DOI: 10.1002/chem.

201504810

Full Paper

& Carboxylic Acid Adsorption

Structure-Dependent Anchoring of Organic Molecules to

Atomically Defined Oxide Surfaces: Phthalic Acid on Co3O4(111),
CoO(100), and CoO(111)
Tao Xu,[a] Matthias Schwarz,[a] Kristin Werner,[a] Susanne Mohr,[a] Max Amende,[a] and
Jrg Libuda*[a, b]

Abstract: We have performed a model study to explore the

influence of surface structure on the anchoring of organic
molecules on oxide materials. Specifically, we have investigated the adsorption of phthalic acid (PA) on three different,
well-ordered, and atomically defined cobalt oxide surfaces,
namely 1) Co3O4(111), 2) CoO(111), and 3) CoO(100) on
Ir(100). PA was deposited by physical vapor deposition
(PVD). The formation of the PA films and interfacial reactions
were monitored in situ during growth by isothermal time-resolved IR reflection absorption spectroscopy (TR-IRAS) under
ultrahigh vacuum (UHV) conditions. We observed a pronounced structure dependence on the three surfaces with
three distinctively different binding geometries and characteristic differences depending on the temperature and coverage. 1) PA initially binds to Co3O4(111) through the formation of a chelating bis-carboxylate with the molecular plane
oriented perpendicularly to the surface. Similar species were
observed both at low temperature (130 K) and at room tem-

perature (300 K). With increasing exposure, chelating monocarboxylates became more abundant and partially replaced
the bis-carboxylate. 2) PA binds to CoO(100) in the form of
a bridging bis-carboxylate for low coverage. Upon prolonged deposition of PA at low temperature, the bis-carboxylates were converted into mono-carboxylate species. In
contrast, the bis-carboxylate layer was very stable at 300 K.
3) For CoO(111) we observed a temperature-dependent
change in the adsorption mechanism. Although PA binds as
a mono-carboxylate in a bridging bidentate fashion at low
temperature (130 K), a strongly distorted bis-carboxylate was
formed at 300 K, possibly as a result of temperature-dependent restructuring of the surface. The results show that the
adsorption geometry of PA depends on the atomic structure
of the oxide surface. The structure dependence can be rationalized by the different arrangements of cobalt ions at
the three surfaces.

Introduction

growth and structure of the film and, therefore, their electronic

and chemical properties (see, for example, ref. [3]).
Although the general behavior of these anchoring groups is
well understood, very little is known of the mechanisms of interaction on the atomic scale. What is the structure of the
bonding site at the surface and how does it influence the stability, geometry, energetics, and formation kinetics of the organic film? It is the intrinsic complexity of such interfaces and
the associated experimental challenges that are the main reasons for this lack of understanding.
Atomically well-defined oxide surfaces can be prepared following a surface science approach under ultrahigh vacuum
(UHV) conditions. A number of oxide materials and crystallographic orientations are available either as a bulk single-crystal
surface or in the form of ordered thin films on metal single
crystals (see, for example, ref. [4]). The thin-film approach prevents charging when working with photoelectron spectroscopy or scanning tunneling microscopy (STM). In addition, it
allows the stoichiometry and surface structure to be varied, at
least in some cases. In this work we used the thin-film approach to study molecular anchoring as a function of the surface structure.

Organic thin films on oxide surfaces have great potential in

photovoltaics[1] and molecular electronics.[2] For such applications, the organic entities are often bound to the interface
through anchoring groups, for example, carboxylic acids, phosphonates, or hydroxy groups. Knowledge of the associated
binding mechanisms, kinetics, and energetics is essential for
understanding the growth and structure formation processes
at organic oxide interfaces. They are the key to controlling the
[a] T. Xu, M. Schwarz, K. Werner, S. Mohr, M. Amende, Prof. J. Libuda
Lehrstuhl fr Physikalische Chemie II
Friedrich-Alexander-Universitt Erlangen-Nrnberg
Egerlandstrae 3, 91058 Erlangen (Germany)
Fax: (+ 49) 9131-8527308
E-mail: joerg.libuda@fau.de
[b] Prof. J. Libuda
Erlangen Catalysis Resource Center
and Interdisciplinary Center Interface Controlled Processes
Friedrich-Alexander-Universitt Erlangen-Nrnberg
Egerlandstrae 3, 91058 Erlangen (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201504810. Colored figures are available.
Chem. Eur. J. 2016, 22, 1 14

These are not the final page numbers!

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper
Specifically, we scrutinized cobalt oxides, which exhibit interesting catalytic[5] and magnetic[6] properties. As a multivalent
metal, cobalt forms oxide phases with different stoichiometries.
Of these, the most important are CoO and Co3O4. The structural and chemical properties of their surfaces depend on the direction of surface plane.[7] Most remarkably, differently ordered
CoO and Co3O4 thin-film systems are available by growth on
metallic substrates, specifically on Ir(100). Many of these systems have been structurally characterized in great detail by
Heinz and Hammer, both by STM and low-energy electron diffraction (LEED) I-V analysis (see their review[8] and references
therein). In this work, we scrutinized three different surfaces,
Co3O4(111), CoO(111), and CoO(100). All the films were grown
under UHV by the reactive deposition of cobalt onto Ir(100) in
an oxygen atmosphere and subsequent annealing.[8]
The atomic structures of the three surfaces are very different. The Co3O4(111) surface is very similar to that of a bulk-terminated spinel. The surface is terminated by coordinatively unsaturated cobalt ions occupying tetrahedral sites of the bulk
lattice.[9] Similarly to Co3O4(111), the surface of CoO(111) is
polar in nature. To compensate for the polarity, the structure
switches from the bulk rock-salt structure to a wurtzite structure close to the surface.[10] The cobalt ions are embedded in
the subsurface and are less accessible to adsorbates. Finally,
the structure of CoO(100) is identical to that of bulk rock salt.[8]
The surface is nonpolar and cobalt and oxygen ions are exposed in a similar fashion. This toolbox of structures makes it
possible to probe the role of surface structure on the interactions with anchoring groups.
Among the different anchors,[3e, 11] the carboxylic acid group
is one of the most common. Self-assembled monolayers
(SAMs) of large organic molecules with carboxy anchoring
groups have been prepared on many oxide surfaces, mostly
from solution.[3f, 12] After deprotonation the carboxylate attaches to the surface metal cations in the form of a monodentate,
bridging or chelating bidentate, or distorted bidentate species

