Ultraviolet Photodissociation Dynamics of the 1-Propenyl Radical

Article pubs.acs.org/JPCA

Ultraviolet Photodissociation Dynamics of the 1‑Propenyl Radical Michael Lucas and Yu Song Department of Chemistry, University of California at Riverside, Riverside, California 92521, United States

Jingsong Zhang* Department of Chemistry and Air Pollution Research Center, University of California at Riverside, Riverside, California 92521, United States

Christopher Brazier Department of Chemistry and Biochemistry, California State University, Long Beach Long Beach, California 90840, United States

Paul L. Houston School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332 Department of Chemical Biology, Cornell University, Ithaca, New York 14850, United States

Joel M. Bowman Department of Chemistry and Cherry L. Emerson Center for Scientific Computation, Emory University, Atlanta, Georgia 30322, United States ABSTRACT: Ultraviolet (UV) photodissociation dynamics of jet-cooled 1-propenyl radical (CHCHCH3) were investigated at the photolysis wavelengths from 224 to 248 nm using high-n Rydberg atom time-of-flight (HRTOF) technique. The 1-propenyl radicals were produced from 193 nm photolysis of 1-chloropropene and 1-bromopropene precursors. The photofragment yield (PFY) spectra of the H atom product have a broad peak centered at 230 nm. The H + C3H4 product translational energy P(ET) distribution’s peak near ∼8 kcal/mol, and the fraction of average translational energy in the total available energy, ⟨f T⟩, is nearly a constant of ∼0.12 from 224 to 248 nm. The H atom product has an isotropic angular distribution with the anisotropy parameter β ≈ 0. Quasiclassical trajectory calculations were also carried out using an ab initio ground-state potential energy surface for dissociation of 1-propenyl at the excitation energy of 124 kcal/mol (230 nm). The calculated branching ratios are 60% to the methyl + acetylene products, 16% to H + propyne, 4% to H + allene, and 1% to H + cyclopropene. The experimental and calculated P(ET) distributions of the H + C3H4 products at 230 nm are in a qualitative agreement, suggesting that the H + propyne dissociation is the main H atom product channel. The calculated dissociation time scale on the ground electronic state is ∼1 ps, shorter than but close to the time scale of >10 ps for the overall UV photodissociation implied by the isotropic H atom product angular distribution. The UV photodissociation mechanism of 1-propenyl can be described as unimolecular decomposition of hot 1-propenyl radical on the ground electronic state following internal conversion from the electronically excited states of 1-propenyl.

■

INTRODUCTION C3H5 radicals have four isomers, allyl (CH2CHCH2), 1-propenyl (CHCHCH3), 2-propenyl (CH2CCH3), and cyclo-propyl (c-C3H5). Among the three linear isomers, allyl is a resonancestabilized radical and the smallest π-conjugated system with an odd number of π-electrons, while 1-propenyl and 2-propenyl are β- and α-CH3-substituted vinyl radicals, respectively. The C3H5 radicals are important in hydrocarbon combustion.1−5 Direct H atom abstraction reactions of propene to form the linear C3H5 radicals compete with addition reactions of propene in © XXXX American Chemical Society

combustion conditions and become more important at higher temperatures.6,7 The allyl radical is a precursor in the formation of aromatic hydrocarbons, polycyclic aromatic hydrocarbons, Special Issue: Piergiorgio Casavecchia and Antonio Lagana Festschrift Received: January 31, 2016 Revised: March 4, 2016

A

DOI: 10.1021/acs.jpca.6b01056 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 1. Potential energy diagram of the C3H5 system. All the energies are referenced to that of the most stable isomer, allyl radical (global minimum, GM). The 1-propenyl, 2-propenyl, and cyclo-propyl isomers are labeled as LM2 (local minimum 2), LM1, and LM3, respectively, as in ref 21. Three H-loss dissociation channels and two CH3-loss channels are shown. The energetics and pathways are based on the theoretical calculations in ref 8 and 21, with zero point energy corrections. The vertical excitation energies of the 1-propenyl electronic excited states are based on ref 20 and are relative to 1-propenyl (LM2) at 24.0 kcal/mol. The bracket next to the excited electronic states indicates the photolysis wavelengths from 224 to 248 nm (starting from 1-propenyl, LM2) in this study.

and soot,2−4 while isomerization and unimolecular dissociation of the 1-propenyl and 2-propenyl radicals are involved in the allyl kinetics in combustion.5,8 The electronic excited states and spectroscopy of the allyl radical have been extensively studied experimentally and theoretically,9−19 while the studies on the 1-propenyl and 2-propenyl radicals are limited.20 The first electronic absorption band of allyl at 370−410 nm was assigned to the Ã 2B1 ← X̃ 2A2 (valence n ← π and π* ← n) transition.9 A strong ultraviolet (UV) electronic absorption band with diffuse structure was observed in the 210−250 nm region,10−15 due to Rydberg ← n excitations to four Rydberg electronic excited states, B̃ 2A1(3s) (band origin 249.7 nm), C̃ 2B2(3py) (240.63 nm), D̃ 2A1(3pz) (236.8 nm), and Ẽ 2B1(3px) (∼230.9 nm) of allyl.14,15 The transition to the Ẽ 2B1(3px) state contributes most of the oscillator strength in the UV electronic band of allyl.14,15 There have been no experimental studies on the electronic spectroscopy of the 1-propenyl and 2-propenyl radicals so far, while there was only one theoretical study using the equation-of-motion coupled cluster methods (EOM-CCSD).20 The electronic excited states of 1-propenyl and 2-propenyl are similar to those of the vinyl radical with the first three electronic excitations being valence n ← π, π* ← n, and π* ← π, followed by the Rydberg 3s ← n and 3p ← n excitations. The theory predicted that the first few electronic excited states of 1-propenyl are Ã 2A″ (n ← π, vertical excitation energy 408 nm), B̃ 2A″ (π* ← n, 242 nm), C̃ 2A′ (π* ← π, 221 nm), and D̃ 2A′ (3s ← n, 202 nm).20 Among these states, B̃ 2A″ and C̃ 2A′ were predicted to have higher oscillator strengths.20 Among the four C3H5 isomers, the photodissociation of the allyl radical has been extensively investigated. Figure 1 shows the energetics and dissociation pathways of the C3H5 isomers based on the theoretical work of Chen et al.21 and Narendrapurapu et al.8 Three H-loss dissociation channels, allene (CH2CCH2) + H, propyne (CH3CCH) + H, and cyclopropene (c-C3H4) + H, and two methyl loss channels, CH3 + C2H2 (acetylene) and CH3 + CCH2(vinylidene), are available. The photodissociation dynamics of allyl via the Ã 2B1 state has

been examined in the near-UV region of 420−380 nm,22,23 and via the B̃ 2A1, C̃ 2B2, and Ẽ 2B1 states in the UV region of 249− 216 nm.22,24−26 In the near-UV region, the H-loss product channels of allyl have a product translational energy distribution P(ET) peaking near zero and an isotropic angular distribution. It was proposed that the Ã 2B1 state of allyl decays into the ground state and the hot allyl radical dissociates to allene + H directly and to propyne + H via isomerization to the 1- and 2-propenyl radicals with the direct loss of the central H atom to form allene as the major product channel.22,23 In the UV region from 249 to 216 nm, both the H atom loss and CH3-loss channels were experimentally observed and confirmed by trajectory calculations21,22,24−27 with the H-loss channel being more important and a branching ratio of the H loss to the CH3 loss of 95:5 at 248 nm.21,22,25 The UV photodissociation mechanism of allyl was hot radical dissociation on the ground electronic state following internal conversion from the UV excited electronic states.21,22,24,26 The product translational energy release of the H-loss product channel is modest with the P(ET) distributions peaking near 8.5 kcal/mol and the fraction of the average translational energy in the total excess energy, ⟨f T⟩, being ∼0.2 from 249 to 216 nm.26 The H atom product angular distribution is isotropic. H atoms are produced mainly from direct dissociation of allyl to allene + H, while isomerization to 1-propenyl or 2-propenyl is required to produce propyne + H, with an allene + H and propyne + H ratio of 6:1 indicated by the trajectory calculations for an excitation energy of 115 kcal/mol (equivalent to the 248 nm photoexcitation energy).21 The P(ET) of the CH3 + C2H2 channel peaked at 16 kcal/mol.22 The CH3 elimination channel requires a 1,3 sigmatropic H shift to 1-propenyl or sequential 1,2 H shifts from allyl to 2-propenyl to 1-propenyl.21,25,27 As the intermediate in the unimolecular dissociation pathways of allyl to the propyne + H and CH3 elimination channels, 1-propenyl has the same dissociation channels as allyl (Figure 1),8,21 C−C bond fission to produce C2H2 + CH3, C−H bond fission to produce propyne + H, and isomerization to allyl or 2-propenyl B


Article

The Journal of Physical Chemistry A radicals followed by C−H bond fission to allene + H or to cyclopropene + H.

toward a microchannel plate (MCP) detector positioned perpendicular to the molecular beam and were field-ionized in front of the detector and detected. The nominal flight length was 37.02 cm, which was calibrated by 236 nm photodissociation of HBr with the well-known HBr dissociation energy and spin− orbit splitting energy of Br(2P3/2) and Br(2P1/2). The ion signals were amplified by a fast preamplifier, and the H atom TOF spectra were recorded and averaged using a multichannel scaler. Quasiclassical trajectory calculations were carried out using the ab initio ground-state potential energy surface and method previously reported21 with the assumption that the electronically excited 1-propenyl radical undergoes internal conversion to the ground-state surface before the unimolecular dissociation. The trajectories were initiated with a starting geometry that corresponded to LM2, the 1-propenyl CH3−CH−CH structure. Because the zero-point energy of LM2 is 14261.6 cm−1, the total energy was chosen to be 57739.9 cm−1 above LM2, which corresponded to a photon wavelength of 230 nm for excitation of the 1-propenyl radical from its lowest vibrational level. A total of 2200 trajectories were performed using steps of 0.125 fs, as previously described.21 The relative translational energy was determined from the final frame of each trajectory. The total reaction rate was determined as follows: a histogram, P(i), of the individual reaction times was constructed, giving the number of trajectories that reacted (to any product) in the time bin, i, of about 0.65 ps width. The number of trajectories left unreacted at any time were counted as N(i) = Ntot − Σi P(i). Finally, ln[N(i)] was plotted as a function of time and determined the slope, which was equal to the reaction rate constant.

CHCHCH3 → C2H 2 + CH3, ΔHrxn,0 = 23.1 kcal/mol → propyne + H, ΔHrxn,0 = 30.5 kcal/mol → allene + H,

ΔHrxn,0 = 31.6 kcal/mol

→ c−C3H4 + H, ΔHrxn,0 = 53.9 kcal/mol

There has been no photodissociation study on the 1-propenyl radical so far. Niedzielski et al. investigated dissociation and isomerization of 1- and 2-propenyl radicals chemically activated by hot H atom addition to allene and propyne in a gas cell.28 At an excitation energy of 23 kcal/mol above the H + allene or H + propyne products, 1-propenyl was found to mainly undergo C−C bond cleavage to yield CH3 + acetylene and to a lesser extent isomerize to the allyl radical. Morton et al. studied the unimolecular dissociation of the hot 1-propenyl radical from the secondary dissociation of cis-1-bromopropene at 193 nm.29,30 They found that the dominate channel is methyl elimination and the H elimination channel competes with the methyl elimination at higher energies. The current work investigates the UV photodissociation dynamics of jet-cooled 1-propenyl radical for the first time. In the UV photolysis wavelength region from 224 to 248 nm, the H atom photofragment yield (PFY) spectra (i.e., action spectra) of the 1-propenyl radical are obtained, identifying its UV absorption feature for the first time. The H atom time-of-flight (TOF) spectra using different precursors are recorded and transformed to the product translational energy distributions. The H atom product angular distributions are obtained. The unimolecular dissociation rate of the 1-propenyl is measured with the pump−probe delay in the UV region. Quasi-classical trajectory calculations are also carried out to compare the dynamic information from the experiment and to provide more insight into the dissociation dynamics of the 1-propenyl radical.

