The highly reactive benzhydryl cation isolated and stabilized in water ice

Article abstract of DOI:10.1021/ja507894x

Diphenylcarbene (DPC) shows a triplet ground-state lying approximately 3 kcal/mol below the lowest singlet state. Under the conditions of matrix isolation at 25 K, DPC reacts with single water molecules embedded in solid argon and switches its ground state from triplet to singlet by forming a strong hydrogen bond. The complex between DPC and water is only metastable, and even at 3 K the carbene center slowly inserts into the OH bond of water to form benzhydryl alcohol via quantum chemical tunneling. Surprisingly, if DPC is generated in amorphous water ice at 3 K, it is protonated instantaneously to give the benzhydryl cation. Under these conditions, the benzhydryl cation is stable, and warming to temperatures above 50 K is required to produce benzhydryl alcohol. Thus, for the first time, a highly electrophilic and extremely reactive secondary carbenium ion can be isolated in a neutral, nucleophilic environment avoiding superacidic conditions.

Full text of DOI:10.1021/ja507894x

Article
pubs.acs.org/JACS
The Highly Reactive Benzhydryl Cation Isolated and Stabilized in
Water Ice
Paolo Costa,† Miguel Fernandez-Oliva,‡ Elsa Sanchez-Garcia,*,‡ and Wolfram Sander*,†
†Lehrstuhl fur Organische Chemie II, Ruhr-Universitat Bochum, 44780 Bochum, Germany
̈
̈
‡Max-Planck-Institut fur Kohlenforschung, 45470 Mulheim an der Ruhr, Germany
̈
̈
S
* Supporting Information
ABSTRACT: Diphenylcarbene (DPC) shows a triplet ground-state lying
approximately 3 kcal/mol below the lowest singlet state. Under the
conditions of matrix isolation at 25 K, DPC reacts with single water
molecules embedded in solid argon and switches its ground state from
triplet to singlet by forming a strong hydrogen bond. The complex between
DPC and water is only metastable, and even at 3 K the carbene center
slowly inserts into the OH bond of water to form benzhydryl alcohol via
quantum chemical tunneling. Surprisingly, if DPC is generated in
amorphous water ice at 3 K, it is protonated instantaneously to give the
benzhydryl cation. Under these conditions, the benzhydryl cation is stable,
and warming to temperatures above 50 K is required to produce benzhydryl alcohol. Thus, for the first time, a highly electrophilic
and extremely reactive secondary carbenium ion can be isolated in a neutral, nucleophilic environment avoiding superacidic
conditions.
INTRODUCTION
■
Carbenium ions are among the most important reactive
intermediates in chemistry. They play key roles in a very
large number of reactions in the laboratory and in industrial
chemical production.1,2 The electrophilicity, and thus the
reactivity, of carbenium ions stretches over many orders of
magnitude, depending on the substituents at the positively
charged carbon atom.3−6 The benzhydryl cation 2 is a
prototype of a strongly electrophilic secondary cation that in
nucleophilic solvents, such as alcohols, is extremely short-lived
(lifetimes in the order of picoseconds).7 Only in superacidic,
non-nucleophilic solvents, cation 2 is stable enough for
spectroscopic characterization. Thus, in concentrated sulfuric
acid, benzhydryl alcohol 3 is protonated and water eliminated
to form a yellow solution of 2 with an absorption maximum at
Kohler et al. investigated the protonation of 1 using
femtosecond absorption spectroscopy.7 In acetonitrile and
cyclohexane, a transient species with an absorption maximum at
370 nm was assigned to singlet carbene S-1. The intersystem
crossing (ISC) rates for the formation of triplet T-1 was
determined to 340 ps in acetonitrile and to 110 ps in
cyclohexane, in excellent agreement with previous measure-
ments by Eisenthal et al. using picosecond laser-induced
fluorescence.12,13 In neat methanol, S-1 is protonated on an
ultrafast time scale, and 2 is formed with a time constant of only
9 ps.7 This is the fastest known intermolecular proton transfer
to a carbon atom. Under these conditions, cation 2 is very
short-lived and decays with a time constant of 31 ps by reaction
with one of the surrounding methanol molecules.