(see Figure 1a). The stability of these species should critically

depend on the arrangement and accessibility of the cations.
However, experimental evidence of atomically defined surfaces
to prove this hypothesis is pending, mainly because there are
very few ordered oxides that allow the surface structure to be
changed whilst keeping a high degree of ordering and control
at an atomic level.
In this work we studied the mechanism of binding to the
cobalt oxide surfaces by using in situ time-resolved IR reflection absorption spectroscopy (TR-IRAS). The method is particularly appropriate because it provides information on both the
anchoring mechanism and the molecular orientation. The
latter may be straightforwardly derived from the metal surface
selection rule (MSSR),[13] which is valid for thin oxide films on
metallic substrates. We chose phthalic acid (PA, benzene-1,2-dicarboxylic acid, o-C6H4(COOH)2) as a test molecule. With two
carboxy anchoring groups, phthalic acid is particularly sensitive
to the surface structure: The two carboxy groups may either
bind through a single carboxy group (mono-carboxylate) or
through both (bis-carboxylate) (see Figure 1b). In addition,
each carboxy group may bind in different geometries, as
shown in Figure 1a (see ref. [14] and references therein).
So far there have only been a few surface science studies of
PA at oxide surfaces. The most detailed picture has been obtained for terephthalic acid (benzene-1,4-dicarboxylic acid, pC4H6(COOH)2) on TiO2(110)-(1  1) (see, for example,
ref. [3d, 15]). Here, the molecule binds through a single bidentate carboxylate to two five-fold coordinated titanium ions.
The molecules adopt a nearly upright geometry, but tilt to
form dimers and the aromatic ring rotates with respect to the
carboxylate unit. STM and X-ray photoelectron spectroscopic
(XPS) analyses of isophthalic acid (benzene-1,3-dicarboxylic
acid, m-C4H6(COOH)2) on copper-modified Au(111), on the
other hand, suggest that both carboxy groups can bind to the
surface with the aromatic plane being forced to a smaller tilting angle with respect to the surface.[16]

Figure 1. Possible binding geometries of a) a carboxylate group on a metaloxide surface and b) phthalic acid carboxylates on a metal oxide surface.

&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper
er and better-ordered CoO(100) films, a second reactive deposition
of cobalt in O2 was conducted over 18 min while keeping the
sample at a low temperature (< 50 8C). After that, the CoO(100)
film was annealed at 600 8C for 10 min. The as-prepared CoO(100)
thin film showed a sharp (1  1) pattern in LEED.

In this work we have shown that PA can bind to cobalt

oxides in many different geometries depending on the coverage, temperature, and, most importantly, the surface structure.
These findings underline the importance of surface structure
for the anchoring of molecular films onto oxide surfaces.

Organic thin-film deposition: PA (SigmaAldrich, 99.9 %) was applied by using a home-built Knudsen source that can be separated
from the UHV chamber by a gate valve. The evaporator was baked
overnight before first usage. PA was placed into a glass reservoir
with an orifice. For degassing, the glass reservoir containing the PA
was heated at 350 K for at least 30 min while being pumped
through a separate pumping line. Subsequently, the gate valve to
the chamber was opened. For deposition, the evaporator was
placed in front of the sample and the sample shutter was opened.
The evaporation temperature was adjusted to ensure a constant
deposition rate during the time scale of the dosing experiment.
The deposition rate was estimated to be around 0.03 mL min1 on
Co3O4(111) and CoO(111) and 0.08 mL min1 on CoO(100), as determined by low-temperature adsorption experiments in which we
could distinguish monolayers from multilayers.

Experimental Section
All measurements were conducted in a UHV system (base pressure
of 1.0  1010 mbar), which has been described in detail elsewhere.[17] The UHV system consists of a preparation chamber for
sample cleaning and preparation, and a measurement chamber for
the IRAS experiments. The latter is equipped with an FTIR spectrometer (Bruker VERTEX 80v) connected by differentially pumped
KBr windows. The spectra were recorded with a resolution of
4 cm1. During the adsorption measurements, IR spectra were continuously acquired with at a rate of 30 s per spectrum.
Preparation of Co3O4(111)/Ir(100): A Co3O4(111) thin film was prepared by reactive deposition of cobalt atoms in an O2 atmosphere
according to the method described in the literature,[9] but with
small modifications of parameters such as the partial pressure of
O2, annealing temperature, and the amount of cobalt. The detailed
procedure is as follows: First, the Ir(100) single crystal (MaTeck) was
cleaned by cycles of Ar + sputtering (1.8 keV, room temperature,
1 h; Ar, Linde, 6.0) and annealing (1100 8C, 3 min) until a clear LEED
pattern (room temperature) of the Ir(100)-(5  1) reconstructed surface could be seen. Secondly, the Ir(100)-(5  1) surface was heated
to 1000 8C in 5  108 mbar O2 (Linde, 5.0) for 3 min and cooled to
room temperature in O2 atmosphere. This led to an Ir(100)-(2  1)O
reconstructed surface that shows a clear (2  1) pattern in LEED.
Starting with the Ir(100)-(2  1)O, cobalt was evaporated onto the
surface (cooled to temperatures below 0 8C) by using a commercial
electron beam evaporator (Focus EFM3, 2 mm cobalt rod, Alfa
Aesar 99,995 %, Goodfellow 99.99 %) in an atmosphere of 1.0 
106 mbar O2 for 18 min. The evaporation rate of cobalt was determined to be 2  min1 by means of a quartz crystal microbalance.
After the growth, the film was annealed in O2 (1.0  106 mbar) at
250 8C for 2 min and then under UHV at 430 8C for 5 min. The film
was analyzed qualitatively by comparing the LEED I-V curves with
the literature.[9]

Results and Discussion

PA adsorption on Co3O4(111)/Ir(100) at 130 K
To study the adsorption of PA on Co3O4(111), the freshly prepared Co3O4(111) film was cooled to 130 K. After the acquisition of a reference spectrum, the surface was exposed to PA
for 60 min, and IRAS spectra were recorded continuously at
a rate of 30 s per spectrum. We show a region of the IRAS
spectra recorded at 1 min intervals in Figure 2a. In Figure 2b
we compare selected spectra of the sample with sub-monolayer coverage, slightly above the monolayer saturation coverage,
and at high coverage, which corresponds to a multilayer film.
At the beginning of the deposition, a prominent peak appears at around 1419 cm1, with other features becoming visible at longer deposition times. After about 10 min, a broad
feature appears at around 1700 cm1, which grows in intensity
and becomes the most prominent band (1710 cm1) by the
end of the experiment. In addition, other peaks are observed
at 1855, 1790, 1775, and 1548 cm1. All the features observed
in the mono- and multilayer regions are summarized in Table 1
together with their assignments based on the literature[18] and
our previous work.[19]
In the sub-monolayer region, the band at 1419 cm1 can be
assigned to the OCO symmetric stretching vibration whereas
the OCO asymmetric mode at 1548 cm1 is hardly visible. At
higher coverage, the intense band at 1710 cm1 can be attributed to the C=O stretching mode of the free carboxylic acid
group.[18f, 19] Most of the weaker bands can be assigned to the
stretching and deformation modes of the hydrocarbon unit.
However, there are three additional peaks centered at 1775,
1790, and 1853 cm1 that cannot be attributed to any vibrational modes of PA. In fact, they correspond very well to those
of phthalic acid anhydride (PAA).[18g, 20] In our previous work on
PA on MgO(100)/Ag(100), we observed that a small fraction of
PAA was formed upon PVD of PA and is co-adsorbed onto the
surface at low temperature.[19]

