Gas-phase preparation of carbonic acid and its monomethyl ester

Article abstract of DOI:10.1002/anie.201406969

Carbonic acid (H₂CO₃), an essential molecule of life (e.g., as bicarbonate buffer), has been well characterized in solution and in the solid state, but for a long time, it has eluded its spectral characterization in the gas phase owing to a lack of convenient preparation methods; microwave spectra were recorded only recently. Here we present a novel and general method for the preparation of H₂CO₃ and its monomethyl ester (CH₃OCO₂H) through the gas-phase pyrolysis of di-tert-butyl and tert-butyl methyl carbonate, respectively. H₂CO₃ and CH₃OCO₂H were trapped in noble-gas matrices at 8 K, and their infrared spectra match those computed at high levels of theory [focal point analysis beyond CCSD(T)/cc-pVQZ] very well. Whereas the spectra also perfectly agree with those of the vapor phase above the b-polymorph of H₂CO₃, this is not true for the previously reported a -polymorph. Instead, the vapor phase above a-H₂CO₃ corresponds to CH₃OCO₂H, which sheds new light on the research that has been conducted on molecular H₂CO₃ over the last decades.

Full text of DOI:10.1002/anie.201406969

.
Angewandte
Communications
DOI: 10.1002/anie.201406969
Carbonic Acid
Hot Paper
Gas-Phase Preparation of Carbonic Acid and Its Monomethyl Ester**
Hans Peter Reisenauer, J. Philipp Wagner, and Peter R. Schreiner*
Dedicated to Professor Helmut Schwarz
Abstract: Carbonic acid (H2CO3), an essential molecule of life
(e.g., as bicarbonate buffer), has been well characterized in
solution and in the solid state, but for a long time, it has eluded
its spectral characterization in the gas phase owing to a lack of
convenient preparation methods; microwave spectra were
recorded only recently. Here we present a novel and general
method for the preparation of H2CO3 and its monomethyl ester
(CH3OCO2H) through the gas-phase pyrolysis of di-tert-butyl
and tert-butyl methyl carbonate, respectively. H2CO3 and
CH3OCO2H were trapped in noble-gas matrices at 8 K, and
their infrared spectra match those computed at high levels of
theory [focal point analysis beyond CCSD(T)/cc-pVQZ] very
well. Whereas the spectra also perfectly agree with those of the
vapor phase above the b-polymorph of H2CO3, this is not true
for the previously reported a-polymorph. Instead, the vapor
phase above a-H2CO3 corresponds to CH3OCO2H, which
sheds new light on the research that has been conducted on
molecular H2CO3 over the last decades.
by protonation of (bi)carbonate salts,[6] and in D2O solution
through ultrafast deuteration of bicarbonate.[7] Yet, isolated
1 was long considered not to be detectable in the gas phase,
although the first gas-phase preparation and mass-spectro-
metric identification of 1 were already achieved in 1987
through the thermal decomposition of ammonium bicarbon-
ate (NH4HCO3).[8] This approach, however, could not be
successfully adapted to produce a sufficiently high concen-
tration of 1 in the gas phase for spectroscopic character-
ization. The energetically higher lying cis–trans conformer
(1ct, Scheme 1) was first prepared through an electric
C
arbonic acid (H2CO3, 1) is an essential molecule of life,
whose mere existence, however, was debated for a long time.
It is the key intermediate in the exhalation of CO2 and part of
the natural carbonate buffer, which stabilizes, for instance, the
pH value of blood. Similarly, 1 is an intermediate in the
acidification of the oceans through the uptake of CO2 from
the atmosphere despite its rapid decomposition into CO2 and
H2O,[1] in particular, in aqueous solutions.[2] As such, 1 is also
likely to be a key species to be considered in CO2 sequestra-
tion technologies. Carbonic acid has been suggested to be
present in extraterrestrial ices and interstellar regions,
including the surface of Mars,[3] that contain both water and
carbon dioxide,[4] and the high-energy irradiation of CO2/H2O
(ice) mixtures[5] provides access to bulk carbonic acid.
Furthermore, 1 has been prepared on cold surfaces, in glasses
Scheme 1. Proposed mechanism for the high-vacuum flash pyrolysis
(HVFP) of di-tert-butyl carbonate. a) Trapping of the cis–cis (1cc) and
cis–trans (1ct) conformers of carbonic acid as the pyrolysis products
(isobutene also forms as an extrusion product) in a solid argon matrix
at 8 K. b) Analogous preparation of the cis–cis (2cc) and cis–trans (2ct)
conformers of carbonic acid monomethyl ester from tert-butyl methyl
carbonate through HVFP and trapping in the matrix.
discharge of a CO2/Ar mixture, which was then passed
through water before injection into a pulsed discharge nozzle,
and it was identified by Fourier-transform microwave spec-
troscopy[9] in 2009. The most favorable conformer, namely
1cc, was only characterized in 2011 by utilizing a similar
technique.[10]
Herein, we present a novel approach for the gas-phase
preparation of isolated 1 through unimolecular ester pyrolysis
of alkyl carbonates, which had been suggested theoretically,[11]
but not practically realized. Our pyrolysis procedure is
general and also amenable to the preparation of alkyl
monoesters of 1, for instance, carbonic acid monomethyl
ester (methyl hydrogen carbonate, 2, CH3OCO2H), which had
not been prepared in the gas phase before. We will also show
in the following that the matrix-isolated species derived from
[*] Dr. H. P. Reisenauer, M. Sc. J. P. Wagner, Prof. Dr. P. R. Schreiner
Institute of Organic Chemistry, Justus-Liebig-University
Heinrich-Buff-Ring 58, 35392 Giessen (Germany)
E-mail: prs@uni-giessen.de
[**] This research was supported by the Deutsche Forschungsgemein-
schaft (Schr597/18-1). J.P.W. is grateful to the Fonds der Chem-
ischen Industrie for a fellowship. We thank Alexander Vaughn and
Wesley D. Allen for sharing their Perl script for the FPA computa-
tions. We thank Hinrich Grothe for sending us an electronic version
of Jꢀrgen Bernard’s PhD thesis,[13] and Hinrich Grothe as well as
Thomas Loerting for permission (May 04, 2014) to reproduce
Figure S4 of Ref. [16]. This work was first presented at the EUCHEM
Conference on Stereochemistry (Bꢀrgenstock) on May 04, 2014.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201406969.
11766
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771

Angewandte
Chemie
the sublimation of the so-called a-polymorph of 1,[12] which
was prepared by protonation of KHCO3 with HCl in
methanol,[4a] indeed is not 1, but rather the monomethyl
ester 2.[13] Although carbonic acid monoalkyl esters were
recently suggested to occur in alcoholic beverages,[14] they are
uncommon and have only recently come into focus as key
intermediates in CO2 sequestration as they can ideally form
from the reactions of alcohols with CO2.[15]
1 can effectively compete with that for CO2 extrusion
(34.4 kcalmolꢀ1) from the intermediate carbonic acid mono-
tert-butyl ester (Figure S8) to give tert-butyl alcohol. This is
not the case for other pyrolytic leaving groups for which
alcohol formation dominates (e.g., for carbonic acid diethyl
ester, Figure S6).[11]
Thus, 1 can readily be prepared through ester pyrolysis
from either carbonic acid di-tert-butyl ester (Figure 1) or from
As we will demonstrate, only
the
matrix-isolated
species
derived from the vapor phase
above the b-polymorph of 1[16] is
consistent with the spectroscopic
and theoretical data for 1 pre-
pared by the alternative route
presented here. The present
study will aid in the unequivocal
identification of 1 and 2 in clouds
as well as in extraterrestrial envi-
ronments, and it will allow fur-
ther studies on these and closely
related species owing to the ease
of preparation that is offered by
the present experimental proce-
dure (Scheme 1).
Owing to the proposed,[11]
yet experimentally unreported
route to 1 through ester pyroly-
sis, we surmised that the previ-
ously not considered elimination
of isobutene from di-tert-butyl
carbonate, for example (Sche-
Figure 1. Irradiation-induced isomerization of 1cc and 1ct, prepared by HVFP of di-tert-butyl carbonate.
a) CCSD(T)/cc-pVTZ computed spectra of 1cc (blue) and 1ct (red). b) Difference of the two spectra
recorded after 50 min irradiation with l=2454 nm and after additional 28 min radiation with
l=2108 nm. The splittings of several IR bands are due to matrix effects (multiple sites). For the
me 1a), a starting material that preparation of 1 from Boc2O, see Figure S2.
is readily prepared from the
4-dimethylaminopyridine-cata-
lyzed reaction of di-tert-butyl dicarbonate (Boc2O; Boc = tert-
butoxycarbonyl) with tert-butyl alcohol, would be suitable. As
documented in the Supporting Information (Figure S1),
commercially available Boc2O can also be used directly for
these pyrolyses, but this process requires higher temperatures
(8508C), leading to the formation of somewhat larger
amounts of side products. Di-tert-butyl carbonate was evapo-
rated from a pre-cooled storage bulb at ꢀ458C and passed
through a quartz pyrolysis tube (7108C), after which the
pyrolysis products were co-condensed with a large excess of
argon (introduced by a separate jet) on the surface of the 8 K
matrix window; several experiments were performed to
determine the optimal pyrolysis temperature. Alternatively,
a pre-mixed highly diluted gaseous mixture (dilution in argon
1:1000) was passed directly through the hot pyrolysis tube at
7108C, and the pyrolysis products were condensed on the
surface of the 8 K matrix window. Under these conditions,
most of the olefin elimination reactions were not complete,
and unreacted precursors were present in the matrices;
secondary decomposition reactions to CO2, H2O, and tert-
butyl alcohol were unavoidable (Figure S2). Utilizing density
functional theory (DFT) computations at the M06-2X/cc-
pVTZ level of theory to study this pyrolysis reaction, we
found that the barrier (35.2 kcalmolꢀ1) for the formation of
Boc2O (Figure S2), as evident from the excellent agreement
of the experimental and the unscaled CCSD(T)/cc-pVTZ
computed infrared bands for the two observed rotamers 1cc
and 1ct, which can be interconverted with monochromatic
near infrared (NIR) light from an optical parametric oscil-
lator (OPO) by selective excitation of overtones or combina-
tion modes (Figure 1).
Our infrared band positions agree very well with those
previously assigned to 1cc and 1ct of matrix-isolated “b-
H2CO3”[16] (for detailed comparisons, see Figure 3 as well as
Tables S7 and S8). Additionally, we provide several previ-
ously unreported IR bands, including the very strong absorp-
ꢀ
tions of the O H out-of-plane deformation mode at
569.9 cmꢀ1 for 1cc as well as 472.3 cmꢀ1 for 1ct, which will
aid in the detection of 1 in other environments.
The hitherto unknown isomer 1tt cannot be prepared
under these conditions as it is much higher in energy
(+ 10.1 kcalmolꢀ1) than the most favorable structure 1cc
according to our high-level computations (Figure 2); the
isomerization barrier to 1ct is only + 1.8 kcalmolꢀ1. The very
good agreement of the geometrical parameters of the
computed structures with those determined by microwave
spectroscopy[9,10] (the experimental and computed rotational
constants agree within 0.5%, Table S1) lends confidence to
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11767

.
Angewandte
Communications
two conformers of the a- and b-poly-
morphs, respectively (Figure 3, see
below).
In our view, the two different sets of
carbonyl absorptions in the matrix-isola-
tion IR spectra from sublimation of the a-
and b-polymorphs of 1 and the unprece-
dented behavior upon recondensation
can only be due to two different com-
pounds constituting a- and b-H2CO3. As
a-H2CO3 was prepared from a methanolic
solution of KHCO3 and HCl,[4a,12b,16] we
surmised that carbonic acid monomethyl
ester (2, Scheme 1b) formed instead of
1 under such typical esterification con-
ditions. This conjecture is supported by
the notion that b-H2CO3 is transformed
into a-H2CO3 in acidic methanolic solu-
tions.[4a,19] As there are only very few
reports on the preparation of 2 in solu-
tion, for example from protonation of
Figure 2. Potential energy hypersurface for the rotational interconversion of the rotamers of
1 with relative energies DH0 (in kcalmolꢀ1) computed from convergent focal-point analyses
extrapolated to the basis-set limit. The bond lengths [ꢁ] and angles [8] given for 1 are ground-
state optimum geometrical parameters for the CCSD(T)/cc-pVQZ optimized structures in the
given symmetries. For comparison, the experimental geometrical parameters determined from
Fourier-transform microwave spectroscopy for 1ct[9] and 1cc[10] are given in parentheses.
Carbon dark gray, hydrogen light blue, oxygen red.
CH3OCO2 Na+ with HCl or H2O,[20] we
ꢀ
had to devise a strategy for the prepara-
tion of 2 in the gas phase, so that it can be
the quality of the geometry optimizations. The relative
energies (DH0) computed from the convergent focal-point
analyses (FPA)[17] and extrapolated to the basis set limit (for
details, see the Supporting Information) provide the most
accurate assessment of the energies of the stationary points of
1 (Figure 2) to date, and they slightly revise previously
published values.[9,10] Hence, we find that 1tt displays overall
C2 symmetry, which gives a 0.8 kcalmolꢀ1 lower relative
energy, and that the previously computed C2v geometry[9,10] is
not a minimum. The 1ct!1cc isomerization barrier is also
reduced from 8.8 to 7.9 kcalmolꢀ1.[9]
condensed into a noble-gas matrix to allow detailed compar-
isons of its IR spectrum with that of 1.