Recently, we described the reaction of triplet T-1 with single
molecules of CH3OH in argon matrices doped with 0.5−1% of
methanol.14 The singlet state S-1 is highly basic and forms a
very strong hydrogen bond with methanol, while T-1, similar to
doublet radicals,15−18 is only a weak hydrogen-bond acceptor.
Since the stabilization of S-1 compared to T-1 in the methanol
complex is larger than the S−T gap in 1, the spin ground state
of the carbene−methanol complex is shifted from triplet to
singlet. Consequently, the singlet carbene complex S-1···
HOCH3 is formed. This complex is metastable and even at 3
442 nm.8 The pKR value of 3 in sulfuric acid was determined
+
to −13.3, which reveals the high electrophilicity of 2, and
explains its extremely short lifetime in nucleophilic solvents.
Therefore, the spectroscopic characterization of 2 and related
cations requires either superacidic solvents or ultrafast time-
resolved techniques.
The most general route to 2 and related carbenium ions is
the heterolysis of a C−X bond in a benzhydryl derivative or
proton transfer to a singlet carbene such as diphenylcarbene S-
1. Evidence that methanol is able to efficiently protonate
carbene 1 was presented by Kirmse et al. more than 50 years
ago.9,10 Later, Kirmse, Kilian, and Steenken were able to
provide spectroscopic evidence for the formation of 2 by
protonation of 1 using picosecond UV−vis absorption
spectroscopy.11
Received: August 1, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
K slowly reacts to ether 4, the formal product of the OH-
insertion, via quantum chemical tunneling (QMT).
To explore the scope of this solvent-induced spin-flip, we
investigated the reaction of 1 with water under similar
conditions. In addition, we generated 1 in amorphous water
ice at cryogenic temperatures. To our surprise, amorphous
water ice is not only able to protonate 1 at temperatures as low
as 3 K but also to kinetically stabilize the benzhydryl cation 2
and to suppress ion pair recombination. For the first time, a
highly electrophilic and extremely reactive secondary carbe-
nium ion can be isolated in a neutral, nucleophilic environment
avoiding superacidic media.
Figure 1. IR spectra showing the formation of the complex S-1···H2O.
(a) T-1 in argon doped with 1% H2O at 3 K. (b) Difference IR
spectrum showing changes after annealing for 10 min at 25 K. Bands
pointing downward are disappearing and assigned to T-1 and H2O,
bands pointing upward are appearing and assigned to the complex S-
1···H2O. (c) Difference IR spectrum showing changes after 17 h at 3
K. Bands pointing downward are assigned to the complex S-1···H2O,
bands pointing upward are assigned to 3. (d) 3 matrix isolated in argon
at 3 K.
RESULTS AND DISCUSSIONS
■
Reaction of DPC with Single Water Molecules. Triplet
carbene T-1 is obtained by photolysis of diphenyldiazomethane
(DPDM) matrix isolated in argon doped with 0.5−1% water at
3 K. The IR spectrum of such a matrix shows the expected
absorptions of water and of triplet carbene T-1, while reaction
products are not observed. If such a matrix is annealed for
several minutes at 25 K, water is diffusing and can now interact
with the carbene. This results in the formation of a strongly
hydrogen-bonded complex between the singlet carbene S-1 and
one water molecule (Scheme 1, Figure 1). The IR bands of S-1
phase and in simulated argon matrices at the QM/MM level of
theory (BLYP-D3/def2-TZVP//CHARMM, Figure 2). In the
Scheme 1. Reaction of Diphenylcarbene 1 with Water
Figure 2. S-T gaps (kcal/mol) of 1 and its most stable complexes with
water computed at the BLYP-D3/def2-TZVP level of theory (singlet
energies in red, triplet energies in green, all energies including ZPE
correction). Values in blue correspond to the QM/MM (BLYP-D3/
def2-TZVP//CHARMM) simulations of the system in a box of argon
atoms. The red and green boxes represent the range of energies
calculated for different matrix sites obtained from several dynamic
snapshots.
in the water complex are identical to that of the methanol
complex described previously.14 The strongest band of S-1 at
1327.8 cm−1 is assigned to the asym. C−C−C str. vibration of
the carbene center. On 13C labeling of the carbene center, this
band is red-shifted by −21.9 cm−1, in agreement with
predictions from DFT calculations (see Figure S5a).
The UV−vis spectra also show the formation of the singlet
carbene S-1 on annealing of a 1% water-doped argon matrix
containing the triplet carbene T-1 at 25 K: the characteristic
absorption of T-1 in the visible region (λmax = 457 nm) strongly
decreases in intensity, and a new UV band with λmax = 360 nm
appears simultaneously (Figure S6). This is in good agreement
with the observation of a transient band with λmax = 370 nm (in
acetonitrile or cyclohexane as solvent) assigned to S-1 by
Kohler et al. using femtosecond transient absorption spectros-
copy.7
gas phase, the S-T gap of 1 is calculated to 2.8 kcal/mol (3.3
kcal/mol including ZPE correction) which is close to the 2.9
kcal/mol found in a single point CCSD(T)/cc-pVTZ
calculation at the same geometry. In argon, the S-T gap is
reduced to 1.5 kcal/mol (QM/MM) which reflects the stronger
stabilization of the highly polar S-1 by solvation with argon
compared to the unpolar T-1. The S-T gap is inverted by the
interaction of the carbene with a single water molecule, with the
singlet state predicted to be more stable than the triplet by 1.6
kcal/mol in the gas phase and even by 2.4 kcal/mol in argon.
These findings are corroborated by DFT calculations (BLYP-
D3/def2-TZVP) of 1 and the 1···H2O complexes in the gas
B
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The complex S-1···H2O is only metastable and at temper-
atures between 3 and 12 K rearranges with a rate of 5.8 0.2 ×
10−6 to benzhydryl alcohol 3. In this temperature range, the
rearrangement is independent of temperature, which strongly
indicates a tunneling reaction. With D2O a kinetic isotope effect
(KIE) of 3 is observed. Compared to the methanol complex,14
the tunneling rate in the water complex is approximately 1
order of magnitude slower, and the KIE is significantly smaller
(KIE = 5 in the methanol complex). This is in accordance with
the calculations, which predict an activation barrier for the
rearrangement of the water complex of 16.7 kcal/mol (at the
BLYP-D3/def2-TZVP level, 14.1 kcal/mol when including ZPE
correction, 12.6 kcal/mol for single point CCSD(T)/cc-pVTZ
at the same geometry, Figure 3). This barrier is significantly
benzhydryl cation 2 (Figure 4).7,8,11,22,23 The 435 nm
absorption is persistent at low temperature and only at
Figure 4. UV−vis spectra showing the photochemistry of DPDM
isolated in LDA ice at 8 K. (a) DPDM in LDA ice. (b) 3 h irradiation
with λ = 530 nm. The band with λmax = 435 nm is assigned to cation 2.
(c) 10 min irradiation with λ = 450 nm. Bands with λmax = 330 and 257
nm are assigned to radical 5 and alcohol 3, respectively.
annealing above 40 K slowly disappears. The singlet carbene
S-1 is not observed under these conditions, and the EPR
spectrum shows only trace amounts of triplet carbene T-1 (see
Figure S11).