Preparation of CoO(111)/Ir(100): The CoO(111) film was prepared

from a Co3O4(111) thin film prepared by the procedure described
above. Simple heating of the Co3O4(111) film under UHV leads to
loss of oxygen and the transformation to CoO(111).[10b] In our experiments, Co3O4(111) thin films (cobalt amount equivalent of 36 
cobalt metal) were heated at 620 8C for 5 min. The quality of the
CoO(111) thin film was verified by LEED.[10b]
Preparation of CoO(100)/Co/Ir(100): To obtain the CoO(100) thin
film, a metastable Ir(100)-(1  1) reconstructed surface was first prepared. To this end, we started with a clean Ir(100)-(5  1) surface
and prepared the Ir(100)-(2  1)O reconstructed surface following
the procedure described above for the Co3O4(111) film. Then, the
sample was heated at 275 8C in 1  107 mbar H2 (Linde, 5.3) for
1 min and then heated under UHV at 275 8C for 1 min before cooling to room temperature. This procedure yielded an Ir(100)-(1  1)
surface characterized by a sharp (1  1) pattern in LEED. Subsequently, metallic cobalt was deposited onto the surface (50 8C) at
a rate of 2  min1 for 6 min. After preparation of the cobalt buffer
layer, the sample was cooled to a temperatures below 50 8C.
Then the cobalt was deposited reactively (2  min1) in 4 
107 mbar O2 for 3 min. Subsequently, the film was annealed at
100 8C for 1 min to afford an ordered CoO(100) structure. For thickChem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper
band at 1710 cm1 remains nearly invisible. In region II (10
30 min), the band at 1419 cm1 approaches saturation and the
band at 1710 cm1 starts to grow more quickly. Finally, in region III (from 30 min), the intensity of the band at1419 cm1 remains constant whereas the peak at 1710 cm1 continues to
increase at a constant rate.
We rationalize the origin of the three growth regions in Figure 2c as follows. At low coverage, both carboxy groups are
deprotonated upon adsorption and anchor to the surface.
Therefore no n(C=O) band at 1710 cm1 can be observed. Instead, we observe the ns(OCO) band of the surface-bound carboxylate dominating at 1419 cm1. This implies that the PA is
bound to the Co3O4(111) surface through both carboxylate
groups, that is, it forms a bis-carboxylate at the surface. In region II, the intensity of the free carboxylic acid C=O band increases whereas the intensity of the bands associated with carboxylate decreases. There are two possible scenarios that can
explain this behavior. Either there are intact PA molecules coadsorbing as bis-carboxylate species are continuously formed
or, alternatively, PA may be adsorbed in the form of mono-carboxylates. Later we will show that free carboxylic acid groups
are formed even at temperatures well above the desorption
temperature of PA. This observation supports the formation of
surface-anchored mono-carboxylates. We suggest that the limited surface mobility of the bis-carboxylate leads to the formation of a disordered monolayer in which isolated sites persist.
PA can bind to the latter only as a mono-carboxylate. Finally,
a multilayer PA film grows in region III and no further surface
carboxylates are formed.
Detailed information on the molecular orientation of the surface carboxylate can be derived from IRAS-active vibrations of
the hydrocarbon backbone. First, we observe that the intensity
of the CH out-of-plane deformation mode at 743 cm1 is very
low for sub-monolayer coverage. If we take into account the
fact that the dipole moment of this mode is perpendicular to
the molecular plane and use the MSSR, which states that only
the component of the dynamic dipole moment perpendicular
to the surface contributes to IR absorption, we can conclude
that the aromatic ring is oriented nearly perpendicularly to the
surface. Secondly, we observe that the asymmetric stretching
mode nas(OCO) at around 1548 cm1 is nearly invisible in the
sub-monolayer region. This observation suggests that the carboxylate units adsorb in a nearly symmetric bidentate geometry with little distortion. This is in contrast to our previous
work on MgO(100) in which a substantial distortion of the bisbidentate carboxylate was observed.[19] Note that the out-ofplane signal at 743 cm1 can be observed in the multilayer region III, which indicates a smaller average tilting angle of the
aromatic ring with respect to the surface in the multilayer.
A schematic model of the adsorption geometry is displayed
in Figure 3a. We have used the surface structure of the
Co3O4(111) films previously established by LEED I-V analysis.[9]
The internal molecular bond lengths correspond to those in
the free molecule. As outlined above, the IRAS experiments indicate adsorption of a practically symmetric bis-carboxylate
with the aromatic ring oriented nearly perpendicularly to the
surface. Note that the Co3O4(111) surface is terminated by Co2 +

Figure 2. a) IR spectra recorded during the deposition of PA onto

Co3O4(111)/Ir(100) for 60 min at 130 K. b) Comparison of a sub-monolayer
spectrum (5 min of deposition), a multilayer spectrum (60 min of deposition), and the spectrum of a thicker multilayer film (obtained at a higher
dosing rate). c) Integrated peak intensities at 1419 cm1 (bound carboxylate)
and 1710 cm1 (free carboxylic acid) as a function of deposition time.

To investigate the growth of the film in more detail, the intensities of the bands at 1419 and 1710 cm1 were analyzed as
a function of deposition time. The result is shown in Figure 2c.
Three regions of growth can be distinguished in this plot (Figure 2c, IIII). In region I (0 to 10 min), the carboxylate band at
1419 cm1 rapidly gains in intensity whereas the free acid
&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper
Table 1. Vibrational frequencies and assignments for phthalic acid on Co3O4(111)/Ir(100), CoO(111)/Ir(100), and CoO(100)/Ir(100).[a]
CoO(100)
CoO(111)
Co3O4(111)
Assignment[b]
Monolayer [cm1] Multilayer [cm1] Monolayer [cm1] Multilayer [cm1] Monolayer [cm1] Multilayer [cm1]
743

1415

1309
1417
1497

1552
1585
1602
1712
1776
1793
1853

1416
1495
1550

1700

743
1262
1312
1416
1495

743
1259
1312
1419
1494

1419
1494
1548

1583
1601
1712
1776
1793
1855

1583
1603
1710
1775
1790
1853

g(CH)ring + g(CC)ring
ns(COC)
n(CO) + n(CC) + d(OH)
nsym(OCO)
n(CCring) + nCCring + n(CH)
nasym(OCO)
n(CC)ring
n(CC)ring
n(C=O)
nasym(C=O)
d(CH) + d(CC)ring
nsym(C=O)

Comment

phthalic anhydride

phthalic anhydride
phthalic anhydride
phthalic anhydride

[a] n = stretching, g = out-of-plane bending, and d = in-plane bending. [b] See ref. [18].

In the IR spectra we observe one prominent peak at

1425 cm1, which is the most intense feature over the complete deposition time. Other peaks are observed at 1494, 1548,
and 1703 cm1. Comparing the spectra in Figure 4b, we can
conclude that the changes in relative intensities and positions
of the bands are moderate over the deposition time. The exception is the broad band at around 1703 cm1, which is not
present in the first spectrum, but is clearly seen in the last. In
Figure 4c, we display the integrated peak areas of the bands at
1425 and 1703 cm1 as a function of deposition time. We have
subdivided the deposition experiment into two regions. In the
first, the region from the beginning to 30 min deposition time,
we observe an increase in intensity and the saturation of the
dominant band at 1425 cm1, whereas the peak at 1703 cm1
remains invisible. At deposition times exceeding 30 min, the
peak at 1425 cm1decreases slowly in intensity whereas the
feature at 1703 cm1 grows continuously.
On the basis of the much lower intensities of the bands and
the saturation behavior, we conclude that at 300 K only
a single monolayer of PA is adsorbed onto Co3O4(111)/Ir(100).
For the peak assignments, refer to Table 1. Only slight frequency shifts of the IR signals are observed in comparison with the
deposition at low temperature. The dominant bands are the
OCO symmetric stretching mode at 1425 cm1, the antisymmetric OCO mode at 1548 cm1, the ring mode at 1494 cm1,
and the C=O band of the free carboxylic acid group at
1703 cm1.
In accord with the discussion above, we attribute the spectral features in region I to the formation of a symmetric bis-bidentate carboxylate bound to the oxide surface, similar to that
observed for thin-film growth at low temperature. In region II,
the appearance of the band at around 1700 cm1 indicates the
formation of free carboxylic acid groups. As the experiment
was performed well above the multilayer desorption temperature and the free carboxylic acid can be observed on the surface up to much higher temperatures (data not shown), we exclude the co-adsorption of unbound PA. Consequently, we
have assigned the free carboxylic acid band to the very slow
formation of singly bound PA species, that is, a mono-carboxylate. The decrease in intensity of the band at 1425 cm1 sug-

Figure 3. Schematic model of a PA molecule adsorbed onto a) Co3O4(111)

b) CoO(111), and c) CoO(100) surfaces at 130 K. The atoms of PA are depicted
as follows: oxygen = white, carbon = light gray, and hydrogen = dark gray.

cations with a Co2 + Co2 + distance of about 5.7 . The distance

between the two oxygen ions in the carboxylate group
(COO), however, is only about 2.3 . This implies that a single
carboxylic acid group cannot bridge two surface Co2 + ions.
Therefore, the only symmetric adsorption geometry for PA that
is compatible with the structure of the Co3O4(111) surface is
a chelating bis-carboxylate, as illustrated in Figure 3a.
PA adsorption on Co3O4(111)/Ir(100) at 300 K
Next, we investigated the growth of PA on Co3O4(111)/Ir(100)
at 300 K by using the experimental procedure described
above. The IR spectra recorded as a function of exposure time
are shown in Figure 4a, the spectra of samples with low and
high PA coverage are compared in Figure 4b, and the intensities of the symmetric OCO band of the bound carboxylate
(1425 cm1) and the C=O stretching band of the free carboxylic
acid (1703 cm1) are plotted in Figure 4c.
Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper
through two anchoring groups. Finally, we note that within
the signal-to-noise ratio of the experiment, no out-of-plane
peak at 743 cm1 is observed, which indicates that the PA is
adsorbed with the molecular plane oriented nearly perpendicularly to the surface.
PA adsorption on CoO(111)/Ir(100) at 130 K
Next, we investigated the growth of PA on CoO(111)/Ir(100) at
130 K by using the experimental procedures described above.
The IR spectra as a function of exposure time are shown in Figure 5a, the IR spectra recorded for sub-monolayer, monolayer,
and multilayer coverages are compared in Figure 5b, and the
intensities of the carboxylate (1416 cm1) and free carboxylic
acid (1712 cm1) bands as a function of time are plotted in Figure 5c.
At low PA exposure, the most prominent bands appear at
1416, 1495, 1550, and 1700 cm1. For longer deposition times,
further peaks appear at 1855, 1793, 1776, and 1312 cm1. For
a detailed assignment, please refer to Table 1 and the discussions above.
Again, we have subdivided the deposition of PA into two regions of growth (see Figure 5c). In region I, from the beginning
of the deposition to approximately 30 min deposition time,
the symmetric OCO stretching band at 1416 cm1 increases in
intensity and finally approaches saturation. At the same time,
the C=O band of the free carboxylic acid groups at 1712 cm1
increases almost linearly. This behavior is very different from
that observed for growth on Co3O4(111); it implies that right
from the beginning of the deposition, the surface concentrations of both the free and bound carboxylic acid groups increase linearly, which suggests the formation of mono-carboxylates even at low coverages. With increasing coverage, the intensity of the carboxylate band grows more slowly, which indicates an increasing fraction of intact PA molecules. Finally, in
region II (> 30 min), no further surface-bound carboxylates are
formed and a multilayer film of intact PA starts to grow.
The atomic structure of the CoO(111) surface is well known
from previous LEED I-V analyses, in which it was shown that
the polar surface undergoes reconstruction to form a wurtzitetype top layer (see Figure 3b[10b]). The Co2 + ions at the surface
are arranged in a hexagonal lattice and are separated by a distance of 3.0 , which is similar to that of bulk rock salt. These
Co2 + ions are located, however, slightly below the terminating
layer of O2 ions. From a steric perspective this may restrict
access to the Co2 + ions and therefore limit the range of possible binding geometries of the carboxylate groups. We propose
that this steric restriction prevents the PA from binding
through both carboxylate units. For the adsorbed carboxylate,
both adsorption geometries, bridging and chelating, would be
possible from a geometric point of view and both geometries
would also be compatible with the IR data. Previous experimental and theoretical studies of different carboxylic acids on
metal oxide surfaces showed, however, that the bridging carboxylate binding mode is typically preferred from an energetic
point of view (see, for example, formic acid on ZnO(1010),
rutile TiO2(110),[21] anatase TiO2(110),[22] and CeO2(111),[23] and

Figure 4. a) IR spectra recorded during the adsorption and growth of PA on

Co3O4(111)/Ir(100) at 300 K. b) Comparison of the IR spectra recorded after
1 and 60 min of deposition. c) Integrated peak intensities at 1425 cm1
(bound carboxylate) and 1703 cm1 (free carboxylic acid) as a function of
deposition time.

gests that, upon continuous dosing, a portion of the bis-carboxylates is slowly converted into the mono-carboxylate species. This interpretation is supported by the change in shape
of the antisymmetric OCO band at 1548 cm1 in region II. We
suggest that the formation of two singly bound mono-carboxylates is energetically preferred at certain surface binding sites
because of the strain that is induced when the PA binds
&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper
PA adsorption on CoO(111)/Ir(100) at 300 K
Next, we explored the growth of PA on CoO(111)/Ir(100) at
300 K by using the experimental procedure described above.
The IR spectra as a function of exposure time are shown in Figure 6a, the IRAS spectra recorded at low and high PA coverage
are compared in Figure 6 b, and the intensities of the symmet-

Figure 5. a) IR spectra recorded during the deposition of PA onto CoO(111)/

Ir(100) for 60 min at 130 K. b) Comparison of the sub-monolayer spectrum
(2 min of deposition), a multilayer spectrum (60 min of deposition), and the
spectrum of a thicker multilayer film (obtained at higher dosing rate). c) Integrated peak intensities at 1416 cm1 (bound carboxylate) and 1712 cm1
(free carboxylic acid) as a function of deposition time.

benzoic acid on rutile TiO2(110)[15] or MgO(100)[19]). Therefore

we propose that at 130 K, PA adsorbs on CoO(111) as a bridging
mono-carboxylate. Some possible adsorption geometries are
schematically depicted in Figure 3b. It is worth noting that at
p p
300 K a more complex and roughened ( 3  3)R 308 structure is formed.[10b] Nevertheless, Figure 3b should convey the
image of the bonding of a mono-carboxylate on the surface.
Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

Figure 6. a) IR spectra recorded during the adsorption and growth of PA on

CoO(111)/Ir(100) at 300 K. b) Comparison of the IR spectra recorded after 2
and 60 min of deposition. c) Integrated peak intensities at 1428 cm1 (bound
carboxylate) and 1700 cm1 (free carboxylic acid) as a function of deposition
time.

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper
ric OCO band of the bound carboxylate (1428 cm1) and the
C=O stretching band of the free carboxylic acid (1700 cm1) as
a function of time are plotted in Figure 6c.
As discussed above, no PA multilayer was formed at 300 K
and adsorption occurs in the monolayer region only. We observe the symmetric OCO band of the bound carboxylate peak
at 1428 cm1 and this band is the most prominent throughout
the deposition experiment. Smaller bands are observed at
1493, 1550, and 1700 cm1. As discussed above (see Table 1),
these features have been assigned to the ring mode of PA
(1493 cm1), the antisymmetric OCO mode (1550 cm1) of the
bound carboxylate, and the C=O band (1700 cm1) of the free
carboxylic acid group.
As indicated in Figure 6c, we have subdivided the experiment into two regions of growth. In region I, up to a deposition
time of approximately 45 min, the symmetric OCO mode
(1428 cm1) increases in intensity and finally saturates. For
longer PA exposures, no further changes are observed. Interestingly, the free carboxylic acid band at 1700 cm1 remains
very weak throughout the whole experiment, which clearly indicates that the PA is adsorbed as a bis-carboxylate. This behavior is in sharp contrast to that observed at 130 K; at this
temperature singly bound mono-carboxylates are formed at
low coverage. At 300 K, the band of the free carboxylic acid
group at 1700 cm1 is observed at long PA exposure only, but
remains very weak. We associate this weak band with the formation of a small fraction of mono-carboxylates at defects and
vacancies in the adsorbate layer. The out-of-plane CH deformation mode at 743 cm1 is practically absent, which indicates
that the aromatic plane is oriented nearly perpendicularly to
the surface. Of special interest is the asymmetric OCO peak at
1550 cm1. Note that this band is absent in perfectly symmetric
carboxylates and that it increases in intensity with increasing
asymmetry of the OCO bridge. The relatively high intensity of
this feature indicates the formation of a strongly distorted carboxylate species.
The most surprising finding for this system is the exclusive
formation of the bis-carboxylate at room temperature, whereas
at low temperature the mono-carboxylate is preferred. The dependence on temperature implies that the bis-carboxylate is
formed in an activated process. Although the nature of this
process is not understood yet, we may speculate on its origin
based on the structural properties of the CoO(111) film. In the
previous section, we argued that the formation of the monocarboxylate is driven by steric hindrance due to the presence
of oxygen ions in the top layer. Following this argument, we
propose that the activated formation of the bis-carboxylate at
300 K is associated with a restructuring of the surface that fosters the accessibility of the second carboxylate group. This hypothesis is supported by the fact that the CoO(111) film is
known to undergo a structural phase transition close to room
temperature.[10b] The phase transition leads to a roughened
surface on which we would expect the formation of a more
strongly distorted bridging carboxylates species, as is indeed
observed in the experiment.

&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

PA adsorption on CoO(100)/Ir(100) at 130 K

Next, we investigated the growth of PA on CoO(100)/Ir(100) at
130 K by using the experimental procedures described above.
It needs to be stated again that the evaporation rate used
here was about 1.5 times higher than that used in the previous
experiments. The IR spectra are shown as a function of exposure time in Figure 7a, the IR spectra recorded of samples with
sub-monolayer and multilayer coverage are compared in Figure 7b, and the intensities of the carboxylate (1417 cm1) and
free carboxylic acid (1712 cm1) bands as a function of time
are plotted in Figure 7c.
For the peak assignments, again refer to Table 1. Similarly, as
discussed for PA deposition on Co3O4/Ir(100) and CoO(111)/
Ir(100) at low temperature, we observe the characteristic features of free PA, surface-bound carboxylates, and traces of
phthalic acid anhydride. We have subdivided the growth into
three regions (Figure 7). In region I (up to 15 min), a strong
symmetric OCO peak at 1417 cm1 appears and finally saturates, whereas the carboxylic acid band at 1712 cm1 is practically absent. This observation indicates a strong preference for
the formation of bis-carboxylates. The corresponding spectrum
in Figure 7b (upper trace) is particularly simple. Neither the
asymmetric OCO stretching mode of the carboxylate nor the
out-of-plane CH deformation mode of the organic ring is observed. This shows that the PA forms a nearly perfectly symmetric bridging bis-carboxylate adsorbed with the aromatic
ring oriented perpendicularly to the surface.
In region II (deposition times of 1530 min), the symmetric
OCO peak decreases whereas the C=O band of the free carboxylic acid increases with a similar slope. This suggests the
formation of a mono-carboxylate. The substantial loss of intensity of the OCO band in this region can be explained either by
the conversion of bis-carboxylates into mono-carboxylates or
by the dynamic coupling of the surface-bound PA to PA in the
multilayer. As we do not observe similar effects for the other
surfaces, however, we conclude that a substantial fraction of
the bis-carboxylates is converted into singly bound species. In
region III (beyond 30 min), we observe the linear growth of
a PA multilayer film. Some loss of the perpendicular alignment
in the growing multilayer film is indicated by the small out-ofplane peak at 743 cm1 at higher coverage (see Figure 7b).
The adsorption geometry of PA on CoO(100) at low coverage and low temperature is schematically shown in Figure 3c.
CoO(100) has a nonpolar surface terminated by both Co2 + and
O2 ions.[8] In contrast to Co3O4(111), the distance between the
cobalt ions at the surface is relatively small (3.0 ), which permits the formation of a bridging carboxylate. In contrast to
CoO(111), the Co2 + ions are easily accessible from a steric
point of view. Thus, there is little structural restriction of the
adsorption of anchoring molecules and the energetically most
favorable adsorption geometry, the bridging bis-bidentate
phthalate, can be realized. At high PA coverage, the formation
of a densely packed layer of mono-carboxylates becomes more
favorable and, as a result, one carboxylate group again detaches from the surface.

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper

Figure 7. a) IR spectra recorded during the deposition of PA on CoO(100)/

Ir(100) for 60 min at 130 K. b) Comparison of a sub-monolayer spectrum
(3 min of deposition) and a multilayer spectrum (60 min of deposition). c) Integrated peak intensities at 1417 cm1 (bound carboxylate) and
1712 cm1band (free carboxylic acid) as a function of deposition time.

Figure 8. a) IR spectra recorded during the adsorption and growth of PA on

CoO(100)/Ir(100) at 300 K. b) Comparison of the IR spectra recorded after
1 and 60 min of deposition. c) Integrated peak intensities at 1429 cm1
(bound carboxylate) and 1700 cm1 (free carboxylic acid) as a function of
deposition time.

PA adsorption on CoO(100)/Ir(100) at 300 K

(1700 cm1) bands are plotted as a function of time in Figure 8c.

The overall band intensities observed in Figure 8 a are compatible with the formation of a monolayer. In the first region
of growth (025 min), the intensity of the symmetric OCO
band (1429 cm1) increases and finally saturates. In this region
the C=O band of the free carboxylic acid (1700 cm1) peak is

Finally, we investigated the growth of PA on CoO(100)/Ir(100)

at 300 K by using the experimental procedures described
above. The IR spectra as a function of exposure time are
shown in Figure 8a, the IR spectra of sub-monolayer and multilayer coverage are compared in Figure 8b, and the intensities
of the carboxylate (1429 cm1) and free carboxylic acid
Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper
nearly invisible, which indicates that the PA adsorbs nearly exclusively in the form of a bis-carboxylate. The symmetric OCO
stretching band of the carboxylate at 1429 cm1 is the most
prominent feature throughout the deposition (see Table 1 for
details). As for the deposition at 130 K, the intensities of the
carboxylate mode at 1547 cm1 and the out-of-plane CH deformation mode remain weak, which indicates the formation of
a rather symmetric bridging bis-carboxylate perpendicular to
the surface.
In region II (from 25 min), no further change is observed in
the intensity of the symmetric OCO mode. However, a small
C=O peak originating from free carboxylic acid groups grows
in a stepwise fashion for deposition times between 25 and
35 min. The appearance of the C=O peak indicates the formation of a small fraction of mono-carboxylate. We assign this
effect to adsorption at defect sites on the surface or in the adsorbate layer; the two carboxylic acid groups of PA are unable
to bind under these circumstances. Note, we do not observe
the conversion of the bis-carboxylate into the mono-carboxylate, as was found, for example, for PA on Co3O4(111) or
CoO(100) at low temperature. We suggest that the strong
binding of PA to the oxide surface in the form of a bridging
bis-carboxylate is also associated with a higher activation barrier for its conversion into the mono-carboxylate. As a result, the
desorption of molecularly adsorbed PA is preferred and the formation of the mono-carboxylate is kinetically hindered.
Discussion of the structure dependency of PA adsorption on
cobalt oxide surfaces

Figure 9. a) Comparison of the IR spectra of a sub-monolayer and multilayer

of PA on Co3O4(111), CoO(111), and CoO(100) at 130 K. b) Comparison of the
integrated peak intensities of the carboxylate band (Ins(OCO)), the free acid
band (In(C=O)), the sum of the carboxylate and free acid, and the intensity
ratio Ins(OCO)/(In(C=O) + Ins(OCO)).

Finally, we compare the interaction and growth of PA on the

three different cobalt oxide surfaces and discuss the possible
role of the surface structure of Co3O4(111), CoO(111), and
CoO(100). A direct comparison of selected spectra at low and
high coverages and of the evolution of intensity of the carboxylate and acid bands are shown for the three surfaces in
Figure 9. The integrated peak areas of the carboxylate (Ins(OCO))
and acid (In(C=O)) bands as well as the intensity ratio Ins(OCO)/(In(C=
O) + Ins(OCO)) are shown in Figure 9b as a function of exposure
time. Because of the different amplitudes and orientations of
the dynamic dipole moments, the intensity ratio does not
quantitatively reflect the relative abundance of the species,
but allows us to follow the coverage-dependent trends in
a qualitative fashion.
At low temperature we can clearly differentiate between
mono- and multilayer growth (Figure 9a). For both Co3O4(111)
and CoO(100) we observe the formation of a bis-carboxylate at
low coverage, that is, a phthalate species that binds to the surface through both carboxylic acid groups. For both surfaces,
the low intensity of the asymmetric OCO stretching mode indicates that the carboxylate groups are practically undistorted,
and the low intensity of the CH out-of-plane deformation
mode indicates that the aromatic ring is oriented nearly perpendicular to the surface. Also, the band positions are very
similar. Inspection of the surface structure suggests, however,
that the binding geometries must be quite different: For
Co3O4(111) with a large CoCo distance of 5.7 , a chelating
&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

bis-carboxylate must be formed, whereas the formation of

a bridging bis-carboxylate is possible for the CoO(100) surface,
which has a CoCo distance of only 3.0 .
For longer PA exposure times we observe the competing
formation of mono-carboxylates and the conversion of bis-carboxylates into mono-carboxylates (Figure 9b). The latter process is most pronounced on CoO(100). We speculate that the
high density of surface Co2 + ions and their good accessibility
provide the flexibility that is required to facilitate this restructuring process.
A completely different adsorption behavior is observed for
CoO(111). In this case the formation of mono-carboxylates is
observed from the lowest coverages. This is reflected by the
Ins(OCO)/(Ins(OCO) + Ins(OCO)) ratio at low coverage, which is much
smaller than for the other two surfaces (Figure 9b). Although
the density of the Co2 + ions is very similar in CoO(100) and
CoO(111), the main difference between the two surfaces is the
fact that in the latter the Co2 + ions are located slightly below
the terminating layer of O2 ions. We assume that the oxygen
ions restrict the accessibility of the surface cobalt ions, thereby
enforcing a molecular orientation that does not permit the
second carboxylic acid group to interact with the surface. Also,
10

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper
the asymmetric stretching mode of the carboxylate is found to
be more pronounced for CoO(111), which indicates that a substantially distorted bridging carboxylate is formed.
Clear structure dependencies are also observed for the adsorption of PA at room temperature (Figure 10). At 300 K, only
chemisorbed monolayer films are stable. Interestingly, the
structure dependencies and interaction mechanisms are different to those at low temperature. On Co3O4(111) and CoO(100),
bis-carboxylates are formed exclusively during the initial stages
of deposition. Based on the surface structures of the films, we
again assume the formation of chelating species on Co3O4(111)
and of bridging species on CoO(100). Upon prolonged dosing,
the bis-carboxylates on Co3O4(111) are slowly converted into
mono-carboxylates. The mechanism of this process is not yet
clear. A possible explanation involves the slow desorption of
more weakly bound chelating species, the re-adsorption of PA,
and, finally, the formation of mono-carboxylates at vacancies in
the PA film. On the CoO(100) surface, on the other hand,
a very stable bridging bis-carboxylate is formed. Even upon
prolonged PA exposure, only a very small fraction of mono-carboxylate species is formed on this film, which we attribute to
adsorption at defects in the film, which do not accommodate
doubly linked bis-carboxylates.

The most surprising observation is the temperature-dependent change in the anchoring mechanism observed for the
CoO(111) surface. Although we find mono-carboxylates at low
temperature, only bis-carboxylates are formed at room temperature. This observation implies that there is a substantial activation barrier for the formation of the bis-carboxylate species.
As we would not expect such a high barrier for a simple deprotonation reaction, it appears likely that bis-carboxylate formation requires a restructuring of the surface, thereby making
the Co2 + ions more easily accessible for the carboxylate
groups. Indeed, a structural phase transition has been observed in CoO(111) and the high intensity of the asymmetric
OCO mode supports the proposal of a strongly distorted bridging carboxylate on a more corrugated surface.

Conclusion
We have studied the adsorption and growth of phthalic acid
deposited by physical vapor deposition (PVD) onto well-ordered Co3O4(111), CoO(111), and CoO(100) thin films on Ir(100).
We have monitored the interfacial reactions in situ during film
growth by using time-resolved IRAS under UHV conditions.
The observations are schematically illustrated in Figure 11 and
are summarized as follows.

PA on Co3O4(111)
At low temperature (130 K) and low coverage, PA binds to the
Co3O4(111) surface forming a chelating bis-carboxylate with the
molecular plane oriented perpendicularly to the surface. With
increasing coverage, the formation of a chelating mono-carboxylate is preferred, before multilayers of physisorbed PA
start to grow on top of the chemisorbed monolayer. Upon
PVD of PA, a small fraction of phthalic acid anhydride is
formed that is co-deposited on all three oxide surfaces.
At 300 K, only a chemisorbed monolayer of PA is stable on
Co3O4(111). At low exposure the PA binds to the oxide surface
as a chelating bis-carboxylate. On continuing exposure, the
chelating bis-carboxylate is slowly replaced by an increasing
fraction of chelating mono-carboxylate.

PA on CoO(100)
At low temperature (130 K) and low coverage, PA binds to the
CoO(100) surface in a bridging bis-carboxylate mode with the
molecular plane oriented perpendicularly to the surface. With
increasing coverage, the bridging bis-carboxylate species is
converted into a bridging mono-carboxylate before finally the
PA multilayer starts to grow. Upon deposition at 300 K, a very
stable monolayer of bridging bis-carboxylates is formed. Even
upon prolonged exposure to PA, only a small fraction of
mono-carboxylate is co-adsorbed, possibly onto defect sites in
the film and on the surface.

Figure 10. a) Comparison of the IR spectra of a sub-monolayer and saturated

monolayer of PA on Co3O4(111), CoO(111), and CoO(100) at 300 K. b) Comparison of the integrated peak intensities of the carboxylate band (Ins(OCO)),
the free acid band (In(C=O)), the sum of the carboxylate and free acid, and the
intensity ratio Ins(OCO)/(In(C=O) + Ins(OCO)).
Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

11

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper

Figure 11. Schematic summary of binding geometries of PA on the three cobalt oxide surfaces at different temperatures and coverages.

Acknowledgements

PA on CoO(111)
For PA on CoO(111), a temperature-dependent change in the
adsorption mechanism was observed. At low temperature
(130 K) and low coverage, PA binds to CoO(111) in the form of
a mono-carboxylate with a distorted bridging geometry and its
molecular plane oriented nearly perpendicular to the surface.
At 300 K, the interaction mechanism changes and PA is adsorbed in the form of a strongly distorted bis-carboxylate. We
attribute this effect to a temperature-dependent restructuring
of the surface. The bis-carboxylate film is stable and only small
fractions of mono-carboxylates are co-adsorbed upon prolonged exposure to PA.
These results show that the adsorption geometry and adsorption kinetics of PA on cobalt oxide are strongly dependent
upon the surface structure. This structure dependence can be
rationalized on the basis of the density and accessibility of the
cations at the oxide surface. The results show that the structure of the surface is important for the anchoring of molecular
films. Both a detailed understanding of the interaction mechanisms and a high level of control of the structure of the surface
is essential to tailor hybrid interfaces between organic films
and oxide surfaces.
&

&

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

This project was financially supported by the Deutsche Forschungsgemeinschaft (DFG) within the Research Unit FOR
1878 funCOS Functional Molecular Structures on Complex
Oxide Surfaces. Additional support is acknowledged from the
Excellence Cluster Engineering of Advanced Materials within
the framework of the excellence initiative. The authors also acknowledge additional support by the Deutsche Forschungsgemeinschaft within the DACH Project COMCAT, the European
Commission (chipCAT, Grant Agreement No. 310191), and
COST Action CM1104 Reducible oxide chemistry, structure and
functions. T.X. gratefully acknowledges support by way of
a Ph. D. grant from the China Scholarship Council (CSC).
Keywords: adsorption carboxylic acids cobalt IR
spectroscopy thin films

[1] a) J. S. Kim, J. H. Park, J. H. Lee, J. Jo, D. Y. Kim, K. Cho, Appl. Phys. Lett.
2007, 91, 112111; b) K.-J. Jiang, K. Manseki, Y.-H. Yu, N. Masaki, K. Suzuki,
Y.-l. Song, S. Yanagida, Adv. Funct. Mater. 2009, 19, 2481 2485; c) O. K.
Varghese, M. Paulose, C. A. Grimes, Nat. Nanotechnol. 2009, 4, 592 597.

12

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

These are not the final page numbers!