Fortunately, our HVFP protocol for the preparation of 1 is
general and thus also applicable to the synthesis of 2 from tert-
butyl methyl carbonate (Scheme 1b). Hence, this constitutes
the first gas-phase preparation of 2, allowing a direct compar-
ison with the rotamer spectra of 1. As found for the parent
acid, 2 also displays two readily identifiable rotamers that
differ in the position of the OH bond cis (2cc) or trans (2ct)
=
relative to the C O moiety. We have not found evidence for
the formation of conformers with a trans orientation of the
methyl group, in line with the facts that 2tc and 2tt are 3.5 and
10.7 kcalmolꢀ1 higher in energy than 2cc at the MP2/cc-
pVQZ level of theory, and that their barriers for rearrange-
ment to the lower-lying cis-methyl isomers are only 7.0 and
2.5 kcalmolꢀ1, respectively. Whereas the acid moieties of the
lowest-lying conformers 2cc and 1cc are structurally very
Our IR spectral data perfectly agree with the band
assignments for 1 generated through vapor deposition from b-
H2CO3, but they are significantly different to those assigned
to a-H2CO3;[4a,6,12b,16,18] note that the signal intensities herein
reported for 1 are an order of magnitude larger, which
simplifies their assignment. Our results shed new light on an
earlier report that states “that not only a-H2CO3 can sublime
and recondense as a-H2CO3 but also b-H2CO3 can sublime
and recondense as b-H2CO3”.[16] In our view, this astounding
feature would imply that the bulk structural information
about the polymorphic origin is retained upon sublimation
into individual molecules in the gas phase. It is important to
note that the a and b polymorphs first introduced were only
tentatively assigned;[6] the crystal structures of the two
polymorphs remain unsolved, but structural predictions
were made for a-H2CO3.[1] The IR spectra of matrix-isolated
1 sublimed and re-condensed from a-H2CO3 were interpreted
as evidence for a 10:1 mixture of the 1cc and 1ct conformers
in addition to small amounts of carbonic acid dimers, CO2,
=
similar, the difference for the C O IR absorptions amounts to
Dn = 12.8 cmꢀ1 (Figure 3, Tables S7–S12). This band separa-
tion is much too large to be due to experimental errors, it
rather is indicative of two different chemical entities.
Visual inspection of the overlaid published spectra[16] of
the matrix-isolated condensed vapor phases of the a- and b-
H2CO3 polymorphs (Figure 3b) with our individual IR
spectra for 1 (Figure 3a) and 2 (Figure 3c) indicates that in
spectrum (b) in Figure 3b, for the condensed vapor of a-
H2CO3, the bands for the cis–trans isomers 2cc and 2ct of the
monomethyl ester can be clearly seen (for full spectra, see
Figure S4). The agreement of the data of the present work
and those of Figure S4 of Ref. [16] for a-H2CO3 is within less
than 0.5 cmꢀ1 (for exact tabulated comparisons, see Tables S9
and S10). The spectral data also agree very well with our
unscaled harmonic CCSD(T)/cc-pVTZ computations for 2ct
and 2cc (Tables S7–S12). The difference of only 0.5 cmꢀ1
between Dn (experiment) = 50.3 cmꢀ1 and Dn (theory) =
and H2O.[12b] At the same time, comparisons of the n(C O)
=
region of carbonic acid vapor isolated in argon after
sublimation of a- and b-H2CO3 (Figure S4 in Ref. [16]) very
clearly show two different sets of carbonyl absorptions for the
11768 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771

Angewandte
Chemie
for the interpretation that “Based on isotopic substitution
experiments and the fact that the product from reaction in
aqueous solution (b-H2CO3) can be converted into the
product obtained by reaction in methanolic solution (a-
H2CO3), we conclude that the two reaction products corre-
spond to two polymorphic modifications of carbonic acid. For
both products alternative assignments have been discussed
and rejected”.[21] As no reagents were added in the sublima-
tion of a-H2CO3, we also conclude that all stud-
ies[4a,6,12,16,18,19,21] relating to the “a-polymorph” in fact must
be assigned to 2 instead. To chemically and spectroscopically
unequivocally confirm that we have indeed made 2 and that
its spectral identity is significantly different from 1, we also
prepared 2 with a perdeuterated methyl group ([D3]-2)
through pyrolysis of di-tert-butyl [D3]-methyl carbonate
under otherwise identical conditions to compare the exper-
imental and computed isotopic shifts exerted on the vibra-
tional bands through deuteration of the methyl group
(Tables S11–S14). As the agreement of the vibrational shifts
is exceptionally good (on average within ꢁ 2 cmꢀ1), both in
sign (i.e., blue- or red-shift) and relative magnitude
(Tables S13 and S14), a methyl group must be present, and
there is no doubt that we have positively identified 2 and
[D3]-2.
In summary, we have presented the first gas-phase
preparation of carbonic acid (1) and its monoalkyl esters, as
exemplified for the monomethyl ester (2), through unim-
olecular ester pyrolysis from readily available dialkyl carbo-
nates. The unequivocal spectral assignments of both 1 and 2,
including isotopic shifts for the deuterated precursor of 2, and
the perfect matching with high-level ab initio computations
allowed us to re-interpret the identity of the molecules
derived from the vapor phase above the a-polymorph of solid
carbonic acid, which must be, as shown in the present work, be
re-assigned to the monoester 2. This casts a new light on the
gas-phase chemistry of 1 and will help in its identification in
environmental, atmospheric, and astrophysical environments.
Figure 3. Comparison of the carbonyl region of the FT-IR spectra.
a) FT-IR spectrum of the matrix-isolated (Ar, 8 K) pyrolysis products of
di-tert-butyl-carbonate (deposition time: 2 h, 7108C, *=co-fragment
isobutene). b) Figure S4 of Ref. [16] (reproduced with permission from
Experimental Section
Preparation of di-tert-butyl carbonate: In a 100 mL flask with
a magnetic stir bar, di-tert-butyl dicarbonate (5 g, 0.023 mol), tert-
butyl alcohol (6.8 g, 0.092 mol), 4-dimethylaminopyridine (1.4 g,
0.011 mol), and triethylamine (4.8 mL, 0.034 mol) were dissolved in
dichloromethane (50 mL). Then, sodium hydride (1.10 g, 0.046 mol)
was added in portions under intense gas evolution. The mixture was
stirred vigorously for two days. Under ice cooling, first water was
added to the reaction mixture and then 0.5m citric acid. The reaction
was extracted several times with diethyl ether, and the combined
organic layers were washed with 0.5m citric acid, water, and brine.
The organic phase was dried with sodium sulfate, and the solvent was
removed by rotary evaporation. The crude product was distilled
under vacuum giving 0.48 g (0.003 mol, 12%) of a colorless oil that
crystallized after some time. 1H NMR (200 MHz, CDCl3): d =
1.43 ppm (s, 18H). 13C NMR (50 MHz, CDCl3): d = 28.0, 81.0,
152.2 ppm. ESI microTOF, positive mode: m/z calcd for
C9H18NaO3: 197.1154 [M+Na]+; found: 197.1154.
=
the authors) with the caption “Comparison of the n(C O) region of
carbonic acid vapor isolated in argon after sublimation of a) b-H2CO3
and b) a-H2CO3”. c) Matrix-isolated pyrolysis products (Ar, 8 K) of tert-
butyl methyl carbonate (deposition time: 2 h, 6508C). *=co-fragment
isobutene, #=traces of formaldehyde; IR bands of unreacted starting
material have been subtracted for clarity. The carbonyl absorptions are
split because of matrix effects.
49.8 cmꢀ1 for the carbonyl absorptions of 2ct and 2cc
underlines our spectral assignments. It is also evident that in
spectrum (a) of Figure 3b, the condensed vapor of b-H2CO3
corresponds to the bands for 1cc and 1ct that agree equally
well with our computations.
Our results strongly suggest that the condensed vapors
that were thought to be originating from a-H2CO3 rather
correspond to the monomethyl ester 2. We found no evidence
Preparation of tert-butyl methyl carbonate: In a 100 mL flask with
a magnetic stir bar, di-tert butyl dicarbonate (3 g, 0.014 mol), 4-
dimethylaminopyridine (0.17 g, 0.001 mol), and triethylamine
(5.1 mL, 0.037 mol) were dissolved in methanol (30 mL). The mixture
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11769

.
Angewandte
Communications
was heated at reflux for 24 h. Then, the reaction was quenched with
0.5m citric acid. Thereafter, the reaction was extracted several times
with dichloromethane, and the combined organic phases were washed
with 0.5m citric acid, water, and brine. The organic phase was dried
with sodium sulfate, and the solvent was removed by rotary
evaporation. The obtained liquid was distilled under atmospheric
pressure to yield 0.73 g (0.005 mol, 39%) of a colorless liquid.
1H NMR (200 MHz, CDCl3): d = 3.63 (s, 3H), 1.41 ppm (s, 9H).
13C NMR (50 MHz, CDCl3): d = 27.9, 54.0, 82.2, 154.3 ppm. ESI
microTOF, positive mode: m/z calcd for C6H12NaO3: 155.0684
[M+Na]+; found: 155.0672.
(ethyl methyl carbonate: 8508C; tert-butyl methyl carbonate: 700–
7508C; di-tert-butyl carbonate (bis(2-methyl-2-propanyl) carbonate):
7108C; di-tert-butyl dicarbonate: 8508C; see also Figure S2). Com-
mercial di-tert-butyl dicarbonate (Sigma–Aldrich) was used without
further purification.
Electronic structure computations: We utilized single-reference
coupled-cluster theory with all single and double excitations (CCSD)
and perturbatively treated triple excitations [CCSD(T)] for the
benchmarking of the carbonic acid conformer energy landscape.[22]
Geometries were optimized at the CCSD(T)/cc-pVQZ level of theory
within the frozen core (FC) approximation (no deleted virtual
orbitals). We applied the focal-point analysis (FPA) of Allen and co-
workers,[17,23] targeting the CCSD(T) complete basis set (CBS) limit.
Herein, the treatment of the electron correlation and the quality of
the basis set are systematically enlarged and thus improved. To assure
smooth convergence towards the CBS limit, we employed Dunningꢀs
correlation-consistent basis set families cc-pVXZ and cc-pCVXZ
(with X = D(2), T(3), Q(4), 5).[24] The SCF energy was extrapolated
with a Feller-type[25] scheme; the dynamic electron correlation energy
was extrapolated with a two-point Helgaker[26] power law. As only
valence electrons are correlated in the focal point tables, the core
correlation was estimated by the following correction:
Preparation of tert-butyl [D3]-methyl carbonate: In a 100 mL
flask with
a magnetic stir bar, di-tert-butyl dicarbonate (6 g,
0.027 mol), [D4]-methanol (1.97 g, 0.056 mol), 4-dimethylaminopyr-
idine (0.33 g, 0.003 mol), and triethylamine (10.2 mL, 0.074 mol) were
dissolved in dichloromethane (60 mL). The mixture was stirred for
five days and then heated at reflux for four hours. The reaction was
quenched with 0.5m citric acid. Thereafter, the reaction was extracted
several times with dichloromethane, and the combined organic layers
were washed with 0.5m citric acid, water, and brine. The organic phase
was dried with sodium sulfate, and the solvent was removed by rotary
evaporation. The obtained liquid was distilled under atmospheric
pressure to yield 0.33 g (0.002 mol, 9%) of a colorless liquid. 1H NMR
(600 MHz, CDCl3): d = 1.37 ppm (s, 9H). 13C NMR (150 MHz,
CDCl3): d = 27.6, 52.9 (sept, J = 22.3 Hz), 81.6, 154.0 ppm. ESI
microTOF, positive mode: m/z calcd for C6H9D3NaO3: 158.0872
[M+Na]+; found: 158.0872.
ccꢀpCVTZ
DEcore ¼ EccꢀpCVTZ
AEꢀCCSDðTÞ ꢀ E
ð1Þ
FCꢀCCSDðTÞ
Relativistic corrections to the energy, including mass velocity
contributions and one-electron Darwin terms (MVD1), were com-
puted at the CCSD(T)/cc-pVTZ level of theory.[27] Additionally, the
diagonal Born–Oppenheimer correction (DBOC) was included in the
final focal-point energies; the latter one was computed at the HF/cc-
pVTZ level of theory.[28] Finally, the CCSD(T)/cc-pVTZ zero-point
vibrational energies (ZPVE) were added to the corresponding
electronic energies. Vibrational frequency computations were also
used to determine the nature of each stationary state (ground or
transition structure). All ab initio computations were performed with
the CFOUR program package employing analytic first and second
derivatives.[29]
For the pyrolyses of the carbonates only a density functional
theory (DFT) approach was possible. The M06-2X functional was
chosen in combination with a cc-pVTZ basis set.[30] All DFT
computations were performed with the Gaussian09 electronic
structure code.[31]
Matrix-isolation studies: The cryostat used for the matrix-
isolation studies was a Leybold RDK 10-320/RW-2 closed-cycle
refrigerator system, whose temperature was controlled by a Leybold
LTC 60 Si-diode temperature controller. The cryostat was equipped
with CsI windows for IR and BaF2 windows for UV/Vis measure-
ments. The noble-gas deposition rates of the matrices were controlled
by an MKS 11798 gas-flow controller that was set to 1–2 sccm. IR
spectra were recorded with a Bruker Vertex 70 FTIR spectrometer
(4500–300 cmꢀ1, resolution 0.7 cmꢀ1). Whenever necessary to avoid
unwanted photochemistry induced by the IR light from the light
source of the spectrometer, spectra were recorded using long-pass IR
filters (l > 4.5 mm). For the combination of high-vacuum flash
pyrolysis (HVFP) with matrix isolation, a small home-built water-
cooled oven that was directly connected to the vacuum shroud of the
cryostat was used. The pyrolysis zone consisted of a resistively heated
completely empty quartz tube (inner diameter 8 mm, length of
heating zone 50 mm). The temperature was controlled by a Ni/CrNi
thermocouple. The precursors were evaporated from a pre-cooled
storage bulb (ethyl methyl carbonate: ꢀ708C; tert-butyl methyl
carbonate: ꢀ508C; di-tert-butyl carbonate: ꢀ458C; di-tert-butyl
dicarbonate: 10–188C) into the quartz pyrolysis tube. Immediately
after leaving the tube, at a distance of approximately 50 mm, the
pyrolysis products were co-condensed with a large excess of argon,
which was introduced by a separate jet on the surface of the 8 K
matrix window. Alternatively, gaseous mixtures (argon, 1:1000) were
prepared for all precursors, except for di-tert-butyl dicarbonate, and
subjected to pyrolysis. For broad-band NIR irradiation of matrix-
isolated samples, light emitted from the IR source (globar) of the
FTIR spectrometer or alternatively light from a high-pressure
mercury lamp (HBO 200, Osram) filtered by a long-pass filter (l >
1100 nm) was used. An optical parametric oscillator (GWU OPO
versaScan 280 MB, pump laser: Spectra-Physics Quanta Ray
Nd:YAG LAB-170-10, 355 nm) was used for narrow-band NIR
irradiation (line width 4 cmꢀ1). Specific cis–trans isomerizations were
induced by narrow-band NIR irradiation using an OPO laser.
Difference spectra of these NIR light-induced isomerization reactions
were particularly well suited for the specific elaboration the IR bands
of the rotamers of carbonic acid. Several experiments were performed
to determine the optimal pyrolysis temperature with respect to the
yield of the carbonic acid or the carbonic acid monomethyl ester
Received: July 7, 2014
Published online: September 3, 2014
Keywords: ab initio calculations · ester pyrolysis ·
.
IR spectroscopy · matrix isolation
[1] S. K. Reddy, S. Balasubramanian, Chem. Commun. 2014, 50, 503.
[2] T. Loerting, C. Tautermann, R. T. Kroemer, I. Kohl, A.
Hallbrucker, E. Mayer, K. R. Liedl, Angew. Chem. Int. Ed.
2000, 39, 891; Angew. Chem. 2000, 112, 919.
[3] G. Strazzulla, J. R. Brucato, G. Cimino, M. E. Palumbo, Planet.
Space Sci. 1996, 44, 1447.
[4] a) W. Hage, K. R. Liedl, A. Hallbrucker, E. Mayer, Science 1998,
279, 1332; b) Y. Oba, N. Watanabe, A. Kouchi, T. Hama, V.
Pirronello, Astrophys. J. 2010, 722, 1598; c) W. Zheng, R. I.
Kaiser, Chem. Phys. Lett. 2007, 450, 55.
[5] M. H. Moore, R. K. Khanna, Spectrochim. Acta Part A 1991, 47,
255.
[6] W. Hage, A. Hallbrucker, E. Mayer, J. Am. Chem. Soc. 1993, 115,
8427.
[7] K. Adamczyk, M. Prꢁmont-Schwarz, D. Pines, E. Pines, E. T. J.
Nibbering, Science 2009, 326, 1690.
11770 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771

Angewandte
Chemie
[8] J. K. Terlouw, C. B. Lebrilla, H. Schwarz, Angew. Chem. Int. Ed.
Engl. 1987, 26, 354; Angew. Chem. 1987, 99, 352.
[9] T. Mori, K. Suma, Y. Sumiyoshi, Y. Endo, J. Chem. Phys. 2009,
130, 204308.
[10] T. Mori, K. Suma, Y. Sumiyoshi, Y. Endo, J. Chem. Phys. 2011,
134, 044319.
[11] G. Bucher, Eur. J. Org. Chem. 2010, 1070.
[12] a) K. Winkel, W. Hage, T. Loerting, S. L. Price, E. Mayer, J. Am.
Chem. Soc. 2007, 129, 13863; b) J. Bernard, M. Seidl, I. Kohl,
K. R. Liedl, E. Mayer, ꢂ. Gꢃlvez, H. Grothe, T. Loerting,
Angew. Chem. Int. Ed. 2011, 50, 1939; Angew. Chem. 2011, 123,
1981; c) C. Mitterdorfer, J. Bernard, F. Klauser, K. Winkel, I.
Kohl, K. R. Liedl, H. Grothe, E. Mayer, T. Loerting, J. Raman
Spectrosc. 2012, 43, 108.
[13] J. Bernard, PhD Thesis, University of Innsbruck (Austria), 2014.
This thesis was sent to the authors on May 9, 2014. J. Barnard
concludes “…the formation of the monomethyl carbonic acid
appears possible.” (p. 106) and “To shed further light on the
nature of the different species, more experiments need to be
done.” (p. 107).
[21] W. Hage, A. Hallbrucker, E. Mayer, J. Mol. Struct. 1997, 408,
527.
[22] a) K. Raghavachari, G. W. Trucks, J. A. Pople, M. Head-Gordon,
Chem. Phys. Lett. 1989, 157, 479; b) R. J. Bartlett, J. D. Watts,
S. A. Kucharski, J. Noga, Chem. Phys. Lett. 1990, 165, 513;
c) R. J. Bartlett, J. D. Watts, S. A. Kucharski, J. Noga, Chem.
Phys. Lett. 1990, 167, 609; d) T. D. Crawford, H. F. Schaefer, in
Reviews in Computational Chemistry, Wiley, Hoboken, 2007,
p. 33.
[23] A. G. Csꢃszꢃr, G. Tarczay, M. L. Leininger, O. L. Polyansky, J.
Tennyson, W. D. Allen in Spectroscopy from Space, Vol. 20
(Eds.: J. Demaison, K. Sarka), Kluwer, Dordrecht, 2001, p. 317.
[24] a) J. T. H. Dunning, J. Chem. Phys. 1989, 90, 1007; b) D. E.
Woon, T. H. Dunning, Jr., J. Chem. Phys. 1995, 103, 4572.
[25] D. Feller, J. Chem. Phys. 1993, 98, 7059.
[26] a) A. Halkier, T. Helgaker, P. Jørgensen, W. Klopper, H. Koch, J.
Olsen, A. K. Wilson, Chem. Phys. Lett. 1998, 286, 243; b) T.
Helgaker, W. Klopper, H. Koch, J. Noga, J. Chem. Phys. 1997,
106, 9639.
[27] W. Klopper, J. Comput. Chem. 1997, 18, 20.
[14] M. R. Rossi, D. T. R. Vidal, C. L. do Lago, Food Chem. 2012,
133, 352.
[28] N. C. Handy, Y. Yamaguchi, H. F. Schaefer, J. Chem. Phys. 1986,
84, 4481.
[15] a) S. Minakata, I. Sasaki, T. Ide, Angew. Chem. Int. Ed. 2010, 49,
1309; Angew. Chem. 2010, 122, 1331; b) K. N. West, C. Wheeler,
J. P. McCarney, K. N. Griffith, D. Bush, C. L. Liotta, C. A.
Eckert, J. Phys. Chem. A 2001, 105, 3947.
[16] J. Bernard, R. G. Huber, K. R. Liedl, H. Grothe, T. Loerting, J.
Am. Chem. Soc. 2013, 135, 7732.
[17] A. G. Csꢃszꢃr, W. D. Allen, H. F. Schaefer, J. Chem. Phys. 1998,
108, 9751.
[18] W. Hage, A. Hallbrucker, E. Mayer, J. Chem. Soc. Faraday Trans.
1996, 92, 3183.
[29] J. F. Stanton, J. Gauss, M. E. Harding, P. G. Szalay, CFOUR,
a quantum chemical program package, with contributions from
A. A. Auer, R. J. Bartlett, U. Benedikt, C. Berger, D. E.
Bernholdt, Y. J. Bomble, L. Cheng, O. Christiansen, M. Heckert,
O. Heun, C. Huber, T.-C. Jagau, D. Jonsson, J. Jusꢁlius, K. Klein,
W. J. Lauderdale, D. A. Matthews, T. Metzroth, D. P. OꢀNeill,
D. R. Price, E. Prochnow, K. Ruud, F. Schiffmann, W. Schwal-
bach, S. Stopkowicz, A. Tajti, J. Vꢃzquez, F. Wang, J. D. Watts,
and the integral packages MOLECULE (J. Almlçf and P. R.
Taylor), PROPS (P. R. Taylor), ABACUS (T. Helgaker,
H. J. Aa. Jensen, P. Jørgensen, and J. Olsen), and ECP routines
by A. V. Mitin and C. van Wꢄllen, 2012.
[19] W. Hage, A. Hallbrucker, E. Mayer, J. Chem. Soc. Faraday Trans.
1995, 91, 2823.
[20] a) G. Gattow, W. Behrendt, Angew. Chem. Int. Ed. Engl. 1972,
11, 534; Angew. Chem. 1972, 84, 549; b) A. Dibenedetto, M.
Aresta, P. Giannoccaro, C. Pastore, I. Pꢃpai, G. Schubert, Eur. J.
Inorg. Chem. 2006, 2006, 908.
[30] Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157.
[31] M. J. T. Frisch et al., Gaussian09, D.01, Wallington CT, 2009.
Angew. Chem. Int. Ed. 2014, 53, 11766 –11771
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11771