Temperature-dependent electronic and vibrational structure of the 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide room-temperature ionic liquid surface: a study with XPS, UPS, MIES, and HREELS.
ABSTRACT The near-surface structure of the room-temperature ionic liquid 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide has been investigated as a function of temperature between 100 and 620 K. We used a combination of photoelectron spectroscopies (XPS and UPS), metastable induced electron spectroscopy (MIES), and high-resolution electron energy loss spectroscopy (HREELS). The valence band and HREELS spectra are interpreted on the basis of density functional theory (DFT) calculations. At room temperature, the most pronounced structures in the HREELS, UPS, and MIES spectra are related to the CF3 group in the anion. Spectral changes observed at 100 K are interpreted as a change of the molecular orientation at the outermost surface, when the temperature is lowered. At elevated temperatures, early volatilization, starting at 350 K, is observed under reduced pressure.
- [show abstract] [hide abstract]
ABSTRACT: We apply electron spectroscopy methods with different surface sensitivities to elucidate the DOS of the surface and the near-surface region of [XMIm]Cl (X=octyl, hexyl, butyl, and ethyl alkyl chain) ionic liquids. Using metastable induced electron spectroscopy (MIES) we are able to detect the density of states in front of the outermost surface, whereas ultraviolet and X-ray photoelectron spectroscopy (UPS and XPS) measurements provide lower surface sensitivity. The assignment of certain structures in the valence band spectra to particular atoms/functional groups of the ionic liquid based on DFT calculations and the reconstruction of PES spectra enables us to obtain information on the dominating groups at the surface, or in other words, on the molecular/ionic arrangement and orientation at the surface. From angular resolved XPS it is concluded that the alkyl chains dominate at the outermost surface. In agreement with this a decreasing chlorine signal is observed in the UPS spectra for ionic liquids with increasing alkyl chain length. The analysis of the MIES data shows that in case of [OMIm]Cl--in contrast to UPS and XPS--no Cl-induced features are visible in the MIES spectra at all and that the MIES spectra are dominated by the [OMIm](+) alkyl chain.ChemPhysChem 04/2012; 13(7):1718-24. · 3.35 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Differences in the surface structure of protic ionic liquids (ILs) with three different cations and a common anion; ethyl-, propyl- and 2-hydroxyethyl- (or ethanol-) ammonium nitrate (EAN, PAN and EtAN, respectively) have been observed by neutral impact collision ion scattering spectroscopy (NICISS) and metastable induced electron spectroscopy/ultraviolet photoelectron spectroscopy (MIES/UPS). NICISS is used to determine the concentration depth profiles of the elements in each IL and it reveals an enrichment of cation alkyl chains of PAN and EtAN in the outermost layer compared to EAN, and a corresponding depletion of nitrate from the outermost layer of the EtAN surface. MIES probes the molecular orbitals of only the species in the outermost layer of a sample and confirms that, while both the anion and the cation are present to some degree at the surface of all three ILs, the cation is enriched to a greater extent at the surface of PAN and EtAN compared to EAN.Physical Chemistry Chemical Physics 10/2012; · 3.83 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Ultraviolet photoelectron spectroscopy (UPS), work function measurements, low energy electron diffraction (LEED) and scanning tunnelling microscopy (STM) have been used to study the adsorption and desorption of 1-ethyl-3-methylimidazolium bis[(trifluoromethyl)sulfonyl]imide, [C(2)C(1)Im][Tf(2)N], on the (1×2) clean surface reconstruction of Au(110) in the temperature range 100-674 K. The ionic liquid adsorbed without decomposition, and desorbed without leaving any residue on the surface. For adsorption at room temperature a monolayer of strongly bound ionic liquid was formed with four interface states visible in UP spectra. STM at 100 K showed that the monolayer consisted of well-ordered rows of adsorbed ionic liquid aligned parallel to the close packed rows of surface gold atoms (the  direction) with a separation of ×2 (the same as the clean surface reconstruction) between the rows in the orthogonal  direction. Multilayer adsorption at room temperature occurred by droplet formation followed by smoothing of the droplets to a layered morphology with time. Heating caused multilayer desorption at temperatures in the 363-383 K range, followed by partial monolayer desorption at 548 K to produce a Au(110)-(1×3) reconstructed surface with sub-monolayer domains of ionic liquid. Desorption of the remaining ionic liquid at 600 K caused the gold surface to reconstruct back to the clean (1×2) reconstruction.Physical Chemistry Chemical Physics 03/2012; 14(17):6054-66. · 3.83 Impact Factor
Temperature-Dependent Electronic and Vibrational Structure of the
1-Ethyl-3-methylimidazolium Bis(trifluoromethylsulfonyl)amide Room-Temperature Ionic
Liquid Surface: A Study with XPS, UPS, MIES, and HREELS†
S. Krischok,*,‡M. Eremtchenko,‡,|M. Himmerlich,‡P. Lorenz,‡J. Uhlig,‡A. Neumann,‡
R. O 2 ttking,‡W. J. D. Beenken,‡O. Ho 1fft,§S. Bahr,§V. Kempter,§and J. A. Schaefer‡
Institut fu ¨r Physik and Institut fu ¨r Mikro- und Nanotechnologien, Technische UniVersita ¨t Ilmenau,
P.O. Box 100565, D-98684 Ilmenau, Germany, and Institut fu ¨r Physik und Physikalische Technologien,
Technische UniVersita ¨t Clausthal, Leibnizstr. 4, D-38678 Clausthal-Zellerfeld, Germany
ReceiVed: October 30, 2006; In Final Form: March 3, 2007
The near-surface structure of the room-temperature ionic liquid 1-ethyl-3-methylimidazolium bis(trifluoro-
methylsulfonyl)amide has been investigated as a function of temperature between 100 and 620 K. We used
a combination of photoelectron spectroscopies (XPS and UPS), metastable induced electron spectroscopy
(MIES), and high-resolution electron energy loss spectroscopy (HREELS). The valence band and HREELS
spectra are interpreted on the basis of density functional theory (DFT) calculations. At room temperature, the
most pronounced structures in the HREELS, UPS, and MIES spectra are related to the CF3 group in the
anion. Spectral changes observed at 100 K are interpreted as a change of the molecular orientation at the
outermost surface, when the temperature is lowered. At elevated temperatures, early volatilization, starting at
350 K, is observed under reduced pressure.
Room-temperature ionic liquids (RT-ILs) have attracted much
attention for their excellent properties, namely they remain in
liquid phase over a wide temperature range, have very low vapor
pressure at RT, are chemically inert, and possess high heat
capacities.1,2These properties make them good candidates for
use in many fields.2,3They may be used as “green” solvents,2
heat reservoirs,4as well as electrolytes in electrochemical appli-
cations,5homogeneous catalysis,2,6,7and dye sensitized solar
cells.8In order to refine the performance of RT-ILs, detailed
knowledge of the geometric, electronic, and vibrational structure
of RT-ILs as a function of temperature appears indispensable.
The need for vacuum-based studies employing surface analytical
techniques has been pointed out recently.9So far, only a few
of these studies have appeared very recently.1,10-15
A typical representative of RT-ILs is 1-ethyl-3-methylimi-
dazolium bis(trifluoromethylsulfonyl)amide ([EMIM][Tf2N]).16,17
In two previous studies, we have applied ultraviolet and X-ray
photoelectron spectroscopy (UPS (He I and II), XPS) as well
as metastable induced electron spectroscopy (MIES),13partially
supported by density functional theory (DFT) calculations,14to
the study of the electronic structure of [EMIM][Tf2N] at room
temperature. In the present study, we have investigated the
surface-near structure using the above-mentioned techniques as
a function of temperature. In addition, we report the temperature-
and depth-dependence of the vibrational spectrum for the
[EMIM][Tf2N] film at 100 and 300 K as measured with high-
resolution electron energy loss spectroscopy (HREELS). We
utilize quantum-chemical calculations based on DFT in order
to analyze the electronic and vibrational spectra.
2. Experimental and Theoretical Methods
2.1. Sample Preparation. A polycrystalline Au film of ∼250
nm thickness deposited on Si(100), separated by a Ti adhesion
layer, serves as sample support. The samples were prepared by
depositing one droplet of the ultrapure RT-IL onto the Au
substrate, and were, after careful outgassing in a load lock
system, introduced into the UHV chamber. The prepared
[EMIM][Tf2N] samples have very low vapor pressure, and no
change of the base pressure could be detected during the
measurements performed at RT. After the introduction of the
RT-IL into the UHV chamber, the samples were characterized
by XPS. The survey spectra indicate that a closed RT-IL film
was successfully prepared. From the absence of any substrate
(Au) related spectral feature in the XPS spectra, we estimate a
film thickness of more than 10 nm.
Since the ionic liquid can be damaged by photons and
electrons, special attention has been paid to minimize the
irradiation time. In both the HREELS and the combined MIES/
UPS measurements, the applied flux was kept rather low.
Moreover, we have carefully checked that the spectra taken after
warming up to RT are virtually identical to the original data
obtained at RT. Thus, we tend to exclude beam damage to be
responsible for the observed spectral changes. Auxiliary experi-
ments show that metastable atoms, low-energy photons, and
electrons with an kinetic energy of ∼10 eV cause, in contrast
to high-energy photons and electrons (>1keV), only minor
changes in the valence band and vibrational spectra, even if
comparably high fluxes are applied.
2.2. Experimental Methods. This study is a continuation of
our previous activities.13,14Briefly, it combines photoelectron
spectroscopy (ultraviolet photoelectron spectroscopy, UPS (He
I and II) and monochromated X-ray photoelectron spectroscopy,
XPS (Al KR)), metastable induced electron spectroscopy
(MIES), and high-resolution electron energy loss spectroscopy
(HREELS) performed in different UHV systems (base pressure
†Part of the special issue “Physical Chemistry of Ionic Liquids”.
* Corresponding author. Tel.: +49-3677-69-3405. Fax: +49-3677-69-
3365. E-mail address: firstname.lastname@example.org.
‡Technische Universita ¨t Ilmenau.
§Technische Universita ¨t Clausthal.
J. Phys. Chem. B 2007, 111, 4801-4806
10.1021/jp067136p CCC: $37.00© 2007 American Chemical Society
Published on Web 05/03/2007
below 2 × 10-10Torr). All chambers are equipped with a LN2-
cooling system enabling measurements down to approximately
100 K. Our monochromated XPS measurements feature an
energy resolution below 0.6 eV (Ag(3d5/2) at 15 eV pass energy),
whereas a resolution below 250 meV was employed for all UPS
and MIES measurements presented here. For HREELS a Delta
0.5 spectrometer (Ibach type) was applied, having an intrinsic
energy resolution of ∼40 cm-1(5 meV) for the device settings
used in the presented experiments. A more detailed introduction
into the experimental setups can be found elsewhere.13,14
Special attention must be paid to avoid charge-up effects,
well-known for inorganic salts, during the application of any
electron spectroscopies. For this reason, the probe beam currents
were chosen low enough to avoid features attributable to typical
charging phenomena (see below). Special care is required when
performing measurements at reduced temperatures where the
conductivity of the IL is reduced drastically (see sections 3.1
2.3. DFT Calculations. In order to interpret the experimental
results, we have performed electronic structure and vibrational
frequency calculations for an [EMIM][Tf2N] ion pair as well
as for the individual ions using DFT as implemented in Gaussian
03.18We utilized Becke’s B3LYP functional19and a 6-311G
basis set with (3d, 2pd) polarization functions. We analyze the
molecular Kohn-Sham orbitals of all states contributing to the
electronic density of states (DOS) in the energy range accessible
to UPS and MIES. Though Koopman’s theorem is not valid
for DFT, the DOS of occupied molecular states, as measured
by UPS and MIES, is well determined for Janak’s theorem20
and the hybrid character of the B3LYP functional, which
partially includes the Hartree-Fock exchange term.21-24
In order to simulate the HREELS spectra both IR-dipoles
and Raman-activities for the electronic ground state are calcu-
lated. Here we analyzed the respective normal modes. Details
to selection rules may be found in ref 25.
3.1. Core Level Spectra (XPS). In Figure 1, the C(1s) and
N(1s) emission as observed by XPS are presented as a function
of temperature. At room temperature, the C(1s), N(1s), O(1s),
F(1s), and S(2s) emissions display well-resolved and narrow
spectral features. Neither emission from the substrate nor from
unexpected elements could be detected. Furthermore, the relative
peak areas agree well with the stoichiometry of the molecule.
It is important to note that the data presented in Figure 1 have
been taken under rising temperature. From the similarity of the
RT spectrum with spectra directly taken at RT after introducing
the film into UHV (see ref 13), we conclude that even in the
presented XPS spectra beam damage plays a minor role. A
detailed discussion of the XPS spectra typical for the [EMIM]-
[Tf2N] film at room temperature can be found elsewhere.13At
lower temperatures, the following changes in the spectra are
observed. Between 200 K and room temperature, the structures
are narrow and no significant change in their line width is
registered, whereas at temperatures between 150 and 200 K, a
broadening of all core level peaks is observed, which prevents
the separation of the different chemical states (see Figure 1).
Furthermore, a reduction of the kinetic energy of all core levels
toward lower values is observed at low temperatures (up to 10
eV at 120 K).
We attribute this behavior to charging of the sample surface
due to a reduction of the IL film conductivity when lowering
the temperature.12,26As an example, the behavior of [EMIM]
EtSO4during cooling below the solidification point has been
studied with XPS and TOF-SIMS.12Upon solidification, the
material behaved as an electrical insulator, and charge compensa-
tion was required to allow for spectra to be recorded. In this
context, it should be further mentioned that the [EMIM][Tf2N]
undergoes a phase transition in the investigated temperature
3.2. Valence Band Spectra (UPS, MIES). Room-tempera-
ture valence band spectra of [EMIM][Tf2N] obtained using vari-
ous electron spectroscopy techniques can be found in refs 13
and 14. Upon variation of information depth using different exci-
tation sources, no significant spectral changes can be observed.14
At low temperatures, strong changes in the MIES and UPS
spectra are observed. The changes of the valence band spectra
occurring during temperature reduction to 127 K were reported
by us in ref 13. In contrast to our XPS measurements, we tend
to exclude charging during the application of MIES and UPS
(with low intensities as provided by our MIES/UPS-source),
because low intensities of the probe beams are employed. Fur-
thermore, none of the features typical for charging are observed.
In particular, no shift of any of the spectral features occurs nor
do they become smeared out. Furthermore, a continuous
decrease of the onset of the spectra at low kinetic energies would
be seen, which is not the case. Nevertheless, we will base our
discussion of the low temperature behavior mainly on our
HREELS results in combination with DFT calculations.
The MIES and UPS results from annealing the film up to
620 K are presented in Figure 2. Above ∼350 K, the spectra
change with increasing temperature. We notice an intensity
decrease of the double peak at ∼11 eV. Additionally, together
Figure 1. Temperature dependence of the [EMIM][Tf2N] core level
spectra upon cooling measured using monochromated Al KR radiation;
(a) C(1s) and (b) N(1s). For visualization the spectra are displayed in
binding energy, where the position of the spectra was shifted to the
energy observed at 300 K.
4802 J. Phys. Chem. B, Vol. 111, No. 18, 2007
Krischok et al.
with the disappearance of the structure at ∼11 eV, the structure
at ∼6 eV changes, too.
Above 500 K, the broad structure in UPS, centered around 7
eV, loses intensity and three sharp features at ∼6, ∼4, and ∼3
eV, respectively, start to grow continuously with increasing
temperature. These structures are well-known for an Au surface.
In the same temperature range, the structures typical for [EMIM]-
[Tf2N] disappear in MIES, and the spectra become structureless.
The observed MIES spectra are typical for a high work function
metal such as Au.
3.3. Vibrational Spectra (HREELS). In order to get more
information on eventual structural changes occurring during
cooling, we have applied vibrational spectroscopy. In Figure 3,
HREELS spectra of [EMIM][Tf2N] are shown at 300 and 100
K sample temperature, respectively. For both temperatures,
spectra were collected at different primary beam energies E0
(2.5, 20, and 80 eV). At room temperature (Figure 3a), besides
the elastic peak, loss structures are visible at ∼580 cm-1(marked
as I in Figure 3), ∼1200 cm-1(II), ∼2400 cm-1(2 × 1200
cm-1), and ∼2980 cm-1(III) in the investigated spectral range
(0-4000 cm-1). These main features are clearly visible for all
used primary electron energies. Some less pronounced and not
as well resolved structures (e.g. at 1400 , 2500, and 3100 cm-1)
are present in the spectra, too. The observed FWHM of the
elastic peak decreases from 425 cm-1for a primary beam energy
of E0 ) 2.5 eV to 220 cm-1for E0 ) 80 eV. Notably, the
resolution of the applied device settings (40 cm-1) is much better
than the observed width of the elastic peak.
At 100 K, the spectra show characteristic changes as
compared to the room temperature measurements (see Figure
3). All peaks, the elastic and all observed loss features, narrow.
However, they are still considerably wider than the intrinsic
resolution of the apparatus. The observed changes of the width
of the elastic peak (FWHM) with primary beam energy (135
cm-1at E0) 2.5 eV and 160 cm-1at E0) 20 eV) are less
pronounced. The observed loss structures at 100 K are more
complex compared to the spectra taken at 300 K, and the spectra
now change drastically with beam energy. Compared to the
HREELS spectra at 300 K, in the spectra at 100 K additional
loss structures at 810 (IV), 990 (V), 1430 (VI), 2120 (VII), and
2350 cm-1(VIII) are now visible. These new features are
most pronounced for low primary beam energies. With increas-
ing primary beam energy the spectra start to resemble those for
3.4. Electronic DOS Obtained by DFT. For the interpreta-
tion of the valence band spectra, we employ quantum-chemical
DFT calculations of the electronic DOS for the single ions
[EMIM]+and [Tf2N]-as well as for the [EMIM][Tf2N] ion
pair complex. A detailed analysis of the corresponding Kohn-
Sham molecular orbitals (MO) has been made previously.14In
Figure 4a, the total electronic DOS is compared with the UPS
and MIES measurements carried out at RT. For optimal
comparison with experiment, a Gaussian broadening of 0.8 eV
(FWHM) and a shift of 1.35 eV toward lower binding energies
was applied to the DOS. Good qualitative agreement exists with
the energetic positions of most features seen in the spectra
(except for the spectral region around 14 eV, which is strongly
influenced by secondary electron emission). A quantitative fit
of the shape of the spectra cannot be expected because the
Figure 2. Temperature dependence of the valence band upon heating
measured using (a) UPS (He I) and (b) MIES.
Figure 3. HREELS spectra of [EMIM][Tf2N] at (a) 300 K and (b)
100 K using different primary beam energies. The magnified spectra
have been subject to a smoothing procedure for better visibility of the
The [EMIM][Tf2N] RT-IL Surface
J. Phys. Chem. B, Vol. 111, No. 18, 2007 4803
intensities of the different spectroscopic methods do not depend
on the DOS only.
As previously,14we find that all states from the highest
occupied state at 6.7 eV (5.3 eV including shift) down to those
of a binding energy of 9.0 eV (7.6 eV) belong to the [Tf2N]-
ion in the complex. The MO analysis shows furthermore that
these states are particularly delocalized over the bis-sufonyl-
amide core. The highest occupied state of the [EMIM]+ion in
the complex, which is mainly localized on the π orbital of the
imidiazole ring, appears at 9.1 eV (7.7 eV). This means that in
the complex it is significantly shifted to lower binding energies
compared to the isolated [EMIM]+ion, where we found the
HOMO binding energy at 11.9 eV. This difference is obviously
caused by the negative charge of the [Tf2N]-counterion of the
complex. On the other hand, for the highest occupied state of
the isolated [Tf2N]-ion, we calculated a binding energy of 3.9
eV compared to 6.7 eV (5.3 eV) in the complex. According to
our calculations, the second prominent peak, located at ∼11.4
eV (10.0 eV; assigned as 2pF in Figure 2), contains states
attributed to the nonbonding orbitals of the fluorine atoms in
the [Tf2N]-ion. Notably, for the isolated [Tf2N]-ion MOs of
the same character are found for two states peaking the DOS at
8.2 eV, which is about the energy shift found for the HOMOs
3.5. DFT Calculations of the Vibrational Structure. In
order to interpret the HREELS spectra, we have calculated the
vibrational DOS (Figure 4b) by DFT and analyzed the corre-
sponding normal-modes. This enables us to interpret the
HREELS spectra on a molecular level. We find good agreement
between the background-corrected RT-HREELS spectrum and
the calculated vibrational DOS, if we use a Gaussian-broadening
of 120 cm-1(FWHM). Furthermore, each mode is weighted
by the respective IR- and Raman-activity in a ratio of 2:1, which
emerges from the best fit of the predominantly IR-active mode
at 1200 cm-1and the predominantly Raman-active peak at 3090
cm-1, respectively. Double-loss peaks have been taken into
account as well using a weighting factor of 0.125 for the best
fit of the most prominent peak at 2400 cm-1(2 × 1200 cm-1).
Thus, the calculated vibrational spectra match the HREELS data
qualitatively. The most pronounced vibrational structures be-
tween ∼1000 and ∼1400 cm-1are dominated by coupled
antisymmetric OdSdO and CF3 stretching modes in the
[Tf2N]-ion (see Figure 3a peak I and II), and, indeed are found
to possess the highest IR-activity. Corresponding symmetric
modes, which we expect between 500 and 650 cm-1, are much
lower in dipole activity and consequently not resolved in the
As the low-temperature HREELS spectra reveals, the vibra-
tional spectra of the [EMIM][Tf2N] film are depth-dependent
at 100 K. As mentioned above, spectra with high primary beam
energy are basically the same as for RT (see Figure 3b peaks
I-III). For the HREELS spectrum with E0) 2.5 eV, however,
we need a more sophisticated way of fitting the calculated DOS
to the experimental spectrum (see Figure 4c). The key to such
a fit is the orientation of the dipole derivative vector of the
vibrational modes and their allocation onto the two ions, the
[EMIM]+and the [Tf2N]-. We found that there are normal
modes with dipole derivative vector perpendicular to the
imidazole ring of the [EMIM]+ion, contributing to the peaks
IV and VI, which are additionally found in the low-temperature
(T ) 100 K) HREELS spectrum for E0) 2.5 eV.
4.1. Electronic and Vibrational Structure at RT. As already
reported previously, the VB data obtained by MIES, UPS, and
XPS are rather similar and agree well with DFT calculations14
(see Figure 4a). Thus, a variation of the information depth does
not lead to significant changes in the spectra, indicating that, at
room temperature, both species are present at the outermost
surface without a preferential molecular orientation. In this
study, this interpretation is further supported by vibrational
spectroscopy (HREELS). The data presented in Figure 3 do not
change significantly if the primary beam energy is varied, as
far as the observed loss features are concerned. Since the probing
depth is dependent on primary beam energy according to
d ∝ E0
different information depth. Thus, we conclude that at room
temperature the surface electronic and vibrational structure is
very similar to the bulk structure.
However, the FWHM of the elastic peak is in the range of
100-500 cm-1is strongly dependent on primary beam energy.
In all cases it is much higher than the device resolution for the
applied settings. At room temperature the FWHM decreases
monotonously from 420 cm-1at E0) 2.5 eV to 220 cm-1at
E0) 80 eV with increasing primary beam energy. At present,
we are not able to give a detailed interpretation for this
observation. However, a similar behavior is found for semi-
conductor surfaces with a depth dependent charge carrier
concentration.28,29Similar effects might be responsible here, too.
However, proof of the existence of free carriers in IL requires
additional investigations related to calculations based on
dielectric theory. Note, that the situation is even more compli-
cated here, since the FWHM could also be influenced by
changes in the viscosity and conductivity of the ionic liquid,
especially when the temperature is varied. Consequently, the
following discussion will be based solely on the observed
∼0.5(ref 29), the presented sets of spectra illustrate the
Figure 4. (a) MIES and UPS (He I) measurements of [EMIM][Tf2N]
at 300 K compared to DFT calculations of the total density of occupied
states. (b) Background-corrected HREELS spectrum at 300 K (primary
electron beam energy E0 ) 2.5 eV) compared to the vibrational
calculations of the gas-phase ion pair. (c) Low temperature HREELS
spectrum (T ) 100 K, E0 ) 2.5 eV) compared to vibrational modes
possessing dipole activities perpendicular to the central [EMIM]+ring
(for details see text).
4804 J. Phys. Chem. B, Vol. 111, No. 18, 2007
Krischok et al.
4.2. Vibrational Structure at Low T. The decrease of the
RT-IL film temperature has strong impact on the surface
vibrational structure; the low-temperature HREELS data differ
remarkably from the RT data. Especially, with low primary
beam energy and high surface sensitivity, the well pronounced
(CF3and antisymmetric OdSdO related) vibrational modes are
almost absent, whereas the modes around 800 cm-1, expected
according to our DFT calculations, are present. The latter
vibrations are associated to the imidazolium-H bending modes
and to the [Tf2N]-OdSdO symmetric stretching modes. As
already pointed out, the information depth becomes larger with
increasing beam energy. We observe that the spectra become
increasingly similar to the RT data. This behavior may be
simulated by considering a reorientation of the ion pairs in the
outermost surface region. In addition, we have to take the well-
known dipole selection rules (e.g., in ref 25) into account, where
only those vibrations, having oscillating dipole moments
perpendicular to the surface, contribute to the specular regime.
In our qualitative model, we considered the reorientation of the
ion pairs with the [EMIM]+ring lying flat and either the anion
or the cation as the outermost group. By removing the vibrations
with oscillating dipole moments parallel to the surface, we can
qualitatively explain the disappearance of the modes around
1100 cm-1and the appearance of the modes around 800 cm-1
in the HREELS spectra (Figure 4c). In addition, in the
comparison of HREELS spectra and our frequency fit, the
[Tf2N]--related signatures matching the selection rules appear
less pronounced but are still present. Summarizing, the tem-
perature dependence of the HREELS spectra may be under-
stood consistently only if changes in the molecular order close
to the surface are assumed. Such changes in the molecular
order should influence the valence band structure of the sur-
face and thus result in changes of the MIES and UPS spec-
tra. Even if other origins cannot completely ruled out, such
changes at low temperatures are indeed observed as already
4.3. High-Temperature Results. We find a strong de-
crease of the intensity from both cation and anion, starting
around 350 K. Upon heating, changes in the low binding
energy part of the MIES and UPS spectra are observed be-
sides the disappearance of the 2pF emission (already ob-
served upon cooling from room temperature down to 127 K,
ref 13). In UPS, emission originating from the Au substrate
becomes visible. The MIES spectra differ now from the
corresponding UPS data due to a change in the interaction
process. Metastable He atoms now interact mainly via the Auger
neutralization process, as is anticipated for a high work function
metal such as Au. Thus, the appearance of the substrate emission
during heating signals desorption of the material. Above 530
K, both the MIES and UPS spectra are practically identical with
those of the neat Au substrate.
On the other hand, the temperature for the onset of the IL
decomposition is reported to be between 727 and 742 K.30Thus,
decomposition can be excluded as the reason for the observed
intensity decrease, and our results offer a very direct proof that
early volatilization of [EMIM][Tf2N] takes place under reduced
pressure. This finding lends strong support to the recent
observations in refs 16, 17, and 31 of an early volatilization of
[EMIM][Tf2N] under reduced pressure. In particular, a vapor
pressure of 1.24 × 10-4Torr (as measured with the integral
effusion Knudsen method at a residual pressure of 10-5Torr)
was reported at 455.2 K; the absence of IL decomposition during
the evaporation process was established by infrared spectros-
On the basis of UPS, MIES, and HREELS, backed-up by
DFT calculations, we have investigated the surface structure of
the room-temperature ionic liquid [EMIM][Tf2N] between 100
and 620 K. We found that at RT the surface structure is similar
to that of the bulk. At low temperatures (below 200 K), the
conductivity of the film decreases drastically. According to
HREELS, the surface structure at 100 K differs remarkably from
that of the bulk. The presented data lead to the conclusion that,
at low temperatures, the molecules at the surface-vacuum
interface are not randomly oriented anymore. For heating the
sample under UHV conditions, our MIES and UPS data give
evidence that volatilization of [EMIM][Tf2N] starts around
Acknowledgment. We are grateful to F. Endres (Clausthal
University of Technology) for the supply of the RT-IL samples
and his continuous interest in this work.
References and Notes
(1) Yoshimura, D.; Yokoyama, T.; Nishi, T.; Ishii, H.; Ozawa, R.;
Hamaguchi, H.; Seki, K. J. Electron Spectrosc. Relat. Phenom. 2005, 144-
(2) Wasserscheid, P.; Keim, W. Angew. Chem. 2000, 112, 3926.
(3) Binnemans, K. Chem. ReV. 2005, 105, 4148.
(4) Crosthwaite, J. M.; Muldoon, M. J.; Dixon, J. K.; Anderson, J. L.;
Brennecke, J. F. J. Chem. Thermodyn. 2005, 37, 559.
(5) Endres, F.; El Abedin, S. Z. Phys. Chem. Chem. Phys. 2006, 8,
(6) Ko ¨lle, P.; Dronskowski, R. Inorg. Chem. 2004, 43, 2803.
(7) Welton, T. Chem. ReV. 1999, 99, 2071.
(8) Pinilla, C.; Del Popolo, M. G.; Lynden-Bell, R. M.; Kohanoff, J.
J. Phys. Chem. B 2005, 109, 17922.
(9) Dupont, J.; Suarez, P. A. Z. Phys. Chem. Chem. Phys. 2006, 8,
(10) Caporali, S.; Bardi, U.; Lavacchi, A. J. Electron Spectrosc. Relat.
Phenom. 2006, 151, 4.
(11) Smith, E. F.; Villar-Garcia, I. J.; Briggs, D.; License P. Chem.
Commun. 2005, 5633.
(12) Smith, E. F.; Rutten, F. J. M.; Villar-Garcia, I. J.; Briggs, D.;
License, P. Langmuir 2006, 22, 9386.
(13) Ho ¨fft, O.; Bahr, S.; Himmerlich, M.; Krischok, S.; Schaefer, J. A.;
Kempter, V. Langmuir 2006, 22, 7120.
(14) Krischok, S.; O ¨ttking, R.; Beenken, W. J. D.; Himmerlich, M.;
Lorenz, P.; Ho ¨fft, O.; Bahr, S.; Kempter, V.; Schaefer, J. A. Z. Phys. Chem.
2006, 220, 1407.
(15) Gottfried, J. M.; Maier, F.; Rossa, J.; Gerhard, D.; Schulz, P. S.;
Wasserscheid, P.; Steinru ¨ck, H.-P. Z. Phys. Chem. 2006, 220, 1439.
(16) Rebelo, L. P. N.; Lopes, J. N. C.; Esperanca, J. M. S. S.; Filipe, E.
J. Phys. Chem. B 2005, 109, 6040.
(17) Earle, M. J.; Esperanca, J. M. S. S.; Gilea, M. A.; Lopes, J. N. C.;
Rebelo, L. P. N.; Magee, J. W.; Seddon, K. R.; Widegren, J. A. Nature
2006, 439, 831.
(18) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(19) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(20) Janak, J. F. Phys. ReV. B 1978, 18, 7165.
(21) Politzer, P.; Abu-Awwad, F. Theor. Chem. Acc. 1998, 99, 83.
The [EMIM][Tf2N] RT-IL Surface
J. Phys. Chem. B, Vol. 111, No. 18, 2007 4805