Full Paper
[2] a) H. Klauk, U. Zschieschang, J. Pflaum, M. Halik, Nature 2007, 445, 745
748; b) E. C. P. Smits, S. G. J. Mathijssen, P. A. van Hal, S. Setayesh, T. C. T.
Geuns, K. A. H. A. Mutsaers, E. Cantatore, H. J. Wondergem, O. Werzer, R.
Resel, M. Kemerink, S. Kirchmeyer, A. M. Muzafarov, S. A. Ponomarenko,
B. de Boer, P. W. M. Blom, D. M. de Leeuw, Nature 2008, 455, 956 959.
[3] a) Y. Joseph, M. Whn, A. Niklewski, W. Ranke, W. Weiss, C. Wll, R.
Schlgl, Phys. Chem. Chem. Phys. 2000, 2, 5314 5319; b) S. C. Li, J. G.
Wang, P. Jacobson, X. Q. Gong, A. Selloni, U. Diebold, J. Am. Chem. Soc.
2009, 131, 980 984; c) P. Rahe, M. Nimmrich, A. Nefedov, M. Naboka, C.
Woll, A. Kuhnle, J. Phys. Chem. C 2009, 113, 17471 17478; d) J. S. Prauzner-Bechcicki, S. Godlewski, A. Tekiel, P. Cyganik, J. Budzioch, M. Szymonski, J. Phys. Chem. C 2009, 113, 9309 9315; e) L. M. Liu, S. C. Li,
H. Z. Cheng, U. Diebold, A. Selloni, J. Am. Chem. Soc. 2011, 133, 7816
7823; f) P. Rahe, M. Kittelmann, J. L. Neff, M. Nimmrich, M. Reichling, P.
Maass, A. Khnle, Adv. Mater. 2013, 25, 3948 3956; g) S. Godlewski, M.
Szymonski, Int. J. Mol. Sci. 2013, 14, 2946 2966.
[4] a) J. Libuda, F. Winkelmann, M. Bumer, H. J. Freund, T. Bertrams, H.
Neddermeyer, K. Mller, Surf. Sci. 1994, 318, 61 73; b) H.-J. Freund, H.
Kuhlenbeck, V. Staemmler, Rep. Prog. Phys. 1996, 59, 283; c) H.-J.
Freund, G. Pacchioni, Chem. Soc. Rev. 2008, 37, 2224 2242; d) M.
Sobota, I. Nikiforidis, W. Hieringer, N. Paape, M. Happel, H.-P. Steinrck,
A. Grling, P. Wasserscheid, M. Laurin, J. r. Libuda, Langmuir 2010, 26,
7199 7207.
[5] a) J. Jansson, J. Catal. 2000, 194, 55 60; b) H. K. Lin, H. C. Chiu, H. C.
Tsai, S. H. Chien, C. B. Wang, Catal. Lett. 2003, 88, 169 174; c) F. Jiao, H.
Frei, Angew. Chem. Int. Ed. 2009, 48, 1841 1844; Angew. Chem. 2009,
121, 1873 1876; d) B. Solsona, I. Vazquez, T. Garcia, T. E. Davies, S. H.
Taylor, Catal. Lett. 2007, 116, 116 121.
[6] a) H. Yoshikawa, K. Hayashida, Y. Kozuka, A. Horiguchi, K. Awaga, S.
Bandow, S. Iijima, Appl. Phys. Lett. 2004, 85, 5287 5289; b) G. X. Wang,
H. Liu, J. Horvat, B. Wang, S. Z. Qiao, J. Park, H. Ahn, Chem. Eur. J. 2010,
16, 11020 11027.
[7] a) X. W. Xie, W. J. Shen, Nanoscale 2009, 1, 50 60; b) K. Song, E. Cho,
Y. M. Kang, Acs Catal. 2015, 5, 5116 5122.
[8] K. Heinz, L. Hammer, J. Phys. Condens. Matter 2013, 25, 173001.
[9] W. Meyer, K. Biedermann, M. Gubo, L. Hammer, K. Heinz, J. Phys. Condens. Matter 2008, 20, 265011.
[10] a) W. Meyer, D. Hock, K. Biedermann, M. Gubo, S. Muller, L. Hammer, K.
Heinz, Phys. Rev. Lett. 2008, 101, 016103; b) W. Meyer, K. Biedermann, M.
Gubo, L. Hammer, K. Heinz, Phys. Rev. B 2009, 79, 4.
[11] a) T. Gillich, E. M. Benetti, E. Rakhmatullina, R. Konradi, W. Li, A. Zhang,
A. D. Schlter, M. Textor, J. Am. Chem. Soc. 2011, 133, 10940 10950;

Chem. Eur. J. 2016, 22, 1 14

www.chemeurj.org

These are not the final page numbers!

[12]
[13]
[14]
[15]
[16]

[17]
[18]

[19]
[20]

[21]
[22]
[23]

b) G. Guerrero, P. H. Mutin, A. Vioux, Chem. Mater. 2001, 13, 4367 4373;

c) C. Xu, K. Xu, H. Gu, R. Zheng, H. Liu, X. Zhang, Z. Guo, B. Xu, J. Am.
Chem. Soc. 2004, 126, 9938 9939.
S. P. Pujari, L. Scheres, A. T. M. Marcelis, H. Zuilhof, Angew. Chem. Int. Ed.
2014, 53, 6322 6356; Angew. Chem. 2014, 126, 6438 6474.
F. Hoffmann, Surf. Sci. Rep. 1983, 3, 107.
M. Buchholz, Q. Li, H. Noei, A. Nefedov, Y. Wang, M. Muhler, K. Fink, C.
Wll, Top. Catal. 2015, 58, 174 183.
M. Buchholz, M. Xu, H. Noei, P. Weidler, A. Nefedov, K. Fink, Y. Wang, C.
Wll, Surf. Sci. 2016, 643, 117 123.
a) I. Cebula, C. Shen, M. Buck, Angew. Chem. Int. Ed. 2010, 49, 6220
6223; Angew. Chem. 2010, 122, 6356 6360; b) C. Shen, I. Cebula, C.
Brown, J. L. Zhao, M. Zharnikov, M. Buck, Chem. Sci. 2012, 3, 1858
1865.
A. Desikusumastuti, T. Staudt, M. Happel, M. Laurin, J. Libuda, J. Catal.
2008, 260, 315 328.
a) L. Colombo, V. Volovsek, M. Lepostollec, J. Raman Spectrosc. 1984, 15,
252 256; b) B. G. Frederick, M. R. Ashton, N. V. Richardson, T. S. Jones,
Surf. Sci. 1993, 292, 33 46; c) B. G. Frederick, F. M. Leibsle, S. Haq, N. V.
Richardson, Surf. Rev. Lett. 1996, 03, 1523 1546; d) K. D. Dobson, A. J.
McQuillan, Spectrochim. Acta Part A 2000, 56, 557 565; e) J. van den
Brand, O. Blajiev, P. C. J. Beentjes, H. Terryn, J. H. W. de Wit, Langmuir
2004, 20, 6308 6317; f) J. F. Arenas, J. I. Marcos, Spectrochim. Acta Part
A 1980, 36, 1075 1081; g) S. Haq, R. C. Bainbridge, B. G. Frederick, N. V.
Richardson, J. Phys. Chem. B 1998, 102, 8807 8815.
T. Xu, S. Mohr, M. Amende, M. Laurin, T. Dopper, A. Gorling, J. Libuda, J.
Phys. Chem. C 2015, 119, 26968 26979.
NIST Mass Spec Data Center, S.E. Stein, director, Infrared Spectra in
NIST Chemistry WebBook, NIST Standard Reference Database Number 69,
Eds. P.J. Linstrom and W.G. Mallard, National Institute of Standards and
Technology, Gaithersburg MD, 20899, http://webbook.nist.gov, (retrieved
February 22, 2016). .
B. E. Hayden, A. King, M. A. Newton, J. Phys. Chem. B 1999, 103, 203
208.
M. Xu, H. Noei, M. Buchholz, M. Muhler, C. Wll, Y. Wang, Catal. Today
2012, 182, 12 15.
W. O. Gordon, Y. Xu, D. R. Mullins, S. H. Overbury, Phys. Chem. Chem.
Phys. 2009, 11, 11171 11183.

Received: November 30, 2015

Published online on && &&, 0000

13

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper

FULL PAPER
& Carboxylic Acid Adsorption
T. Xu, M. Schwarz, K. Werner, S. Mohr,
M. Amende, J. Libuda*
&& &&
Structure-Dependent Anchoring of
Organic Molecules to Atomically
Defined Oxide Surfaces: Phthalic Acid
on Co3O4(111), CoO(100), and
CoO(111)

&

&

Chem. Eur. J. 2016, 22, 1 14

Surface structure matters: The adsorption geometry of phthalic acid is dependent on the atomic structure of the
cobalt oxide surface. The structure dependence can be rationalized by the

different arrangements of the cobalt

ions at the three surfaces (see figure;
IRAS = IR reflection absorption spectroscopy).

14

 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

These are not the final page numbers!