Stephanie Ten Brinck1, Celine Nieuwland1, Angela van der Werf1, Richard M P Veenboer1, Harold Linnartz2, F Matthias Bickelhaupt1,3, Célia Fonseca Guerra1,4. 1. Department of Theoretical Chemistry, Amsterdam Institute of Molecular and Life Sciences (AIMMS), Amsterdam Center for Multiscale Modeling (ACMM), Vrije Universiteit Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands. 2. Laboratory for Astrophysics, Leiden Observatory, Leiden University, P.O. Box 9513, 2300 RA Leiden, The Netherlands. 3. Institute for Molecules and Materials (IMM), Radboud University, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands. 4. Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands.
Abstract
It has been experimentally observed that water-ice-embedded polycyclic aromatic hydrocarbons (PAHs) form radical cations when exposed to vacuum UV irradiation, whereas ammonia-embedded PAHs lead to the formation of radical anions. In this study, we explain this phenomenon by investigating the fundamental electronic differences between water and ammonia, the implications of these differences on the PAH-water and PAH-ammonia interaction, and the possible ionization pathways in these complexes using density functional theory (DFT) computations. In the framework of the Kohn-Sham molecular orbital (MO) theory, we show that the ionic state of the PAH photoproducts results from the degree of occupied-occupied MO mixing between the PAHs and the matrix molecules. When interacting with the PAH, the lone pair-type highest occupied molecular orbital (HOMO) of water has poor orbital overlap and is too low in energy to mix with the filled π-orbitals of the PAH. As the lone-pair HOMO of ammonia is significantly higher in energy and has better overlap with filled π-orbitals of the PAH, the subsequent Pauli repulsion leads to mixed MOs with both PAH and ammonia character. By time-dependent DFT calculations, we demonstrate that the formation of mixed PAH-ammonia MOs opens alternative charge-transfer excitation pathways as now electronic density from ammonia can be transferred to unoccupied PAH levels, yielding anionic PAHs. As this pathway is much less available for water-embedded PAHs, charge transfer mainly occurs from localized PAH MOs to mixed PAH-water virtual levels, leading to cationic PAHs.
It has been experimentally observed that water-ice-embedded polycyclic aromatic hydrocarbons (PAHs) form radical cations when exposed to vacuum UV irradiation, whereas ammonia-embedded PAHs lead to the formation of radical anions. In this study, we explain this phenomenon by investigating the fundamental electronic differences between water and ammonia, the implications of these differences on the PAH-water and PAH-ammonia interaction, and the possible ionization pathways in these complexes using density functional theory (DFT) computations. In the framework of the Kohn-Sham molecular orbital (MO) theory, we show that the ionic state of the PAH photoproducts results from the degree of occupied-occupied MO mixing between the PAHs and the matrix molecules. When interacting with the PAH, the lone pair-type highest occupied molecular orbital (HOMO) of water has poor orbital overlap and is too low in energy to mix with the filled π-orbitals of the PAH. As the lone-pair HOMO of ammonia is significantly higher in energy and has better overlap with filled π-orbitals of the PAH, the subsequent Pauli repulsion leads to mixed MOs with both PAH and ammonia character. By time-dependent DFT calculations, we demonstrate that the formation of mixed PAH-ammonia MOs opens alternative charge-transfer excitation pathways as now electronic density from ammonia can be transferred to unoccupied PAH levels, yielding anionic PAHs. As this pathway is much less available for water-embedded PAHs, charge transfer mainly occurs from localized PAH MOs to mixed PAH-water virtual levels, leading to cationic PAHs.
The
presence of polycyclic aromatic hydrocarbons (PAHs) in the
interstellar medium has been proven through the presence of strong
infrared (IR) emission bands between 3.3 and 11.3 μm.[1] For quite some time, these IR emission features
could not be explained and therefore were named the unidentified IR
features. In the 80s, these features were hypothesized to originate
from various stretching and bending modes of vibrationally excited
hydrocarbon systems, such as PAHs, after excitation by photons from
the interstellar radiation field.[1b−1d,1g,2] Nowadays, the existence of interstellar
PAHs is generally accepted, specifically after the recent unambiguous
astronomical identification of a number of aromatic species.[3] Organic molecules such as PAHs play an important
role in interstellar chemistry. They are formed in the outflows of
dying stars, contribute upon fragmentation to the molecular inventory
in space, and are expected to freeze out onto cold dust grains. Here,
they get embedded in an ice matrix comprising mainly water but that
also comprises other species that are expected to form upon the solid-state
hydrogenation of atomic precursors, such as ammonia.[4,5] Although water–ice is the main component of interstellar
ice analogues, observations with the Spitzer Space Telescope showed
that ammonia can be present in water-rich ices, with an abundance
typically 10–20 times lower with respect to water.[6] Upon exposure to high-energy irradiation, reactions
of PAHs or PAH derivatives with the surrounding matrix are expected
to occur. In fact, the vacuum UV (VUV) irradiation and particle bombardment
of these interstellar ices produces complex organic molecules that
possibly played a vital role in the origin of life.[7]Previous in situ spectroscopic studies
have shown
that PAHs embedded in a water–ice matrix are readily ionized.[8,9] For example, Gudipati and Allamandola showed that 4-methylpyrene,
naphthalene, and quaterrylene in water–ice can be ionized by
Lyman-α (10.2 eV) irradiation to their cationic forms.[8] Similarly, Kofman et al. and
Bouwman et al. observed the formation of cationic
triphenylene and pyrene, respectively, in water–ice at 20 K
when exposed to VUV broadband irradiation (120–160 nm) by recording
the time-resolved electronic spectra of these PAHs as a function of
fluence.[9,10] In the study of Bouwman et al., the authors state that the photolysis of pyrene is the result of
direct single-photon ionization of the neutral species. In 2012, however,
Cuylle et al. showed that both pyrene and benzo[ghi]perylene embedded in ammonia–ice become anionic
when exposed to Lyman-α irradiation.[11] The formation of anionic PAHs becomes more prevalent as the ratio
of ammonia to water in mixed ices increases. Cuylle et al. hypothesized that an electron can be transferred to the PAH through
ammonia-related photoproducts, and therefore, anionic PAHs are formed.
They did, however, not provide experimental evidence for how this
process occurs.The purpose of our study is to achieve an understanding
of the
underlying physical factors responsible for the observation that upon
VUV irradiation, the PAHs in water–ice turn into their radical
cations, whereas the PAHs in ammonia–ice rather transform into
the corresponding radical anions. Under interstellar and laboratory
conditions, PAHs are fully surrounded by matrix molecules. Herein,
we aim at disentangling the origin of the observed phenomena from
other bulk effects and investigate if, already at the level of the
interaction between a single representative PAH and a single matrix
molecule,[12] the two opposite tendencies
can be recovered and related to a causal physical mechanism.To understand the effect of the matrix on the photochemical behavior
of the PAH in its essence, we need to go back to the fundamental difference
in the interaction between one PAH molecule and one water versus one ammonia molecule. We first quantum-chemically
analyze the fundamental differences between the electronic structures
of water and ammonia in relation to those of three representative
PAHs, benzene (Ben, C6H6), pyrene
(Py, C16H10), and benzo[ghi]perylene (BgP, C22H12), using Kohn–Sham molecular orbital (MO) theory, as contained
in Kohn–Sham density functional theory (DFT). Second, we analyze
how these differences affect the nature of the PAH–water versus PAH–ammonia interaction with respect to the
MO electronic spectra of the resulting PAH–matrix complexes
and the associated potential ionization routes. Third, we compute
the actual excitations and their oscillator strength using time-dependent
DFT (TDDFT).Here, we anticipate that we find fundamental differences
between
water and ammonia molecules that translate into opposite charge-transfer
excitations in the corresponding complexes with PAHs, leading preferentially
to PAH-to-water excitation (i.e., PAH cation formation
in the case of water) and ammonia-to-PAH excitation (i.e., PAH anion formation in the case of ammonia).
Computational
Details
General Procedure
Calculations were
carried out with the Amsterdam Density Functional (ADF) program, version
2013.01.[13] The generalized gradient approximation
functional BLYP was employed to find the equilibrium structures of
all species using the basis set TZ2P, a small frozen core, and Grimme’s
DFT-D3 correction with Becke–Johnson (BJ) damping to correct
for dispersion effects.[14] The TZ2P basis
set is of valence triple-ζ quality and has been augmented by
two polarization functions.[15] This basis
set is known from previous work to yield relatively small basis-set
superposition errors that can be safely neglected; see ref (15b). Scalar relativistic
corrections were included using the zeroth-order regular approximation
(ZORA) to allow for the possible expansion of this work in the future.[16,17] Numerical integration was performed with the Becke grid set to “verygood”.[18] All geometry optimizations were performed in
the gas phase and without imposing geometric constraints. The equilibrium
geometries were confirmed to be in a (local) minimum-energy state
using vibrational frequency analyses (i.e., no imaginary
frequencies were found).
Fock Matrix-Symmetrized
Fragment Orbitals
(FMATSFOs)
In the ADF program, the electronic structure of
molecular systems, such as our PAH–matrix complexes, can be
analyzed in terms of the interaction between two meaningfully chosen
fragments and the bonding mechanism between the corresponding fragment
MOs. The effective orbital energies of fragments can and, in general,
do change once these fragments are at their positions in the final
complex and experience each other’s electrostatic potential.[19,20] Often, but not always, the effective fragment MO energies are stabilized
when they are exposed to the potential of the other fragment. Note
that this is the actual starting situation from which orbital interactions
take place. These shifted effective fragment orbital energies can
be computed in ADF using the keyword FMATSFO.[19]
TDDFT Computations
Possible charge-transfer
excitations between the PAHs and matrix molecules were investigated
using TDDFT.[21] All optically allowed (i.e., singlet–singlet) vertical excitations were
calculated for the PAH–water and the PAH–ammonia complexes.
The range-separated functional CAMY-B3LYP was used, along with a TZ2P
basis set.[22] Relativistic effects were
accounted for using ZORA.[16,17] The CAMY-B3LYP functional
has shown to provide accurate results for charge-transfer excitations
in naphthalene derivatives and is therefore a suitable functional
to study the PAHs in this work.[23] For more
detailed information on the CAMY-B3LYP functional, we refer to ref (22b).
Results and Discussion
MO Analysis of Ammonia,
Water, and PAHs
Understanding of the ionization pathways
of PAHs in water and ammonia–ice
starts with the quantum chemical analysis of the electronic structure
of the two different-matrix molecules (H2O and NH3). In this analysis, we focus on the highest occupied MOs (HOMOs)
and lowest unoccupied MOs (LUMOs) in particular since these levels
are most likely involved in electronic excitations. Figure shows the optimized geometries
and the HOMO and LUMO orbitals and corresponding energies of water
and ammonia.
Figure 1
HOMO and LUMO energy levels (in eV) of water (W) and
ammonia (A), along with their isosurfaces (at 0.05 au),
calculated at the ZORA-BLYP-D3(BJ)/TZ2P level of theory in the gas
phase.
HOMO and LUMO energy levels (in eV) of water (W) and
ammonia (A), along with their isosurfaces (at 0.05 au),
calculated at the ZORA-BLYP-D3(BJ)/TZ2P level of theory in the gas
phase.The HOMO of ammonia (HOMO) is at a relatively
high energy (−6.1 eV) compared to the HOMO of water (HOMO, −7.2 eV), resulting in an energy difference
between the two HOMOs of 1.1 eV. These HOMOs represent the lone pair
(LP) of ammonia and the highest-energy LP of water (the HOMO –
1 of water, i.e., the other LP, lies at −9.2
eV). In contrast to the HOMOs, the energy difference between the LUMOs
of the two matrix molecules is relatively small, only 0.4 eV, where
the LUMO of ammonia lies at −0.5 eV and the LUMO of water at
−0.9 eV. Qualitatively, similar results are obtained when the
water and ammonia molecule are in a medium of other like molecules
(the HOMO–LUMO gaps increase by ca. 0.5 eV;
see Figure S1).Next, we compare
the relative positioning of the HOMO and LUMO
of ammonia and water to the frontier MOs of pyrene (Py) and benzo[ghi]perylene (BgP), which
are representative for the PAHs used in the experiments by Cuylle et al. [see Figure for the energy levels of Py, W,
and A (left) and BgP, W, and A (right)].[11] The HOMO and LUMO
have π- and π*-like characters, respectively, for both Py and BgP. The smallest intermolecular HOMO–LUMO
gap with regard to water and ammonia is encircled in red (see Figure ). The energy gap
between the HOMO and the LUMO is smaller than the energy gap between the HOMO and the LUMO due to the
higher-lying HOMO. For ammonia, it amounts
to 3.7 and 3.5 eV between HOMO and, respectively,
the LUMO and LUMO. For water, the values are 4.8 and 4.6 eV between the HOMO and, respectively, the LUMOs of Py and BgP. The energy differences between the HOMO of the PAHs and
the LUMOs of ammonia and water are less pronounced as the LUMOs of
the matrix molecules differ less. The smallest energy gap here is
found between the HOMO and the LUMO, being 4.1 and 4.0 eV for Py and BgP, respectively (see also Figure ).
Figure 2
(a) Molecular structures of pyrene (Py) and benzo[ghi]perylene (BgP). (b)
HOMO–LUMO energy
levels (in eV) of water (W) and ammonia (A) compared to Py and (c) BgP, calculated
at the ZORA-BLYP-D3(BJ)/TZ2P level of theory in the gas phase. The
smallest intermolecular energy gaps are encircled in red.
(a) Molecular structures of pyrene (Py) and benzo[ghi]perylene (BgP). (b)
HOMO–LUMO energy
levels (in eV) of water (W) and ammonia (A) compared to Py and (c) BgP, calculated
at the ZORA-BLYP-D3(BJ)/TZ2P level of theory in the gas phase. The
smallest intermolecular energy gaps are encircled in red.Let us consider an electron transfer in these model systems
assuming
that the electron transfer occurs via the smallest
energy gap (see encircled gaps in Figure ). In a PAH–water complex, the electron
is excited over the smaller energy gap from the PAH to water (4.1
and 4.0 eV for Py and BgP) than from water
to the PAH (4.8 and 4.6 eV for Py and BgP). This suggests that it is more likely to form cationic PAHs in
a PAH–water complex than the anionic species. In the case of
PAH–ammonia complexes, it is the other way around. The electron
is more easily excited over the smaller energy gap from ammonia to
the PAH (3.7 and 3.5 eV for Py and BgP,
respectively) than from the PAH to ammonia (4.5 and 4.4 eV for Py and BgP, respectively). Here, the suggestion
is that it is more likely to form anionic PAHs in a PAH–ammonia
complex than cationic PAHs. These findings are consistent with the
observations of Cuylle et al. in their in
situ experiments.[11]The
ground-state energetics of the individual molecules give a
good indication of the electronic differences between ammonia and
water and how the relative HOMO and LUMO energies of ammonia, water,
and the PAHs are responsible for the size of the intermolecular energy
gaps. Note, however, that this simplified picture is an incomplete
representation as charge-transfer excitations that occur in interstellar
ices involve PAHs embedded in them, thus interacting with ammonia
and water. The individual ground-state HOMO–LUMO levels do
not take into account any change in the MOs as a result of the interaction
between the PAHs and the matrix molecules. This coupling or mixing
of the PAH levels with the matrix electronic structure is crucial
for understanding the actual factors determining the transition probability
for PAHs in water–ice versus that of PAHs
in ammonia–ice (vide infra).
MO Interaction Diagrams of PAH–Matrix
Complexes
In the previous section, we analyzed the ground-state
frontier MOs of the individual PAH, water, and ammonia molecules.
For a basic analysis of the MOs of small molecules, we refer the reader
to the book “Orbital Interactions in Chemistry” by T.
A. Albright et al., see ref (24). Next, we quantify the
interactions in the PAH–ammonia and PAH–water complexes
to see how the HOMO–LUMO levels change as a result of the interaction.
Interacting PAH–matrix complexes were created by placing a
single matrix molecule above the π-electronic system of the
aromatic rings and subsequently optimizing the complex (see Figure ). For each of the
final six complexes (labeled BenW, BenA, PyW, PyA, BgPW, and BgPA), energy minima were found. In the optimized geometries, the ammonia
and water molecules direct the N–H or O–H bond(s) toward
the π-system of the PAH, see Figure . The Cartesian coordinates of the optimized
PAH–matrix complexes can be found in the Supporting Information. Note that we find several local minima
for each PAH–matrix combination. These complexes are very similar,
both energetically and in terms of their interaction mechanism but
only differ in the exact position at which the matrix molecule binds
to the π-system of the PAH. Herein, we highlight one of the
obtained complexes per PAH–matrix combination that clearly
demonstrates the fundamental differences between the PAH–ammonia
and PAH–water complexes. For the geometries of the other computed
PAH–matrix complex minima, the reader is referred to Figure S3.
Figure 3
Optimized PAH–matrix complexes
(top view and side view)
with the calculated bond energy shown in parentheses in kcal mol–1, calculated at the ZORA-CAMY-B3LYP-D3(BJ)/TZ2P//ZORA-BLYP-D3(BJ)/TZ2P
level of theory. For the bond energy calculated fully with ZORA-BLYP-D3(BJ)/TZ2P,
see Figure S2.
Optimized PAH–matrix complexes
(top view and side view)
with the calculated bond energy shown in parentheses in kcal mol–1, calculated at the ZORA-CAMY-B3LYP-D3(BJ)/TZ2P//ZORA-BLYP-D3(BJ)/TZ2P
level of theory. For the bond energy calculated fully with ZORA-BLYP-D3(BJ)/TZ2P,
see Figure S2.The interaction between the PAH and matrix molecule when forming
the PAH–matrix complexes is small, with the bond energies ranging
between −2.27 and −3.27 kcal mol–1, which can be considered as a weak N(H)···π
or O(H)···π hydrogen bond interaction. In general,
the bond energy for the PAH–water complexes (−3.02 to
−3.27 kcal mol–1) is more stabilizing than
that for the PAH–ammonia complexes (−2.27 to −2.66
kcal mol–1).Next, the interaction between
the PAH and the matrix molecules
in the PAH–matrix complexes is further investigated. The Py–matrix complexes are highlighted here to provide
general insights on the PAH–matrix interaction as similar results
were found for the Ben–matrix and BgP–matrix complexes. Additional orbital interaction figures
for the Ben–matrix, Py–matrix,
and BgP–matrix complexes can be found in Figures
S4–S9 of the Supporting Information.The interaction between the occupied MOs of pyrene and the
matrix
molecules water and ammonia is given in Figure . The extent of mixing between the PAH and
matrix-molecule orbitals is determined not only by the differences
in energy between the orbitals of the PAHs and the matrix molecules
but also for an important extent by the orbital overlap. Now, we first
inspect how much the extent of mixing is in the case of a water and
an ammonia matrix. For the PyW complex (Figure a), there is very little-to-no
mixing that occurs in the occupied MOs of the PyW complex,
with roughly 98% of the HOMO – 2 originating from the HOMO – 2,
100% of the HOMO – 3 originating
from the HOMO – 3, and 98% of the
HOMO – 4 originating from the HOMO. Thus, the occupied π-orbitals of pyrene
(HOMO – 2 and HOMO – 3) and the LP of water (HOMO) mix minimally and remain mostly localized on the original
fragments.
Figure 4
Orbital interaction diagram of occupied fragment MOs (FMOs) of
(a) PyA and (b) PyW, with FMO contributions
(in %) and visualization of the overall complex MOs (isosurface at
0.03 au), calculated at the ZORA-CAMY-B3LYP/TZ2P level of theory.
The FMO energies are calculated in the field of the other fragment,
as described in Computational DetailsSection , and reported
in Table S3. For a full description of
the occupied MOs, see Figures S6 and S7.
Orbital interaction diagram of occupied fragment MOs (FMOs) of
(a) PyA and (b) PyW, with FMO contributions
(in %) and visualization of the overall complex MOs (isosurface at
0.03 au), calculated at the ZORA-CAMY-B3LYP/TZ2P level of theory.
The FMO energies are calculated in the field of the other fragment,
as described in Computational DetailsSection , and reported
in Table S3. For a full description of
the occupied MOs, see Figures S6 and S7.In contrast, the LP of ammonia
(HOMO)
and a π-orbital of pyrene (HOMO – 2) mix strongly in the PyA complex (Figure b). Two delocalized MOs (HOMO
– 2 and HOMO – 3) are formed, with roughly 14% of the total MO originating
from one monomer while the remaining 85–86% originating from
the other monomer (see Figure b). It is a manifestation of the Pauli exclusion principle
due to the overlap between occupied orbitals of both monomers (see Tables S1 and S2). The HOMO – 3 is a bonding combination, and the HOMO –
2 is an antibonding combination between
the HOMO – 2 and the HOMO. This difference in interaction between the PAH–water
and PAH–ammonia complexes can be attributed to the HOMO energies
of the two matrix molecules and orbital overlap between HOMO of the
matrix molecules (ammonia or water) and π-orbitals of the PAH.Now, we return to the rationale behind the stronger PAH–matrix
orbital mixing in the case of ammonia compared to water. The amount
of mixing between an orbital on one monomer and an orbital on the
other monomer is approximately proportional to the overlap between
the two orbitals on both monomers S (more precisely,
it is proportional to the matrix element between the two orbitals)
and inversely proportional to the energy difference between those
orbitals, Δε. The absolute energy gap between the HOMO of ammonia and the HOMO – 2 (0.24 eV) is much smaller than the gap between the
HOMO of water and the HOMO – 2 (0.84 eV), as can be seen in Figure . Furthermore, the spatial
orientation of the LP of ammonia relative to the PAH allows for better
orbital overlap with the PAH MOs in comparison to water (see also Figures and 3). The reason for this is that the ammonia LP is essentially
the nitrogen 2p (with a slight admixture
of hydrogen 1s character), and due to ammonia’s trigonal pyramidal
geometry, as it points with its N–H bond(s) to the PAH, the
LP also points directly to the π-system of the PAH and, thus,
overlaps well. In contrast, the oxygen 2p-type LP is arranged parallel
to the PAH surface as the water molecule points with its O–H
bond(s) to the PAH (see also Figures and 3). This orbital does not
point with a lobe toward the PAH and therefore has a smaller overlap
with the π-system. For example, in the PyA complex,
the orbital overlap between the LP of ammonia (HOMO) and the π-orbital of pyrene (HOMO – 2) is larger (S = 0.0266) than the
overlap between the LP of water (HOMO) and
the π-orbital of pyrene (HOMO – 2) in the PyW complex, which is only 0.0175 (see Table S2). The orbital overlap between the HOMO and HOMO – 3 is even 1 order of magnitude smaller (S = 0.0017).
These smaller orbital overlaps explain why the MOs of pyrene do not
mix significantly with the HOMO.The
strong mixing in the PyA complex and the poor
mixing in the PyW complex is a clear fundamental difference
between these two complexes. As will be shown in the following section,
the formation of these mixed PAH–ammonia MOs plays an important
role in the available charge-transfer pathways.
Charge-Transfer Excitations in Interacting
PAH–Matrix Complexes
In this section, we analyze the
possible charge-transfer excitation pathways in the PAH–matrix
complexes (Figure ) using TDDFT calculations. All charge-transfer excitations below
10.2 eV (Lyman-α radiation) with an oscillator strength greater
than 0.2 are considered in the exploration of the possibility of charge
transfer upon VUV irradiation. The calculated charge-transfer excitations
are reported in Table for the complexes of Figure . See Table S4 and Figure S11 for
the charge-transfer excitations of other local minima of these PAH–matrix
complexes. In this section, we focus mostly on the Py–matrix complexes as these complexes clearly explain the difference
between PAH–water and PAH–ammonia complexes. All relevant
MOs involved in the Py–matrix excitations are
shown in Figure .
The MOs of the Ben–matrix and BgP–matrix complexes can be found in Figure S10.
Table 1
Charge-Transfer Excitations (in eV)
in PAH−Matrix Complexes Calculated at the ZORA-CAMY-B3LYP/TZ2P
Level of Theorya
complexes
transition (eV)
oscillator strength
weight (%)
from the MO (complex)
to the MO (complex)
character of charge-transfer direction
BenW
7.09
0.443
18.9
HOMO – 1BenW
LUMO + 2BenW
πPAH → σPAH* +
BenA
7.11
0.610
26.4
HOMO – 2BenA
LUMO + 1BenA
πPAH & LPNH3 →
πPAH*
20.3
HOMOBenA
LUMO + 1BenA
πPAH & LPNH3 →
πPAH*
7.10
0.547
25.7
HOMO – 2BenA
LUMOBenA
πPAH & LPNH3 →
πPAH*
19.3
HOMOBenA
LUMOBenA
πPAH & LPNH3 →
πPAH*
PyW
7.35
0.599
12.2
HOMO – 2PyW
LUMO + 4PyW
πPAH → σPAH* +
6.70
0.544
29.3
HOMO – 1PyW
LUMO + 4PyW
πPAH → σPAH* +
7.21
0.418
5.3
HOMO – 2PyW
LUMO + 4PyW
πPAH → σPAH* +
6.66
0.259
57
HOMO – 1PyW
LUMO + 4PyW
πPAH → σPAH* +
PyA
7.30
0.780
62.3
HOMO – 3PyA
LUMO + 3PyA
πPAH & LPNH3 →
πPAH*
6.76
0.675
85.9
HOMO – 3PyA
LUMO + 2PyA
πPAH & LPNH3 →
πPAH*
BgPW
6.53
1.053
5.7
HOMO – 1BgPW
LUMO + 5BgPW
πPAH → σPAH* +
3.7
HOMOBgPW
LUMO + 8BgPW
πPAH → σPAH* +
BgPA
6.95
0.742
31.0
HOMO – 3BgPA
LUMO + 4BgPA
πPAH & LPNH3 →
πPAH*
6.58
0.443
26.7
HOMO – 3BgPA
LUMO + 3BgPA
πPAH & LPNH3 →
πPAH*
8.4
HOMO – 2BgPA
LUMO + 3BgPA
πPAH & LPNH3 →
πPAH*
6.46
0.353
38.6
HOMO – 2BgPA
LUMO + 3BgPA
πPAH & LPNH3 →
πPAH*
6.60
0.268
15.1
HOMO – 3BgPA
LUMO + 3BgPA
πPAH & LPNH3 →
πPAH*
6.13
0.205
23.3
HOMO – 2BgPA
LUMO + 2BgPA
πPAH & LPNH3 →
πPAH*
Lone pairs are abbreviated as LP.
The MOs involved in the charge-transfer excitations (columns 5 and
6) can be found in Figure .
Figure 5
Relevant MOs involved in charge-transfer excitations for PyW (top) and PyA (bottom) with isosurfaces (at
0.03 au) calculated at the CAMY-B3LYP/TZ2P level of theory. Occupied
MOs are red and blue, while unoccupied MOs are cyan and orange. Optimized
structures can be found in Figure . The situation is similar for BenW, BenA, BgPW, and BgPA; see Figures S12 and S13.
Relevant MOs involved in charge-transfer excitations for PyW (top) and PyA (bottom) with isosurfaces (at
0.03 au) calculated at the CAMY-B3LYP/TZ2P level of theory. Occupied
MOs are red and blue, while unoccupied MOs are cyan and orange. Optimized
structures can be found in Figure . The situation is similar for BenW, BenA, BgPW, and BgPA; see Figures S12 and S13.Lone pairs are abbreviated as LP.
The MOs involved in the charge-transfer excitations (columns 5 and
6) can be found in Figure .In the PyW complex, charge-transfer excitations can
occur from the localized π-orbitals of pyrene (HOMO –
1 or HOMO – 2) to a delocalized unoccupied MO consisting of the mixed PAH–water
σ*-orbital with a large coefficient on the water molecule (LUMO
+ 4), see Figure . This direction of charge-transfer excitations
suggests that pyrene donates electronic density to water and therefore
can become positively charged upon irradiation.Interestingly,
the direction of charge-transfer excitations in
the PyA complex is different from the water-embedded
variant PyW. In the PyA complex, charge-transfer
excitations are observed from a mixed MO with ammonia and pyrene π-orbital
character (HOMO – 3) to two different
unoccupied MOs with a π*-like character localized on the pyrene
molecule (LUMO + 2 or LUMO + 3), see Figure . This direction of charge-transfer excitation leads to a
shift of charge density from the occupied mixed MO of the complex
toward the unoccupied orbitals of pyrene and indicates a net electronic
density flow from ammonia to pyrene, leading to a negatively charged
PAH upon irradiation. These results on the dimers of PAH–water
and PAH–ammonia show the same trends as the experimental findings
by Cuylle et al., who reported the formation of anionic
PAHs in an ammonia matrix, while cationic PAHs were formed in a water
matrix.[11]The results for the BenW and BgPW complexes
are similar to those for the PyW complex: charge-transfer
excitations are observed from localized π-orbitals (HOMO –
1 for BenW and HOMO or HOMO – 1 for BgPW) toward mixed PAH–water σ*-MOs (LUMO + 2 for BenW and LUMO + 5 or LUMO + 8 for BgPW). This suggests the formation of cationic PAH radicals
upon the irradiation of PAH–water complexes. For the BenA and BgPA complexes, all observed charge-transfer
excitations originate from mixed MOs of the LP of ammonia and the
π-orbitals of the aromatic molecule (HOMO or HOMO – 2 for BenA and HOMO – 2 or HOMO –
3 for BgPA) toward localized
π*-orbitals of the aromatic molecule (LUMO or LUMO + 1 for BenA and LUMO + 2, LUMO + 3, or LUMO + 4 for BgPA). Accordingly, this direction of charge transfer suggests the formation
of anionic PAH radicals. These results are recovered in all other
PAH–ammonia complexes, which all go with predominant ammonia
to PAH excitations. In the case of other PAH–water complexes,
the situation is not always that consistent, and in some cases, besides
PAH-to-water excitations, there also exist water-to-PAH excitations.In the study by Cuylle et al., it was shown that
changing the ratio of water/ammonia in the ice matrix gives rise to
a turning point in the ionization direction, that is, raising the
ratio water/ammonia led to the formation of cationic PAHs over anionic
PAHs. Our TDDFT findings nicely explain this phenomenon: raising the
ratio water/ammonia implies strengthening the mechanism that leads
to PAH cation formation while lowering the ratio water/ammonia corresponds
to reinforcing the mechanism of PAH anion formation. Note however
that our archetypal model systems do not allow for a prediction of
the water/ammonia ratio at which the turning point occurs.
Conclusions
We have quantum-chemically developed a
rationale for the experimental
observation that water–ice-embedded PAHs become cationic when
exposed to high-energy irradiation, whereas ammonia–ice-embedded
PAHs lead to the formation of anionic PAH photoproducts. Interestingly,
this turns out to be not so much an average medium effect but rather
the consequence of discrete quantum chemical interactions that are
already effective upon bonding one matrix molecule in PAH–water
and PAH–ammonia complexes.Our analyses, based on Kohn–Sham
MO theory and TDDFT, reveal
that the nitrogen LP-type HOMO of ammonia is optimally oriented toward
the aromatic π-system as this matrix molecule binds with its
N–H bond(s) to the PAH. This orientation, and the relatively
high energy of the ammonia LP HOMO, guarantees good overlap and strong
mixing between occupied ammonia and PAH orbitals. Indeed, our TDDFT
analyses confirm that excitations from occupied orbitals of the PAH–ammonia
complex have a strong component of charge transfer from the ammonia
matrix into virtual orbitals that are largely on the PAH. In other
words, this generates PAH radical anions.At variance, the oxygen
LP-type HOMO of water is suboptimally oriented,
namely, in parallel to the aromatic π-system as this matrix
molecule binds with its O–H bond(s) to the PAH. This orientation,
and the somewhat lower energy of the water LP HOMO, impair the buildup
of overlap and the mixing between occupied water and PAH orbitals.
This yields a situation in which excitations take place from occupied
orbitals that remain mainly localized on the PAH into virtuals with
water and PAH character, as perceived from TDDFT analyses. This comes
down to generating PAH radical cations.
Authors: Brett A McGuire; Andrew M Burkhardt; Sergei Kalenskii; Christopher N Shingledecker; Anthony J Remijan; Eric Herbst; Michael C McCarthy Journal: Science Date: 2018-01-12 Impact factor: 47.728
Authors: J Cernicharo; M Agúndez; C Cabezas; B Tercero; N Marcelino; J R Pardo; P de Vicente Journal: Astron Astrophys Date: 2021-05-13 Impact factor: 5.802