Probing a gas/liquid acid-base reaction by X-ray photoelectron spectroscopy

Article abstract of DOI:10.1002/anie.201304115

A mechanistic study: In a Bronsted acid/base reaction, gaseous triflic acid (OTfH) transfers its proton to the basic Cl^- anion of the ionic liquid [C₈C₁Im]Cl forming volatile HCl and the IL [C₈C₁Im][TfO]. This anion exchange reaction is monitored in the near-surface region quantitatively by X-ray photoelectron spectroscopy as a function of time (see picture; BE=binding energy). Copyright

Full text of DOI:10.1002/anie.201304115

.
Angewandte
Communications
DOI: 10.1002/anie.201304115
Ionic Liquids
Probing a Gas/Liquid Acid–Base Reaction by X-ray Photoelectron
Spectroscopy**
Inga Niedermaier, Nicola Taccardi, Peter Wasserscheid, Florian Maier,* and
Hans-Peter Steinrꢀck*
One of the most fundamental issues in reaction chemistry is
the identification of the reaction mechanism, possible inter-
mediates, or by-products and the reaction kinetics. Numerous
analytical methods are routinely employed to monitor the
progress of reactions and to extract information on the
underlying molecular mechanisms. These methods include
NMR, UV/Vis, and IR spectroscopy, mass spectrometry, and
calorimetric techniques. X-ray photoelectron spectroscopy
(XPS) is not traditionally used for reaction monitoring,
although it has the potential to complement these techniques
and is even superior in certain respects.
Unlike many other spectroscopic tools, XPS is quantita-
tive and enables direct, in situ observation of nearly all
elements present in a sample, except hydrogen and helium.
Changes of the chemical environment are reflected by
changes in the corresponding core-level binding energy, the
so-called chemical shifts. Furthermore, XPS is sensitive to the
sample surface with a typical information depth at the
nanometer scale. This is particularly helpful when studying
heterogeneous reactions of gas-phase species with reactive
surfaces. It is this interface between both reactants which
mainly determines the reaction behavior. Thus, the inves-
tigation of processes at reactive interfaces is of fundamental
importance.
ultrathin adsorbate layers at monolayer or submonolayer
coverage. Often the otherwise liquid or gaseous species
remain adsorbed on the surface only at low temperatures. The
investigation of macroscopic amounts of conventional liquids
is impossible because of their comparatively high vapor
pressure.
Ionic liquids (ILs) are an exception to this restriction, as
UHV-based investigations like XPS can be performed even
on macroscopic amounts at ambient or elevated temper-
atures.[11,12] This is possible because of their extremely low
vapor pressure,[13,14] which, together with other interesting
physico-chemical properties like conductivity[15] and consid-
erable thermal stability[16] has given them a wide scientific
attention within the last few decades. ILs are salts composed
solely of ions, which are liquid below 1008C.
Using ILs, we have recently shown that it is possible to
make a huge step forward in the study of reactions by XPS,
overcoming the restrictions to solid–adsorbate surface reac-
tions.[17,18] We investigated a liquid-phase reaction, namely the
alkylation of an amine, under solvent-free conditions in
a macroscopic liquid film by tethering the reactive centers to
ionic head groups, thereby combining reactivity and low
volatility. Apart from monitoring the course of the reaction
over time by following changes in peak intensities associated
with the participating species, it was possible to identify side
reactions and product yield. By using angle-resolved XPS
(ARXPS) under grazing electron emission, that is, with
increased surface sensitivity, molecular orientation and seg-
regation effects at the outermost surface layers were deter-
mined.[18]
Within the last 20 years, many contributions have been
made towards the understanding of reactions on catalytically
active metal or metal oxide surfaces by using real-time XPS or
other surface-sensitive techniques.[1–10] As most of these
techniques require ultra-high vacuum (UHV) conditions,
their application is restricted to solid surfaces covered with
Herein, we have taken this concept one step further by
investigating a Brønsted acid–base reaction of a gaseous
strong acid with the anion of the ionic liquid acting as a base.