■

RESULTS The jet-cooled 1-propenyl radicals in the molecular beam from both the 1-chloropropene and 1-bromopropene precursor were characterized by VUV photoionization TOF mass spectrometry. The production of the C3H5 radicals from 193 nm photolysis of the precursors was identified at m/z 41 in the mass spectra. The nascent C−Br dissociation product translational energy release distribution of the 1-bromopropene precursor at 193 nm indicated that approximately 50% of 1-propenyl radicals produced were stable and the other 50% undergo dissociation or isomerization followed by dissociation.29,30 The stable 1-propenyl radicals were then cooled to low temperature in the supersonic expansion in the molecular beam. The resonanceenhanced multiphoton ionization (REMPI) spectrum of the allyl radical was searched for15,26 and the absence of any signal indicated that no allyl radicals were produced or stabilized from isomerization of the photolytically generated 1-propenyl radicals in the beam. The TOF spectra of the H atom products from the photodissociation of 1-propenyl were measured in the photolysis wavelength region of 224−248 nm with the photolysis laser polarization parallel and perpendicular to the flight path, respectively. Backgrounds have been properly removed, as described in previous studies.26,34 Figure 2 shows the net H atom product TOF spectra of 1-propenyl radical from the 1-chloropropene precursor at four UV photolysis wavelengths 226, 230, 232, and 240 nm, respectively. The TOF spectra using the 1-bromopropene precursor were also taken and they were similar to those from the 1-chloropropene precursor. The photolysis laser power was kept in the range of 0.2−1.3 mJ/pulse; the photolysis laser power dependence was measured, indicating little or no multiphoton dissociation contribution to the H atom product signals. The H atom PFY spectra of 1-propenyl were obtained from integrating the net HRTOF spectra as a function of the UV

■

EXPERIMENTAL AND COMPUTATIONAL METHOD The HRTOF technique and experimental setup have been described in previous studies.26,31−34 The precursors used to generate the 1-propenyl radical were 1-chloropropene (98%, MP Biomedicals) and 1-bromopropene (95%, Acros Organics). 1-Bromopropene has been used as a precursor for the 1-propenyl radical in previous studies.29 A gas mixture of ∼2% precursor seeded in Ar (at a total pressure of ∼120 kPa) was expanded from a pulsed valve (General Valve, Series 9) through a 2 mm diameter nozzle; the precursor in expanded gas mixture was photolyzed right in front of the nozzle by 193 nm radiation from an ArF excimer laser, producing a pulsed 1-propenyl radical beam. The 1-propenyl radical beam was characterized using 121.6 nm vacuum ultraviolet (VUV) photoionization TOF mass spectrometry (TOFMS).26,34 The 1-propenyl radicals were photodissociated by a slightly focused UV photolysis laser radiation (at 220−250 nm, 0.2−1.3 mJ/pulse, line width 0.3 cm−1). A Fresnel-Rhomb achromatic λ/2 plate was used to rotate the polarization of the photolysis radiation for the H atom product angular distribution measurements. The H atoms produced from the 1-propenyl photodissociation were tagged by two-color resonant excitation from 12S to 22P via the H atom Lyman-α transition at 121.57 nm and then further to a high-n Rydberg state by UV radiation at 366.24 nm. A small fraction of the metastable Rydberg H atoms drifted with their nascent velocities C


Article


The net H atom TOF spectra of the jet-cooled 1-propenyl photodissociation are transformed to total center-of-mass (CM) P(ET)s.26,31−34 The CM translational energy of the products, ET, is converted from the H atom flight time tH using the following equation 2 ⎛ ⎛ mH ⎞ mH ⎞⎛ L ⎞ ⎟⎟E H = 1 mH⎜⎜1 + ⎟⎟⎜ ⎟ E T = ⎜⎜1 + 2 ⎝ mC3H4 ⎠ mC3H4 ⎠⎝ t H ⎠ ⎝

(1)

where EH is the laboratory translational energies of the H atom photofragment and L is the flight length. The resultant P(ET) distributions from the photodissociation of 1-propenyl at 226, 230, 232, and 240 nm using the 1-chloropropene precursor are shown in Figure 4 as examples. The P(ET) distributions have a

Figure 2. H atom TOF spectra in the photodissociation of jet-cooled 1-propenyl radical at photolysis wavelengths 226, 230, 232, and 240 nm, respectively. The 1-propenyl radical was produced from 193 nm photolysis of 1-chloropropene precursor. The TOF spectra were net H atom TOF with the 193 nm photolysis radiation on minus off. The photolysis radiation polarization was parallel to the TOF axis (the H atom product angular distribution was shown to be isotropic). Each TOF spectrum was scaled to its maximum intensity.

excitation wavelength. To correct for the variation and drift of the experimental conditions, the H atom signals at 230 nm (for 1-bromopropene precursor) and 232 nm (for 1-chloropropene precursor) were used as references and were monitored after every three measurements at other wavelengths. The H atom intensities at other photolysis wavelengths were scaled to those at 230 or 232 nm and normalized for the photolysis laser power. Figure 3 shows the H atom PFY spectra at 224−248 nm with

Figure 4. CM P(ET) distributions of the 1-propenyl radical at 226, 230, 232, and 240 nm, respectively. Each distribution was normalized to its maximum. The vertical lines indicate the maximum translational energies of the three H atom production channels. The energies are based on the theoretical dissociation energies of the three channels in refs 8 and 21

broad feature that peaks at low translational energy ∼8 kcal/mol and extends to the high energy of ∼75 kcal/mol. The maximum available energy for the propyne + H, allene + H, and c-C3H4 + H channels are indicated in Figure 4. The translational energy release of the H atom loss dissociation channel of 1-propenyl is modest; for example, the average product CM translational energy ⟨ET⟩ is 13.4 kcal/mol at 230 nm and 12.9 kcal/mol at 232 nm, and the fraction of ⟨ET⟩ in the total available energy is ⟨f T⟩ ≈ 0.14 for both wavelengths, assuming for the lowest energy product channel propyne + H. The P(ET) distributions at other photolysis wavelengths from 224 to 248 nm are similar to those at 230 and 232 nm, and the ⟨f T⟩ values are in the range of 0.13−0.15 with an average of 0.12 and nearly a constant value from 224 to 248 nm (Figure 5). The ⟨f T⟩ values using the 1-bromopropene precursor are also obtained at the photolysis wavelengths 224−248 nm and they are similar to but slightly smaller than those from the 1-chloropropene precursor, with an average of 0.11 from 225 to 248 nm (Figure 5). The H atom product angular distributions in the UV photodissociation of 1-propenyl were examined using linearly polarized laser radiation. Figure 6 shows the H atom TOF spectra from the 1-propenyl photodissociation by linearly polarized 230 nm photolysis radiation. The polarization direction of the photolysis