The IR spectrum of cation 2 in LDA ice nicely matches the
calculated gas-phase spectrum of 2 (Figure 5). The line-
Figure 3. Intrinsic reaction coordinate (IRC) for the rearrangement of
the complex S-1···H2O to benzhydryl alcohol 3 calculated at the
BLYP-D3/def2-TZVP level of theory. Energies are given in kcal/mol,
and the CCSD(T)/cc-pVTZ//BLYP-D3/def2-TZVP calculated acti-
vation barrier Ea is shown in red. The structures and some bond
distances of S-1···H2O, 3, and the transition state are shown in the
diagram. In addition, a structure along the IRC reminiscent to a
contact ion pair between cation 2 and OH− is presented.
larger than that calculated for the methanol complex. The ion
pair 2···OH− is neither experimentally observed in the argon
matrix nor predicted by theory in the gas phase (Figure 3) to be
an intermediate in this rearrangement.
Reaction of DPC in Amorphous Water Ice. It was
tempting to use a more polar medium than argon in order to
stabilize the 2···OH− ion pair. We therefore studied the
reaction of 1 in low density amorphous (LDA) water ice.19
LDA ice is characterized by the absence of long-range periodic
structures, and its structure is more similar to liquid water than
to crystalline ice. However, while in liquid water hydrogen
bonds are exchanged on a ps time scale, LDA ice shows
ultraslow reorientation dynamics even at temperatures close to
the glass−liquid transition,20,21 and at 3 K all this motion is
expected to be completely frozen.
Diphenyldiazomethane (DPDM) was slowly sublimed and
trapped with a large excess of water on a spectroscopic window
at 50 K. The LDA ice matrix formed was subsequently cooled
to 3 K and photolyzed with visible light (530 nm). During the
photolysis, the ice matrix turned intense yellow due to a strong
absorption with a maximum at 435 nm. Band position and
shape of this band are identical to that reported for the
Figure 5. IR spectra in LDA ice showing the reaction of the
benzhydryl cation 2 at 100 K. (a) 2 calculated at the BLYP-D3/def2-
TZVP level of theory (scaled by 1.007, line width set to approximate
experimental spectrum). (b) Difference IR spectrum: bands pointing
downward are disappearing upon warming from 3 to 100 K and
assigned to 2. (c) Benzhydryl alcohol 3, matrix-isolated in LDA ice at 3
K.
broadening in the experimental spectrum is attributed to the
inhomogeneity of the matrix cages in LDA ice. To confirm the
assignment of the IR bands of 2, we also generated 13C-2 with a
13C-labeled cationic center. A strong band at 1522 cm−1
exhibits a large 13C isotopic shift of −17.5 cm−1 and is assigned
to the asymmetric C−C−C stretching vibration across the
cationic center. The other absorptions show only small 13C
isotopic shifts as predicted by the BLYP-D3 calculations. In
D2O LDA ice, the deuterium transfer to the singlet carbene S-1
results in the formation of the deuterated carbenium ion d1-2.
C
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Again, the calculated isotopic shifts are in excellent agreement
with the predictions from the BLYP-D3 calculations.
Interestingly, νas C−C−C in d1-12C-2 shows almost the same
isotopic shift as in 13C-2.
carbenes should be even stronger in solvents of higher polarity
and polarizability.