Apart from extending the possibilities for XPS to study
reaction and diffusion phenomena in very fast acid-base
reactions at the gas/liquid interface, the applied model
reaction excludes the occurrence of side reactions.
[*] I. Niedermaier, Dr. F. Maier, Prof. Dr. H.-P. Steinrꢀck
Lehrstuhl fꢀr Physikalische Chemie II
Friedrich-Alexander-Universitꢁt Erlangen-Nꢀrnberg
Egerlandstraße 3, 91058 Erlangen (Germany)
E-mail: florian.maier@fau.de
hans-peter.steinrueck@fau.de
In detail, we investigated the acid–base reaction (see
Scheme 1) between 1-methyl-3-octyl-imidazolium chloride
([C8C1Im]Cl) and triflic acid (TfOH). During this reaction,
Dr. N. Taccardi, Prof. Dr. P. Wasserscheid
Lehrstuhl fꢀr Chemische Reaktionstechnik
Friedrich-Alexander-Universitꢁt Erlangen-Nꢀrnberg
Egerlandstraße 3, 91058 Erlangen (Germany)
[**] This work has been supported by the DFG through SPP 1191 (grant
numbers STE 620/7-3 and WA 1615/8-3) and by the Excellence
Cluster “Engineering of Advanced Materials” granted to the
University of Erlangen-Nuremberg. I.N. acknowledges financial
support from the Faculty of Science of the university. We also
acknowledge Assma Benkada and Martin Demleitner for their
helpful contributions.
Scheme 1. Reaction of [C8C1Im]Cl with volatile triflic acid TfOH to form
a new IL, [C8C1Im][TfO], and HCl that evaporates under UHV con-
ditions.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201304115.
8904
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 8904 –8907

Angewandte
Chemie
a proton is transferred from TfOH to the basic chloride ion of
the ionic liquid, forming HCl and a new anion, triflate
([TfO]ꢀ). The reaction must be driven both by acid strength
and, as it is carried out under vacuum conditions in the XPS
chamber, by the differences in vapor pressure (pvap(TfOH)
ꢁ 101 hPa at 298 K,[19] pvap(HCl, pure) ꢁ 104 hPa at 295 K).[20]
Overall, because of the evaporation of the reaction product
HCl, the reaction could also be classified as an anion
exchange reaction, where the anion of the IL [C8C1Im]Cl is
substituted with [TfO]ꢀ to form the IL [C8C1Im][TfO]. Such
a type of ion exchange reaction has frequently been used to
synthesize ionic liquids with weakly basic anions from their
corresponding chloride salts by addition of a strong acid.[21]
The reaction was carried out by dosing gaseous TfOH
(through a leak valve connected to a reservoir with the liquid
acid; see the Supporting Information for further experimental
details) at room temperature onto the surface of a macro-
scopic film of [C8C1Im]Cl (thickness ꢁ 0.2 mm) at a TfOH
partial pressure of 1 ꢀ 10ꢀ6 mbar (base pressure: 5 ꢀ
10ꢀ10 mbar). Figure 1a shows the overall changes in an XP
overview spectrum before (0 minutes) and after 170 minutes
of TfOH dosage, and Figure 1b displays all relevant core-
level regions before dosing and after 80 and 170 minutes total
dosage time. Prior to TfOH exposure, only the expected C 1s,
N 1s, Cl 2s, and Cl 2p signals of [C8C1Im]Cl were observed,
with the expected intensities. Upon TfOH dosage, new signals
emerged around 688.5, 532.0, 231.9, and 169.7 eV, which are
attributed to photoelectrons emitted from the F 1s, O 1s, S 2s
and S 2p orbitals, respectively, of the newly formed [TfO]ꢀ
anion, along with the corresponding Auger signals. In the C 1s
region, a new peak arose at 292.6 eV because of the CF3
groups of [TfO]ꢀ. At the same time, both the Cl 2s and Cl 2p
signals of the chloride ion decreased in intensity.
After 80 minutes dosage, the initial Cl 2p intensity drop-
ped down to about 50%, and concomitantly, the signals of the
triflate anion increased to about 50% intensity of pure
[C8C1Im][TfO]. This indicates that (within the information
depth of XPS) about half of the Clꢀ anions of [C8C1Im]Cl
reacted with TfOH to [C8C1Im][TfO] and volatile HCl. After
170 minutes, Clꢀ signals virtually disappeared in the overview
spectrum, indicating nearly complete conversion of
[C8C1Im]Cl to [C8C1Im][TfO]. From the corresponding
spectrum of the Cl 2p region in Figure 1b, the remaining
amount of Clꢀ is quantified to ꢁ 9% of its original intensity.
Since no other Cl species, such as covalently bound Cl or Cl2
are observed in the Cl 2p region, we conclude that the proton
transfer/anion exchange is the only process taking place. The
calculated atom percentages of all elements after 170 minutes
of TfOH dosage are—apart from a small reduction in [TfO]ꢀ
signals and the additional Cl signals—in perfect agreement
with the nominal values for pure [C8C1Im][TfO] (gray spectra
in Figure 1b) within the experimental error (ꢂ 2% in relative
intensities), clearly pointing towards near completion of the
reaction.
Figure 1. a) Overview spectra before (0 minutes, top) and after
170 minutes (bottom) of TfOH dosage. b) Region spectra before
(0 minutes), after 80 minutes and after 170 minutes of total TfOH
dosage time (pTfOH =1ꢂ10ꢀ6 mbar). Core-level spectra of pure
[C8C1Im][TfO] (gray) are shown for comparison. Gas dosage and XPS
recording were performed at room temperature under normal emis-
sion conditions (information depth ꢁ7–9 nm).[12]
atoms of the octyl chain bound to carbon and hydrogen only,
and the peak at ꢁ 286 eV (Chetero) refers to the five carbon
atoms bound to nitrogen of the imidazolium ring. The Calkyl
,
the Chetero, and also the N 1s signals solely originate from the
cation of the IL, which is not involved in the acid–base
reaction. Nevertheless, these signals decrease by 12–14%
after 170 minutes of dosage time. This decrease directly
reflects the overall change of the IL density, because of the
exchange of the small Clꢀ anion by the considerably larger
[TfO]ꢀ anion (molecular volumes of the neat ILs see Kolbeck
et al.).[22]
Further information can be extracted from the spectral
changes of the C 1s and N 1s core levels. In the C 1s region
three peaks are observed: the signal at 292.6 eV (Canion) stems
from the [TfO]ꢀ anion formed in the course of the reaction.
The peak at 285.0 eV (Calkyl) is attributed to the seven carbon
Angew. Chem. Int. Ed. 2013, 52, 8904 –8907
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8905

.
Angewandte
Communications
The second prominent change in the cation signals is the
shift of the N 1s and Chetero peaks towards higher binding
energy by approximately 0.2 eV after 80 minutes of gas
exposure, and by an additional 0.1 eV shift after 170 minutes
total dosage time. These shifts reflect changes in the mean
anionic environment of the imidazolium head group in course
of the reaction. Such anion-dependent effects were system-
atically studied by Cremer et al.[23] The here observed N 1s
and Chetero peak positions for [C8C1Im]Cl (before the reaction)
and [C8C1Im][TfO] (after the reaction) coincide with the
values reported there.[23] To finally prove that the reaction
product is indeed [C8C1Im][TfO], we included the spectra of
pure [C8C1Im][TfO] in Figure 1b (gray spectra at the bottom).
It is evident that the spectra are identical to those of
[C8C1Im]Cl after reaction with TfOH. Thus, all our findings
clearly demonstrate that, at least within the information depth
of XPS (7–9 nm), nearly complete conversion to
[C8C1Im][TfO] has indeed taken place for this acid–base
reaction.
As next step, we show that this conversion was restricted
to the near surface region of the macroscopic IL film. After
170 minutes exposure time to TfOH, the reacted IL was first
stored under UHV for 30 h and, thereafter, heated to 370 K
for one hour to decrease viscosity and increase ion mobility.
After the first step, that is, leaving the IL in UHV, the spectra
remained unchanged within the margin of error (not shown).
However, after heating, distinct changes occurred, as shown
in Figure 2.
evaporates directly upon formation and is pumped off the
vacuum system. Thus, HCl liberated in the previous ion
exchange cannot account for the reappearance of the chloride
signals given the applied UHV conditions for a period of
several hours.
In conclusion, we succeeded to monitor a Brønsted-type
acid–base reaction of gaseous triflic acid (TfOH) and the
basic chloride anion of the IL [C8C1Im]Cl in situ by XPS. This
gas/liquid reaction occurs at the surface of the IL and is nearly
complete in the surface-near region after 170 minutes of
TfOH dosage at 1 ꢀ 10ꢀ6 mbar. The progress of the reaction
was deduced from the decrease of the chloride signal and by
simultaneous appearance of the F 1s, O 1s, Canion and S 2p
signals of the triflate ionic liquid. The corresponding spectra
after the reaction coincide with the spectra of pure
[C8C1Im][TfO]. Furthermore, intensity changes in N 1s,
C
hetero, and Calkyl intensities reflect changes in the IL density
because of exchange of the small Clꢀ anion by the much more
bulky [TfO]ꢀ anion. This is accompanied by changes in charge
transfer from the anion to the cation in the case of
[C8C1Im][TfO] (i.e. after reaction) which can be deduced
from Chetero and N 1s peak shifts. Heating of the IL layer after
the reaction shows a reappearance of the reactant [C8C1Im]Cl
which indicates that under the applied conditions the reaction
is restricted to the near surface region, and that because of the
increased ion mobility at elevated temperatures, an intermix-
ing of [TfO]ꢀ ions at the surface with unreacted Clꢀ ions from
the bulk occurs.
These results constitute the first example of an in situ XPS
study of a gas/liquid acid–base reaction and open a new route
towards characterizing such reactions in great detail. In
contrast to our earlier liquid–liquid alkylation studies,[17,18] the
absence of additional side reactions and the possibility to
apply one of the reactants in a well-defined manner allow
detailed investigations on the reaction mechanisms, reaction
kinetics, and diffusion processes, particularly in temperature-
dependent experiments. In the future, angle-resolved XPS
measurements will give access to additional information on
the surface composition and on the orientation of the
molecules at the gas/liquid interface. It is also anticipated
that by comparing the reaction rate of several volatile acids
with [C8C1Im]Cl (or any other ionic liquid with a basic anion)
the acidity of such acids in the corresponding ionic liquids can
be determined in an indicator-free manner. Given the high
importance of acidic ionic liquids and acidic IL mixtures in
catalysis, such studies will be also of significant practical
relevance.
Figure 2. Comparison of core-level spectra taken under normal emis-
sion of [C8C1Im]Cl exposed for 170 minutes to TfOH at 1ꢂ10ꢀ6 mbar
partial pressure forming mainly [C8C1Im][TfO] (top) and after heating
the reacted IL to 370 K for 60 minutes (bottom).
Received: May 13, 2013
Published online: August 2, 2013
The peaks attributed to the [TfO]ꢀ anion, namely F 1s,
O 1s, Canion, and S 2p, decreased to 47 ꢂ 8% of their maximum
intensities and simultaneously, the Cl 2p doublet reappeared.
This effect is attributed to diffusion of Clꢀ anions from the IL
bulk region, where no reaction with TfOH occurred under the
applied reaction conditions, to the near-surface region (i.e. to
within the information depth of our XPS setup of about 7–
9 nm). The back reaction, that is, proton transfer to re-form
Clꢀ is ruled out, as the gaseous product HCl irreversibly
Keywords: acid–base reactions · ionic liquids ·
reaction mechanisms · X-ray photoelectron spectroscopy
.
[1] R. Denecke, Appl. Phys. A 2005, 80, 977 – 986.
[2] R. Streber, C. Papp, M. P. A. Lorenz, A. Bayer, R. Denecke, H.-
P. Steinrꢁck, Angew. Chem. 2009, 121, 9925 – 9929; Angew.
Chem. Int. Ed. 2009, 48, 9743 – 9746.
8906 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 8904 –8907

Angewandte
Chemie
[3] W. F. Egelhoff, Core-level binding energy shifts at surfaces and in
solids, Springer, Amsterdam, 1987.
[15] D. R. Macfarlane, M. Forsyth, P. C. Howlett, J. M. Pringle, J. Sun,
G. Annat, W. Neil, E. I. Izgorodina, Acc. Chem. Res. 2007, 40,
1165 – 1173.
[16] P. Wasserscheid, T. Welton, Ionic liquids in synthesis, Wiley-
VCH, Weinheim, 2008.
[4] D. P. Woodruff, T. A. Delchar, Modern techniques of surface
science, Cambridge University Press, Cambridge, 1986.
[5] W. Schattke, M. A. Van Hove, Solid-state photoemission and
related methods: theory and experiment, Wiley-VCH, Weinheim,
2003.
[6] S. Hꢁfner, Photoelectron spectroscopy: principles and applica-
tions, rev. and enlarged, 3rd ed., Springer, Berlin, 2003.
[7] C. S. Fadley, Surf. Interface Anal. 2008, 40, 1579 – 1605.
[8] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomben, J.
Chastain, R. C. King, Jr., Physical Electronics Incorporation,
Handbook of X-ray photoelectron spectroscopy: a reference
book of standard spectra for identification and interpretation of
XPS data, Physical Electronics, Eden Prairie, 1995.
[9] D. Briggs, J. T. Grant, Surface analysis by Auger and X-ray
photoelectron spectroscopy, IM, Chichester, 2003.
[17] C. Kolbeck, I. Niedermaier, N. Taccardi, P. S. Schulz, F. Maier, P.
Wasserscheid, H.-P. Steinrꢁck, Angew. Chem. 2012, 124, 2664 –
2667; Angew. Chem. Int. Ed. 2012, 51, 2610 – 2613.
[18] I. Niedermaier, C. Kolbeck, N. Taccardi, P. S. Schulz, J. Li, T.
Drewello, P. Wasserscheid, H.-P. Steinrꢁck, F. Maier, Chem-
PhysChem 2012, 13, 1725 – 1735.
[19] Sigma Aldrich web page: trifluoromethanesulfonic acid, purum,
ꢃ 98.0% (T) j CF3SO3H j Sigma – Aldrich, URL: http://www.
sigmaaldrich.com/catalog/product/fluka/91738?lang =
de&region = DE, accessed 23.04.2013.
[20] Handbook of Chemistry and Physics, 93rd ed. (Ed.: W. M.
Haynes), CRC Press, Boca Raton, 2012 – 2013.
[10] A. Baraldi, G. Comelli, S. Lizzit, M. Kiskinova, G. Paolucci, Surf.
Sci. Rep. 2003, 49, 169 – 224.
[21] J. Fuller, R. T. Carlin, H. C. Delong, D. Haworth, J. Chem. Soc.
Chem. Commun. 1994, 299 – 300.
[11] K. R. J. Lovelock, I. J. Villar-Garcia, F. Maier, H.-P. Steinrꢁck, P.
Licence, Chem. Rev. 2010, 110, 5158 – 5190.
[12] H.-P. Steinrꢁck, Phys Chem. Chem. Phys 2012, 14, 5010 – 5029.
[13] P. Wasserscheid, Nature 2006, 439, 797 – 797.
[14] M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea, J. N. C. Lopes,
L. P. N. Rebelo, J. W. Magee, K. R. Seddon, J. A. Widegren,
Nature 2006, 439, 831 – 834.
[22] C. Kolbeck, J. Lehmann, K. R. J. Lovelock, T. Cremer, N. Paape,
P. Wasserscheid, A. P. Froba, F. Maier, H.-P. Steinrꢁck, J. Phys.
Chem. B 2010, 114, 17025 – 17036.
[23] T. Cremer, C. Kolbeck, K. R. J. Lovelock, N. Paape, R. Wçlfel,
P. S. Schulz, P. Wasserscheid, H. Weber, J. Thar, B. Kirchner, F.
Maier, H.-P. Steinrꢁck, Chem. Eur. J. 2010, 16, 9018 – 9033.
Angew. Chem. Int. Ed. 2013, 52, 8904 –8907
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8907