Figure 3. H atom product yield (PFY) spectra of 1-propenyl as a function of photolysis excitation energy in the region of 224−248 nm. Open circles (○) represent the integrated HRTOF signals of 1-propenyl using the 1-chloropropene precursor, while the filled circles (•) represent those from the 1-bromopropene precursor. The signals were normalized to the same photolysis laser power and scaled to the maximum intensity. The spectra are offset for clarity. The error bars indicate the 95% confidence limit from multiple measurements.

both precursors. Both PFY spectra show a similar trend peaking at 230 nm, thus supporting that the H atom product in the TOF spectra originates from the photodissociation of the same species, the 1-propenyl radical. These H atom PFY spectra reveal the UV absorption feature of the 1-propenyl radical for the first time. D


Article


1-propenyl radicals with their electronic transition dipole moment parallel to the electric E of the polarized light. The photofragment angular distribution is given by I(θ) = (1/4π)[1 + βP2(cos θ)], where β is the anisotropy parameter (−1 ≤ β ≤ 2), θ is the angle between the electric vector of the polarized laser radiation E and the recoiling velocity vector of the H atom product (the direction of detection of the TOF axis), and P2(cos θ) is the second Legendre polynomial. Using this equation and the H atom TOF spectra at 230 nm, an anisotropy parameter β ≈ 0 is derived (Figure 6). The H atom product yield time profile in the UV photolysis of 1-propenyl was investigated by varying the delay time between the photolysis laser and probe laser. The time profile was obtained by integrating the HRTOF spectra as a function of the photolysis-probe delay time. The time profile can be used to obtain the microcanonical rate of the unimolecular dissociation of 1-propenyl. The initial rise is determined by the rate of H atom formation from 1-propenyl, while the decay of signal is because of the H atom flight out of the interaction region between the two beams. Figure 7 shows the time profiles of 1-propenyl from the

Figure 5. Photolysis wavelength and fraction of average translational energy release in the total available energy, ⟨f T⟩, in the UV photodissociation of the 1-propenyl radical using both the 1-chloropropene and 1-bromopropene precursors. The average translational energies are calculated from the experimental P(ET) distributions. The total available energy at each photolysis wavelength is derived from the corresponding photon energy and the dissociation energy of the lowest energy H + propyne product channel. The error bars indicate the 95% confidence limit from multiple measurements.

Figure 7. H atom product signal of the 1-propenyl radical as a function of photolysis and probe laser delay time using the 1-chloropropene and 1-bromopropene precursors, respectively. The signals were obtained by integrating the HRTOF spectra at various photolysis-probe delay times. The photolysis wavelength of the 1-propenyl radical from the 1-chloropropene precursor is 230 nm and 236 nm for the 1-bropropene precursor. The solid lines are the fits. See the text for details.

1-chloropropene precursor at 230 nm photolysis radiation and from the 1-bromopropene precursor at 236 nm. The time profiles of the H atom signals, SH(t), are fitted using an expression based on the previous work from Chen’s group to estimate the unimolecular dissociation rates of the 1-propenyl radical35

Figure 6. H atom TOF spectra of 230 nm photodissociation of 1-propenyl, with the polarization E vector of the photolysis radiation (a) perpendicular (θ = 90°) and (b) parallel (θ = 0°) to the TOF axis. The sharp peaks in (a) are due to photolysis of the HCl byproduct produced in the 193 nm photolysis of the 1-chloropropene precursor. (c) The anisotropy parameter β is plotted as a function of H atom timeof-flight. The β parameter stays close to the limit of an isotropic angular distribution.

⎡ ⎤ ⎢ ⎥ 1 SH(t ) = N[1 − exp(−kHt )]⎢ ⎥ (t − a) ⎤ ⎡ ⎢⎣ exp⎣ b ⎦ + 1 ⎥⎦

relative to the TOF axis is indicated in the figure. The H atom TOF spectra are basically identical, showing an isotropic angular distribution. The linearly polarized light preferentially excites the

(2)

where kH(t) is the unimolecular dissociation rate constant for the H atom formation from 1-propenyl, and a and b are constants that describe the width of the plateau region and the decay E


Article


Figure 8. Experimental P(ET) distribution in comparison with those from the trajectory calculations for 1-propenyl photodissociation at 230 nm. The calculated distributions from the three H atom production channels are shown along with the total distribution.

Figure 9. Number of trajectories for H atom loss from different sites in the H + propyne and H + allene product channels, respectively. The 5 atoms of the 1-propenyl radical are labeled in the figure. The geometry of the 1-propenyl radical (LM2) is from ref 21

of the signal. The fittings shown in solid lines in Figure 8 give dissociation rate constants of kH ≥ 1.0 × 108 s−1 at 230 nm and 1.6 × 108 s−1 at 236 nm, respectively. These dissociation rate constant values provide the lower limit as the measurements are limited by the laser pulse time width of ∼10 ns. Under the assumption that the electronically excited 1-propenyl radical undergoes internal conversion and then unimolecular dissociation on the ground-state surface, a total of 2200 trajectories were calculated from the 1-propenyl structure (LM2) with a total internal energy of 57739.9 cm−1 (43478.3 cm−1 or 124.3 kcal/mol above the lowest vibrational level of 1-propenyl, equivalent to the photoexcitation wavelength of 230 nm). Of these trajectories, 4% produced H + allene, 16% gave H + propyne, 1% gave H + cyclopropene, and 60% gave methyl + acetylene. A recent trajectory calculation study of rotationally cold 1-propenyl (LM2) dissociation at a higher excitation energy of 133.2 kcal/mol showed consistent results, with branching ratios of 2% for H + allene, 23% for H + propylene, and 75% for methyl + acetylene.36 The product translational energy distributions for the H-loss channels were obtained and plotted in Figure 8, showing a predominant contribution from the H + propyne channel. The calculated translational energy release is modest with the higher intensity around 10−20 kcal/mol and ⟨ET⟩ ≈ 17.6 kcal/mol and ⟨f T⟩ ≈ 0.19. The trajectory calculations identify the probabilities and mechanisms of H atom loss from different sites, as shown in Figure 9. In the H + propyne channel, removal of the vinylic H1 on the central C1 atom is dominant to that of H2 on the terminal C2 atom. In the minor H + allene channel, the removal of the central vinylic H1 is still more probable than that of the other four H atoms. The total reaction rate was determined in the trajectory calculations to be 1.3 × 1012 s−1 or a lifetime of 0.8 ps. For comparison, RRKM calculations have been carried out in the current work using theoretical energetics and configurations of the transition states in the 1-propenyl reaction pathways.21,37 The total RRKM dissociation lifetime of 1-propenyl at the 124 kcal/mol excitation energy is estimated to be ∼0.3 ps. The RRKM calculations also show that the CH3 loss and H + propyne are the main dissociation channels of 1-propenyl and isomerization to 2-propenyl and allyl is much smaller (∼10% of the H + propyne channel), which is in qualitative agreement with the trajectory calculations and consistent with the lower dissociation energetics of the CH3 + C2H2 and H + propyne product channels (Figure 1).

■

DISCUSSION The UV photodissociation of the 1-propenyl radical in the wavelength range of 224−248 nm was investigated experimentally for the first time. The H atom products from the H + C3H4 dissociation channels were observed directly from the H atom TOF spectra. The H atom PFY spectra recorded using two different precursors of 1-propenyl are similar and reveal a broad UV absorption feature peaking at 230 nm, while no previously reported UV absorption spectrum of 1-propenyl is available for comparison. Koziol et al. calculated the electronic excited states of 1-propenyl as a CH3-substituted vinyl. Two excited states, B̃ 2A″ (π* ← n, vertical excitation energy at 242 nm) and C̃ 2A′ (π* ← π, 221 nm), have their vertical electronic excitation energies in the studied region of 224−248 nm and are also predicted to have higher oscillator strengths.20 The observed UV absorption feature could then be due to either or both of these two electronic excited states. Note that the H atom PFY spectrum represents only part of the UV absorption spectrum of 1-propenyl, as other product channels such as CH3 loss can be significant. As discussed next, the electronically excited 1-propenyl radical likely decays via internal conversion to the ground electronic state and then undergoes unimolecular dissociation on the ground-state surface. Our trajectory and RRKM calculations for the ground-state dissociation of 1-propenyl show that the CH3 loss is the dominant dissociation channel and H + propyne is the main H-loss product channel. The angular distribution of the H atom product from the UV photodissociation of the 1-propenyl radical is isotropic (Figure 6). The isotropic distribution indicates that the time scale of the overall UV photodissociation of 1-propenyl is longer than the rotational period of the 1-propenyl radical. The rotational temperature of the 1-propenyl radical in the molecular beam is expected to around 15 K based on the previous study on the allyl radical.26 Assuming this temperature, the average rotational period of the 1-propenyl radical is estimated to be in the range of 10 ps. The pump−probe experiments in this study were limited by the 10 ns pulsed laser temporal resolution and can only show that the lower limit of the dissociation rate of the 1-propenyl radical at excitation wavelengths 230 and 236 nm is on the order of ∼108 s−1 (Figure 7). The angular distribution and pump− probe delay experiments suggest that the UV photodissociation time scale of the 1-propenyl radical is longer than ∼10 ps. F


Article


The calculated P(ET) distribution has higher kinetic energy release and is more repulsive than the experimental one. The trajectory calculations also provide the P(ET) distributions for individual H-loss channels, which show that the H + propyne channel is dominant, the H + allene channel is significant, but the H + cyclopropene channel is of minor importance. This result is consistent with the reaction barrier for these three H-loss channels (Figure 1) with the c-C3H4 + H channel being the highest energy channel and requiring isomerization to allyl and then ring closing to form cyclo-propyl (c-C3H5). The dominant H + propyne channel has an exit barrier of 5.4 kcal/mol from 1-propenyl, and the peak of ∼8 kcal/mol in the experimental P(ET) distribution is consistent with this barrier. The calculated P(ET) distribution peaks at higher energies, indicating that the calculated ground-state potential energy surface for the H + propyne dissociation channel might be more repulsive than the actual surface. It is unclear why there are two peaks in the calculated P(ET) distribution; better statistics in the trajectory calculations might smooth out the P(ET) distribution. The trajectory calculations also identify the sites of the H atom loss and provide more insights into the dissociation dynamics (Figure 9). In the dominant H + propyne channel, removal of the vinylic H (H1) on the central C1 atom is the main pathway, mostly due to direct C−H bond fission in 1-propenyl (Figure 1). The number of trajectories of loss of H2 on the terminal C2 atom is about one-fifth of that of H1; this suggests that 1-propenyl undergoes 1,2 H shift to 2-propenyl which then loses either H1 or H2 to form propyne. The large H1 to H2 ratio indicates that direct C−H bond fission in 1-propenyl (over the 35.9 kcal/mol energy barrier) is more important than H-elimination from 2-propenyl after isomerization of 1-propenyl (over the 40.5 kcal/mol barrier) (Figure 1). In the minor H + allene channel, the removal of the central vinylic H1 is still more probable than that of the other four H atoms. The loss of H1 to form allene requires 1,3 sigmatropic H shift of one of the methyl H atoms (H3, H4, or H5) from 1-propenyl to allyl (over the 40.2 kcal/mol barrier) and then loss of the central H1 atom. The loss of H3, H4, and H5 suggest 1,2 H shift from 1-propenyl to 2-propenyl followed by loss of the methyl H atoms and/or sequential 1,2 H shifts to allyl and then H loss on the central C atom to H + allene. The loss of H2 is probably due to H1 and H2 scrambling in the 1,2 H shift isomerization between 1-propenyl and 2-propenyl and H2 and H3, H4, H5 scrambling in the 1,3 sigmatropic H shift isomerization between 1-propenyl and allyl, followed by the above dissociation mechanisms to H + allene. Note that direct methyl elimination from 1-propenyl is the lowest energy channel and has the lowest energy barrier among all the pathways of 1-propenyl. Indeed, the trajectory calculations show that this channel is more important than all the H atom channels. However, our current instrument and study could only provide measurements of the H-loss product channels. Overall, the UV photodissociation of 1-propenyl radical is similar to that of the allyl isomer. Although the initial electronic excitation of allyl involves the 3s and 3p Rydberg states and that of 1-propenyl with most likely the valence π* state, both allyl and 1-propenyl undergo internal conversion and proceed to unimolecular dissociation on the ground electronic state. Consequently, the energetics and isomerization pathways determine the outcomes of the overall photodissociation. Although the H atom product P(ET) distribution and angular distribution of the allyl radical and the 1-propenyl radical are similar, there are significant differences in their unimolecular dissociation dynamics. The allyl radical, as the most stable C3H5 isomer, dissociates

The trajectory calculations determined that the total reaction rate of 1-propenyl dissociation on the ground-state surface at the equivalent excitation energy of 230 nm was 1.3 × 1012 s−1 or a lifetime of 0.8 ps, close to the range of the time scale implied by the isotropic product angular distribution. Note that this calculated rate or lifetime is only for the unimolecular dissociation of 1-propenyl once it has reached the ground-state electronic surface after internal conversion and does not include the time for decay from the excited electronic states to the ground electronic state. In earlier studies on the electronically excited states of the allyl radical using picosecond time-resolved pump− probe photoelectron spectroscopy, the decay time of the B̃ 2A1 and C̃ 2B2 excited states of allyl in the 250−238 nm region were shown to range from 20 to 9 ps.38,39 It is plausible that the excited states of 1-propenyl in the UV region have similar decay times. The overall UV photodissociation time scale of >10 ps implied by the isotropic angular distribution could then be consistent with the combined time scale of the decay of the electronic excited states (∼10 ps) and the unimolecular dissociation on the ground electronic state (∼1 ps). Another uncertainty in the current experiment is that the degree of rotational excitation of the 1-propenyl radical was not actually measured and was not well established. If present in the 1-propenyl radical, high rotational excitation might both decrease the rate of dissociation and decrease the anisotropy, as observed in earlier investigations.36,40−44 The CM P(ET) distributions for the H + C3H4 channel show small product translational energy release. The P(ET) distributions have the shape of a typical statistical distribution in unimolecular dissociation, peaking at a low energy of ∼8 kcal/mol and extending to high kinetic energy, approaching the maximum available energy for the cyclopropene + H channel in a onephoton dissociation process of 1-propenyl. The onsets of the other two H-elimination channels, propyne + H and allene + H, are at higher energies (Figure 4). The experimental P(ET) distributions at various excitation wavelengths have similar shape and do not change significantly with increasing excitation energy (with only a small amount of increase in the translational energy release). The ⟨f T⟩ values stay around 0.12 for the photolysis wavelength region of 224−248 nm (Figure 5). The observed P(ET) distributions support the dissociation mechanism of unimolecular decomposition of the hot 1-propenyl radical on the ground electronic state following internal conversion of the excited state. The experimental P(ET) distribution alone cannot differentiate the contributions from the three H-loss dissociation channels. As shown in Figure 1, in addition to direct dissociation of 1-propenyl to CH3 + C2H2 and H + propyne isomerization of 1-propenyl to 2-propenyl and allyl could compete and lead to the H + allene products. The trajectory calculations can help identify the branching ratios of the H-loss channels. Our trajectory calculations, along with the RRKM calculations, show that the H + propyne channel is the dominant H-loss channel, indicating that isomerization of 1-propenyl is not important and could not compete with the direct H-loss dissociation. This is consistent with the higher isomerization barrier than the direct H-loss barrier of 1-propenyl (Figure 1). The P(ET) distribution from the trajectory calculations for the H atom dissociation of 1-propenyl at the equivalent excitation of 230 nm is in qualitative agreement with the experimental P(ET) distribution at 230 nm (Figure 8). The overall H atom P(ET) distribution by the trajectories shows a peak around 10 kcal/mol, along with another peak near 20 kcal/mol, and extends to high kinetic energy of ∼60 kcal/mol. G



■

mainly to H + allene in a simple C−H bond rupture and to a lesser extent to H + propyne after isomerization to 1-propenyl and 2-propenyl over higher barriers; the CH3 loss channel is a minor channel as it also requires isomerization to 1-propenyl.8,21,22,24−26 The 1-propenyl radical, 24.0 kcal/mol higher in energy than allyl, can dissociate to H + propyne in a simple C−H bond rupture and to a lesser extent to H + allene after isomerization to 2-propenyl and allyl; furthermore, 1-propenyl can directly dissociate to CH3 + C2H2 with the lowest energy barrier, making the CH3 elimination (as indicated by the trajectory calculations) the main dissociation channel of 1-propenyl. Because of its lowest energy and higher dissociation and isomerization barriers, the allyl radical in the ground electronic state has a lifetime of 16 ps at 115 kcal/mol internal excitation energy (by trajectory calculations).21 1-Propenyl, due to its higher energy and lower dissociation barriers, dissociates faster in the ground electronic state and has a lifetime of ∼1 ps at the excitation energy of 124 kcal/mol (230 nm).

REFERENCES

(1) Hansen, N.; Miller, J. A.; Taatjes, C. A.; Wang, J.; Cool, T. A.; Law, M. E.; Westmoreland, P. R. Photoionization Mass Spectrometric Studies and Modeling of Fuel-Rich Allene and Propyne Flames. Proc. Combust. Inst. 2007, 31, 1157−1164. (2) Miller, J. A.; Klippenstein, S. J.; Georgievskii, Y.; Harding, L. B.; Allen, W. D.; Simmonett, A. C. Reactions between Resonance-Stabilized Radicals: Propargyl + Allyl. J. Phys. Chem. A 2010, 114, 4881−4890. (3) Zhang, H. R.; Eddings, E. G.; Sarofirn, A. F. A Journey from nHeptane to Liquid Transportation Fuels. 1. The Role of the Allylic Radical and its Related Species in Aromatic Precursor Chemistry. Energy Fuels 2008, 22, 945−953. (4) McEnally, C. S.; Pfefferle, L. D.; Atakan, B.; Kohse-Hoeinghaus, K. Aromatic Hydrocarbon Formation Mechanisms in Flames: Progress Towards Closing the Fuel Gap. Prog. Energy Combust. Sci. 2006, 32, 247−294. (5) Fernandes, R. X.; Giri, B. R.; Hippler, H.; Kachiani, C.; Striebel, F. Shock Wave Study on the Thermal Unimolecular Decomposition of Allyl Radicals. J. Phys. Chem. A 2005, 109, 1063−1070. (6) Huynh, L. K.; Zhang, H. R.; Zhang, S.; Eddings, E.; Sarofim, A.; Law, M. E.; Westmoreland, P. R.; Truong, T. N. Kinetics of Enol Formation from Reaction of OH with Propene. J. Phys. Chem. A 2009, 113, 3177−3185. (7) Kislov, V. V.; Mebel, A. M.; Aguilera-Iparraguirre, J.; Green, W. H. Reaction of Phenyl Radical with Propylene as a Possible Source of Indene and Other Polycyclic Aromatic Hydrocarbons: An Ab Initio/ RRKM-ME Study. J. Phys. Chem. A 2012, 116, 4176−4191. (8) Narendrapurapu, B. S.; Simmonett, A. C.; Schaefer, H. F., III; Miller, J. A.; Klippenstein, S. J. Combustion Chemistry: Important Features of the C3H5 Potential Energy Surface, Including Allyl Radical, Propargyl + H2, Allene + H, and Eight Transition States. J. Phys. Chem. A 2011, 115, 14209−14214. (9) Currie, C. L.; Ramsay, D. A. Electronic Absorption Spectrum and Dissociation Energy of Allyl Radical. J. Chem. Phys. 1966, 45, 488−491. (10) Callear, A. B.; Lee, H. K. Electronic Spectra of Free Allyl Radical and Some of its Simple Derivatives. Trans. Faraday Soc. 1968, 64, 308− 316. (11) Jenkin, M. E.; Murrells, T. P.; Shalliker, S. J.; Hayman, G. D. Kinetics and Product Study of the Self-Reactions of Allyl and Allyl PeroxyRadicals at 296 K. J. Chem. Soc., Faraday Trans. 1993, 89, 433−446. (12) Blush, J. A.; Minsek, D. W.; Chen, P. Electronic Spectrum of Allyl and Allyl-d5 Radicals - the B[12A1]-X[12A2], C[22B1]-X[12A2], and D[12B2]-X[12A2] Band Systems. J. Phys. Chem. 1992, 96, 10150−10154. (13) Minsek, D. W.; Blush, J. A.; Chen, P. 1 + 1 Resonant Multiphoton Ionization Spectrum of the Allyl Radical - Rotational Structure in the C[22B1]-X[12A2] Origin Band. J. Phys. Chem. 1992, 96, 2025−2027. (14) Gasser, M.; Schulenburg, A. M.; Dietiker, P. M.; Bach, A.; Merkt, F.; Chen, P. Single-Photon and Resonance-Enhanced Multiphoton Threshold Ionization of the Allyl Radical. J. Chem. Phys. 2009, 131, 014304−1−8. (15) Gasser, M.; Frey, J. A.; Hostettler, J. M.; Bach, A.; Chen, P. Vibronic Structure of the 3s and 3p Rydberg States of the Allyl Radical. J. Phys. Chem. A 2010, 114, 4704−4711. (16) Ha, T.-K.; Baumann, H.; Oth, J. F. M. Ab Initio Cl Study of the Electronic Spectrum of the Allyl Radical (CH2-CH-CH2). J. Chem. Phys. 1986, 85, 1438−1442. (17) Aquilante, F.; Jensen, K. P.; Roos, B. O. The Allyl Radical Revisited: a Theoretical Study of the Electronic Spectrum. Chem. Phys. Lett. 2003, 380, 689−698. (18) Matsika, S.; Yarkony, D. R. Beyond Two-State Conical Intersections. Three-State Conical Intersections in Low Symmetry Molecules: The Allyl Radical. J. Am. Chem. Soc. 2003, 125, 10672−10676. (19) Mach, T. J.; King, R. A.; Crawford, T. D. A Coupled Cluster Benchmark Study of the Electronic Spectrum of the Allyl Radical. J. Phys. Chem. A 2010, 114, 8852−8857. (20) Koziol, L.; Levchenko, S. V.; Krylov, A. I. Beyond Vinyl: Electronic Structure of Unsaturated Propen-1-yl, Propen-2-yl, 1-Buten2-yl, and trans-2-Buten-2-yl Hydrocarbon Radicals. J. Phys. Chem. A 2006, 110, 2746−2758.

■

CONCLUSION The H atom product channel in the UV photodissociation of jet-cooled 1-propenyl radicals is studied at the photolysis wavelengths from 224 to 248 nm using the HRTOF technique. In this region, the H atom PFY spectra have a broad feature peaking at 230 nm and show the UV absorption feature of 1-propenyl for the first time. The H + C3H4 P(ET) distributions peak near ∼8 kcal/mol, and the fraction of average translational energy in the total excess energy, ⟨f T⟩, is nearly a constant of ∼0.12 from 224 to 248 nm. The H atom product has an isotropic angular distribution. Quasi-classical trajectory calculations were also carried out for the excitation energy of 230 nm. The calculations give the products branching ratios of 60% to methyl + acetylene, 16% to H + propyne, 4% to H + allene, and 1% to H + cyclopropene, indicating that H + propyne dissociation is the predominant H atom product channel and isomerization of 1-propenyl to 2-propenyl and allyl is not significant. The calculated P(ET) distribution of the H + C3H4 products at 230 nm is in a qualitative agreement with the experimental distribution but has a larger kinetic energy release, suggesting that the region of the H-loss channels on the calculated potential energy surface is somewhat too repulsive. The calculated dissociation time scale is ∼1 ps, shorter than but close to the time scale of ∼10 ps for the overall UV photodissociation implied by the isotropic H atom product angular distribution. The UV photodissociation mechanism of 1-propenyl is consistent with internal conversion from the electronically excited state of 1-propenyl followed by unimolecular decomposition of the hot radical on the ground electronic state.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported by the U.S. National Science Foundation (CHE-1214157). The trajectory calculations were supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award Number DE-FG02-97ER14782. H


Article


(43) Szpunar, D. E.; Morton, M. L.; Butler, L. J.; Regan, P. M. Primary and Secondary Dissociation from Allyl Iodide Excited at 193 nm: Centrifugal Effects in the Unimolecular Dissociation of the Allyl Radical. J. Phys. Chem. B 2002, 106, 8086−8095. (44) Szpunar, D. E.; Liu, Y.; McCullagh, M. J.; Butler, L. J.; Shu, J. Photodissociation of Allyl-d2 Iodide Excited at 193 nm: Stability of Highly Rotationally Excited H2CDCH2 Radicals to C-D Fission. J. Chem. Phys. 2003, 119, 5078−5084.

(21) Chen, C.; Braams, B.; Lee, D. Y.; Bowman, J. M.; Houston, P. L.; Stranges, D. The Dynamics of Allyl Radical Dissociation. J. Phys. Chem. A 2011, 115, 6797−6804. (22) Stranges, D.; Stemmler, M.; Yang, X.; Chesko, J. D.; Suits, A. G.; Lee, Y. T. UV Photodissociation Dynamics of Allyl Radical by Photofragment Translational Spectroscopy. J. Chem. Phys. 1998, 109, 5372−5382. (23) Castiglioni, L.; Bach, A.; Chen, P. Spectroscopy and Dynamics of A2B1 Allyl Radical. Phys. Chem. Chem. Phys. 2006, 8, 2591−2598. (24) Deyerl, H. J.; Fischer, I.; Chen, P. Photodissociation Dynamics of the Allyl Radical. J. Chem. Phys. 1999, 110, 1450−1462. (25) Stranges, D.; O’Keeffe, P.; Scotti, G.; Di Santo, R.; Houston, P. L. Competing Sigmatropic Shift Rearrangements in Excited Allyl Radicals. J. Chem. Phys. 2008, 128, 151101. (26) Song, Y.; Lucas, M.; Alcaraz, M.; Zhang, J.; Brazier, C. Ultraviolet Photodissociation Dynamics of the Allyl Radical via the B̃ 2A1(3s), C̃ 2B2(3py), and Ẽ 2B1(3px) Electronic Excited States. J. Phys. Chem. A 2015, 119, 12318−12328. (27) Chen, C.; Braams, B.; Lee, D. Y.; Bowman, J. M.; Houston, P. L.; Stranges, D. Evidence for Vinylidene Production in the Photodissociation of the Allyl Radical. J. Phys. Chem. Lett. 2010, 1, 1875−1880. (28) Niedzielski, J.; Gawlowski, J.; Gierczak, T. Dissociation and isomerization of excited C3H5 radicals in the gas phase. J. Photochem. 1983, 21, 195−206. (29) Morton, M. L.; Miller, J. L.; Butler, L. J.; Qi, F. Dissociation Channels of the 1-Propenyl Radical and Its Photolytic Precursor cis-1Bromopropene. J. Phys. Chem. A 2002, 106, 10831−10842. (30) Miller, J. L.; Morton, M. L.; Butler, L. J.; Qi, F.; Krisch, M. J.; Shu, J. Dissociation Channels of the 1-Propenyl Radical and Its Photolytic Precursor cis-1-Bromopropene. J. Phys. Chem. A 2002, 106, 10965− 10967. (31) Amaral, G.; Xu, K.; Zhang, J. UV Photodissociation Dynamics of Ethyl Radical via the A2A’(3s) State. J. Chem. Phys. 2001, 114, 5164− 5169. (32) Xu, K.; Amaral, G.; Zhang, J. Photodissociation Dynamics of Ethanol at 193.3 nm: The H-atom Channel and Ethoxy Vibrational Distribution. J. Chem. Phys. 1999, 111, 6271−6282. (33) Zheng, X.; Wu, J.; Song, Y.; Zhang, J. Ultraviolet photodissociation of the SD radical in vibrationally ground and excited states. Phys. Chem. Chem. Phys. 2009, 11, 4761. (34) Song, Y.; Zheng, X.; Lucas, M.; Zhang, J. Ultraviolet Photodissociation Dynamics of the Benzyl Radical. Phys. Chem. Chem. Phys. 2011, 13, 8296−8305. (35) Deyerl, H. J.; Fischer, I.; Chen, P. Photodissociation Dynamics of the Propargyl Radical. J. Chem. Phys. 1999, 111, 3441−3448. (36) Conte, R.; Houston, P. L.; Bowman, J. M. Trajectory Study of Energy Transfer and Unimolecular Dissociation of Highly Excited Allyl with Argon. J. Phys. Chem. A 2014, 118, 7742−7757. (37) Davis, S. G.; Law, C. K.; Wang, H. Propyne Pyrolysis in a Flow Reactor: An Experimental, RRKM, and Detailed Kinetic Modeling Study. J. Phys. Chem. A 1999, 103, 5889−5899. (38) Schultz, T.; Fischer, I. The Nonradiative Decay of the Allyl Radical Excited B2A1 State Studied by Picosecond Time-Resolved Photoelectron Spectroscopy. J. Chem. Phys. 1997, 107, 8197−8200. (39) Schultz, T.; Fischer, I. Time-Resolved Photoelectron Spectroscopy of the Allyl Radical: The Lifetimes of the Ultraviolet Bands. J. Chem. Phys. 1998, 109, 5812−5822. (40) Ratliff, B. J.; Womack, C. C.; Tang, X. N.; Landau, W. M.; Butler, L. J.; Szpunar, D. E. Modeling the Rovibrationally Excited C2H4OH Radicals from the Photodissociation of 2-Bromoethanol at 193 nm. J. Phys. Chem. A 2010, 114, 4934−4945. (41) Womack, C. C.; Ratliff, B. J.; Butler, L. J.; Lee, S.-H.; Lin, J. J.-M. Photoproduct Channels from BrCD2CD2OH at 193 nm and the HDO + Vinyl Products from the CD2CD2OH Radical Intermediate. J. Phys. Chem. A 2012, 116, 6394−6407. (42) McKown, B. G.; Ceriotti, M.; Womack, C. C.; Kamarchik, E.; Butler, L. J.; Bowman, J. M. Effects of High Angular Momentum on the Unimolecular Dissociation of CD2CD2OH: Theory and Comparisons with Experiment. J. Phys. Chem. A 2013, 117, 10951−10963. I


Ultraviolet Photodissociation Dynamics of the 1-Propenyl Radical

Recommend Documents