Communications
=
monomer. The intense n(C O) is observed at very similar
positions in small argon-isolated monocarboxylic acids such
as formic acid (1767 cmꢀ1), acetic acid (1779 cmꢀ1), and
propionic acid (1777 cmꢀ1).[30] Bands at 1826 cmꢀ1, 1797 cmꢀ1,
and 1714 cmꢀ1 show an intensity of about one-tenth of the
most intense band at 1776 cmꢀ1. A predicted shift between the
C2v and Cs modes of + 45 cmꢀ1 compared to the observed
+ 50 cmꢀ1 suggests an assignment of the 1826 cmꢀ1 band as
for the Cs monomer and the C2h dimer are found in Table 1.
These appear at a fraction of roughly 10% here, but may be
the main components under different conditions in nature.
These are the positions to be employed for the search of
carbonic acid in astrophysical environments.
We emphasize that our spectra do not show the rotational
transitions observable in high-resolution infrared spectra
recorded in astrophysical environments, for instance, by the
Infrared Space Observatory (ISO), because rotation is
prevented in the matrix cages. Nevertheless, the spectra
indicate the exact position of the Q branch and the respective
center of the zero gap. Unidentified bands may be attribut-
able to carbonic acid fundamentals on the basis of our data.
We, thus, suggest including our data in spectroscopy databases
for molecules of astrophysical or atmospheric interest, for
example, the Cologne Database for Molecular Spectroscopy
(CDMS).[33] In many astrophysical environments solid car-
bonic acid may form from CO2 and H2O upon irradiation and
then undergo many sublimation and recondensation cycles
without decomposition. We expect gaseous carbonic acid to
be present, for example, in the atmosphere of Mars or Venus
and in cometary comae or tails once the comet reaches a
position sufficiently close to the sun, where the temperature
rises beyond 200 K. In principle, one or the other of these four
=
n(C O) of the Cs monomer. In acetic acid the most stable
cyclic dimer shows a band at 1720 cmꢀ1 [31]
,
very close to the
band at 1722 cmꢀ1, which we have assigned to the cyclic
carbonic acid dimer in Table 1. After assignment of the
species C2h, Cs, and C2v some very weak features (< 1% of the
intensity of the C2v absorptions) remain unexplained; they
may, for example, arise from traces of an open dimer[32] or an
H2CO3·H2O adduct. Such species were also observed in
matrix-isolated mixtures of water and carboxylic acids.[30,31]
The spectral region at 1000–1500 cmꢀ1 is the most difficult to
interpret since in this region three fundamental modes,
ꢀ
namely asymmetric and symmetric C (OH) stretching and
in-plane deformation modes, overlap, and since some of these
modes are only weakly IR active. In the spectral region at
roughly 750–820 cmꢀ1 again this intensity ratio of 1:10:1 is
observed for the bands at 808/794/785 cmꢀ1, which are
assigned to the out-of-plane deformation of carbonic acid in
the C2h/C2v/Cs conformations, respectively. This represents, to
the best of our knowledge, the first experimental proof for the
existence of dimeric carbonic acid in the gas phase, and also
for the existence of more than one monomer conformation.
An important piece of evidence for the success of the
isolation of carbonic acid from the gas phase is provided by
checking what remains on the gold mirror after removal of the
matrix. Heating matrix-isolated carbonic acid to 180 K results
in desorption of argon, and to some extent carbonic acid is
dragged away from the surface along with argon. However,
some carbonic acid molecules remain on the surface even
after removal of argon. The spectrum of the sample remaining
after evaporation of the matrix is shown in Figure 1b. This
spectrum exhibits a remarkable similarity to the spectrum of
the as-produced thin film of carbonic acid shown in Figure 1a.
There is only one band missing in Figure 1b (at 584 cmꢀ1),
which is explained by the fact that the available detector in
Vienna had no sensitivity below 600 cmꢀ1. This spectrum can
be clearly attributed to crystalline carbonic acid. That is, the
isolated carbonic acid molecules form larger aggregates upon
removal of argon and finally arrange themselves to produce
the same polymorph produced also after removal of the
solvent methanol from dissolved carbonic acid.[2,22,23] The
main difference between the two spectra is that the absorb-
ance is roughly a factor 100 lower in Figure 1b, which
corresponds to the reduced thickness of the film. The
morphology and the texture of the respective crystallites
and clusters are in turn responsible for the additional
differences in the two spectra.
=
modes may also be caused by other species; the C O
stretching mode of monocarboxylic acids such as formic
acid, for instance, is also found at 5.6 mm.[34] However, these
do not show any intense bands in the vicinity of 6.9 mm,
8.6 mm, and 12.7 mm, and so the presence of carbonic acid can
be distinguished from the presence of monocarboxylic acids
by employing the set of four monomer marker bands in high-
resolution IR spectra.
Experimental Section
Matrix-isolation spectroscopy is a technique aimed at obtaining pure
vibrational spectra of nonrotating molecules by trapping them in an
inert and transparent matrix of argon or neon. In the ideal case there
is no interaction between the matrix material and the trapped species.
In reality, there is a weak interaction resulting in a slight blueshift of
individual absorptions. In addition, there may be different geo-
metrical types of cages, resulting in matrix-induced band splittings.
These shifts are typically on the order of a few wavenumbers
compared to the gas-phase spectrum. Spectra obtained in neon
matrices are considered to most closely resemble gas-phase spectra.
Similarly, the probably most common matrix material, argon, is
known to result in only a small deviation from gas-phase spectra.
Comparison between argon and neon matrix reveals the particular
impact of the matrix and explains the respective splitting of some
bands. Our matrix-isolation study was done in the ultrahigh-vacuum
chamber in Vienna (see Figure 1 in the Supporting Information),
which was previously employed for successfully isolating reactive
species such as halogen oxides.[20,21] Solid carbonic acid was produced
for the purpose of matrix isolation as a micrometer-thin film on IR-
transparent windows, typically CsI, in Innsbruck using the low-
temperature technique developed in the 1990s in this labora-
tory.[2,22,23] The IR spectrum depicted in Figure 1a was recorded on
the carbonic acid sample after production and coincides with the
literature spectrum,[2,22,23] which implies that the thin film is crystal-
line. The sample was then transported immersed in liquid nitrogen to
Vienna, and the matrix-isolation procedure was performed as
described in the Supporting Information by subliming solid carbonic
The matrix data allow firm assignment of the gas-phase
bands at (3608 ꢁ 30) cmꢀ1 [(2.77 ꢁ 0.02) mm], (1776 ꢁ 7) cmꢀ1
[(5.63 ꢁ 0.03) mm],
(1452 ꢁ 4) cmꢀ1
[(6.89 ꢁ 0.02) mm],
(1182 ꢁ 5) cmꢀ1 [(8.46 ꢁ 0.03) mm], and (794 ꢁ 4) cmꢀ1
[(12.59 ꢁ 0.06) mm] to the C2v monomer. The band positions
1942
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 1939 –1943
Bernard, Juergen