In a series of publications, Gudipati et al. have shown that the
radical cations of aromatic hydrocarbons are easily obtained by
visible light photoionization of the corresponding hydrocarbons
in LDA ice.26−29 Closed-shell cations such as the phenyl, allyl,
or benzyl cation could be generated and isolated by ionization
of matrix-isolated radicals using vacuum ultraviolet or X-ray
irradiation.30−32 Our experiments establish an alternative route
to the isolation of cations in a neutral environment by
protonation of carbenes in LDA ice. At temperatures below 40
K, the benzhydryl cation 2 is entirely stable and does not react
with the water matrix. The IR spectrum of 2 in LDA ice is
nicely matched by gas-phase calculations (Figure 5), indicating
only small solvent effects. Obviously, the OH− produced during
the proton transfer from water is stabilized and immobilized in
ice enough not to recombine with the cation 2, despite the high
exothermicity of this reaction. It is important to note that 2 in
neutral LDA ice is kinetically but not thermodynamically
stabilized. The equilibrium in neutral water is very far on the
To study the interactions between cation 2 and the water
matrix, gas-phase calculations of 2 with a cluster of five water
molecules were carried out. In addition, QM and QM/MM
molecular dynamics simulations (QM MD and QM/MM MD)
with 2 and five water molecules as QM region were performed
in the gas phase (BLYP-D3/def2-SVP) and in a box of explicit
water molecules (BLYP-D3/def2-SVP//CHARMM). Interest-
ingly, cation 2 seems not to form a hydrogen bond with the
surrounding water. The C−H stretching vibration at the
cationic center is predicted to be a very weak band at 3066
cm−1 which is slightly blue-shifted to 3073 cm−1 in the presence
of five water molecules in the gas phase and to 3071 cm−1 in
the water box. νas C−C−C is predicted to show a blue-shift of
+6 cm−1 by interacting with five water molecules and +9 cm−1
in the water box. Thus, there are no specific strong interactions
between 2 and the surrounding water, which rationalizes why
gas-phase calculations of 2 nicely match the experimental
spectra obtained in LDA water ice.
Cation 2 shows an interesting and exceptional photo-
chemistry. Irradiation into the strong visible absorption at
435 nm using a 450 nm LED results in the rapid bleaching of
the yellow color of the matrix. A new strong band with a
maximum at 330 nm showing a weak vibrational progression is
formed concomitantly (Figure 4c), which is assigned to the
benzhydryl radical 5 by comparison with literature spectra.23
The EPR spectrum recorded after the 530 nm photolysis of
DPDM in LDA ice deposited on a copper rod shows only
traces of DPC T-1 (Figure S11), unlike the spectrum obtained
in solid argon, which shows very intense absorptions of T-1.
After the 450 nm irradiation, the EPR spectrum shows very
intense radical signals, whereas the traces of T-1 remain almost
unchanged. This is in line with the assumption that the EPR
silent 2 is the source of the radicals. Due to the line-broadening
in LDA ice it is difficult to assign the radical signals, but it is
tempting to assume that part of the signal is caused by the
benzhydryl radical 5. The OH radical could not be observed in
these experiments.
+
side of benzhydryl alcohol 3, as indicated by the pKR value of
−13.3. In contrast, in superacidic media carbocations are
thermodynamically stable, since the very low pH shifts the
equilibria toward the cations. The high reactivity of cation 2 is
therefore maintained in LDA ice, which explains the
unexpected photochemical electron transfer to produce the
benzhydryl radical 5. In superacidic media we expect radical 5
to be rapidly oxidized to give cation 2.
In summary, we here describe a novel technique to isolate
and kinetically stabilize the highly reactive secondary carbenium
ion 2 in a neutral, very clean environment, avoiding superacidic,
highly ionic media. This provides easy access to the IR and
UV−vis spectra of 2, and in addition its photochemistry can be
investigated without referring to ultrashort time-resolved
spectroscopy. The series of experiments described here allow
us to synthesize and spectroscopically characterize some of the
most important reactive intermediates within the same family
of compounds: diphenylcarbene in its triplet ground state T-1,
the singlet state S-1 stabilized by hydrogen bonding, the
corresponding carbenium ion 2 obtained via proton transfer
from water to S-1, and finally the radical 5 produced via
electron transfer from 2. This paves the way to explore the
chemistry of these species in unprecedented detail.
CONCLUSION
■
The reaction of carbene 1 with single molecules of water under
the conditions of matrix isolation is very similar to the reaction
of 1 with methanol: a strongly hydrogen-bonded complex with
a singlet ground state is formed (Scheme 1, Figure 1). This
gives us the unique opportunity to record IR and UV−vis
spectra of the singlet state of diphenylcarbene S-1. As expected,
the spectra of S-1 in its water complex are almost identical to
that of S-1 in the methanol complex described by us
previously.14 Both complexes react via quantum chemical
tunneling24,25 to give the products of the formal OH insertion.
However, the tunneling of the water complex is about 1 order
of magnitude slower, which is in line with the higher activation
barrier calculated for the OH insertion in water compared to
methanol.
EXPERIMENTAL SECTION
■
Preparation of Low Density Amorphous Ice (LDA). Ultrapure
water was degassed by successive freeze−thaw and held at room
temperature. The water vapors were allowed to co-deposit together
with DPDM into a high-vacuum system in different substrates (Cu
rod, CsI and sapphire windows, respectively, for EPR, FT-IR, and
UV−vis spectroscopy) precooled at 50 K. After finishing the co-
deposition the matrices generated were cooled at the lowest
temperature possible. The deposition rate of the water vapors was
controlled by a fine metering valve.
Matrix Isolation Spectroscopy. Matrix isolation experiments
were performed by standard techniques using Sumitomo Heavy
industries two-staged closed-cycle helium cryostats (cooling power 1
W at 4 K) to obtain temperatures around 3 K. Water was degassed
several times before deposition. The matrices were generated by co-
deposition of DPDM and 1% of water with a large excess of argon
(Messer Griesheim, 99.99%) on top of different substrates (Cu rod,
CsI, and sapphire windows, respectively, for EPR, FT-IR, and UV−vis
spectroscopy) cooled at 3 K. A flow rate of approximately 1.80 sccm
was used for the deposition of the matrix. Diphenylcarbene T-1 was
generated by photolysis of DPDM at 3 K using a 500 W high-pressure
We therefore predict that for carbenes with a S-T gap below
5 kcal/mol, the singlet state should become ground state by
hydrogen bonding to water, alcohols, or other molecules with
similar hydrogen-bond donor capability. Since S-1 is stabilized
compared to T-1 even by the weakly solvating argon, we also
expect that the stabilization of hydrogen-bonded complexes of
D
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
mercury lamp and a cutoff filter with 50% transmission at 530 nm.
Irradiation at λ = 450 nm was performed by using LED source. After
annealing at 25 K for 10 min the matrices were cooled back to 3 K.
The experiments in LDA ice matrix were performed in the same
manner by replacing the argon with water. FTIR spectra were recorded
in the range between 400 and 4000 cm−1 with 0.5 cm−1 resolution.
Matrix EPR spectra were recorded with a Bruker ELEXSYS 500 X-
band spectrometer. Matrix UV−vis spectra were recorded with a
Varian Cary 5000 UV−vis-NIR spectrophotometer in the range of
200−800 nm with a resolution of 0.1 nm.
Computational Details. All gas-phase DFT geometry optimiza-
tions and frequency calculations were carried out using the BLYP
functional33−35 with D3 empirical dispersion correction36 and the
def2-TZVP basis set. The TURBOMOLE program (version 6.4)37 was
employed.
QM MD, QM/MM MD simulations and QM/MM optimizations
were performed using the program ChemShell38,39 as an interface to
TURBOMOLE (version 6.4) and CHARMM 31b1.40 QM MD
simulations were conducted at the BLYP-D3/def2-SVP level of theory,
while QM/MM MD simulations were carried out at the BLYP-D3/
def2-SVP//CHARMM level (10 ps) with a time step of 2 fs under
NVT conditions at temperatures of 25 K (Ar simulations), 3 K (gas
phase and water simulations), and 200 K (water simulations). The
electrostatic interactions were treated using the PME method with a
cutoff distance of 1.2 nm.41 For the QM/MM MD simulations, the
molecules of the QM regions (see below) were placed in boxes of
argon or water (15 Å box padding). Before the QM/MM MD
production runs, the QM atoms were kept frozen, while the MM
regions were allowed to move freely for 5 ns under NPT conditions.
A total of 12 QM MD and QM/MM MD simulations of the
following systems were performed: (1) QM MD: S-1···5H2O and T-
1···5H2O in the gas phase at 3 K; and (2) QM/MM MD: S-1, T-1, S-
1···H2O, and T-1···H2O (QM regions) in a box of explicit Ar at 25 K;
S-1···H2O and T-1···H2O (QM regions) in a box of explicit water at 3
K; with S-1···5H2O and T-1···5H2O as QM regions in a water box at 3
K; with S-1···5H2O as QM region in a box of explicit water at 200 K
and with benzhydryl cation-2 as QM region in a box of water at 3 K.
Ten snapshots were taken from each of the MD simulations and
reoptimized at the BLYP-D3/def2-TZVP//CHARMM level of theory.
Zero-point energies were computed based on harmonic frequency
calculations.
REFERENCES
■
(1) Olah, G. A. J. Org. Chem. 2001, 66, 5943.
(2) Olah, G. A. Angew. Chem., Int. Ed. 1973, 12, 173.
(3) Phan, T. B.; Breugst, M.; Mayr, H. Angew. Chem., Int. Ed. 2006,
45, 3869.
(4) Mayr, H.; Ofial, A. R. Angew. Chem., Int. Ed. 2006, 45, 1844.
(5) Phan, T. B.; Breugst, M.; Mayr, H. Angew. Chem. 2006, 118,
3954.
(6) Mayr, H.; Ofial, A. R. Angew. Chem. 2006, 118, 1876.
(7) Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002, 124,
6428.
(8) Deno, N. C.; Jaruzelski, J. J.; Schriesheim, A. J. Am. Chem. Soc.
1955, 77, 3044.
(9) Kirmse, W. Justus Liebigs Ann. Chem. 1963, 666, 9.
(10) Kirmse, W. Angew. Chem. 1963, 75, 678.
(11) Kirmse, W.; Kilian, J.; Steenken, S. J. Am. Chem. Soc. 1990, 112,
6399.
(12) Eisenthal, K. B.; Turro, N. J.; Aikawa, M.; Butcher, J. A., Jr.;
DuPuy, C.; Hefferon, G.; Hetherington, W.; Korenowski, G. M.;
McAuliffe, M. J. J. Am. Chem. Soc. 1980, 102, 6563.
(13) Eisenthal, K. B.; Moss, R. A.; Turro, N. J. Science 1984, 225,
1439.
(14) Costa, P.; Sander, W. Angew. Chem., Int. Ed. 2014, 53, 5122.
(15) Crespo-Otero, R.; Bravo-Rodriguez, K.; Roy, S.; Benighaus, T.;
Thiel, W.; Sander, W.; Sanchez-Garcia, E. ChemPhysChem 2013, 14,
805.
(16) Mardyukov, A.; Sanchez-Garcia, E.; Crespo-Otero, R.; Sander,
W. Angew. Chem., Int. Ed. Engl. 2009, 48, 4804.
(17) Crespo-Otero, R.; Sanchez-Garcia, E.; Suardiaz, R.; Montero, L.
A.; Sander, W. Chem. Phys. 2008, 353, 193.
(18) Sander, W.; Roy, S.; Polyak, I.; Ramirez-Anguita, J. M.; Sanchez-
Garcia, E. J. Am. Chem. Soc. 2012, 134, 8222.
(19) Tulk, C. A.; Benmore, C. J.; Urquidi, J.; Klug, D. D.; Neuefeind,
J.; Tomberli, B.; Egelstaff, P. A. Science 2002, 297, 1320.
(20) Shalit, A.; Perakis, F.; Hamm, P. J. Phys. Chem. B 2013, 117,
15512.
(21) Low, F.; Amann-Winkel, K.; Loerting, T.; Fujara, F.; Geil, B.
Phys. Chem. Chem. Phys. 2013, 15, 9308.
(22) Ammer, J.; Sailer, C. F.; Riedle, E.; Mayr, H. J. Am. Chem. Soc.
2012, 134, 11481.
(23) Sailer, C. F.; Thallmair, S.; Fingerhut, B. P.; Nolte, C.; Ammer,
J.; Mayr, H.; Pugliesi, I.; de Vivie-Riedle, R.; Riedle, E. ChemPhysChem
2013, 14, 1423.
(24) Ley, D.; Gerbig, D.; Schreiner, P. R. Org. Biomol. Chem. 2012,
10, 3781.
(25) McMahon, R. J. Science 2003, 299, 833.
(26) Gudipati, M. S. J. Phys. Chem. A 2004, 108, 4412.
(27) Gudipati, M. S.; Allamandola, L. J. Astrophys. J. 2004, 615, L177.
(28) Gudipati, M. S.; Allamandola, L. J. J. Phys. Chem. A 2006, 110,
9020.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis of precursors, extended computational methods,
additional spectroscopic data (figures and tables) of the
experiments in argon and water matrix, optimized structures
of computed compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
(29) Gudipati, M. S.; Allamandola, L. J. Astrophys. J. 2006, 638, 286.
(30) Winkler, M.; Sander, W. Angew. Chem., Int. Ed. 2000, 39, 2014.
(31) Winkler, M.; Sander, W. J. Org. Chem. 2006, 71, 6357.
(32) Misic, V.; Piech, K.; Bally, T. J. Am. Chem. Soc. 2013, 135, 8625.
(33) Becke, A. D. Phys. Rev. A 1988, 38, 3098.
(34) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(35) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett.
1989, 157, 200.
AUTHOR INFORMATION
■
Corresponding Authors
wolfram.sander@rub.de
esanchez@kofo.mpg.de
Notes
The authors declare no competing financial interest.
(36) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys.
2010, 132, 154104.
ACKNOWLEDGMENTS
(37) Ahlrichs, R.; Bar, M.; Haser, M.; Horn, H.; Kolmel, C. Chem.
̈
̈
̈
■
Phys. Lett. 1989, 162, 165.
This work was supported by the Cluster of Excellence RESOLV
(EXC 1069) funded by the Deutsche Forschungsgemeinschaft
(DFG). E.S.-G. acknowledges a Liebig Stipend of the Fonds der
Chemischen Industrie and the support of the Collaborative
Research Center SFB1093 funded by the DFG. We thank Dr.
Murthy Gudipati for sharing his expertise in generating
amorphous water ice matrices.
(38) Chemshell, a Computational Chemistry Shell (www.chemshell.
org).
(39) Sherwood, P.; de Vries, A. H.; Guest, M. F.; Schreckenbach, G.;
Catlow, C. R. A.; French, S. A.; Sokol, A. A.; Bromley, S. T.; Thiel, W.;
Turner, A. J.; Billeter, S.; Terstegen, F.; Thiel, S.; Kendrick, J.; Rogers,
S. C.; Casci, J.; Watson, M.; King, F.; Karlsen, E.; Sjovoll, M.; Fahmi,
A.; Schafer, A.; Lennartz, C. J. Mol. Struct. THEOCHEM 2003, 632, 1.
E
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(40) Brooks, B. R.; Brooks, C. L., III; Mackerell, A. D., Jr.; Nilsson,
L.; Petrella, R. J.; Roux, B.; Won, Y.; Archontis, G.; Bartels, C.;
Boresch, S.; Caflisch, A.; Caves, L.; Cui, Q.; Dinner, A. R.; Feig, M.;
Fischer, S.; Gao, J.; Hodoscek, M.; Im, W.; Kuczera, K.; Lazaridis, T.;
Ma, J.; Ovchinnikov, V.; Paci, E.; Pastor, R. W.; Post, C. B.; Pu, J. Z.;
Schaefer, M.; Tidor, B.; Venable, R. M.; Woodcock, H. L.; Wu, X.;
Yang, W.; York, D. M.; Karplus, M. J. Comput. Chem. 2009, 30, 1545.
(41) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.;
Pedersen, L. G. J. Chem. Phys. 1995, 103, 8577.
F
dx.doi.org/10.1021/ja507894x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX