported in the literature, from
which we cite three articles in which the reader can find
most of the relevant references [8–10]. The majority of
these previous reports are room temperature studies, thus
neglecting any low-temperature effects. In order to close
this gap, we have performed measurement between 50
and 300 K with synchrotron radiation (BESSY-II,
U49-II-PGM-1) in the C1s and N1s excitation range for
SAMs with aliphatic and aromatic backbones. Here we
summarize the main results of these studies. For details of
sample preparation, data acquisition or data evaluation
the reader is referred to [11,12].
For the above temperature and energy range we find
some electronically induced reactions which are nearly
temperature independent, and others which strongly de-
pend on the sample temperature. Nearly temperature in-
dependent radiation effects are those which do not in-
volve transport of large fragments; abstraction of H atoms
is the best example (see [10] for a description of the
mechanism of electron stimulated H abstraction; although
primary photons are applied in our study, secondary elec-
trons with energies similar to those of [10] will stimulate
the majority of abstraction processes even in our case, see
below). Abstraction of H atoms leaves bond vacancies
which recombine either by double bond formation (if at
the same molecule) or by cross-linking (if at neighboring
particles) [8]. The consequence of cross-linking will be-
come obvious from results below, whereas double bond
formation directly shows up in x-ray absorption spectro-
scopy (XAS).
Figure 2 shows XAS data for the C1s range from
Au–S–(CH2)15–CH3 SAMs. Apart from resonances at
288, 293 and 300 eV which are due to Rydberg, (C–H)-�*
and (C–C)-�* excitations, respectively [8], an additional
peak appears and grows at 285 eV for extended photon ir-
radiation. This maximum has previously been assigned to
double bond-derived [C1s]�* states formed as a result of
radiation induced hydrogen abstraction [8,11]. Compar-
ing data at 50 and 300 K, no significant temperature de-
pendence is observed for the growth of this maximum as a
function of photon exposure (Fig. 2 [11]). At 50 K it
grows even slightly faster than at 300 K, due to rapid ma-
terial loss at high temperatures, see below.
Compared with double bond formation after abstrac-
tion of hydrogen atoms, a completely different tempera-
ture dependence is obtained for the desorption of large
Electronically induced modification of thin layers on surfaces
Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7 681
Endgroup
Substrate
Headgroup
Backbone
Fig. 1. Structure elements of self-assembled monolayers (see
also [4,5]).
290 300 310290 300 310
300 K
PEY
1.8
3.6
5.3
7.0
8.7
10
50 K
12
10
7.7
5.2
2.7
PEY
Exposure
[10 /cm ]
16 2
h�
Exposure
[10 /cm ]
16 2
h�
h , eV� h , eV�
Fig. 2. C1s XAS for Au–S–(CH2)15–CH3 SAMs, at 50 and
300 K (for Az polarization, i.e., E-vector perpendicular to the
surface normal, after [11]).
hydrocarbon fragments by irradiation with soft x-ray pho-
tons. Figure 3 shows photon stimulated desorption (PSD)
of neutral particles recorded with a mass spectrometer
[13] from mercaptoheptadecanenitrile SAMs on Au.
At 300 K, PSD of large fragments is much more in-
tense than at 50 K. We observe CnHx fragments up to n = 9
which is the upper limit of our mass spectrometer. At 50 K
only fragments up to n = 5 are observable, all but those for
n = 1 with much smaller amplitudes than at 300 K. De-
sorption of larger fragments requires their diffusive trans-
port to the vacuum interface of the film which is hindered
at cryogenic conditions by the lack of thermal stimula-
tion. We believe that the suppression of electronically
stimulated desorption of large fragments from organic
films at low temperatures is a phenomenon of general va-
lidity. It has been observed not only for thiolate bonded
SAMs on Au and Ag, but also for alkyls on silicon [11]
and diamond [14], and for phosphonate bonded SAMs on
silicon oxide [15].
We emphasize that the data of Fig. 3 have been ob-
tained under �-resonant excitation of the CN endgroup,
i.e., the primary excitation energy has been allocated to a
specific bond of the molecule. Although this primary
[N1s]�* state is not dissociative, population of highly re-
pulsive final states is expected upon core decay. The
products of these dissociation processes of the CN end-
group, namely C and N atoms, appear in the mass spec-
trum (Fig. 3). These site-specific signals which are due to
direct photon induced bond breaking do not depend on
temperature as expected. However, they are smaller than
other desorption maxima even at 50 K. This indicates that
the major part of the desorption signal from such layers is
due to unspecific, not site selective excitations by second-
ary electrons, even for site selective primary excitations
of states with large cross section such as the [N1s]�* re-
sonance [16].
Irradiation induced material loss can also be moni-
tored by x-ray photoelectron spectroscopy (XPS), e.g., by
recording the film induced attenuation of the XPS signal
from the substrate as a function of irradiation dose and
film temperature. From such data cross-section values and
information on the saturation behavior of beam damage
can easily be derived. Figure 4 shows initial desorption
cross-section values for thiolate bonded aliphatic
(CH3–(CH2)11–S–Au: C12/Au) and aromatic SAMs
(C6H5–C6H4–S–Au: BPT/Au).
Integral PSD cross-sections (we continue using the
term «PSD», although «ESD» contributes substantially,
as shown above) are about 2·10–17 cm2 at 50 K, for
C12/Au as well as for BPT/Au. Between 50 and 300 K
they increase by a factor of 5 for C12/Au and 3 for
BPT/Au. Both layers show saturation behavior, i.e., a dra-
matic reduction of beam effects for large exposures. This
has previously been explained as due to cross-linking as a
consequence of hydrogen abstraction. The cross linking
network stabilizes the film against further radiation attack
by hindering material transport. On the other hand, it en-
ables horizontal delocalization of electronic excitations,
thus reducing the probability of excitation localization at
682 Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7
U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, and M. Zharnikov
20 40 60 80 100 120
NC
[N1s]�* excitation of CN
Au-S-(CH )2 16-CN
50 K
300 K
Mass, amu
P
S
D
si
g
n
al
Fig. 3. PSD of neutrals by resonant [N1s]�* excitation from
Au–S–(CH2)16–CN at 50 K and at 300 K (Az-light).
50
50
100
100
150
150
200
200
250
250
300
300
0
2
4
6
8
10
Au4f
C12/Au
T, K
T, K
2
3
4
5
6
Au4f
BPT/Au
C
ro
ss
-s
ec
ti
o
n
,
1
0
cm
–
1
7
C
ro
ss
-s
ec
ti
o
n
,
1
0
cm
–
1
7
Fig. 4. Initial desorption cross-sections from (top) C12/Au
(h� = 250 eV) and (bottom) BPT/Au (h� = 310 eV).
an individual bond. This localization, however, is a pre-
requisite for the coupling of the electronic excitation to
nuclear motion. At room temperature, this cross-linking
effect is much more efficient for aromatic SAMs than for
aliphatic layers, see Fig. 5, in perfect agreement with pre-
vious experiments [8].
We obviously encounter a competition between cross-
linking and material loss. Cross-linking prevents further
material loss, but rapid material loss makes the film so po-
rous that cross-linking is hindered. From the above re-
sults we expect low temperatures to favor cross-linking
compared with material loss, because cross-linking re-
sults from temperature independent hydrogen abstraction
(Fig. 2), whereas material loss is enhanced by thermally
stimulated diffusive transport (Figs. 3 and 4). The experi-
mental results depicted in Fig. 5 agree with these expec-
tations. The lower the temperature, the more efficient
the cross-linking induced limitation of material loss. For
aliphatic films with their large desorption cross-sections
at room temperature (cf. Fig. 3) this effect is most pro-
nounced and is of great practical importance. At room
temperature stabilization by cross-linking is efficient only
for aromatic SAMs, because for aliphatic SAMs material
loss is too fast at that temperature. Under cryogenic con-
ditions, however, aliphatic SAMs can be stabilized by ra-
diation as well. Using temperature as a process parameter,
compact and interconnected, or porous aliphatic films can
easily be prepared by radiation. We believe this result to
be of great importance for many applications, particularly
for resist technology. Keeping in mind that cross-linking
stabilizes not only against radiative but also against che-
mical attack, we foresee the process temperature as a tool
to tailor negative or positive resist behavior from identi-
cal films.
Using thiolate bonded SAMs as an example, we finally
focus on radiation induced modifications of the head-
group. In previous studies it has been shown that irradia-
tion by electrons or photons breaks the thiolate bond,
leading either to dialkylsulfide species if sulfur termi-
nated hydrocarbon fragments diffuse from the substrate
and get trapped by dangling bonds formed by hydrogen
abstraction, or to accumulation of atomic sulfur at the
substrate. Both species can be well discriminated in XPS.
Film temperature modifies this branching (Fig. 6: C12/Au;
results for BPT/Au are qualitatively similar).
The irradiation induced loss of thiolate species is
clearly enhanced at 300 K; it is paralleled by the gain of
the dialkylsulfide species. The build-up of atomic sulfur,
a process in competition with the dialkylsulfide formation
but of much smaller cross-section, is enhanced at low
temperature, when the diffusive transport of larger frag-
ments from the metal interface — a necessity for the dial-
kylsulfide formation — is hindered. We emphasize that
negligible changes of the composition of the layer are ob-
tained for heating after irradiation. This clearly indicates
that cooling not simply freezes the structure, but hinders
permanent bond breaking by preventing diffusive trans-
port. This is a result of great importance in particular for
cryospectroscopy: Cryogenic conditions may allow spec-
troscopic investigations of films which under room tem-
perature suffer too rapid degradation. On the other hand,
we can summarize that temperature is a parameter which
steers electronically stimulated modification of all parts
Electronically induced modification of thin layers on surfaces
Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7 683
50
50
100
100
150
150
200
200
250
250
300
300
5
10
15
20
25
30
35
40 C12/Au
R
el
at
iv
e
ch
an
g
e
o
f
th
ic
k
n
es
s,
%
R
el
at
iv
e
ch
an
g
e
o
f
th
ic
k
n
es
s,
%
T, K
T, K
4
5
6
7
8
9
10
BPT/Au
�d/dpristine
�d/dpristine
Fig. 5. Saturation behavior of beam damage as a function of
temperature: Relative change of layer thickness after extended
irradiation for C12/Au (top) and BPT/Au (bottom).
50 100 150 200 250 300
0
10
20
30
40
50
60
70
80 C12/Au
thiolate
dialkylsulfide
atomic S
R
el
at
iv
e
S
2
p
in
te
n
si
ty
,
%
T, K
Fig. 6. Relative amounts of sulfur containing species after ex-
posure to 5·10
16
h��cm
2
.
of thin organic films on surfaces, a result of the connec-
tion of thermal and electronic stimulation.
2.2. Beam induced conversion of water layers on
Ru(001)
Products of beam induced modifications influencing
thermally stimulated reactions is our next topic. Our ex-
ample is chemisorbed water on the close-packed Ru(001)
surface. Chemisorption of water is of interest for many
fields ranging from electrochemistry over fuel cell tech-
nology to biological and medical applications. The che-
misorption system H2O/D2O on Ru(001) is of special fun-
damental interest because it shows an uncommonly large
isotope effect in thermal desorption (TD, Fig. 7).
Apart from multilayer contributions, TD of chemi-
sorbed D2O shows a single maximum for large as well as
small heating rates. For H2O, however, the TD spectrum
is bimodal. The branching between low- and high-tem-
perature TD maxima depends on the heating rate; the cen-
ter of gravity is shifted to the second maximum for slow
heating. The current interpretation of this unusual behav-
ior based on DFT calculations, vibrational spectroscopy
and low-energy electron diffraction (LEED) [19–22] is
that the molecularly chemisorbed water layer correspond-
ing to the low-temperature (= 1st) TD peak is a metastable
state. It can transform into a more strongly bound par-
tially dissociated layer corresponding to the high-tempe-
rature TD maximum which consists of H2O + OH + H
(D2O + OD + D, see, e.g., Fig. 3 of [19]). This transforma-
tion is an activated process; the barrier between the mo-
lecularly chemisorbed and the partially dissociated state
is for D2O higher than the barrier for desorption; for H2O
it is lowered because of the larger zero point energy of the
H–to–O vibration. As a result, two TD maxima are ob-
served for H2O, but only one for D2O.
This subtle balance between two reaction pathways is
efficiently disturbed by small amounts of co-adsorbates,
e.g., oxygen or hydrogen atoms. Clay et al. [18] have
shown that co-adsorbed oxygen in the percent range low-
ers the barrier between the molecular and partially disso-
ciated state with respect to desorption, i.e., increases for
H2O the second TD maximum at the expense of the first
one and induces for D2O the second TD peak which does
not exist for the pure layer. Only 0.09 monolayers (ML)
of preadsorbed oxygen suffice for complete conversion in
TD. Co-adsorbed hydrogen has the inverse effect. Clay’s
data are in qualitative agreement with previous results
from Doering and Madey [23].
The assumptions about the geometry of the molecu-
larly chemisorbed layer have experienced some evolu-
tion. The first investigations assumed a layer consisting
of two types of differently oriented water molecules ar-
ranged in hexagons (Ref. 23, and references therein).
Type I had its hydrogen atoms directed towards the O
atoms of neighboring water molecules, whereas the hyd-
rogen plane of type II was tilted by 90 degrees; one of
the O–H bonds was directed to neighboring water, and
the other one towards the vacuum («H-up», see, e.g.,
Figs. 13–15 in [23]). Essentially, this so-called bilayer
was obtained by cutting one layer out of hexagonal bulk
ice and putting it on a surface. In the meantime this pic-
ture has been modified, including also «H-down» mole-
cules with one hydrogen directed towards the substrate,
which probably is slightly more stable than H-up [19–22].
Very recently it has been proposed that arranging both
above types (for type II exclusively H-down) in chains in-
stead of hexagons could increase the binding energy even
further [22].
The first studies of water on Ru(001) assumed exclu-
sively molecular adsorption. The idea of partial dissocia-
tion was introduced by Feibelman in 2002 [24]. Based on
DFT calculations he concluded that the chemisorption en-
ergy of a molecular layer would be lower than the conden-
sation energy of bulk ice, i.e., the molecular layer would
not wet the surface. As an alternative he suggested the
existence of a more strongly bound partially dissociated
layer consisting of H2O + OH + H (D2O + OD + D).
In [24] a 1:1 ratio was proposed for the H2O and
H2O + OH + H (D2O, D2O + OD + D) species, respec-
tively. In a later publication this ratio had been changed
to 5:3 in order to fit XPS results [25]. The low chemi-
sorption energy of the molecular layer and the resulting
«nonwetting» argument of [24,25] have been questioned
on the basis of more refined DFT calculations and experi-
ments (see, e.g., [22] and references therein). There is
agreement now that for both H2O and D2O the pure layers
before the start of desorption are molecular, i.e., water
does wet the Ru surface. Agreement also exists, however,
684 Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7
U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, and M. Zharnikov
5 K/s
D2O/Ru(001)
120 160 200 240
T, K
0.1 K/s
multilayer
multilayer
H2O/Ru(001)
5 K/s
120 160 200 240
T, K
0.1 K/sM
as
s
1
8
T
D
,
ar
b
.
u
n
it
s
M
as
s
2
0
T
D
,
ar
b
.
u
n
it
s
Fig. 7. TD spectra of D2O (left) and H2O (right) from chemi-
sorbed layers on the Ru(001) surface as a function of the heat-
ing rate (after [17], in perfect agreement with [19]).
that a partially dissociated water layer should be more sta-
ble than the molecular one (see above).
Feibelman’s paper [24] prompted several theoretical
as well as experimental studies on this systems. Two of
them, both utilizing XPS with synchrotron radiation,
came to fully opposite results [25,26]. Reference 25 re-
ported XPS data indicating partial dissociation of H2O
and D2O at 145 K from the beginning, whereas Ref. 26
found no dissociation of D2O, minor dissociation for
H2O, and strong dissociation only for extended exposure
to the synchrotron beam or for adsorption at elevated
sample temperature, in agreement with the above TD re-
sults. In a study partly conducted in our lab we have
shown that very low doses of secondary electrons can in
fact cause rapid dissociation [17]. Due to experimental re-
strictions, it was not possible to irradiate the water layer
homogeneously in this previous study, leading to some ar-
tifacts in the beam effect’s saturation behavior. For the in-
vestigations presented here, we obtained a strictly homo-
geneous irradiation profile by scanning a small spot of
200 eV electrons rapidly over the surface by applying tri-
angular voltages of different frequencies to the x and y de-
flection plates of the electron gun. Monitoring the sample
current on an oscilloscope synchronously with the deflec-
tion allowed exact adjustment of the scan area to the sam-
ple size. From the scan area and the time integrated gun
current we obtained electron exposures with an error of
less than 10%.
Figure 8 shows TD spectra from chemisorbed D2O
layers after irradiation with 200 eV electrons under such
conditions. These layers have been prepared by dosing
D2O beyond saturation of the chemisorbed layer, fol-
lowed by controlled annealing to remove multilayer con-
tributions before starting the electron bombardment (we
will use the term «bilayer» for species obtained by this
preparation procedure). After water desorption was
completed, the samples were heated to 1570 K in order to
remove all residual oxygen before preparing new
bilayers. As observed previously [17], electron bombard-
ment induces a second TD maximum at expense of the
first. As noted previously, this second TD maximum was
completely absent for the pristine D2O bilayer. An elec-
tron dose of ~ 6·1015 e/cm2 sufficed for complete conver-
sion (Fig. 9). This conversion is associated with a loss of
~30% of the total amount of D2O (see TD data in Fig. 9).
Fitting the decay of the first TD peak by a simple expo-
nential we obtain a cross-section value of 10–15cm–2 (dot-
ted line in Fig. 9). This decay is due to conversion into the
second TD maximum and to a lesser extent also to de-
sorption, see Fig. 9.
The main differences of the present investigation com-
pared with our previous study [17] is a larger conversion
cross-section and the finding of complete conversion.
The previously obtained saturation behavior of the con-
version at large exposure obviously was due to an
inhomogeneous irradiation profile.
The value for the conversion and desorption cross-sec-
tion of 10–15cm–2 is very large, even for dissociative elec-
tron attachment, a process which may exhibit large cross-
section values. Considering the strong effect of co-ad-
sorbed oxygen and hydrogen ([18] and above), we con-
sidered possible catalytic conversion activities of the re-
action products produced by irradiation. If such activities
were large, the apparent conversion cross-section would
be a multiple of the cross-section of the primary electroni-
cally stimulated process. We checked this by dosing D2O
after thermal desorption of a previously irradiated layer
without removing residues by flashing to 1570 K. Fi-
gure 10 shows the results. The TD results obtained after
irradiation (full lines), and the TD spectra recorded after
re-dosing of D2O are practically identical, supporting the
above hypothesis. We note that a qualitatively similar ef-
Electronically induced modification of thin layers on surfaces
Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7 685
160 180 200 220 240
6.3
6.3
3.1
3.1
1.3
1.30.6
0.6
0.3
0.3
0
0
Exposure
[10 e/cm ]
15 2
D
O
T
P
D
,
ar
b
.
u
n
it
s
2
T, K
D O TPD
5 K/s
2
1
Fig. 8. Electron induced modifications of TD spectra from a
saturated chemisorbed layer of D2O/Ru(001) as a function of
(homogeneously distributed) electron exposure.
Total amount of D O2
2nd TPD peak
1st TPD peak
R
el
at
iv
e
T
D
p
ea
k
ar
ea
s
Exposure, 10 e /cm
–15 – 2
1.0
0.8
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7
Fig. 9. TD peak areas as a function of electron exposure.
fect has been seen before already by Faradzhev [27] dur-
ing the experiments reported in [17]. As will be shown
below by XPS, irradiation induced perturbation of the hy-
drogen/oxygen balance catalyzes the conversion between
the two TD states.
We note that electron bombardment beyond the dose
necessary for complete conversion of the layer changes
the shape of the TD peaks further (Fig. 11). Apart from an
area reduction due to stimulated desorption of D2O or D
(both alternatives would reduce TD of intact water), we
find additional shoulders at low as well as at high tempe-
rature. Similar effects have been observed for co-adsorp-
tion of large amounts of oxygen (cf. [23] and [28]).
O1s XPS data from D2O bilayers after irradiation and
after TD with a heating rate of 5 K/s up to 260 K are
shown in Fig. 12. Care was taken to minimize beam dam-
age. A small, constant, and preparation independent O1s
contribution stemming from the crystal mount has been
subtracted from all spectra. For pristine D2O bilayers
a single O1s maximum at 533 eV binding energy was
observed in good agreement with [26] (the minimum ex-
posure of 6·1013 eV/cm2 in Fig. 12 corresponds to that by
Al K� photons during data acquisition, i.e., its possible
damage effect constitutes the minimum attainable). After
TD of a pristine bilayer corresponding to zero exposure,
no residual oxygen could be observed (Fig. 12). After
irradiation of a D2O bilayer with 6.3·1015 e/cm2, i.e.,
the dose which induces complete conversion, the O1s
XPS peak became bimodal (Fig. 12). The main peak was
shifted to a lower binding of 532.5 eV, and a second maxi-
mum appeared at 531 eV. The ratio of the two maxima was
~ 5:1.
686 Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7
U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, and M. Zharnikov
140 160 180 200 220 240
D
O
si
g
n
al
2
T, K
0.65·10 e /cm
15 – 2
readsorption
1.3·10 e /cm
15 – 2
readsorption
Fig. 10. TD spectra obtained after electron irradiation (full
lines) and after readsorption of D2O (broken lines). Readsorp-
tion was done by dosing an amount of D2O corresponding to
a full bilayer after the post-irradiation TD spectrum (full line)
was completed. Because no annealing was applied in order not
to disturb peak shapes, small multilayer contributions are vi-
sible at the leading edges of the readsorption spectra.
160 180 200 220 240 260
32
63
6.3
0
T, K
Exposure
[10 e /cm ]
15 – 2
D
O
T
P
D
,
ar
b
.
u
n
it
s
2
D O TPD
5 K/s
2
1
Fig. 11. TD spectra obtained after large homogeneous electron
exposures.
536 534 532 530 528
0
63
6.3
< 0.06
O
1
s
si
g
n
al
,
ar
b
.
u
n
it
s
E , eVB
Exposure
[10 e/cm ]
15 2
Fig. 12. XPS data for the O1s range obtained after electron ir-
radiation (a), and after electron irradiation and TD up to 260 K
(Al K� radiation(b)); smooth lines are fitting results.
We note that the growth of the second O1s peak as well
as the redshift of the first one occurred gradually with in-
creasing electron exposure, without any threshold beha-
vior. We also note that XPS spectra recorded from an ini-
tially completely converted layer after a TD run up to
200 K, i.e., the temperature corresponding to the mini-
mum between the two TD peaks (Fig. 8), showed only
marginal changes (not shown). The high beam effects
(BE) maximum was slightly decreased due to TD in this
temperature range (compare Fig. 8), and the low BE maxi-
mum was increased by ~10%, obviously due to conversion
between the two states. Nevertheless, the ratio of both peak
areas was still ~ 4:1.
This result is surprising with respect to the conver-
sion of molecular to partially dissociated water sketched
above [19–22]. Naively one would assign the first (i.e.,
low-temperature) TD peak to the molecularly absorbed
layer, and the second one to the partially dissociated film.
As a result, one would expect a single O1s peak cor-
responding to intact water for the pristine layer, and a
two peak structure corresponding to the co-existence of
D(H)2O and OD(H) + D(H) for the partially dissoci-
ated layer responsible for the second TD maximum. The
one/two peak structures are observed, but the peak ratio at
the onset of the second TD maximum fits neither the 1:1
ratio of the structure proposed in [24] nor the 5:3 ratio of
the revised version from [25]. Area ratios close to 1:1 are
observed only for electron exposures beyond the electron
dose required for complete TD conversion (Fig. 12).
As mentioned above, no residual oxygen could be de-
tected after TD of pristine D2O layers. After TD of a layer
irradiated until complete conversion (6.3·1015 e/cm2
trace in Fig. 12), a residual oxygen signal at 530.5 eV BE
corresponding to 8% of the initially present oxygen or
~ 0.1 ML was detected, in perfect agreement with the
amount of co-adsorbed oxygen necessary for complete
conversion [18]. Excessive bombardment increases the
residual oxygen signal from 8 to 13% (Fig. 12). This
value is close to the oxygen amount found in [25] after
the desorption of nominally pure D2O and H2O layers.
From the results shown in Figs. 10 and 12 we conclude
that irradiation with a distinct electron dose has the same
effect on the TD spectrum and probably also on the com-
position of the layer as co-adsorption of exactly the
amount of oxygen which is left over after TD of the irradi-
ated layer. The XPS results, however, indicate that the ir-
radiated layer is not simply a combination of D2O, OD
and isolated O; the O1s peak of low BE from the irradi-
ated layer has a larger BE than the residual O obtained af-
ter TD. We obviously do not encounter isolated O atoms
in the irradiated layer but O atoms which are always
correlated with hydrogen. The binding energies for the
post-TD O1s peak and for the ODx peak of the irradiated
layer become more equal for extended irradiation which
also means increased loss of hydrogen (Fig. 12).
In summary, we conclude from our data that the beam
effects observed in our study and previous ones are due to
a beam induced perturbation of the H to O ratio in the ad-
sorbed water layer, in perfect agreement with previous
co-adsorption experiments [18]. Future experiments de-
tecting selectively the amount of H on the surface during
and after TD could test this hypothesis. This perturbation
obviously has a very strong effect on the barrier for par-
tial dissociation of the molecular layer which explains the
very large effective conversion cross-section. Comparing
TD (Fig. 9) and XPS results (Fig. 12 and text), we tenta-
tively conclude that this beam (or co-adsorbate) induced
partial dissociation is by far not complete at the minimum
between the two TD peaks. According to our data only
~20% of the layer would be due to an OD like species.
Further dissociation stabilizing the layer and leading to a
continuously increasing activation energy for desorption
is assumed to occur during TD of the second state. This
would explain the shape of the high-temperature TD peak
which because of recombinative desorption should show
a reaction order not smaller than 1, but in fact exhibits a
slope ratio of leading/trailing edges which is either typi-
cal for a reaction order < 1, or an activation energy which
increases with decreasing coverage. As mentioned above,
future in-situ analysis of the layer composition is required
for final conclusions.
We finally would like to address the point of beam in-
duced artifacts in previous studies. From the results
above it is clear that layer modifications by irradiation
cannot be discriminated from the effect of co-adsorbed
oxygen; such oxygen could be due to incomplete removal
of previously prepared layers or to poor vacuum. On the
other hand, contamination by hydrogen would have the
opposite effect. One could argue that the discrepancies
between our present study and those of [17–22,26] com-
pared with the results reported in [25] are due to hydrogen
contamination in our case and in [17–22,26]. We believe
this not to be the case. All our results have been ob-
tained under very good UHV conditions (base pressure
2·10–11mbar), and all results have been reproduced seve-
ral times. In addition, perfect agreement exists between
our study and those from Refs. 17–22, 26, where compa-
rable.
In summary, chemisorption of water on the Ru(001)
surface was found to be a challenging system for theorists
as well as for experimentalists. For theoreticians because
of the difficulties arising from treating hydrogen bonds
correctly, and for experimentalists because of the extreme
sensitivity of this system to perturbations by coadsorbates
and by irradiation. On the other hand, this system is an
outstanding example for the large effect which the prod-
ucts of a beam induced reaction can have on an intrinsi-
Electronically induced modification of thin layers on surfaces
Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7 687
cally thermally stimulated process like the partial dissoci-
ation of the water molecule.
Acknowledgments
We thank D.L. Allara for supporting us with the sub-
stance for the nitrile terminated SAMs, K.L. Kostov, N.
Faradzhev and T.E. Madey for helpful discussions, and
M. Grunze for support of the SAM related experiments.
Help during the experiments by K. Eberle, R. Schneider,
and the staff of BESSY, in particular O. Schwarzkopf,
H. Pfau, and W. Braun, is gratefully acknowledged. The
investigations have been supported by the Deutsche
Forschungsgemeinschaft (Fe 347/1-3), by the German
BMBF (05 KS4VHA/4 and 05 ES3XBA/5), and by the
German Fonds der Chemischen Industrie.
1. R.D. Levine, Molecular Reaction Dynamics, Cambridge
University Press, Cambridge (2005).
2. One of these exceptions are reactions driven by hot elec-
trons created by intense laser pulses; see, e.g., M. Brand-
byge, P. Hedeg�rd, T.F. Heinz, J.A. Misewich, and D.M.
Newns, Phys. Rev. B52, 6042 (1995).
3. V.E. Cosslett, J. Microscopy 113, 113 (1978); G. Schnei-
der, in: X-Ray Microscopy IV, V.V. Aristov and A.I. Erko
(eds.), Bogorodskii Pechatnik Publishing Company, Cher-
nogolovka (1994), p. 181; G. Schneider, Ultramicroscopy
75, 85 (1998); D. Weiß, G. Schneider, B. Niemann, P. Gutt-
mann, R. Rudolf, and G. Schmahl, Ultramicroscopy 84, 185
(2000).
4. A. Ulman, Chem. Rev. 96, 1533 (1996); F. Schreiber, Prog.
Surf. Sci. 65, 151 (2000).
5. Handbook of Nanotechnology, B. Bushan (ed.), Springer,
Berlin (2004).
6. A. Härtl, E. Schmich, J.A. Garrido, J. Hernando, S.C.R.
Catharino, S. Walter, P. Feulner, A. Kromka, D. Steinmül-
ler, and M. Stutzmann, Nature Mater. 3, 736 (2004).
7. A. Gölzhäuser, W. Geyer, V. Stadler, W. Eck, M. Grunze,
K. Edinger, Th. Weimann, and P. Hinze, J. Vac. Sci. Tech-
nol. B18, 3414 (2000) .
8. M. Zharnikov and M. Grunze, J. Vac. Sci. Technol. B20,
1793 (2002) and references therein.
9. M.A. Huels, P.-C. Dugal, and L. Sanche, J. Chem. Phys.
118, 11168 (2003) and references therein.
10. E. Garand and P.A. Rowntree, J. Phys. Chem. B109, 12927
(2005) and references therein.
11. P. Feulner, T. Niedermayer, K. Eberle, R. Schneider, D.
Menzel, A. Baumer, E. Schmich, A. Shaporenko, Y. Tai,
and M. Zharnikov, Phys. Rev. Lett. 93, 178302 (2004).
12. A. Shaporenko, M. Zharnikov, P. Feulner, and D. Menzel,
J. Phys. Condens. Matter 18, S1677 (2006) .
13. R. Romberg, S.P. Frigo, A. Ogurtsov, P. Feulner, and D.
Menzel, Surf. Sci. 451, 116 (2000).
14. E. Schmich, Diploma Thesis, Technische Universität Mün-
chen (2004).
15. S. Neppl, Diploma Thesis, Technische Universität Mün-
chen (2006).
16. S. Frey, A. Shaporenko, M. Zharnikov, P. Harder, and D.
L. Allara, J. Phys. Chem. B107, 7716 (2003).
17. N. Faradzhev, K.L. Kostov, P. Feulner, T.E. Madey, and
D. Menzel, Chem. Phys. Lett. 415, 165 (2005).
18. C. Clay, S. Haq, and A. Hodgson, Chem. Phys. Lett. 388,
89 (2004).
19. Sheng Meng, E.G. Wang, Ch. Frischkorn, M. Wolf, and
Shiwu Gao, Chem. Phys. Lett. 402, 384 (2005).
20. A. Michaelides, A. Alavi, and D.A. King, J. Am. Chem.
Soc. 125, 2746 (2003).
21. G. Materzanini, G.F. Tantardini, P.J.D. Lindan, and P.
Saalfrank, Phys. Rev. B71, 155414 (2005).
22. S. Haq, C. Clay, G.R. Darling, G. Zimbitas, and A. Hodg-
son, Phys. Rev. B73, 115414 (2006).
23. D.L. Doering and T.E. Madey, Surf. Sci. 123, 305 (1982).
24. P.J. Feibelman, Science 205, 99 (2002).
25. J. Weissenrieder, A. Mikkelsen, J.N. Andersen, P.J. Feibel-
man, and G. Held, Phys. Rev. Lett. 93, 196102 (2004).
26. K. Andersson, A. Nikitin, L.G.M. Pettersson, A. Nilsson,
and H. Ogasawara, Phys. Rev. Lett. 93, 196101 (2004).
27. N. Faradzhev (private communication).
28. M.J. Gladys, A. Mikkelsen, J.N. Andersen, and G. Held,
Chem. Phys. Lett. 414, 311 (2005).
688 Fizika Nizkikh Temperatur, 2007, v. 33, Nos. 6/7
U. Bauer, S. Neppl, D. Menzel, P. Feulner, A. Shaporenko, and M. Zharnikov
|