Photodissociation spectroscopy of the Mg+-acetic acid complex
Yohannes Abate and P. D. Kleiber
Citation: J. Chem. Phys. 125, 184310 (2006); doi: 10.1063/1.2386156
View online: http://dx.doi.org/10.1063/1.2386156
View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v125/i18
Published by the American Institute of Physics.
Related Articles
Product angular distributions in the ultraviolet photodissociation of N2O
J. Chem. Phys. 136, 044314 (2012)
Surface modification by subsurface pressure induced diffusion
Appl. Phys. Lett. 100, 041908 (2012)
Ultraviolet photodissociation dynamics of the phenyl radical
J. Chem. Phys. 136, 044308 (2012)
Infrared spectra and ultraviolet-tunable laser induced photochemistry of matrix-isolated phenol and phenol-d5
J. Chem. Phys. 136, 024505 (2012)
Communication: Manipulating the singlet-triplet equilibrium in organic biradical materials
JCP: BioChem. Phys. 5, 12B401 (2011)
Additional information on J. Chem. Phys.
Journal Homepage: http://jcp.aip.org/
Journal Information: http://jcp.aip.org/about/about_the_journal
Top downloads: http://jcp.aip.org/features/most_downloaded
Information for Authors: http://jcp.aip.org/authors
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
THE JOURNAL OF CHEMICAL PHYSICS 125, 184310 共2006兲
Photodissociation spectroscopy of the Mg+-acetic acid complex
Yohannes Abate and P. D. Kleibera兲
Department of Physics and Astronomy, University of Iowa, Iowa City, Iowa 52242 and Optical Science and
Technology Center, University of Iowa, Iowa City, Iowa 52242
共Received 11 September 2006; accepted 9 October 2006; published online 13 November 2006兲
We have studied the structure and photodissociation of Mg+-acetic acid clusters. Ab initio
calculations suggest four relatively strongly bound ground state isomers for the 关MgC2H4O2兴+
complex. These isomers include the cis and trans forms of the Mg+-acetic acid association complex
with Mg+ bonded to the carbonyl O atom of acetic acid, the Mg+-acetic acid association complex
with Mg+ bonded to the hydroxyl O atom of acetic acid, or to a Mg+-ethenediol association
complex. Photodissociation through the Mg+-based 3p ← 3s absorption bands in the near UV leads
to direct 共nonreactive兲 and reactive dissociation products: Mg+, MgOH+, Mg共H2O兲+, CH3CO+, and
MgCH+3 . At low energies the dominant reactive quenching pathway is through dehydration to
Mg共H2O兲+, but additional reaction channels involving C–H and C–C bond activation are also open
at higher energies. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2386156兴
I. INTRODUCTION
Photodissociation spectroscopy of a weakly bound metal
ion-molecule complex provides a useful tool for probing intermolecular interactions, giving quantitative information
about the structure and bonding of the complex and insight
into the molecular dynamics for both reactive and nonreactive energy disposal pathways.1–11 In recent years our group
has used photodissociation spectroscopy to study a range of
metal ion-hydrocarbon clusters, with particular interest in
probing C–H and C–C bond activation processes.1–6 More
recently we have investigated the spectroscopy and chemical
dynamics of a series of metal ion-aldehyde complexes that
support a metal ion-carbonyl bond, including Mg+- and
Zn+-based clusters with formaldehyde and acetaldehyde.4–6
Here we extend these spectroscopic techniques to the study
of Mg+-acetic acid clusters.
Acetic acid is a primary “volatile organic compound”
共VOC兲 that plays an important role in atmospheric chemistry,
especially affecting the HOx budget of the troposphere.12 The
main sources of atmospheric acetic acid are emission from
biomass burning and through the reactions of peroxy acetyl
radicals with CH3O2 and HO2.12,13 Acetic acid has also been
detected in both comets and in star forming regions of space.
The formation of acetic acid in space may occur through
gas-phase chemistry or heterogeneous processes on grain
surfaces.14 Acetic acid can also be destroyed by photolysis in
the UV, vacuum ultraviolet 共vuv兲, and soft x-ray regions, and
the products of high-energy photodissociation include reactive ions and radicals that can then participate in interstellar
chemistry to form more complex organics.14 Acetic acid can
serve as a source for COOH radicals that are an important
intermediate on the path to simple amino acids such as glycine.
The unimolecular dissociation of acetic acid is complicated and has been the subject of extensive investigation for
a兲
Electronic mail: [email protected]
0021-9606/2006/125共18兲/184310/8/$23.00
many years. The thermal decomposition of acetic acid has
been studied in flow systems,15,16 shock tube
experiments,17,18 and through IR mutiphoton dissociation.19
Thermal decomposition in the gas phase occurs through two
competing processes, dehydration and decarboxylation:
CH3COOH + ⌬ → CH4 + CO2
→CH2CO + H2O.
共1兲
共2兲
Rice-Ramsperger-Kassel-Marcus 共RRKM兲 studies show that
the dehydration channel 共2兲 is favored by ⬃2 : 1 over the
lower energy decarboxylation process 共1兲 at an internal energy of ⬃95 kcal/ mol.16 Theoretical and experimental studies suggest that the dehydration reaction 共2兲 involves two
competitive processes, a direct single step reaction by
H-atom transfer from the methyl to the hydroxyl group and
an indirect two-step process that involves a 1,3 H migration
to a 1,1 ethenediol intermediate, followed by a four-center
elimination of water:16,20,21
H3CCOOH + ⌬ → H2C v C共OH兲2
→ H2C v C v O + H2O.
共3兲
Theoretical activation barriers for all three of these processes
共decarboxylation and the direct and indirect dehydration reactions兲 are similar, lying in the range of
71.8– 76.4 kcal/ mol.20,21 These results are somewhat higher,
but in reasonable agreement with the experimental values
that lie in the range of 65– 70 kcal/ mol.16
The UV and vuv photodissociations of acetic acid have
also received a great deal of attention.14,22–26 Photolysis in
the near UV occurs through a spin allowed ␲* ← n transition
centered in the carbonyl group.26 Photodissociation at
218 nm 共5.69 eV兲 preferentially breaks the C–O bond yielding the OH radical,22
125, 184310-1
© 2006 American Institute of Physics
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-2
J. Chem. Phys. 125, 184310 共2006兲
Y. Abate and P. D. Kleiber
CH3COOH + h␯共218 nm兲 → CH3CO + OH.
共4兲
At somewhat higher energies 共h␯ = 6.2 eV兲 secondary decomposition of the acetyl radical CH3CO to CH3 + CO may
also occur.23 The direct single step scission of the weaker
C–C bond is a minor process at these energies. Even this
result, however, is deceptively complicated. Theoretical studies suggest that the photodissociation involves two competing processes, direct dissociation through 共4兲 over a barrier
of ⬃15 kcal/ mol on the excited singlet surface 共S1兲 and an
indirect two-step process that involves intersystem crossing
through spin-orbit coupling to the excited triplet surface 共T1兲,
followed by dissociation on T1 by C–O or C–C bond
breaking:26
CH3COOH + h␯ → CH3COOH共S1兲 → CH3COOH共T1兲
→ CH3CO + OH
共5兲
or
CH3COOH共T1兲 → CH3 + COOH共or CO + OH兲.
共6兲
Apparently the surface-crossing region for internal conversion to the ground state lies too high in energy to be accessible in the near UV.26 This is consistent with the fact that
energetically favored dehydration and decarboxylation reaction products 共CH2CO + H2O or CH4 + CO2兲 have not been
observed following near UV photodissociation of isolated
acetic acid. In the vuv, above the first ionization energy of
acetic acid 共10.58 eV兲 a range of fragmentation and dissociative ionization channels opens, with the products
CH3CO+, COOH+, and CH+3 especially prominent.25
UV photodissociation of acetic acid in Ar matrices at
193 nm shows that more complex chemistry can occur as a
result of cage effects and secondary reactions. In the matrix
environment the dehydration and decarboxylation reaction
photoproducts are detected, as well as CH3OH + CO and
other products.27
The goal of this work is to investigate the near UV photodissociation spectroscopy of Mg共CH3COOH兲+. In this
weakly bound complex Mg+ serves as a chromophore with
strong metal-centered excitation bands in the near UV that
can efficiently couple energy into the complex. In addition to
the Mg+-centered excitations, low energy charge transfer
bands may be accessible that can open additional dissociation pathways involving the acetic acid ion.28 Mg+ may also
serve as both a catalyst and a reagent, which can open new
reactive dissociation channels for acetic acid at relatively low
energies. One interesting question is whether there are low
energy pathways to the formyl radical COOH that may provide novel routes to glycine formation in molecular clouds or
comets at relatively low photon energies.
II. EXPERIMENTAL ARRANGEMENT
The experimental apparatus and its application to massselected cluster photodissociation spectroscopy measurements have been previously described.1 Mg共CH3COOH兲+
complexes are produced in the supersonic molecular beam
expansion from a laser vaporization source. A small reservoir
containing acetic acid was placed near the pulsed gas valve
FIG. 1. Time-of-flight mass spectrum from the laser vaporization source.
The spectrum consists largely of a series of mass peaks corresponding to
Mg共CH3COOH兲n共H2O兲m+ , labeled 共n , m兲 in the figure. Two additional peaks
are obvious: the peak labeled “a” is at mass of 64 amu and is probably
MgAr+, while the assignment for the peak labeled “b” at mass of 126 amu is
uncertain, though it may be Mg共C4H6O3兲+, Mg+-acetic acid anhydride.
共PGV兲 inlet port and connected by a tee to the carrier gas
flow line. The reservoir is heated to increase the acetic acid
vapor pressure. An Ar gas flow carries acetic acid vapor into
the PGV at a pressure of ⬃50 psi.
The second harmonic of a pulsed Nd:YAG 共yttrium aluminum garnet兲 laser is focused onto the surface of a Mg
metal rod placed at the nozzle of the PGV to generate a laser
plasma. Ion-molecule clusters form in the supersonic expansion from the pulsed gas valve. The vaporization laser timing
is adjusted to overlap the seeded gas expansion from the
PGV in order to achieve a strong stable Mg共CH3COOH兲+
parent cluster ion signal. The gas pulse passes downstream
through a molecular beam skimmer and into a differentially
pumped extraction chamber where ion clusters are pulse extracted and accelerated into the flight tube of an angular reflectron time-of-flight mass spectrometer.
The source mass spectrum, shown in Fig. 1, contains a
series of peaks characteristic of the clusters
关Mg共CH3COOH兲n共H2O兲m兴+ 共n 艋 6 and m = 0 or 1兲. Care was
taken in using relatively low vaporization laser power in order to alleviate fragmentation or reaction in the laser plasma,
and the mass spectrum shows no evidence for significant
fragmentation of acetic acid from the source. A pulsed mass
gate is then used to select the Mg共CH3COOH兲+ parent cluster with mass of 84 amu.
The photodissociation laser beam is generated from the
output of an injection-seeded Nd:YAG laser pumped tunable
optical parametric oscillator 共OPO兲 共Spectra-Physics PRO250/MOPO SL兲 coupled with a Quanta Ray wavelength extension 共WEX兲 system for frequency doubling and mixing.
The OPO covers the visible 424– 690 nm spectral region
with a bandwidth of ⬃0.15 cm−1. The near UV region from
212 to 345 nm is reached by frequency doubling the OPO
output in the WEX. The region from 340 to 419 nm is accessed by mixing the OPO visible output with the Nd:YAG
laser fundamental at 1064 nm in the WEX.
In initial experiments, the laser was focused and time
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-3
Photodissociation of Mg+-acetic acid
delayed to excite the mass-selected parent ion at the turning
point inside the reflectron. This arrangement was used to
obtain the photodissociation action spectra shown in the next
section. However, because our mass gate has limited mass
resolution,
parent
clusters
with
masses
near
Mg共CH3COOH兲+ 共primarily the 25Mg+ and 26Mg+-acetic
acid clusters兲 were also present in the reflectron and could be
photolyzed. Furthermore, several of the observed daughter
ions have masses that differ by only 1 – 3 amu, leading to
confusion in the daughter mass assignments. For these reasons we found it more convenient to carry out the product
branching studies by focusing the laser to cross the molecular beam before the reflectron, catching the parent clusters
“on the fly” 共where the parent clusters of different masses are
spatially well resolved兲. This experimental arrangement has
been discussed previously;1 it allows much better mass resolution and selectivity in cases such as this where mass confusion is a potential problem.
In either arrangement, undissociated parents and daughter fragment ions are reflected and reaccelerated into the
flight tube and mass analyzed in a typical tandem time-offlight arrangement. Ions are detected with a microchannel
plate detector, and the signals are measured with a digital
oscilloscope and gated integrator system, interfaced to a
laboratory personal computer for data collection, storage,
and analysis.
Laser power dependence tests were carried out to check
linearity. However, the daughter signals are fairly weak in
this experiment and we could only verify signal linearity
over a range of ⬃3 – 5 in laser power. Over this range the
signals were linear, consistent with a single photon dissociation process.
In order to verify some of the daughter mass assignments and give insight into the reaction mechanism, we carried out additional studies of the photochemical branching
with d-acetic acid 共CH3COOD兲 obtained from Fisher Scientific. Results from these experiments are summarized below.
III. EXPERIMENTAL RESULTS
Figure 2共a兲 shows a range of the daughter ion mass spectrum for photodissociation of 关MgC2H4O兴+ at a laser wavelength of 250 nm, with resolved daughter ion signals observed at masses of 24 共not shown兲, 39, 41, 42, and 43 amu.
We assign these mass peaks to Mg+, MgCH+3 , MgOH+,
Mg共H2O兲+ and CH2CO+, and CH3CO+ photoproducts, respectively. The action spectrum for Mg+ 共Fig. 3兲 shows three
distinct absorption bands, centered near 344, 335, and
250 nm, respectively. The two lower energy bands are redshifted and the higher energy band is blueshifted from the
Mg+共3s ← 3p兲 atomic resonance at 280 nm. All of these absorptions appear as broad continuum bands, with no evidence for any underlying vibrational structure. The photoproducts at masses of 41 共MgOH+兲, 42 共Mg共H2O兲+
+ CH2CO+兲, and 43 amu 共CH3CO+兲 are also observed in all
three bands and the action spectra for these are similar to Fig.
3. The product at mass of 39 amu, MgCH+3 , however, ap-
J. Chem. Phys. 125, 184310 共2006兲
FIG. 2. Time-of-flight mass spectra for the daughter ions in the mass range
of 39– 43 amu resulting from the photodissociation at 250 nm of 共a兲
Mg共CH3COOH兲+ and 共b兲 Mg共CH3COOD兲+. In each case Mg+ at mass of
24 amu is also observed as a major product.
pears only at higher photolysis energies in the blueshifted
250 nm band. The approximate branching ratios at 335 and
250 nm are shown in Table I.
In isotope substitution experiments using d-acetic acid,
CH3COOD, we observe products at masses of 24, 39, 41, 42,
and 43 amu, that we can assign as Mg+, MgCH+3 , MgOH+,
MgOD+ + CH2CO+, and MgOHD+ + CH3CO+, respectively
关Fig. 2共b兲兴. Again, the product MgCH+3 is observed only in
FIG. 3. Mg+ action spectrum for the photodissociation of Mg共CH3COOH兲+.
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-4
J. Chem. Phys. 125, 184310 共2006兲
Y. Abate and P. D. Kleiber
TABLE I. Product branching ratios in the photodissociation of
Mg共CH3COOH兲+.
Mass 共assigned product兲
24
39
41
42
43
amu
amu
amu
amu
amu
共Mg+兲
共MgCH+3 兲
共MgOH+兲
共Mg共H2O兲+ and CH2CO+兲兲
共CH3CO+兲
335 nm
250 nm
0.30
0
0.27
0.41
0.02
0.34
0.08
0.31
0.21
0.06
the blueshifted 250 nm band at the highest energies. The
additional observation of MgOH+ at 41 amu in the d-acetic
acid case is interesting and suggests the possibility of a second reaction channel involving a methyl H atom to form
MgOH+. Indeed, the MgOH+ signal also shows a different
spectral behavior; while MgOD+ is observed in all three
bands, MgOH+ is not observed in the lowest energy band
centered near 344 nm, but only in the higher energy 335 and
250 nm bands.
IV. ELECTRONIC STRUCTURE CALCULATIONS
A. Ground state structures
In order to guide the interpretation of the experimental
results, we have carried out a series of ab initio electronic
structure calculations, investigating the ground and low-lying
doublet excited states of 关MgC2H4O兴+ using the GAUSSIAN98
package.29 Bare acetic acid, CH3COOH, can be found in
either the trans or cis conformations.30 Previous ab initio
calculations on acetic acid show that the cis geometry is
slightly higher in energy by ⬃5 – 6 kcal/ mol, and give a barrier to trans-cis isomerization of ⬃12 kcal/ mol.26,31 The
trans form has been well studied experimentally, but the cis
form has only recently been observed.30 The cis conformer is
unstable and tunnels back to the trans form on a relatively
short time scale, making it difficult to observe.
We have carried out density functional theory 共DFT兲 calculations of the ground state structures of the 关MgC2H4O兴+
complexes. Calculations at the DFT UB3LYP/ 6-31+ + G*
level identify four strongly bound isomers. Three of these are
shown in Fig. 4. The lowest energy structure 共I-trans兲 is
found in Cs symmetry and can be described as a Mg+-acetic
acid association complex where Mg is bound end-on to the
carbonyl O atom, in a Mg+ – O v COHCH3 geometry. The
Mg–O intermolecular bond dissociation energy in this complex is De共Mg– O兲 = 41.3 kcal/ mol 共1.79 eV兲 and the Mg–O
equilibrium bond length is R共Mg– O兲 = 1.98 Å. The acetic
acid ligand in the complex is relatively undistorted from its
isolated equilibrium trans geometry. The Mg+ – O v C bonding in this complex is analogous to that found earlier by our
group for Mg+-acetaldehyde.5
The second most stable isomer 共I-cis兲 is not shown in
Fig. 4. It corresponds to the analogous Mg-acetic acid association complex, but with acetic acid in the cis conformer.
The energy difference between the cis and trans structures is
very small, with I-cis lying only ⬃0.3 kcal/ mol higher in
energy at the UB3LYP/ 6-31+ + G* level, which is well
within the calculation uncertainty. Optimization calculations
with a larger basis set show the energy difference to be
somewhat larger, ⬃3.1 kcal/ mol at the UB3LYP/ 6-311+
FIG. 4. Optimized structures for selected
stable
isomers
of
关Mg, 2C , 2O , 4H兴+. Bond distances are
shown in angstroms and bond angles
in degrees. The isomers I and II have
Cs symmetry and correspond to
Mg+-acetic acid association complexes. Isomer III has C2v symmetry
and corresponds to the Mg+-ethenediol
association complex.
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-5
J. Chem. Phys. 125, 184310 共2006兲
Photodissociation of Mg+-acetic acid
+ G共3df , 3pd兲 level. It is possible that both isomers might
form in our laser vaporization source, but since the trans
form of isolated acetic acid dominates, we expect that I-trans
will be most abundant in our source.
We have located another stable isomer at somewhat
higher energies, shown as structure II in Fig. 4. This isomer
also corresponds to a Mg+-acetic acid association complex in
Cs symmetry, but one where Mg+ is bonded to the hydroxyl
O atom. In this case the Mg+ – O bond length is R共Mg– O兲
= 2.09 Å. The energy for structure II is 19.3 kcal/ mol
共0.84 eV兲 above the global minimum 共I-trans兲, corresponding to an energy of 22.0 kcal/ mol 共0.95 eV兲 below the
Mg+ + acetic acid entrance channel. This isomer may also
form directly by association in the laser vaporization source
and is likely to be present in our molecular beam, though
probably at lower concentrations owing to its weaker binding.
A final stable isomer III can be viewed as an association
complex between Mg+ and the 1,1 ethenediol CH2C共OH兲2.
In this case, Mg+ is bonded in C2v symmetry to the hydroxyl
groups of the enol as shown in Fig. 4. The Mg+ – O bond
lengths in structure III are R共Mg– O兲 = 2.20 Å. The energy
for structure III lies 1.37 eV above the global minimum 共I兲,
or 0.42 eV below the Mg+ + acetic acid entrance channel. 1,1
ethenediol can form from acetic acid through a 1,3 H-atom
migration, but this has an activation barrier of ⬃74 kcal/ mol
共3.2 eV兲 on the ground state surface of isolated acetic
acid.20,21 For this reason we think it unlikely that this isomer
is formed in our source in any significant concentration; we
include it for discussion here because it may play a significant role as an intermediate in the excited state chemistry of
the complex.
A number of other weakly bound association complexes
of the group 关Mg, C2 , O2 , H4兴+ are likely, for example,
Mg共H2O兲共CH2CO兲+ or Mg共CH4兲共CO2兲+ and others 共Fig. 1兲.
However, these species are only very weakly bound and require reaction or fragmentation of the acetic acid in the laser
vaporization source 共for example, to form CH2O or CO2兲. As
noted above, reactions to form these products have substantial activation barriers, ⬎65 kcal/ mol 共2.8 eV兲. We see no
evidence in the source mass spectrum for any appreciable
fragmentation or reaction of acetic acid from the source. For
example, we see no evidence for Mg共CH2CO兲+ or
Mg共CO2兲+ clusters in the source mass spectrum that would
likely be present if fragmentation or reaction in the source
were significant. While we cannot rule these isomers out, we
think it unlikely that they form in any significant concentration in our laser vaporization source.
B. Excited states
To aid in assigning the absorption bands we have also
carried out UCIS level ab initio calculations of the low-lying
spin doublet excited states in this system. Our results indicate that the metal-based transitions correlating with
Mg+共3p ← 3s兲 excitation will dominate the absorption spectrum 共from any of the stable ground state isomers兲. Calculations show that the Mg+共3p␲共1A⬙ , 2A⬘兲兲 states are more attractive than the ground Mg+共3s␴共1A⬘兲兲 state, and the
TABLE II. The UCIS verticl excitation energies 共in eV兲 from isomers
I共trans兲, II, and III are compared with the experimental band positions. Calculated oscillator strengths are given in parentheses.
Assignment
1A⬙ ← 1A⬘
2A⬘ ← 1A⬘
3A⬘ ← 1A⬘
Structure I-trans
Structure II
Structure III
Experiment
3.51 共0.255兲
3.49 共0.244兲
3.42 共0.248兲
3.49
3.56 共0.256兲
3.65 共0.258兲
3.75 共0.259兲
3.60
5.10 共0.319兲
5.11 共0.354兲
5.10 共0.320兲
4.96
corresponding excitation bands are redshifted from the
asymptotic atomic resonance transition. Of these 1A⬙ is predicted to lie slightly lower with a vertical excitation energy
near 3.5 eV. In contrast the Mg+共3p␴共3A⬘兲 ← 3s␴共1A⬘兲兲
band should be blueshifted and lie near ⬃5.1 eV. The UCIS
vertical excitation energies from isomers I-trans, II, and III
are compared in Table II with the experimental band positions.
In addition to these metal-centered transitions, we may
expect a weak ligand based absorption band corresponding
to the ␲*-n transition in the carbonyl group of acetic acid.
This band is found in the range of 220– 200 nm in isolated
acetic acid,22–24 though the CIS calculation places it at much
higher energies 共␭ ⬍ 170 nm兲 in the complex, which is beyond our laser range. We might also expect a metal-ligand
charge transfer band in the near UV. Unfortunately, the CIS
method is not very reliable for predicting the charge transfer
band position. The difference in ionization energies is ⌬IE
= 关IE共acetic acid兲 − IE共Mg兲兴 = 关10.58− 7.65 eV兴 = 2.93 eV.
However, our experience in similar systems is that the observed charge transfer absorption bands typically lie to the
blue of this asymptotic energy difference by ⬃1 – 2 eV.3,32
共This is because the ligand is typically more polarizable and
the charge more localized in the M +-ligand complex, leading
to a stronger binding than in the charge transfer complex
M-ligand+.兲32 Based on this qualitative argument we might
expect the charge transfer absorption band to lie in the range
of ⬃4 – 5 eV, corresponding to ␭ ⬃ 300− 250 nm. We might
then expect the charge transfer band to mix with the
Mg+共3p␴共3A⬘兲 ← 3s␴共1A⬘兲兲 excitation band.
V. ASSIGNMENT
Based on the electronic structure calculations for the excited states we assign the observed absorption bands to predominantly Mg+-based 3p ← 3s transitions in a Mg+-acetic
acid association complex. In particular, we assign the lowest
energy band centered near 344 nm as 3p␲共1A⬙兲
← 3s␴共1A⬘兲, the intermediate energy band centered near
335 nm as 3p␲共2A⬘兲 ← 3s␴共1A⬘兲, and the blueshifted band
near 250 nm as 3p␴共3A⬘兲 ← 3s␴共1A⬘兲. These results are in
good agreement with the UCIS predicted band positions and
oscillator strengths given in Table II for both structures I and
II. Because the overlapping bands are broad and structureless, and the predicted differences in vertical excitation energy and oscillator strengths between the isomers are small,
we cannot distinguish between these isomers in the spectroscopy. The predicted band positions for isomer III are in less
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-6
J. Chem. Phys. 125, 184310 共2006兲
Y. Abate and P. D. Kleiber
good agreement with the experimental data, particularly in
the energy separation between the red bands.
Possible photodissociation channels can be described
schematically:
Mg共CH3COOH兲+ + h␯
→ Mg+ + CH4 + CO2共− 0.37 eV兲,
共7兲
共0.0 eV兲,
共8兲
→Mg+ + CH3COOH
→Mg共H2O兲+ + CH2CO
共0.25 eV兲,
共9兲
→CH3CO+ + MgOH
共0.46 eV兲,
共10兲
→MgOH+ + CH3CO
共0.97 eV兲,
共11兲
→MgCH+3 + COOH
共1.76 eV兲,
→CH2CO+ + Mg + H2O
共3.34 eV兲.
共12兲
共13兲
In each case the neutral products are inferred to give the
lowest energy product channel consistent with the daughter
ion. Other breakup channels are energetically allowed, either
through direct dissociation or secondary fragmentation. Reactions 共7兲–共13兲 also give estimated threshold energies 共in
eV兲 for the different product channels relative to the entrance
channel, Mg+ + CH3COOH. The relevant heats of formation
have been taken from the NIST database.33 The bond energies for Mg共H2O兲+ and MgCH+3 are taken from Refs. 34 and
35, respectively. The spectroscopic threshold for each channel will be higher by the binding energy of the parent complex, which depends on the assumed parent isomer,
⬃1.79 eV for isomer I and ⬃0.95 eV for isomer II.
Note that the daughter ion in reaction 共13兲, CH2CO+ at
mass of 42 amu, lies quite high in energy, with a spectroscopic threshold ⬎4.3 eV 共⬍288 nm兲. This product has the
same mass as Mg共H2O兲+ and MgOD+. Thus, in the lower
energy redshifted bands we can definitively assign the products at mass of 42 amu as Mg共H2O兲+ and MgOD+ in the
h-acetic acid and d-acetic acid cases, respectively. However,
in the highest energy, blueshifted band the situation is more
ambiguous and we cannot rule out some contribution from
reaction 共13兲 to the observed signals at mass of 42 amu.
VI. DISCUSSION
Energy optimization calculations reveal that the isomers
I, II, and III all lie below the Mg+ + CH3COOH entrance
channel. Thus, any or all of these isomers could be present in
the beam from our vaporization source. Quantifying the isomeric proportions is difficult. However, as noted above we
think it unlikely that there is a significant concentration of
isomer III in the beam owing to the high activation barrier to
form 1,1 ethendiol from acetic acid. The predicted vertical
excitation energies and state separations for isomer III are
also in poor agreement with the experimental data.
However, both structures I and II could easily form by
association in our source from Mg+ and acetic acid. The two
large product signals, Mg+ and MgOH+, could then result
from the photodissociation of parent isomers I and II, corre-
sponding to channels 共8兲 and 共11兲. Of course, we cannot rule
out that intracluster chemistry might play a role in forming
these products, but it would be difficult to unravel. In addition, Mg+ can certainly form by secondary breakup of
MgOH+ or Mg共H2O兲+. These products, Mg+ and MgOH+,
are clearly observed in all three bands, with a branching ratio
⬃1 : 1 that is roughly constant across the absorption spectrum
共Table I兲. If these products result from distinct isomers, we
might expect to find differences in the action spectra. However, the UCIS predicted absorption bands for isomers I and
II lie very close in energy, making such discrimination difficult. We have also attempted to change the isomeric ratio by
changing source conditions. We do not see any evidence for
a change in the action spectra or in the product branching
ratios as a function of source conditions. However, the parent
signals are weak in this experiment and we do not have a
large dynamic range to work with.
As noted earlier, MgOH+ is also observed as a weak
product in the photodissociation of Mg共CH3COOD兲+. This
suggests a second reactive dissociation channel that involves
a methyl H atom. Also as discussed above, MgOH+ is only
observed in the higher energy bands corresponding to the
excited states of 2A⬘ and 3A⬘ character, and not in the lowest
energy 1A⬙ ← 1A⬘ band. This could indicate either an orbital
alignment preference or an activation barrier for reaction
with the methyl C–H bond to form MgOH+. One reaction
scheme could involve a 1,3 H-atom migration from the methyl group to the carbonyl O atom in structure I, leading to
the Mg+-enol intermediate structure III followed by breakup
to MgOH+, i.e.,
Mg+共O v CODCH3兲 + h␯ → Mg+共C共OH兲共OD兲CH2兲
→ 共MgOH兲+共CH2COD兲
→ MgOH+
+ CH2COD共or CH2DCO兲.
共14兲
关Of course, MgOD+ and Mg共HDO兲+ could also form as a
result of the breakup of the Mg+-enol intermediate.兴 Such a
process might occur on the ground state surface following
internal conversion. The activation barrier for this isomerization in free acetic acid has been estimated at ⬃74 kcal/ mol
共⬃3.2 eV兲.20,21 This could explain why this product is only
observed at higher energies, ⬎3.7 eV. Mg+-enol might also
be an important intermediate in the reaction to Mg共H2O兲+
products, as in the dehydration reaction of isolated acetic
acid. Of course, we cannot rule out direct reaction to MgOH+
on the excited state surface as well. Direct Mg+共3p兲* attack
on the methyl C–H bond may be possible in a reaction analogous to the chemical quenching of Mg+* by CH4.2
It is also interesting to note that while reaction 共10兲 is
energetically favored over 共11兲, CH3CO+ is a relatively minor product. Furthermore, the branching to CH3CO+ does
not appreciably increase for excitation in the higher energy
band 共where charge transfer interactions might be expected兲.
Above threshold photoionization of isolated acetic acid
yields CH3CO+, COOH+, and CH2CO+ as major products.25
Even in the blueshifted energy band, we see no significant
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-7
J. Chem. Phys. 125, 184310 共2006兲
Photodissociation of Mg+-acetic acid
enhancement in the CH3CO+ or CH2CO+ signals, and no
evidence for COOH+. Thus, we see no evidence to suggest
that charge transfer chemistry plays a significant role in the
photodissociation of Mg共CH3COOH兲+.
Mg共H2O兲+ is also observed as a major reaction product
in all three absorption bands with ketene as the inferred neutral product, although in the higher energy band both channels 共9兲 and 共13兲 are open and could contribute to the signal.
It is interesting to note that Mg共H2O兲+ is the major product
in the lower energy redshifted bands. In addition, some of the
observed Mg+ could well result from secondary breakup of
nascent Mg共H2O兲+ from the dissociation. Thus, the branching fraction into Mg共H2O兲+ in Table I should be considered a
lower limit. This result is surprising and shows that intracluster chemistry plays a significant role in the photodissociation of Mg共CH3COOH兲+. Dehydration is the major decomposition channel of isolated acetic acid at high energies in the
ground state; our result could suggest that photodissociation
of the complex proceeds through internal conversion to the
ground state surface, followed by the direct single step dehydration reaction by H-atom transfer from the methyl group
to the hydroxyl O atom in structure II,
The reaction product MgCH+3 is also interesting in that it
requires C–C bond activation in the photochemical reaction
共12兲 and offers a low energy pathway to the formyl radical
COOH. As expected from the energetics, this channel should
only be open at higher energies in the blueshifted absorption
band. We observed similar evidence for C–C bond activation
in the photodissociation of Mg+-acetaldehyde through the
analogous Mg+共3p␴兲-like excited state, yielding MgCH+3
+ CHO.5 In these examples, once the channel is energetically
open, C–C bond activation by Mg+ appears to be efficient.
These results might suggest a general orbital alignment preference for the C–C bond activation. The initial Mg+
p␴-orbital alignment is with respect to the Mg–O carbonyl
bond in structure I. However, in an in-plane intermolecular
bend this Mg+ p-orbital alignment may become ␲-like with
respect to the C–C bond. In this symmetry a bond stretch
insertion process facilitated by overlap with the localized
␴*-antibonding orbitals centered on the C–C bond might be
possible.
Mg共OHCOCH3兲+ + h␯ → Mg共H2OCOCH2兲+
VII. SUMMARY
→ Mg共H2O兲共COCH2兲+
→ Mg共H2O兲+ + CH2CO,
共15兲
or an indirect two-step process that proceeds through structure III,
Mg共O v COHCH3兲+ + h␯ → Mg共共OH兲2CCH2兲+
→ Mg共H2O兲共OCCH2兲+
→ Mg共H2O兲+ + CH2CO.
共16兲
If this is correct, it raises some interesting issues. For example, the activation barriers on the ground state surface of
isolated acetic acid, for either the direct or indirect dehydration channels, have been experimentally estimated to lie in
the range of 65– 70 kcal/ mol 共2.8– 3.0 eV兲. We observe
Mg共H2O兲+ as the major reaction product in the 1A⬙ ← 1A⬘
band at photon energies of ⬃3.49 eV. This suggests that either the photon energy is efficiently coupled into the reaction
coordinate or the presence of the Mg+ in the complex lowers
the activation barrier for dehydration on the ground state
surface. It is also intriguing that we see no evidence for
Mg共CO2兲+ product that might result from the competing decarboxylation channel on the ground state surface. We speculate that the complex can relax in the excited state toward a
geometry that is similar to the transition state for dehydration, perhaps the enol form 共III兲, and that the surface crossing
for coupling back to the ground state surface occurs in the
neighborhood of the transition state, favoring dehydration.
At higher energies, in the blueshifted absorption band,
branching into the products at mass of 42 amu 共Mg共H2O兲+
+ CH2CO+兲 decreases. This may be due to increased secondary fragmentation of Mg共H2O兲+, the opening of competitive
channels such as 共12兲, or less efficient coupling to dehydration transition state from 3A⬘.
The photodissociation of acetic acid is an important process in both the Earth atmosphere and in astrophysical environments, generating radical fragments 共such as OH,
CH3CO, COOH兲 that can play an important role in atmospheric and cosmic chemistry. The photoproduct COOH is
particularly interesting since it is an important intermediate
on the path simple amino acids such as glycine. Recent studies of the vuv photodissociation of bare acetic acid have been
carried out and have identified a variety of fragmentation
pathways following excitation in the energy range ⬎11 eV.
However, our results show that acetic acid bound in molecular clusters can be efficiently fragmented at much lower photon energies, ⬃3 eV. In Mg共CH3COOH兲+ clusters the Mg+
serves as both an antenna 共efficiently absorbing photons in
its strong allowed resonance transitions兲 and a reaction partner. It may also act as a catalyst to lower activation barriers
for internal conversion or reaction.
In summary, we have studied the structure and photodissociation of Mg+-acetic acid clusters. Ab initio calculations
suggest four relatively strongly bound ground state isomers
for the 关MgC2H4O2兴+ complex. These isomers include both
the cis and trans forms of the Mg+-acetic acid association
complex with Mg+ bonded to the carbonyl O atom of acetic
acid, the Mg+-acetic acid association complex with Mg+
bonded to the hydroxyl O atom of acetic acid, or to a
Mg+-ethenediol association complex. Photodissociation
through the Mg+-based 3p ← 3s absorption bands in the near
UV is efficient, leading to direct 共nonreactive兲 and reactive
dissociation products: Mg+, MgOH+, Mg共H2O兲+, CH3CO+,
and MgCH+3 . At low energies the dominant reactive quenching pathway is through dehydration to Mg共H2O兲+, but additional reaction channels involving C–H and C–C bond activation are also open and efficient at higher energies.
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions
184310-8
ACKNOWLEDGMENTS
The authors gratefully acknowledge the National Science Foundation and The University of Iowa for support of
this work.
P. D. Kleiber and J. Chen, Int. Rev. Phys. Chem. 17, 1 共1998兲.
Y. C. Cheng, J. Chen, L. N. Ding, P. D. Kleiber, and D. K. Liu, J. Chem.
Phys. 104, 6452 共1996兲; J. Chen, T. H. Wong, K. Montgomery, Y. C.
Cheng, and P. D. Kleiber, ibid. 108, 2285 共1998兲.
3
W.-Y. Lu, T. H. Wong, and P. D. Kleiber, Chem. Phys. Lett. 347, 183
共2001兲; W.-Y. Lu, P. D. Kleiber, M. A. Young, and K.-H. Yang, J. Chem.
Phys. 115, 5823 共2001兲.
4
W.-Y. Lu, T.-H. Wong, Y. Sheng, and P. D. Kleiber, J. Chem. Phys. 117,
6970 共2002兲.
5
W.-Y. Lu and P. D. Kleiber, J. Chem. Phys. 114, 10288 共2001兲; Chem.
Phys. Lett. 338, 291 共2001兲.
6
W.-Y. Lu, Y. Abate, T.-H. Wong, and P. D. Kleiber, J. Phys. Chem. A
108, 10661 共2004兲.
7
J. S. Pilgrim, C. S. Yeh, K. F. Willey, D. L. Robbins, and M. A. Duncan,
Int. Rev. Phys. Chem. 13, 231 共1994兲; M. A. Duncan, Annu. Rev. Phys.
Chem. 48, 69 共1997兲; Int. J. Mass. Spectrom. 200, 545 共2000兲.
8
J. M. Farrar, in Cluster Ions, edited by C. Y. Ng, T. Baer, and I. Powis
共Wiley, New York, 1993兲, p. 243; J. I. Lee, D. C. Sperry, and J. M. Farrar,
J. Chem. Phys. 114, 6180 共2001兲.
9
H. Liu, J. Sun, and S. Yang, J. Phys. Chem. A 107, 5681 共2003兲.
10
D. E. Lessen, R. L. Asher, and P. J. Brucat, in Advances in Metal and
Semiconductor Clusters, edited by M. A. Duncan 共JAI, Greenwich, CT,
1983兲, Vol. 1, p. 267; Int. J. Mass Spectrom. Ion Process. 102, 331
共1990兲.
11
F. Aguirre, J. Husband, C. J. Thompson, K. L. Stringer, and R. B. Metz,
J. Chem. Phys. 116, 4071 共2002兲; R. B. Metz, Int. Rev. Phys. Chem. 23,
79 共2004兲.
12
G. S. Tyndall, R. A. Cox, C. Granier, R. Lesclaux, G. K. Moortgat, M. J.
Pilling, A. R. Ravishankara, and T. J. Wallington, J. Geophys. Res. 106,
157 共2001兲.
13
J. Kesselmeier and M. Staudt, J. Atmos. Chem. 33, 23 共1999兲.
1
2
J. Chem. Phys. 125, 184310 共2006兲
Y. Abate and P. D. Kleiber
14
S. Pilling, A. C. F. Santos, and H. M. Boechat-Roberty, Astron. Astrophys. 449, 1289 共2006兲.
15
P. G. Blake and G. E. Jackson, J. Chem Soc. B 1969, 94 共1969兲.
16
N. I. Butkovskaya, G. Manke II, and D. W. Setser, J. Phys. Chem. 99,
11115 共1995兲.
17
J. C. Mackie and K. R. Doolan, Int. J. Chem. Kinet. 16, 525 共1984兲.
18
K. Saito, T. Sasaki, I. Yoshinobu, and A. Imamura, Chem. Phys. Lett.
170, 385 共1990兲.
19
C. A. Longfellow and Y. T. Lee, J. Phys. Chem. 99, 15532 共1995兲.
20
X. Duan and M. Page, J. Am. Chem. Soc. 117, 5114 共1995兲.
21
M. T. Nguyen, D. Senguupta, G. Raspoet, and L. G. Vanquickenborne, J.
Phys. Chem. 99, 11883 共1995兲.
22
S. S. Hunnicutt, L. D. Waits, and J. A. Guest, J. Phys. Chem. 93, 5188
共1989兲.
23
J. C. Orwutsky and A. P. Baronavski, J. Chem. Phys. 111, 7329 共1999兲.
24
P. D. Naik, H. P. Upadhyaya, A. Kumar, A. V. Sapre, and J. P. Mittal,
Chem. Phys. Lett. 340, 116 共2001兲.
25
S. Leach, M. Schwell, H.-W. Jochims, and J. Baumgartel, Chem. Phys.
321, 171 共2006兲.
26
W.-H. Fang, R.-Z. Liu, X. Zheng, and D. L. Phillips, J. Org. Chem. 67,
8407 共2002兲.
27
E. M. S. Macoas, L. Khriachtchev, R. Fausto, and M. Rasanen, J. Phys.
Chem. A 108, 3380 共2004兲.
28
H. Schwarz, D. H. Williams, and C. Wesdemiotis, J. Am. Chem. Soc.
100, 7052 共1978兲.
29
M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., GAUSSIAN 98, Revision
A.6, Gaussian, Inc., Pittsburgh, PA, 1998.
30
E. M. S. Macoas, L. Khriachtchev, M. Pettersson, R. Fausto, and M.
Rasanen, J. Chem. Phys. 121, 1331 共2004兲.
31
M. L. Senent, Mol. Phys. 99, 1311 共2001兲.
32
K. F. Willey, P. Y. Cheng, K. D. Pearce, and M. A. Duncan, J. Phys.
Chem. 94, 4769 共1990兲.
33
NIST Chemistry Webbook 共2006兲, National Institute of Standards and
Technology Reference Database, available from http://webbook.nist.gov
34
N. F. Dalleska, B. L. Tjelta, and P. B. Armentrout, J. Phys. Chem. 98,
4191 共1994兲.
35
C. W. Bauschlicher and H. Partridge, Chem. Phys. Lett. 181, 129 共1991兲.
Downloaded 07 Feb 2012 to 134.139.237.83. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions