Article pubs.acs.org/JPCA
Dissociative Photoionization and Threshold Photoelectron Spectra of Polycyclic Aromatic Hydrocarbon Fragments: An Imaging Photoelectron Photoion Coincidence (iPEPICO) Study of Four Substituted Benzene Radical Cations Brandi West,† Alicia Sit,† Andras Bodi,‡ Patrick Hemberger,‡ and Paul M. Mayer*,† †
Department of Chemistry, University of Ottawa, Ottawa, K1N 6N5, Canada Molecular Dynamics Group, Paul Scherrer Institute, Villigen 5232, Switzerland
‡
S Supporting Information *
ABSTRACT: Four molecules were investigated by imaging photoelectron photoion coincidence (iPEPICO) spectroscopy: 1-propynylbenzene, indene, ethynylbenzene, and benzocyclobutene. Their threshold photoelectron spectrum was obtained and electronic transitions were assigned by OVGF (outer valence Green’s function) calculations. Vibrational progressions observed in the electronic ground and excited states were simulated by calculating Franck−Condon factors based on the neutral as well as the cation ground and excited state geometries. iPEPICO was used to obtain ion dissociation data in threshold photoionization as a function of photon energy, which were modeled with RRKM theory to extract kinetic parameters for the reactions C9H8+• (1-propynylbezene) → C9H7+ + H (R1); C9H8+• (indene) → C9H7+ + H (R2); C8H8+• (benzocyclobutene) → C8H7+ + H (R3); C8H8+• (benzocyclobutene) → C6H6+ + C2H2 (R4); C8H6+• (1-ethynylbenzene, aka phenylacetylene) → C6H4+ + C2H2 (R5). These results were compared to G3 level calculations. In addition, the enthalpy of formation of the indenyl radical was estimated to be ΔfH°0K = 249 ± 50 kJ mol−1 based on a previously measured IE and a cation ΔfH°0K = 976 kJ mol−1, determined herein.
■
INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) are molecules of astronomical importance due to their potential role as catalysts throughout the interstellar medium (ISM).1 One facet of laboratory PAH studies is to focus on their stability in environments similar to that of the ISM. An example of this is to determine their ability to withstand radiation exposure without fragmentation, as this is a requirement in order to exist long enough to facilitate, for example, catalytic reactions.2 Previously, our lab has investigated the vacuum ultraviolet (VUV) radiation induced dissociative photoionization of small PAH and PAH-like molecules, notably naphthalene,3 anthracene,4 pyrene,5 1,2-dihydronaphthalene, and 9,10-dihydrophenanthrene.6 The most notable dissociation pathways for the unsaturated PAH cations is H loss, with hydrocarbon formation (loss of C2H2 and C4H2) steeply declining with increasing size of the ring system. The dihydro-PAHs lose both H and CH3 radicals. In all cases, there is a significant kinetic shift to the observed onset for dissociation, with pyrene exhibiting fragmentation only 7 eV above the ionization threshold and 2 eV above the thermochemical onset for H loss. This trend suggests that larger PAHs will be photostable when ionized by the intense Lyman-α band in outer space. However, what of the smaller molecules that may act as synthetic precursors to PAHs? In this paper we tackle four molecules that can serve as building blocks for PAHs (benzocyclobutene, ethynylbenzene, indene, © 2014 American Chemical Society
and propynlbenzene (Figure 1)) and explore their dissociation chemistry to gain insight on their photostability. Furthermore, these species are soot precursor molecules in flames,7−9 appear in microreactors,10−13 and being able to identify and quantify these molecules and their ions in such experiments is crucial for elucidating the often complex reaction mechanisms taking place in those environments. One approach to identification is through their photoelectron spectra.12,14,15 Herein, we describe the measurement and interpretation of their threshold photoelectron spectra and dissociative ionization behavior through imaging photoelectron photoion coincidence spectroscopy (iPEPICO).
■
EXPERIMENTAL SECTION The samples (1-propynylbenzene, indene, ethynylbenzene, and benzocyclobutene) were obtained from Sigma-Aldrich, Oakville, Canada, and were used without further purification. Imaging Photoelectron Photoion Coincidence Spectroscopy. iPEPICO experiments were conducted at the VUV beamline of the Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland, using the 600 lines/mm laminar grating as described in detail earlier.16−19 In short, the sample is allowed to Received: August 25, 2014 Revised: October 27, 2014 Published: October 27, 2014 11226
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
which the need to reproduce the observed blue shift at the higher field, too, constrains the RRKM model. If two fragment ions are formed in parallel dissociations, the peak shape analysis of only one fragment yields the overall depletion rate of the parent ion already. The breakdown curves can then be used to determine the branching ratios and assign absolute rate constants to both reaction channels. Two of the samples only exhibited H loss, and it was not possible to fit asymmetric TOF profiles of the fragment ions. In fact, the TOF distribution for the composite peak in the molecular ion region had to be approximated as a sum of Gaussian peaks for the [13C−M]+•, [M]+• and [M−H]+ ions, each with the same TOF full width at half-maximum (fwhm) using the IGOR Pro software package in order to even plot the breakdown diagram.22 That is, the experimental TOF distribution was fitted for the three peaks making up the molecular ion region to quantify the ion abundances. Examples of the fitted TOF distributions for the four molecules are shown in Figure 2. Computational Procedures. Ab Initio Calculations. Neutral and ionic structures were calculated for the four sample molecules as well as the primary daughter ions together with the neutral fragments. Three different sets of calculations were completed using Gaussian 09.23 All species were optimized at the B3LYP/6-311+G(d,p) or ωB97XD/6-311++G(d,p) levels of theory. From these calculations, vibrational frequencies, Hessian matrices, rotational constants, and optimized geometries were extracted to be used in the RRKM fitting and in the Franck− Condon (FC) simulations (see below). Single point energy calculations were performed on the B3LYP/6-311+G(d,p) geometries at the MP2/6-311+G(d,p) and G3 levels of theory. Zero point vibrational energies were scaled by 0.967.24 Outer valence Green’s function calculations (OVGF) were carried out with the cc-pVQZ basis set to predict ionization energies to different cation electronic states.25 Geometries of certain electronic excited ion states were optimized applying time dependent density functional theory (TD-DFT). Franck− Condon factors were calculated by eZspectrum.OSX,26 taking the neutral and ion geometries as well as the Hessian matrices as input. The calculated stick spectra were convoluted with a Gaussian function and shifted in energy to match the experimentally observed 0−0 transition. The baseline shift in the TPES due to autoionization was taken into account using a sigmoid function. The experimental uncertainty of the adiabatic ionization energies (IEa) was taken as the half width at halfmaximum of the 0−0 transition. Breakdown Diagram and TOF Distribution Modeling. Rice−Ramsperger−Kassel−Marcus (RRKM) theory was used to calculate unimolecular dissociation rate constants. The standard equation was used to calculated k(E),
Figure 1. 3D representation of the radical cations studied in this work as calculated using the B3LYP/6-311+G(d,p) level of theory: (A) 1propynylbenzene; (B) indene; (C) benzocyclobutene; (D) 1ethynylbenzene.
diffuse into the ionization region of the iPEPICO apparatus through a 6 mm outer diameter Teflon tube, and the neutral gasphase molecules are single photon ionized via monochromatic VUV synchrotron radiation. The photoelectrons are extracted by a constant electric field and velocity map imaged using a delayline MCP detector where each hit is position and time stamped. Meanwhile, the parent and fragment ions are directed toward a nonimaging multichannel plate (MCP) detector, where each ion’s arrival time is recorded. The electron and ion time stamps are correlated to obtain the time-of-flight (TOF) mass spectrum. Threshold electrons hit the detector in a central spot on the MCP and are selected to study threshold photoionization and to plot the breakdown diagram. Generally, electrons initially formed with nonzero kinetic energy (“hot” electrons) are imaged offcenter because of their translational energy component perpendicular to their flight path. Hot electrons that have no lateral momentum will be imaged at the center of the detector, though, and this contamination is approximated based on the signal in a zone consisting of a small ring around the central spot.20 The TOF mass spectrum corresponding to the ring is subtracted from that of the center spot to obtain the true threshold photoionization signal. The threshold photoelectron spectrum was obtained by the same procedure applied to the electron counts at each photon energy to obtain the threshold electrons signal. The time-of-flight mass spectrometer has two acceleration regions, the first of which has a low draw-out field. Ion dissociating with rate constants between 103 and 107 s−1 will produce fragment ions in this region, which are then nonuniformly accelerated toward the second acceleration region, depending on where in this region they are formed. This results in product ion time-of-flight distributions that are broad and asymmetric. The resulting TOF distributions can be modeled to extract absolute unimolecular decay rate constants (Figure 2).21 Because of the limited mass resolution of the TOF mass spectrometer, the small difference between parent and H-loss daughter ion TOF at m/z − 1 does not allow for peak shape analysis. As an alternative, the residence time in the first acceleration region and thus the time scale of the dissociation can be changed by employing two different draw-out fields, in our case 40 and 120 V cm−1, providing two breakdown diagrams, in
k(E) =
σN ⧧(E − E0) hρ(E)
(1)
where N⧧(E − E0) is the transition state number of states, ρ(E) is the density of states of the dissociating ion, h is Planck’s constant, and σ is the degeneracy of the fragmentation channel. For 1propynylbenzene, σ = 3, representing the three hydrogen atoms from the methyl group. For indene, σ = 2, representing the two hydrogen atoms on the sp3 carbon site. For benzocyclobutene, σ = 4 and σ = 1 for hydrogen atom loss and C2H2 loss, respectively, based on fragmentation of the butene ring only. Finally, for 1ethynylbenzene, σ = 1, representing the loss of C2H2 from the linear ethynyl group. 11227
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Figure 2. Representative TOF fits for (A) 1-propynylbenzene (M+• and [M−H]+), (B) indene (M+• and [M−H]+), (C) benzocyclobutene ([M− C2H2]+•), and (D) 1-ethynylbenzene ([M−C2H2]+•), all at 120 V cm−1 draw-out potential, and each at the quoted photon energy.
(where possible) match the experimental data. The final list of vibrational frequencies (Table S1) and k(E) values (Figure S1) for each reaction can be found in Supporting Information.
The k(E) values, together with the photon energy, the roomtemperature internal energy distribution of the neutral sample, the ionization energy of the molecule, and instrumental parameters are combined to produce a theoretical breakdown diagram and fragment ion TOF distributions using the iPEPICO modeling program by Sztáray et al.27 Both the activation energy and entropy are varied in the k(E) calculation until the resulting model breakdown curve and fragment ion TOF distributions
■
RESULTS AND DISCUSSION
Threshold Photoelectron Spectra. The threshold photoelectron spectra (TPES) are shown in Figures 3−6 from threshold up to 15 eV photon energy, along with the vertical 11228
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Table 2. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for Benzenecyclobutene (C2v Aymmetry)
a
G3 level calculation.
Table 3. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for 1Propynylbenzene (CS Symmetry)
a
Figure 3. Threshold photoelectron spectrum of 1-ethynylbenzene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
G3 level calculation.
Table 4. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for Indene (CS Symmetry)
molecular orbital ionization energies from the OVGF calculations (top panel). Note that the MO numbering used to identify the cation electronic states is based on Hartree−Fock calculations and refers to the outer-valence orbitals only (Tables 1−4). The perturbation treatment in the OVGF calculations can Table 1. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for 1Ethynylbenzene (C2v Symmetry)
a
a
G3 level calculation.
The experimentally determined IE for ethynylbenzene is 8.81 ± 0.01 eV, which is in excellent agreement with the MATI value from Kwon et al. of 8.8195 ± 0.006 eV.28 The ground electronic state of the ion as well as the second and third excited states show vibrational structure indicative of good Franck−Condon overlap between the neutral and ion states and long-lived electronic excited states. Figure 3 (second trace) shows a zoom into the TPES of the ion’s ground electronic state as well as a Franck− Condon simulation (sticks and convolution with a Gaussian function with a fwhm of 19 meV). Unless noted otherwise, all vibrationally active modes were found to be totally symmetric, i.e., of a1, a, or a′ symmetry. The feature at 8.75 eV is assigned to a hot band transition. The C−C−C bending vibrations ν13 and ν12, 469 and 764 cm−1, respectively, are active upon ionization and contribute mainly to the bands at 8.87 and 8.9 eV. Note that Mulliken notation is used for the assignment of the vibrational modes,29 whereas literature MATI spectra were assigned using the Wilson notation.28,30 The band at 8.93 eV is assigned to the
G3 level calculation.
change the ordering of the electronic states. The corresponding molecular orbitals can be found in Figure S2 of Supporting Information. Expansions of TPES regions containing vibrational progressions are also presented in each figure. Since the ground state ionization occurs mostly from the π-system connecting rings and substituents, a moderate geometry change is distributed over the whole carbon skeleton, resulting in wellresolved peaks in the TPES. 11229
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
first overtone of ν13 as well as the fundamental of ν11, a ringbreathing mode (995 cm−1). A H−CC−H scissoring motion (ν9 = 1208 cm−1) and a combination band of ν13 and ν12 account for the band at 8.96 eV. Another three bands above 9 eV are due to ν6, ν5, and a combination of ν13 and ν5 vibrational motions possessing frequencies of 1643 (ν6, HCCH stretch) and 2108 cm−1 (ν5, HCC stretch). The mismatch in the line positions can be due to anharmonicity; however, the underestimation of the intensities of the ν13 mode and the non-negative unresolved background at higher photon energies can also be explained by the autoionization of high lying Rydberg states. These neutral states can couple with energetically similar ion states, placing extra intensity in some transitions (enhanced ionization probability) and even lead to threshold photoelectrons in Franck−Condon gaps. Because of the fact that the FC simulations are based on DFT geometries, an overestimation of the bond length change along the ν13 coordinate upon ionization can also lead to this mismatch. While the first excited state (Ã 2A2) at around 9.5 eV possesses only a broad band structure, implying either a large geometry change upon ionization or a short lifetime of this state, the second (B̃ 2B2, 10.33 eV) and third (C̃ 2B1, 11.02 eV) excited states show well resolved structure in the TPES (Figure 3, third and fourth trace). For the B̃ 2B2 state, corresponding to ionization from the HOMO − 2 of b2 symmetry, a progression of ν5 (1998 cm−1, CC stretch) can explain the main features in the spectrum. Additional vibrational modes, mostly different flavors of C−C−C bending (ν13 = 459, ν12 = 758, and ν11 = 998 cm−1) as well as H−CCH scissoring (ν8 = 1222 cm−1) motions, contribute to the less pronounced features. The third excited (C̃ 2 B1) state exhibits activity in the ν12 and ν11 modes at 741 and 955 cm−1, respectively, which reflects a C−C−C bending and a ring breathing vibration of the benzene ring. Apart from these, another C−C−C bending vibration (ν9 = 1143 cm−1), a H− CC−H scissoring (ν8 = 1193 cm−1), and several C−C stretching vibrations (ν7 = 1488, ν6 = 1591, and ν5 = 2081 cm−1) are also active upon photoionization. The experimentally determined IE for benzocyclobutene is 8.65 ± 0.02 eV (see Figure 4), which is in excellent agreement with another threshold photoelectron spectroscopy value of 8.65 eV, obtained in a pyrolysis experiment using the same apparatus.13 The ground state TPE spectrum in Figure 4 is in good agreement with the literature TPE spectrum and FC simulations. Since all the ground state bands were already assigned to their corresponding vibrational transitions,31 the TPE spectrum will not be further discussed. The experimentally determined IE for propylbenzene is 8.40 ± 0.01 eV, in good agreement with the G3-level calculated value of 8.38 eV. Both are slightly lower than previous photoelectron results of 8.41−8.49 eV.32 The ground state TPES (Figure 5, second trace) displays vibrational structure, which was assigned by a Franck−Condon simulation. A progression of a ring deformation mode, mostly due to a C−C−C bending vibration (ν38 = 408 cm−1) located at the substituent, is responsible for the intense feature at 8.45 eV. Above 8.46 eV, several vibrational modes show activity upon ionization: Two ring-deformation modes (ν33 = 712 cm−1 and ν10 = 1663 cm−1) as well as C−CH3 (ν30 = 969 cm−1), CC (ν9 = 2264 cm−1), and C−H3 (ν8 = 3038 cm−1) stretching vibrations can be assigned in the spectrum. A combination band of ν8 and ν45, a CH3 torsional mode at 45 cm−1, is responsible for the feature at 8.78 eV. The third excited state C̃ 2A at 10.57 eV (see third trace in Figure 5) exhibits vibrations of ν33 = 691 and ν28 = 980 cm−1,
Figure 4. Threshold photoelectron spectrum of benzocyclobutene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower pane. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
Figure 5. Threshold photoelectron spectrum of 1-propynylbenzene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
11230
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
CC−H scissoring modes (ν22 = 1085 cm−1 and ν20 = 1159 cm−1) contribute to the remaining features. In summary, we were able to assign almost all vibrational features in the ground as well as excited states of the ion for phenylacetylene, benzocyclobutene, 1-propynylbenzene, and indene by means of Franck−Condon simulations. Because of the moderate change in geometry upon ionization, mostly ring deformation modes, such as C−C stretch and C−C−C bending vibrations, become active. Those transitions as well as the adiabatic ionizations potentials act as the spectroscopic fingerprint of an isomer, which allows for isomer-selective detection in reactive environments as presented and discussed in detail for indene and propynylbenzene.10,11,13 Furthermore, even transitions into electronic excited ion states exhibit a vibrational fingerprint in some cases, in particular when the geometry change is moderate and the lifetime of the excited electronic state is sufficiently long. These fingerprints can further act as an isomer-specific feature when the lower-energy spectrum is too congested for assignment. Dissociative Photoionization. Five reaction channels were observed for the four molecules:
which are similar to those of the ground state because they involve the in-plane distortion of the benzene ring. The feature at around 10.71 eV is somewhat overestimated in the Franck− Condon simulation and can be assigned to in-plane distortion (ν20 = 1203 cm−1) and H−CC−H scissoring motion (ν19 = 1228 cm−1). C−C stretching vibrations of the ring (ν13 = 1503 cm−1 and ν10 = 1703 cm−1) as well as combination bands contribute to the features between 10.75 and 10.9 eV. The combination band between the C−H stretch (ν8 = 3015 cm−1) of the methyl group and the CH3 torsional mode (ν45 = 39 cm−1) is slightly overestimated in the simulation, and their origin is thus difficult to assign in the experimental spectrum. The experimentally determined IE for indene is 8.14 ± 0.01 eV, in agreement with previous photoelectron results that scatter closely around this value (8.13−8.15 eV).32 Both the ground state of the ion and second excited state show vibrational structure. Several vibrational modes contribute to the ground state TPE spectrum (second trace in Figure 6) of indene, which
C9H8+• (1‐propynylbezene) → C9H 7+ + H
(R1)
C9H8+• (indene) → C9H 7+ + H
(R2)
C8H8+• (benzocyclobutene) → C8H 7+ + H
(R3)
C8H8+• (benzocyclobutene) → C6H6+ + C2H 2
(R4)
C8H6+• (1‐ethynylbenzene) → C6H4 + + C2H 2
(R5)
The proposed fragment ion structures for each reaction are presented in Figure 7. Each pair of similar molecules will be discussed together in order to compare their fragmentation reactions. 1-Propynylbenzene (1) vs Indene (3). Both ions dissociate by H loss (reactions R1 and R2). Figure 8A shows the fitted breakdown curve for R1. A low activation energy of 1.86 ± 0.17 eV, coupled with a negative activation entropy, Δ⧧S1000K = −30 ± 31 J K−1 mol−1, may indicate that an isomerization reaction is the rate-determining step. In addition, the G3 reaction energy for the formation of structure 2 by simple bond cleavage is significantly higher at 2.57 eV (Table 5). The only plausible candidate for this isomer is the indene cation as the lowest energy C9H8+ ion isomer. Thus, it is highly likely that ionized 1-propylnylbenzene dissociates to form the indenyl cation 4. The RRKM E0 translates in this case to an isomerization barrier to ionized indene of ∼85 kJ mol−1 relative to the products. The breakdown diagram for H loss from ionized indene (R2) is shown in Figure 8B. The E0 value of 2.56 ± 0.37 eV is 0.7 eV higher in energy than that of R1, while the Δ⧧S1000K was found to be −2 ± 38 J K−1 mol−1. The G3 reaction energy in this case is very close to the RRKM E0 for the reaction (Table 5), suggesting the lack of a reverse barrier, although the transition state cannot be shown to be loose because of the poorly defined activation entropy. A relaxed potential energy surface scan at the B3LYP/6311++G(d,p) level of theory, however, did not indicate a reverse energy barrier to H loss from indene. Assuming no reverse barrier, the appearance energy of 10.7 ± 0.5 eV (or 1032 ± 48 kJ mol−1, AE = IE(indene) + E0) of the fragment ion from neutral indene (ΔfH°0K = 161.2 ± 2.3 kJ mol−1)32 results in a ΔfH°0K for the indenyl cation of 975 ± 48 kJ mol−1. Combined with a recent, accurate determination of the IE of the indenyl radical of 7.53
Figure 6. Threshold photoelectron spectrum of indene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
can be assigned to two different groups. At roughly 1000 cm−1 (0.124 eV) above the IEa, the spectrum is dominated by C−C−C bending vibrations (ν30 = 392 cm−1, ν29 = 525 cm−1, ν28 = 594 cm−1, and ν26 = 830 cm−1). In the 8.3−8.5 eV energy range, C−C stretching modes (ν11 = 1462 cm−1, ν10 = 1507 cm−1, and ν8 = 1650 cm−1) are mostly active upon ionization. Numerous combination bands as well as overtones of the mentioned vibrational modes contribute to the spectra at higher photon energies and cannot be discriminated from the autoionization background. The second excited state B̃ 2A″ has its origin at 10.29 eV (see third trace in Figure 6) and displays several vibrationally active modes that can be divided into three different groups. Ring deformation modes (ν28 = 592 cm−1 and ν27 = 722 cm−1) can explain the features up to 10.4 eV. Two C−C stretching vibrations (ν23 = 982 cm−1, ν11 = 1476 cm−1) as well as two H− 11231
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Figure 7. Reaction schemes for reactions R1−R5 from (A) 1-propynylbenzene, (B) indene, (C) benzocyclobutene, and (D) 1-ethynylbenzene.
Figure 8. Experimental iPEPICO breakdown curves for the four molecules studied, (A) 1-propynylbenzene, (B) indene, (C) benzocyclobutene, and (D) ethynylbenzene, all taken at 120 V cm−1 draw-out field.
eV,33 this leads to a ΔfH°0K of the indenyl radical of 248 ± 50 kJ mol−1. Benzocyclobutene (5) vs 1-Ethynylbenzene (8). Unlike 1-propynylbenzene and indene, benzocyclobutene (5) has two parallel, low-energy dissociative photoionization channels, dehydrogenation (reaction R3) and the loss of C2H2 (reaction R4). The iPEPICO experiments were conducted in the 11.5−15
eV photon energy range. The fragment ion of R3 is assigned as structure 6, as there are only the two identical sp3 carbon sites in the parent. Reaction R4 is assumed to lead to the benzene cation, as it is the energetically most favored C6H6+• ion. 1Ethynylbenzene (8), on the other hand, only undergoes loss of C2H2 (reaction R5) in the photon energy range studied, yielding the proposed structure 9, benzyne cation. 11232
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Table 5. Energetics Data for Reactions R1−R5 for the Four Molecules Studied. ΔE are the G3-Calculated Dissociation Energies for each Reaction; E0, Δ⧧S1000K, and AE are Determined from the RRKM Fitting of the Experimental Breakdown Curves C9H8+• C9H8+• C8H8+• C8H8+• C8H6+• a
→ → → → →
C9H7+ + H (R1) C9H7+ + H (R2) C8H7+ + H (R3) C6H6+• + C2H2 (R4) C6H4+• + C2H2 (R5)
ΔE (G3)/eV
E0/eV
2.57 2.78 2.33 1.70 4.54
1.86 ± 0.17 2.56 ± 0.37a 2.23 ± 0.10 2.00 ± 0.10 3.94 ± 0.43
Δ⧧S1000K/J K−1 mol−1
AE/eV
−30 ± 31a −2 ± 38a −11 ± 8 −17 ± 9 42 ± 36
10.26 ± 0.27 10.70 ± 0.50 10.88 ± 0.21 10.65 ± 0.27 12.75 ± 0.42
a
Best values obtained by simultaneously fitting breakdown diagrams obtained with ion extraction potentials of 40 and 120 V cm−1.
characterized by a barrier of 85 kJ mol−1 relative to products. The other two molecules were found to ionize and generate C2H2 only at energies above Lyman-α (10.2 eV). These results and those for other sp3-carbon containing PAHs6 indicate that their dissociation thresholds tend to be ∼2 eV lower in energy compared to systems containing only sp2-carbon atoms. However, all have been shown to dissociate above 10.2 eV, indicating that they may indeed survive to act as precursors for larger PAH systems.
Figure 8C shows the breakdown diagram for benzocyclobutene. For reaction R3, the E0 and Δ⧧S1000K values were found to be 2.23 ± 0.10 eV and −11 ± 8 J K−1 mol−1, respectively. The competing channel R4, which dominates at most energies studied, has an E0 of 2.00 ± 0.10 eV and a Δ⧧S1000K value of −17 ± 9 J K−1 mol−1. The low E0 and negative activation entropy are consistent with the fact that loss of C2H2 necessarily involves a rearrangement step leading to the stable benzene ion, which is also identified to be rate-determining. This is also consistent with the fact that the G3 reaction energy is 0.3 eV lower than the RRKM E0 for the reaction (Table 5). Conversely, the energy requirement for reaction R5 is substantially higher at 3.94 ± 0.43 eV (Figure 8D shows the breakdown curve for 1-ethynylbenzene). The activation entropy of Δ⧧S1000K = 42 ± 36 J K−1 mol−1 is indicative of a loose transition state, one that possibly involves an ion−molecule complex, as was previously calculated for the loss of acetylene from the anthracene radical cation. The G3 ΔE for reaction R5 was significantly larger than the derived E0 (4.54 eV compared to 3.94 ± 0.43 eV). While there is a significant degree of variation in the calculated reaction energies as a function of level of theory (Table S2), it is possible that this result also means ionized ethynylbenzene isomerizes prior to dissociation. The only candidate structure that seems reasonable is ionized benzocyclobutadiene, as ethynylbenzene would have to adopt a structure similar to this prior to losing C2H2. However, the fragmentation products are the same in each case, ionized benzyne and C2H2. Thus, isomerization cannot account for the discrepancy between the G3 and RRKM values, since B3LYP/6311+G(d,p) level calculations place ionized ethynyl benzene lower in energy by only 5 kJ mol−1.
■
ASSOCIATED CONTENT
S Supporting Information *
Full authorship for ref 23, Table S1 containing the vibrational frequencies for the processes modeled with eq 1 for both ions, Table S2 containing a comparison of levels of theory used to calculate the energy thresholds for the dissociation reactions in this paper, Figure S1 containing the descriptions for the MOs listed in Tables 1−4, and Figure S2 containing the k(E) data derived from the RRKM fitting of the experimental data. This material is available free of charge via the Internet at http://pubs. acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Phone: 1-613-562-5800, extension 6038. Notes
The authors declare no competing financial interest.
■
■
ACKNOWLEDGMENTS P.M.M. thanks the Natural Sciences and Engineering Research Council of Canada for continuing financial support. Experiments were carried out at the VUV beamline of the Swiss Light Source, Paul Scherrer Institute (PSI). The work was financially supported by the Swiss Federal Office for Energy (BFE Contract 101969/152433).
CONCLUSIONS Threshold photoelectron spectra were obtained for the four target molecules. In general, the measured ground state IEs were in good agreement with literature values and G3-level calculations. Franck−Condon simulations successfully modeled the observed vibrational progressions in both the ground and excited states of the ions, providing not only a fingerprint piece of data for the identification of these species but also confirmation of the suitability of the theoretical method for calculating ion structure and frequencies in both the ground and excited states. The modes observed were due to changes in the carbon skeleton upon excitation, primarily C−C stretching modes and skeletal bending modes. Those transitions, as well as the adiabatic ionizations potentials, act as spectroscopic fingerprints of an isomer which allows for isomer-selective detection in reactive environments. The observed dissociation of each ion was modeled with RRKM theory, and the resulting activation energies and entropies provide information on the barriers present in isomerization reactions prior to fragmentation. It seems likely that 1-propynylbenzene ions isomerize to indene ions prior to dissociation and that the isomerization is
■
REFERENCES
(1) PAHs and the Universe: A Symposium To Celebrate the 25th Anniversary of the PAH Hypothesis; Joblin, C., Tielens, A. G. G. M., Eds.; EAS Publications Series, Vol. 46; EDP Sciences: Les Ulis, France, 2011. (2) Page, V. L.; Snow, T. P.; Bierbaum, V. M. Hydrogenation and Charge States of PAHs in Diffuse Clouds. I. Development of a Model. Astrophys. J., Suppl. Ser. 2001, 132, 233−251. (3) West, B.; Joblin, C.; Blanchet, V.; Bodi, A.; Sztáray, B.; Mayer, P. M. On the Dissociation of the Naphthalene Radical Cation: New iPEPICO and Tandem Mass Spectrometry Results. J. Phys. Chem. A 2012, 116, 10999−11007. (4) West, B.; Sit, A.; Mohamed, S.; Joblin, C.; Blanchet, V.; Bodi, A.; Mayer, P. M. Dissociation of the Anthracene Radical Cation: A Comparative Look at iPEPICO and Collision-Induced Dissociation Mass Spectrometry Results. J. Phys. Chem. A 2014, 118, 9870−9878.
11233
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
(24) NIST. NIST Computational Chemistry Comparison and Benchmark Database, NIST Standard Reference Database Number 101; National Institute of Standards and Technology: Gaithersburg, MD, 2006. (25) Mayer, P. M.; Blanchet, V.; Joblin, C. Threshold Photoelectron Study of Naphthalene, Anthracene, Pyrene, 1,2-Dihydronaphthalene, and 9,10-Dihydroanthracene. J. Chem. Phys. 2011, 134, 244312. (26) Mozhayskiy, V. A.; Krylov, A. I. Jahn−Teller Distortions in the Electronically Excited States of sym-Triazine. Mol. Phys. 2009, 107, 929− 938. (27) Sztáray, B.; Bodi, A.; Baer, T. Modeling Unimolecular Reactions in Photoelectron Photoion Coincidence Experiments. J. Mass Spectrom. 2010, 45, 1233−1245. (28) Kwon, C. H.; Kim, H. L.; Kim, M. S. Vibrational Analysis of Vacuum Ultraviolet Mass-Analyzed Threshold Ionization Spectra of Phenylacetylene and Benzonitrile. J. Phys. Chem. A 2003, 107, 10969− 10975. (29) Report on Notation for the Spectra of Polyatomic Molecules. J. Chem. Phys. 1955, 23, 1997−2011. (30) Pugliesi, I.; Tonge, N. M.; Hornsby, K. E.; Cockett, M. C. R.; Watkins, M. J. An Examination of Structural Characteristics of Phenylacetylene by Vibronic and Rovibronic Simulations of ab Initio Data. Phys. Chem. Chem. Phys. 2007, 9, 5436−5445. (31) Hemberger, P.; Trevitt, A. J.; Gerber, T.; Ross, E.; da Silva, G. Isomer-Specific Product Detection of Gas-Phase Xylyl Radical Rearrangement and Decomposition Using VUV Synchrotron Photoionization. J. Phys. Chem. A 2014, 118, 3593−3604. (32) NIST Standard Reference Database Number 69; National Institute of Standards and Technology: Gaithersburg, MD, 2005. (33) Hemberger, P.; Steinbauer, M.; Schneider, M.; Fischer, I.; Johnson, M.; Bodi, A.; Gerber, T. Photoionization of Three Isomers of the C9H7 Radical. J. Phys. Chem. A 2009, 114, 4698−4703.
(5) West, B.; Useli-Bacchitta, F.; Sabbah, H.; Blanchet, V.; Bodi, A.; Mayer, P. M.; Joblin, C. Photodissociation of Pyrene Cations: Structure and Energetics from C16H10+ to C14+ and Almost Everything in Between. J. Phys. Chem. A 2014, 118, 7824−7831. (6) West, B.; Joblin, C.; Blanchet, V.; Bodi, A.; Sztáray, B.; Mayer, P. M. Dynamics of Hydrogen and Methyl Radical Loss from Ionized DihydroPolycyclic Aromatic Hydrocarbons: A Tandem Mass Spectrometry and Imaging Photoelectron−Photoion Coincidence (iPEPICO) Study of Dihydronaphthalene and Dihydrophenanthrene. J. Phys. Chem. A 2014, 118, 1807−1816. (7) Oßwald, P.; Hemberger, P.; Bierkandt, T.; Akyildiz, E.; Köhler, M.; Bodi, A.; Gerber, T.; Kasper, T. In Situ Flame Chemistry Tracing by Imaging Photoelectron Photoion Coincidence Spectroscopy. Rev. Sci. Instrum. 2014, 85, 025101. (8) Jin, H.; Cuoci, A.; Frassoldati, A.; Faravelli, T.; Wang, Y.; Li, Y.; Qi, F. Experimental and Kinetic Modeling Study of PAH Formation in Methane Coflow Diffusion Flames Doped with n-Butanol. Combust. Flame 2014, 161, 657−670. (9) Hansen, N.; Kasper, T.; Klippenstein, S. J.; Westmoreland, P. R.; Law, M. E.; Taatjes, C. A.; Kohse-Höinghaus, K.; Wang, J.; Cool, T. A. Initial Steps of Aromatic Ring Formation in a Laminar Premixed FuelRich Cyclopentene Flame. J. Phys. Chem. A 2007, 111, 4081−4092. (10) Lang, M.; Holzmeier, F.; Fischer, I.; Hemberger, P. Threshold Photoionization of Fluorenyl, Benzhydryl, Diphenylmethylene, and Their Dimers. J. Phys. Chem. A 2013, 117, 5260−5268. (11) Hemberger, P.; Trevitt, A. J.; Ross, E.; da Silva, G. Direct Observation of Para-Xylylene as the Decomposition Product of the Meta-Xylyl Radical Using VUV Synchrotron Radiation. J. Phys. Chem. Lett. 2013, 4, 2546−2550. (12) Zhang, F.; Kaiser, R. I.; Kislov, V. V.; Mebel, A. M.; Golan, A.; Ahmed, M. A VUV Photoionization Study of the Formation of the Indene Molecule and Its Isomers. J. Phys. Chem. Lett. 2011, 2, 1731− 1735. (13) Hemberger, P.; Trevitt, A. J.; Gerber, T.; Ross, E.; da Silva, G. Isomer-Specific Product Detection of Gas-Phase Xylyl Radical Rearrangement and Decomposition Using VUV Synchrotron Photoionization. J. Phys. Chem. A 2014, 118, 3593−3604. (14) Kislov, V. V.; Mebel, A. M. Ab Initio G3-Type/Statistical Theory Study of the Formation of Indene in Combustion Flames. I. Pathways Involving Benzene and Phenyl Radical. J. Phys. Chem. A 2007, 111, 3922−3931. (15) da Silva, G.; Bozzelli, J. W. Indene Formation from Alkylated Aromatics: Kinetics and Products of the Fulvenallene + Acetylene Reaction. J. Phys. Chem. A 2009, 113, 8971−8978. (16) Bodi, A.; Hemberger, P.; Gerber, T.; Sztáray, B. A New Double Imaging Velocity Focusing Coincidence Experiment: i2PEPICO. Rev. Sci. Instrum. 2012, 83, 083105. (17) Johnson, M.; Bodi, A.; Schulz, L.; Gerber, T. Vacuum Ultraviolet Beamline at the Swiss Light Source for Chemical Dynamics Studies. Nucl. Instr. Methods Phys. Res., Sect. A 2009, 610, 597−603. (18) Bodi, A.; Sztaray, B.; Baer, T.; Johnson, M.; Gerber, T. Data Acquisition Schemes for Continuous Two-Particle Time-of-Flight Coincidence Experiments. Rev. Sci. Instrum. 2007, 78, 084102. (19) Bodi, A.; Johnson, M.; Gerber, T.; Gengeliczki, Z.; Sztaray, B.; Baer, T. Imaging Photoelectron Photoion Coincidence Spectroscopy with Velocity Focusing Electron Optics. Rev. Sci. Instrum. 2009, 80, 034101. (20) Sztáray, B.; Baer, T. Suppression of Hot Electrons in Threshold Photoelectron Photoion Coincidence Spectroscopy Using Velocity Focusing Optics. Rev. Sci. Instrum. 2003, 74, 3763−3768. (21) Baer, T.; Sztaray, B.; Kercher, J. P.; Lago, A. F.; Bodi, A.; Skull, C.; Palathinkal, D. Threshold Photoelectron Photoion Coincidence Studies of Parallel and Sequential Dissociation Reactions. Phys. Chem. Chem. Phys. 2005, 7, 1507−1513. (22) IGOR Pro; WaveMetrics, Inc.: Lake Oswego, OR, 2013. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford CT, 2009. 11234
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
Dissociative Photoionization and Threshold Photoelectron Spectra of Polycyclic Aromatic Hydrocarbon Fragments: An Imaging Photoelectron Photoion Coincidence (iPEPICO) Study of Four Substituted Benzene Radical Cations Brandi West,† Alicia Sit,† Andras Bodi,‡ Patrick Hemberger,‡ and Paul M. Mayer*,† †
Department of Chemistry, University of Ottawa, Ottawa, K1N 6N5, Canada Molecular Dynamics Group, Paul Scherrer Institute, Villigen 5232, Switzerland
‡
S Supporting Information *
ABSTRACT: Four molecules were investigated by imaging photoelectron photoion coincidence (iPEPICO) spectroscopy: 1-propynylbenzene, indene, ethynylbenzene, and benzocyclobutene. Their threshold photoelectron spectrum was obtained and electronic transitions were assigned by OVGF (outer valence Green’s function) calculations. Vibrational progressions observed in the electronic ground and excited states were simulated by calculating Franck−Condon factors based on the neutral as well as the cation ground and excited state geometries. iPEPICO was used to obtain ion dissociation data in threshold photoionization as a function of photon energy, which were modeled with RRKM theory to extract kinetic parameters for the reactions C9H8+• (1-propynylbezene) → C9H7+ + H (R1); C9H8+• (indene) → C9H7+ + H (R2); C8H8+• (benzocyclobutene) → C8H7+ + H (R3); C8H8+• (benzocyclobutene) → C6H6+ + C2H2 (R4); C8H6+• (1-ethynylbenzene, aka phenylacetylene) → C6H4+ + C2H2 (R5). These results were compared to G3 level calculations. In addition, the enthalpy of formation of the indenyl radical was estimated to be ΔfH°0K = 249 ± 50 kJ mol−1 based on a previously measured IE and a cation ΔfH°0K = 976 kJ mol−1, determined herein.
■
INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) are molecules of astronomical importance due to their potential role as catalysts throughout the interstellar medium (ISM).1 One facet of laboratory PAH studies is to focus on their stability in environments similar to that of the ISM. An example of this is to determine their ability to withstand radiation exposure without fragmentation, as this is a requirement in order to exist long enough to facilitate, for example, catalytic reactions.2 Previously, our lab has investigated the vacuum ultraviolet (VUV) radiation induced dissociative photoionization of small PAH and PAH-like molecules, notably naphthalene,3 anthracene,4 pyrene,5 1,2-dihydronaphthalene, and 9,10-dihydrophenanthrene.6 The most notable dissociation pathways for the unsaturated PAH cations is H loss, with hydrocarbon formation (loss of C2H2 and C4H2) steeply declining with increasing size of the ring system. The dihydro-PAHs lose both H and CH3 radicals. In all cases, there is a significant kinetic shift to the observed onset for dissociation, with pyrene exhibiting fragmentation only 7 eV above the ionization threshold and 2 eV above the thermochemical onset for H loss. This trend suggests that larger PAHs will be photostable when ionized by the intense Lyman-α band in outer space. However, what of the smaller molecules that may act as synthetic precursors to PAHs? In this paper we tackle four molecules that can serve as building blocks for PAHs (benzocyclobutene, ethynylbenzene, indene, © 2014 American Chemical Society
and propynlbenzene (Figure 1)) and explore their dissociation chemistry to gain insight on their photostability. Furthermore, these species are soot precursor molecules in flames,7−9 appear in microreactors,10−13 and being able to identify and quantify these molecules and their ions in such experiments is crucial for elucidating the often complex reaction mechanisms taking place in those environments. One approach to identification is through their photoelectron spectra.12,14,15 Herein, we describe the measurement and interpretation of their threshold photoelectron spectra and dissociative ionization behavior through imaging photoelectron photoion coincidence spectroscopy (iPEPICO).
■
EXPERIMENTAL SECTION The samples (1-propynylbenzene, indene, ethynylbenzene, and benzocyclobutene) were obtained from Sigma-Aldrich, Oakville, Canada, and were used without further purification. Imaging Photoelectron Photoion Coincidence Spectroscopy. iPEPICO experiments were conducted at the VUV beamline of the Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland, using the 600 lines/mm laminar grating as described in detail earlier.16−19 In short, the sample is allowed to Received: August 25, 2014 Revised: October 27, 2014 Published: October 27, 2014 11226
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
which the need to reproduce the observed blue shift at the higher field, too, constrains the RRKM model. If two fragment ions are formed in parallel dissociations, the peak shape analysis of only one fragment yields the overall depletion rate of the parent ion already. The breakdown curves can then be used to determine the branching ratios and assign absolute rate constants to both reaction channels. Two of the samples only exhibited H loss, and it was not possible to fit asymmetric TOF profiles of the fragment ions. In fact, the TOF distribution for the composite peak in the molecular ion region had to be approximated as a sum of Gaussian peaks for the [13C−M]+•, [M]+• and [M−H]+ ions, each with the same TOF full width at half-maximum (fwhm) using the IGOR Pro software package in order to even plot the breakdown diagram.22 That is, the experimental TOF distribution was fitted for the three peaks making up the molecular ion region to quantify the ion abundances. Examples of the fitted TOF distributions for the four molecules are shown in Figure 2. Computational Procedures. Ab Initio Calculations. Neutral and ionic structures were calculated for the four sample molecules as well as the primary daughter ions together with the neutral fragments. Three different sets of calculations were completed using Gaussian 09.23 All species were optimized at the B3LYP/6-311+G(d,p) or ωB97XD/6-311++G(d,p) levels of theory. From these calculations, vibrational frequencies, Hessian matrices, rotational constants, and optimized geometries were extracted to be used in the RRKM fitting and in the Franck− Condon (FC) simulations (see below). Single point energy calculations were performed on the B3LYP/6-311+G(d,p) geometries at the MP2/6-311+G(d,p) and G3 levels of theory. Zero point vibrational energies were scaled by 0.967.24 Outer valence Green’s function calculations (OVGF) were carried out with the cc-pVQZ basis set to predict ionization energies to different cation electronic states.25 Geometries of certain electronic excited ion states were optimized applying time dependent density functional theory (TD-DFT). Franck− Condon factors were calculated by eZspectrum.OSX,26 taking the neutral and ion geometries as well as the Hessian matrices as input. The calculated stick spectra were convoluted with a Gaussian function and shifted in energy to match the experimentally observed 0−0 transition. The baseline shift in the TPES due to autoionization was taken into account using a sigmoid function. The experimental uncertainty of the adiabatic ionization energies (IEa) was taken as the half width at halfmaximum of the 0−0 transition. Breakdown Diagram and TOF Distribution Modeling. Rice−Ramsperger−Kassel−Marcus (RRKM) theory was used to calculate unimolecular dissociation rate constants. The standard equation was used to calculated k(E),
Figure 1. 3D representation of the radical cations studied in this work as calculated using the B3LYP/6-311+G(d,p) level of theory: (A) 1propynylbenzene; (B) indene; (C) benzocyclobutene; (D) 1ethynylbenzene.
diffuse into the ionization region of the iPEPICO apparatus through a 6 mm outer diameter Teflon tube, and the neutral gasphase molecules are single photon ionized via monochromatic VUV synchrotron radiation. The photoelectrons are extracted by a constant electric field and velocity map imaged using a delayline MCP detector where each hit is position and time stamped. Meanwhile, the parent and fragment ions are directed toward a nonimaging multichannel plate (MCP) detector, where each ion’s arrival time is recorded. The electron and ion time stamps are correlated to obtain the time-of-flight (TOF) mass spectrum. Threshold electrons hit the detector in a central spot on the MCP and are selected to study threshold photoionization and to plot the breakdown diagram. Generally, electrons initially formed with nonzero kinetic energy (“hot” electrons) are imaged offcenter because of their translational energy component perpendicular to their flight path. Hot electrons that have no lateral momentum will be imaged at the center of the detector, though, and this contamination is approximated based on the signal in a zone consisting of a small ring around the central spot.20 The TOF mass spectrum corresponding to the ring is subtracted from that of the center spot to obtain the true threshold photoionization signal. The threshold photoelectron spectrum was obtained by the same procedure applied to the electron counts at each photon energy to obtain the threshold electrons signal. The time-of-flight mass spectrometer has two acceleration regions, the first of which has a low draw-out field. Ion dissociating with rate constants between 103 and 107 s−1 will produce fragment ions in this region, which are then nonuniformly accelerated toward the second acceleration region, depending on where in this region they are formed. This results in product ion time-of-flight distributions that are broad and asymmetric. The resulting TOF distributions can be modeled to extract absolute unimolecular decay rate constants (Figure 2).21 Because of the limited mass resolution of the TOF mass spectrometer, the small difference between parent and H-loss daughter ion TOF at m/z − 1 does not allow for peak shape analysis. As an alternative, the residence time in the first acceleration region and thus the time scale of the dissociation can be changed by employing two different draw-out fields, in our case 40 and 120 V cm−1, providing two breakdown diagrams, in
k(E) =
σN ⧧(E − E0) hρ(E)
(1)
where N⧧(E − E0) is the transition state number of states, ρ(E) is the density of states of the dissociating ion, h is Planck’s constant, and σ is the degeneracy of the fragmentation channel. For 1propynylbenzene, σ = 3, representing the three hydrogen atoms from the methyl group. For indene, σ = 2, representing the two hydrogen atoms on the sp3 carbon site. For benzocyclobutene, σ = 4 and σ = 1 for hydrogen atom loss and C2H2 loss, respectively, based on fragmentation of the butene ring only. Finally, for 1ethynylbenzene, σ = 1, representing the loss of C2H2 from the linear ethynyl group. 11227
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Figure 2. Representative TOF fits for (A) 1-propynylbenzene (M+• and [M−H]+), (B) indene (M+• and [M−H]+), (C) benzocyclobutene ([M− C2H2]+•), and (D) 1-ethynylbenzene ([M−C2H2]+•), all at 120 V cm−1 draw-out potential, and each at the quoted photon energy.
(where possible) match the experimental data. The final list of vibrational frequencies (Table S1) and k(E) values (Figure S1) for each reaction can be found in Supporting Information.
The k(E) values, together with the photon energy, the roomtemperature internal energy distribution of the neutral sample, the ionization energy of the molecule, and instrumental parameters are combined to produce a theoretical breakdown diagram and fragment ion TOF distributions using the iPEPICO modeling program by Sztáray et al.27 Both the activation energy and entropy are varied in the k(E) calculation until the resulting model breakdown curve and fragment ion TOF distributions
■
RESULTS AND DISCUSSION
Threshold Photoelectron Spectra. The threshold photoelectron spectra (TPES) are shown in Figures 3−6 from threshold up to 15 eV photon energy, along with the vertical 11228
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Table 2. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for Benzenecyclobutene (C2v Aymmetry)
a
G3 level calculation.
Table 3. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for 1Propynylbenzene (CS Symmetry)
a
Figure 3. Threshold photoelectron spectrum of 1-ethynylbenzene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
G3 level calculation.
Table 4. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for Indene (CS Symmetry)
molecular orbital ionization energies from the OVGF calculations (top panel). Note that the MO numbering used to identify the cation electronic states is based on Hartree−Fock calculations and refers to the outer-valence orbitals only (Tables 1−4). The perturbation treatment in the OVGF calculations can Table 1. Comparison of Experimental and OVGF/cc-pVQZ Molecular Orbital Vertical Ionization Energies for 1Ethynylbenzene (C2v Symmetry)
a
a
G3 level calculation.
The experimentally determined IE for ethynylbenzene is 8.81 ± 0.01 eV, which is in excellent agreement with the MATI value from Kwon et al. of 8.8195 ± 0.006 eV.28 The ground electronic state of the ion as well as the second and third excited states show vibrational structure indicative of good Franck−Condon overlap between the neutral and ion states and long-lived electronic excited states. Figure 3 (second trace) shows a zoom into the TPES of the ion’s ground electronic state as well as a Franck− Condon simulation (sticks and convolution with a Gaussian function with a fwhm of 19 meV). Unless noted otherwise, all vibrationally active modes were found to be totally symmetric, i.e., of a1, a, or a′ symmetry. The feature at 8.75 eV is assigned to a hot band transition. The C−C−C bending vibrations ν13 and ν12, 469 and 764 cm−1, respectively, are active upon ionization and contribute mainly to the bands at 8.87 and 8.9 eV. Note that Mulliken notation is used for the assignment of the vibrational modes,29 whereas literature MATI spectra were assigned using the Wilson notation.28,30 The band at 8.93 eV is assigned to the
G3 level calculation.
change the ordering of the electronic states. The corresponding molecular orbitals can be found in Figure S2 of Supporting Information. Expansions of TPES regions containing vibrational progressions are also presented in each figure. Since the ground state ionization occurs mostly from the π-system connecting rings and substituents, a moderate geometry change is distributed over the whole carbon skeleton, resulting in wellresolved peaks in the TPES. 11229
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
first overtone of ν13 as well as the fundamental of ν11, a ringbreathing mode (995 cm−1). A H−CC−H scissoring motion (ν9 = 1208 cm−1) and a combination band of ν13 and ν12 account for the band at 8.96 eV. Another three bands above 9 eV are due to ν6, ν5, and a combination of ν13 and ν5 vibrational motions possessing frequencies of 1643 (ν6, HCCH stretch) and 2108 cm−1 (ν5, HCC stretch). The mismatch in the line positions can be due to anharmonicity; however, the underestimation of the intensities of the ν13 mode and the non-negative unresolved background at higher photon energies can also be explained by the autoionization of high lying Rydberg states. These neutral states can couple with energetically similar ion states, placing extra intensity in some transitions (enhanced ionization probability) and even lead to threshold photoelectrons in Franck−Condon gaps. Because of the fact that the FC simulations are based on DFT geometries, an overestimation of the bond length change along the ν13 coordinate upon ionization can also lead to this mismatch. While the first excited state (Ã 2A2) at around 9.5 eV possesses only a broad band structure, implying either a large geometry change upon ionization or a short lifetime of this state, the second (B̃ 2B2, 10.33 eV) and third (C̃ 2B1, 11.02 eV) excited states show well resolved structure in the TPES (Figure 3, third and fourth trace). For the B̃ 2B2 state, corresponding to ionization from the HOMO − 2 of b2 symmetry, a progression of ν5 (1998 cm−1, CC stretch) can explain the main features in the spectrum. Additional vibrational modes, mostly different flavors of C−C−C bending (ν13 = 459, ν12 = 758, and ν11 = 998 cm−1) as well as H−CCH scissoring (ν8 = 1222 cm−1) motions, contribute to the less pronounced features. The third excited (C̃ 2 B1) state exhibits activity in the ν12 and ν11 modes at 741 and 955 cm−1, respectively, which reflects a C−C−C bending and a ring breathing vibration of the benzene ring. Apart from these, another C−C−C bending vibration (ν9 = 1143 cm−1), a H− CC−H scissoring (ν8 = 1193 cm−1), and several C−C stretching vibrations (ν7 = 1488, ν6 = 1591, and ν5 = 2081 cm−1) are also active upon photoionization. The experimentally determined IE for benzocyclobutene is 8.65 ± 0.02 eV (see Figure 4), which is in excellent agreement with another threshold photoelectron spectroscopy value of 8.65 eV, obtained in a pyrolysis experiment using the same apparatus.13 The ground state TPE spectrum in Figure 4 is in good agreement with the literature TPE spectrum and FC simulations. Since all the ground state bands were already assigned to their corresponding vibrational transitions,31 the TPE spectrum will not be further discussed. The experimentally determined IE for propylbenzene is 8.40 ± 0.01 eV, in good agreement with the G3-level calculated value of 8.38 eV. Both are slightly lower than previous photoelectron results of 8.41−8.49 eV.32 The ground state TPES (Figure 5, second trace) displays vibrational structure, which was assigned by a Franck−Condon simulation. A progression of a ring deformation mode, mostly due to a C−C−C bending vibration (ν38 = 408 cm−1) located at the substituent, is responsible for the intense feature at 8.45 eV. Above 8.46 eV, several vibrational modes show activity upon ionization: Two ring-deformation modes (ν33 = 712 cm−1 and ν10 = 1663 cm−1) as well as C−CH3 (ν30 = 969 cm−1), CC (ν9 = 2264 cm−1), and C−H3 (ν8 = 3038 cm−1) stretching vibrations can be assigned in the spectrum. A combination band of ν8 and ν45, a CH3 torsional mode at 45 cm−1, is responsible for the feature at 8.78 eV. The third excited state C̃ 2A at 10.57 eV (see third trace in Figure 5) exhibits vibrations of ν33 = 691 and ν28 = 980 cm−1,
Figure 4. Threshold photoelectron spectrum of benzocyclobutene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower pane. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
Figure 5. Threshold photoelectron spectrum of 1-propynylbenzene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
11230
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
CC−H scissoring modes (ν22 = 1085 cm−1 and ν20 = 1159 cm−1) contribute to the remaining features. In summary, we were able to assign almost all vibrational features in the ground as well as excited states of the ion for phenylacetylene, benzocyclobutene, 1-propynylbenzene, and indene by means of Franck−Condon simulations. Because of the moderate change in geometry upon ionization, mostly ring deformation modes, such as C−C stretch and C−C−C bending vibrations, become active. Those transitions as well as the adiabatic ionizations potentials act as the spectroscopic fingerprint of an isomer, which allows for isomer-selective detection in reactive environments as presented and discussed in detail for indene and propynylbenzene.10,11,13 Furthermore, even transitions into electronic excited ion states exhibit a vibrational fingerprint in some cases, in particular when the geometry change is moderate and the lifetime of the excited electronic state is sufficiently long. These fingerprints can further act as an isomer-specific feature when the lower-energy spectrum is too congested for assignment. Dissociative Photoionization. Five reaction channels were observed for the four molecules:
which are similar to those of the ground state because they involve the in-plane distortion of the benzene ring. The feature at around 10.71 eV is somewhat overestimated in the Franck− Condon simulation and can be assigned to in-plane distortion (ν20 = 1203 cm−1) and H−CC−H scissoring motion (ν19 = 1228 cm−1). C−C stretching vibrations of the ring (ν13 = 1503 cm−1 and ν10 = 1703 cm−1) as well as combination bands contribute to the features between 10.75 and 10.9 eV. The combination band between the C−H stretch (ν8 = 3015 cm−1) of the methyl group and the CH3 torsional mode (ν45 = 39 cm−1) is slightly overestimated in the simulation, and their origin is thus difficult to assign in the experimental spectrum. The experimentally determined IE for indene is 8.14 ± 0.01 eV, in agreement with previous photoelectron results that scatter closely around this value (8.13−8.15 eV).32 Both the ground state of the ion and second excited state show vibrational structure. Several vibrational modes contribute to the ground state TPE spectrum (second trace in Figure 6) of indene, which
C9H8+• (1‐propynylbezene) → C9H 7+ + H
(R1)
C9H8+• (indene) → C9H 7+ + H
(R2)
C8H8+• (benzocyclobutene) → C8H 7+ + H
(R3)
C8H8+• (benzocyclobutene) → C6H6+ + C2H 2
(R4)
C8H6+• (1‐ethynylbenzene) → C6H4 + + C2H 2
(R5)
The proposed fragment ion structures for each reaction are presented in Figure 7. Each pair of similar molecules will be discussed together in order to compare their fragmentation reactions. 1-Propynylbenzene (1) vs Indene (3). Both ions dissociate by H loss (reactions R1 and R2). Figure 8A shows the fitted breakdown curve for R1. A low activation energy of 1.86 ± 0.17 eV, coupled with a negative activation entropy, Δ⧧S1000K = −30 ± 31 J K−1 mol−1, may indicate that an isomerization reaction is the rate-determining step. In addition, the G3 reaction energy for the formation of structure 2 by simple bond cleavage is significantly higher at 2.57 eV (Table 5). The only plausible candidate for this isomer is the indene cation as the lowest energy C9H8+ ion isomer. Thus, it is highly likely that ionized 1-propylnylbenzene dissociates to form the indenyl cation 4. The RRKM E0 translates in this case to an isomerization barrier to ionized indene of ∼85 kJ mol−1 relative to the products. The breakdown diagram for H loss from ionized indene (R2) is shown in Figure 8B. The E0 value of 2.56 ± 0.37 eV is 0.7 eV higher in energy than that of R1, while the Δ⧧S1000K was found to be −2 ± 38 J K−1 mol−1. The G3 reaction energy in this case is very close to the RRKM E0 for the reaction (Table 5), suggesting the lack of a reverse barrier, although the transition state cannot be shown to be loose because of the poorly defined activation entropy. A relaxed potential energy surface scan at the B3LYP/6311++G(d,p) level of theory, however, did not indicate a reverse energy barrier to H loss from indene. Assuming no reverse barrier, the appearance energy of 10.7 ± 0.5 eV (or 1032 ± 48 kJ mol−1, AE = IE(indene) + E0) of the fragment ion from neutral indene (ΔfH°0K = 161.2 ± 2.3 kJ mol−1)32 results in a ΔfH°0K for the indenyl cation of 975 ± 48 kJ mol−1. Combined with a recent, accurate determination of the IE of the indenyl radical of 7.53
Figure 6. Threshold photoelectron spectrum of indene (top) with OVGF vertical ionization energies shown as red lines. Expansions of regions of the TPES exhibiting vibrational structure are shown in the lower panes. Red line fits to the vibrational structure are the result of Franck−Condon simulations.
can be assigned to two different groups. At roughly 1000 cm−1 (0.124 eV) above the IEa, the spectrum is dominated by C−C−C bending vibrations (ν30 = 392 cm−1, ν29 = 525 cm−1, ν28 = 594 cm−1, and ν26 = 830 cm−1). In the 8.3−8.5 eV energy range, C−C stretching modes (ν11 = 1462 cm−1, ν10 = 1507 cm−1, and ν8 = 1650 cm−1) are mostly active upon ionization. Numerous combination bands as well as overtones of the mentioned vibrational modes contribute to the spectra at higher photon energies and cannot be discriminated from the autoionization background. The second excited state B̃ 2A″ has its origin at 10.29 eV (see third trace in Figure 6) and displays several vibrationally active modes that can be divided into three different groups. Ring deformation modes (ν28 = 592 cm−1 and ν27 = 722 cm−1) can explain the features up to 10.4 eV. Two C−C stretching vibrations (ν23 = 982 cm−1, ν11 = 1476 cm−1) as well as two H− 11231
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Figure 7. Reaction schemes for reactions R1−R5 from (A) 1-propynylbenzene, (B) indene, (C) benzocyclobutene, and (D) 1-ethynylbenzene.
Figure 8. Experimental iPEPICO breakdown curves for the four molecules studied, (A) 1-propynylbenzene, (B) indene, (C) benzocyclobutene, and (D) ethynylbenzene, all taken at 120 V cm−1 draw-out field.
eV,33 this leads to a ΔfH°0K of the indenyl radical of 248 ± 50 kJ mol−1. Benzocyclobutene (5) vs 1-Ethynylbenzene (8). Unlike 1-propynylbenzene and indene, benzocyclobutene (5) has two parallel, low-energy dissociative photoionization channels, dehydrogenation (reaction R3) and the loss of C2H2 (reaction R4). The iPEPICO experiments were conducted in the 11.5−15
eV photon energy range. The fragment ion of R3 is assigned as structure 6, as there are only the two identical sp3 carbon sites in the parent. Reaction R4 is assumed to lead to the benzene cation, as it is the energetically most favored C6H6+• ion. 1Ethynylbenzene (8), on the other hand, only undergoes loss of C2H2 (reaction R5) in the photon energy range studied, yielding the proposed structure 9, benzyne cation. 11232
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
Table 5. Energetics Data for Reactions R1−R5 for the Four Molecules Studied. ΔE are the G3-Calculated Dissociation Energies for each Reaction; E0, Δ⧧S1000K, and AE are Determined from the RRKM Fitting of the Experimental Breakdown Curves C9H8+• C9H8+• C8H8+• C8H8+• C8H6+• a
→ → → → →
C9H7+ + H (R1) C9H7+ + H (R2) C8H7+ + H (R3) C6H6+• + C2H2 (R4) C6H4+• + C2H2 (R5)
ΔE (G3)/eV
E0/eV
2.57 2.78 2.33 1.70 4.54
1.86 ± 0.17 2.56 ± 0.37a 2.23 ± 0.10 2.00 ± 0.10 3.94 ± 0.43
Δ⧧S1000K/J K−1 mol−1
AE/eV
−30 ± 31a −2 ± 38a −11 ± 8 −17 ± 9 42 ± 36
10.26 ± 0.27 10.70 ± 0.50 10.88 ± 0.21 10.65 ± 0.27 12.75 ± 0.42
a
Best values obtained by simultaneously fitting breakdown diagrams obtained with ion extraction potentials of 40 and 120 V cm−1.
characterized by a barrier of 85 kJ mol−1 relative to products. The other two molecules were found to ionize and generate C2H2 only at energies above Lyman-α (10.2 eV). These results and those for other sp3-carbon containing PAHs6 indicate that their dissociation thresholds tend to be ∼2 eV lower in energy compared to systems containing only sp2-carbon atoms. However, all have been shown to dissociate above 10.2 eV, indicating that they may indeed survive to act as precursors for larger PAH systems.
Figure 8C shows the breakdown diagram for benzocyclobutene. For reaction R3, the E0 and Δ⧧S1000K values were found to be 2.23 ± 0.10 eV and −11 ± 8 J K−1 mol−1, respectively. The competing channel R4, which dominates at most energies studied, has an E0 of 2.00 ± 0.10 eV and a Δ⧧S1000K value of −17 ± 9 J K−1 mol−1. The low E0 and negative activation entropy are consistent with the fact that loss of C2H2 necessarily involves a rearrangement step leading to the stable benzene ion, which is also identified to be rate-determining. This is also consistent with the fact that the G3 reaction energy is 0.3 eV lower than the RRKM E0 for the reaction (Table 5). Conversely, the energy requirement for reaction R5 is substantially higher at 3.94 ± 0.43 eV (Figure 8D shows the breakdown curve for 1-ethynylbenzene). The activation entropy of Δ⧧S1000K = 42 ± 36 J K−1 mol−1 is indicative of a loose transition state, one that possibly involves an ion−molecule complex, as was previously calculated for the loss of acetylene from the anthracene radical cation. The G3 ΔE for reaction R5 was significantly larger than the derived E0 (4.54 eV compared to 3.94 ± 0.43 eV). While there is a significant degree of variation in the calculated reaction energies as a function of level of theory (Table S2), it is possible that this result also means ionized ethynylbenzene isomerizes prior to dissociation. The only candidate structure that seems reasonable is ionized benzocyclobutadiene, as ethynylbenzene would have to adopt a structure similar to this prior to losing C2H2. However, the fragmentation products are the same in each case, ionized benzyne and C2H2. Thus, isomerization cannot account for the discrepancy between the G3 and RRKM values, since B3LYP/6311+G(d,p) level calculations place ionized ethynyl benzene lower in energy by only 5 kJ mol−1.
■
ASSOCIATED CONTENT
S Supporting Information *
Full authorship for ref 23, Table S1 containing the vibrational frequencies for the processes modeled with eq 1 for both ions, Table S2 containing a comparison of levels of theory used to calculate the energy thresholds for the dissociation reactions in this paper, Figure S1 containing the descriptions for the MOs listed in Tables 1−4, and Figure S2 containing the k(E) data derived from the RRKM fitting of the experimental data. This material is available free of charge via the Internet at http://pubs. acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Phone: 1-613-562-5800, extension 6038. Notes
The authors declare no competing financial interest.
■
■
ACKNOWLEDGMENTS P.M.M. thanks the Natural Sciences and Engineering Research Council of Canada for continuing financial support. Experiments were carried out at the VUV beamline of the Swiss Light Source, Paul Scherrer Institute (PSI). The work was financially supported by the Swiss Federal Office for Energy (BFE Contract 101969/152433).
CONCLUSIONS Threshold photoelectron spectra were obtained for the four target molecules. In general, the measured ground state IEs were in good agreement with literature values and G3-level calculations. Franck−Condon simulations successfully modeled the observed vibrational progressions in both the ground and excited states of the ions, providing not only a fingerprint piece of data for the identification of these species but also confirmation of the suitability of the theoretical method for calculating ion structure and frequencies in both the ground and excited states. The modes observed were due to changes in the carbon skeleton upon excitation, primarily C−C stretching modes and skeletal bending modes. Those transitions, as well as the adiabatic ionizations potentials, act as spectroscopic fingerprints of an isomer which allows for isomer-selective detection in reactive environments. The observed dissociation of each ion was modeled with RRKM theory, and the resulting activation energies and entropies provide information on the barriers present in isomerization reactions prior to fragmentation. It seems likely that 1-propynylbenzene ions isomerize to indene ions prior to dissociation and that the isomerization is
■
REFERENCES
(1) PAHs and the Universe: A Symposium To Celebrate the 25th Anniversary of the PAH Hypothesis; Joblin, C., Tielens, A. G. G. M., Eds.; EAS Publications Series, Vol. 46; EDP Sciences: Les Ulis, France, 2011. (2) Page, V. L.; Snow, T. P.; Bierbaum, V. M. Hydrogenation and Charge States of PAHs in Diffuse Clouds. I. Development of a Model. Astrophys. J., Suppl. Ser. 2001, 132, 233−251. (3) West, B.; Joblin, C.; Blanchet, V.; Bodi, A.; Sztáray, B.; Mayer, P. M. On the Dissociation of the Naphthalene Radical Cation: New iPEPICO and Tandem Mass Spectrometry Results. J. Phys. Chem. A 2012, 116, 10999−11007. (4) West, B.; Sit, A.; Mohamed, S.; Joblin, C.; Blanchet, V.; Bodi, A.; Mayer, P. M. Dissociation of the Anthracene Radical Cation: A Comparative Look at iPEPICO and Collision-Induced Dissociation Mass Spectrometry Results. J. Phys. Chem. A 2014, 118, 9870−9878.
11233
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
The Journal of Physical Chemistry A
Article
(24) NIST. NIST Computational Chemistry Comparison and Benchmark Database, NIST Standard Reference Database Number 101; National Institute of Standards and Technology: Gaithersburg, MD, 2006. (25) Mayer, P. M.; Blanchet, V.; Joblin, C. Threshold Photoelectron Study of Naphthalene, Anthracene, Pyrene, 1,2-Dihydronaphthalene, and 9,10-Dihydroanthracene. J. Chem. Phys. 2011, 134, 244312. (26) Mozhayskiy, V. A.; Krylov, A. I. Jahn−Teller Distortions in the Electronically Excited States of sym-Triazine. Mol. Phys. 2009, 107, 929− 938. (27) Sztáray, B.; Bodi, A.; Baer, T. Modeling Unimolecular Reactions in Photoelectron Photoion Coincidence Experiments. J. Mass Spectrom. 2010, 45, 1233−1245. (28) Kwon, C. H.; Kim, H. L.; Kim, M. S. Vibrational Analysis of Vacuum Ultraviolet Mass-Analyzed Threshold Ionization Spectra of Phenylacetylene and Benzonitrile. J. Phys. Chem. A 2003, 107, 10969− 10975. (29) Report on Notation for the Spectra of Polyatomic Molecules. J. Chem. Phys. 1955, 23, 1997−2011. (30) Pugliesi, I.; Tonge, N. M.; Hornsby, K. E.; Cockett, M. C. R.; Watkins, M. J. An Examination of Structural Characteristics of Phenylacetylene by Vibronic and Rovibronic Simulations of ab Initio Data. Phys. Chem. Chem. Phys. 2007, 9, 5436−5445. (31) Hemberger, P.; Trevitt, A. J.; Gerber, T.; Ross, E.; da Silva, G. Isomer-Specific Product Detection of Gas-Phase Xylyl Radical Rearrangement and Decomposition Using VUV Synchrotron Photoionization. J. Phys. Chem. A 2014, 118, 3593−3604. (32) NIST Standard Reference Database Number 69; National Institute of Standards and Technology: Gaithersburg, MD, 2005. (33) Hemberger, P.; Steinbauer, M.; Schneider, M.; Fischer, I.; Johnson, M.; Bodi, A.; Gerber, T. Photoionization of Three Isomers of the C9H7 Radical. J. Phys. Chem. A 2009, 114, 4698−4703.
(5) West, B.; Useli-Bacchitta, F.; Sabbah, H.; Blanchet, V.; Bodi, A.; Mayer, P. M.; Joblin, C. Photodissociation of Pyrene Cations: Structure and Energetics from C16H10+ to C14+ and Almost Everything in Between. J. Phys. Chem. A 2014, 118, 7824−7831. (6) West, B.; Joblin, C.; Blanchet, V.; Bodi, A.; Sztáray, B.; Mayer, P. M. Dynamics of Hydrogen and Methyl Radical Loss from Ionized DihydroPolycyclic Aromatic Hydrocarbons: A Tandem Mass Spectrometry and Imaging Photoelectron−Photoion Coincidence (iPEPICO) Study of Dihydronaphthalene and Dihydrophenanthrene. J. Phys. Chem. A 2014, 118, 1807−1816. (7) Oßwald, P.; Hemberger, P.; Bierkandt, T.; Akyildiz, E.; Köhler, M.; Bodi, A.; Gerber, T.; Kasper, T. In Situ Flame Chemistry Tracing by Imaging Photoelectron Photoion Coincidence Spectroscopy. Rev. Sci. Instrum. 2014, 85, 025101. (8) Jin, H.; Cuoci, A.; Frassoldati, A.; Faravelli, T.; Wang, Y.; Li, Y.; Qi, F. Experimental and Kinetic Modeling Study of PAH Formation in Methane Coflow Diffusion Flames Doped with n-Butanol. Combust. Flame 2014, 161, 657−670. (9) Hansen, N.; Kasper, T.; Klippenstein, S. J.; Westmoreland, P. R.; Law, M. E.; Taatjes, C. A.; Kohse-Höinghaus, K.; Wang, J.; Cool, T. A. Initial Steps of Aromatic Ring Formation in a Laminar Premixed FuelRich Cyclopentene Flame. J. Phys. Chem. A 2007, 111, 4081−4092. (10) Lang, M.; Holzmeier, F.; Fischer, I.; Hemberger, P. Threshold Photoionization of Fluorenyl, Benzhydryl, Diphenylmethylene, and Their Dimers. J. Phys. Chem. A 2013, 117, 5260−5268. (11) Hemberger, P.; Trevitt, A. J.; Ross, E.; da Silva, G. Direct Observation of Para-Xylylene as the Decomposition Product of the Meta-Xylyl Radical Using VUV Synchrotron Radiation. J. Phys. Chem. Lett. 2013, 4, 2546−2550. (12) Zhang, F.; Kaiser, R. I.; Kislov, V. V.; Mebel, A. M.; Golan, A.; Ahmed, M. A VUV Photoionization Study of the Formation of the Indene Molecule and Its Isomers. J. Phys. Chem. Lett. 2011, 2, 1731− 1735. (13) Hemberger, P.; Trevitt, A. J.; Gerber, T.; Ross, E.; da Silva, G. Isomer-Specific Product Detection of Gas-Phase Xylyl Radical Rearrangement and Decomposition Using VUV Synchrotron Photoionization. J. Phys. Chem. A 2014, 118, 3593−3604. (14) Kislov, V. V.; Mebel, A. M. Ab Initio G3-Type/Statistical Theory Study of the Formation of Indene in Combustion Flames. I. Pathways Involving Benzene and Phenyl Radical. J. Phys. Chem. A 2007, 111, 3922−3931. (15) da Silva, G.; Bozzelli, J. W. Indene Formation from Alkylated Aromatics: Kinetics and Products of the Fulvenallene + Acetylene Reaction. J. Phys. Chem. A 2009, 113, 8971−8978. (16) Bodi, A.; Hemberger, P.; Gerber, T.; Sztáray, B. A New Double Imaging Velocity Focusing Coincidence Experiment: i2PEPICO. Rev. Sci. Instrum. 2012, 83, 083105. (17) Johnson, M.; Bodi, A.; Schulz, L.; Gerber, T. Vacuum Ultraviolet Beamline at the Swiss Light Source for Chemical Dynamics Studies. Nucl. Instr. Methods Phys. Res., Sect. A 2009, 610, 597−603. (18) Bodi, A.; Sztaray, B.; Baer, T.; Johnson, M.; Gerber, T. Data Acquisition Schemes for Continuous Two-Particle Time-of-Flight Coincidence Experiments. Rev. Sci. Instrum. 2007, 78, 084102. (19) Bodi, A.; Johnson, M.; Gerber, T.; Gengeliczki, Z.; Sztaray, B.; Baer, T. Imaging Photoelectron Photoion Coincidence Spectroscopy with Velocity Focusing Electron Optics. Rev. Sci. Instrum. 2009, 80, 034101. (20) Sztáray, B.; Baer, T. Suppression of Hot Electrons in Threshold Photoelectron Photoion Coincidence Spectroscopy Using Velocity Focusing Optics. Rev. Sci. Instrum. 2003, 74, 3763−3768. (21) Baer, T.; Sztaray, B.; Kercher, J. P.; Lago, A. F.; Bodi, A.; Skull, C.; Palathinkal, D. Threshold Photoelectron Photoion Coincidence Studies of Parallel and Sequential Dissociation Reactions. Phys. Chem. Chem. Phys. 2005, 7, 1507−1513. (22) IGOR Pro; WaveMetrics, Inc.: Lake Oswego, OR, 2013. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford CT, 2009. 11234
dx.doi.org/10.1021/jp5085982 | J. Phys. Chem. A 2014, 118, 11226−11234
