Coulomb Explosion of Dichloroethene Geometric Isomers at 1 PW cm

Jan 18, 2013 - ABSTRACT: Strikingly different Coulomb explosion behavior under intense laser fields is shown between the cis and trans geometric isome...
1 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCA

Coulomb Explosion of Dichloroethene Geometric Isomers at 1 PW cm−2 Tomoyuki Yatsuhashi,*,†,‡ Nobuaki Nakashima,† and Juri Azuma§ †

Department of Chemistry, Graduate School of Science, and §Department of Chemistry, Faculty of Science, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi, Osaka 558-8585 Japan ‡ PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho Kawaguchi, Saitama, 332-0012 Japan ABSTRACT: Strikingly different Coulomb explosion behavior under intense laser fields is shown between the cis and trans geometric isomers of dichloroethene using 40-fs pulses at 0.8 μm. Although the fragment-ion distributions in the mass spectra did not aid in the identification of the geometric and positional isomers of the dichloroethenes, we found that the angular distributions of atomic ions were strongly dependent on the geometric structures. The angular distributions of chlorine ions, carbon ions, and protons were similar between 1,1- and cis-1,2-dichloroethene, whereas trans-1,2-dichloroethene showed a very sharp distribution of chlorine ions and quite different distributions of carbon ions and protons. The origin of the anisotropic ion angular distributions is the geometric selection of molecules in the tunnel-ionization process followed by a Coulomb explosion, although molecules are randomly oriented in the gas phase. The highly charged molecular ions exploded into pieces, and the direction of atomic-ion ejection was strongly correlated with the relative configuration of atoms with respect to the electron-extraction axis, the repulsion with adjacent atomic ions within the molecule, and the degree of the persistence of a molecular frame. We propose herein that the most probable electron-extraction axis by tunneling, which is governed by the configuration of molecular orbitals, is different among three dichloroethene isomers. Because multiple ionization under intense laser fields occurs by sequential tunneling processes, the first ionization step at the leading edge of the laser pulse dominates the further ionization steps. Therefore, the shapes of the highest occupied molecular orbitals and probably the underlying orbitals determine the anisotropic emission of atomic ions that can be used to identify isomers.

■

INTRODUCTION Tunneling ionization is a fundamental and important process under intense laser fields.1 One of the most important features of tunneling with strong laser fields is the geometric selection of target molecules, even though molecules are randomly oriented in the gas phase. The conventional ionization methods involving collisions with highly energetic particles such as electrons, atoms, and molecular ions provide no geometry selection of target molecules, although the efficiency of ionization might depend on the relative orientation in collisions between energetic particles and target molecules in some cases.2 Recent studies of high-harmonic generation3 and Coulomb explosions4 revealed that the geometric selectivity and, thus, the tunneling ionization rate depend on the shape of the highest occupied molecular orbitals (HOMOs)5 rather than the Stark shift under strong laser fields.6,7 Moreover, not only the shape but also the symmetry of molecular orbitals determine the overall ionization dynamics. For example, the suppression of ionization efficiency has been reported8 when a molecular orbital has antibonding character (σu or πg), where the destructive interferences between two subwaves of ionizing electron emerging from the two atomic centers occurs.9 In © 2013 American Chemical Society

addition, electron rescattering, which triggers double ionization and fragmentation dynamics, is also affected by orbital symmetry.10,11 This unique feature of tunneling has been used to explore many aspects of the physics of atoms and molecules, with molecular orbital tomography12 and ultrafast molecular orbital dynamics,13 for example. As the laser intensity becomes higher, multiple ionization followed by a Coulomb explosion occurs, and the anisotropic emission of multiply charged atomic ions has been used to reconstruct the molecular structure before the Coulomb explosion by covariance mapping,14 momentum imaging techniques,15 and simple consideration of kinetic energy releases. The successful reconstruction of the structure of a multiply charged ion is based on the geometric selection by tunneling ionization. In addition to the geometric selection of a single species, the identification of multiple species having the same mass has been studied by using shaped femtosecond laser pulses.16 Mass spectroscopy is a powerful tool in the fields of medicine, Received: October 19, 2012 Revised: January 16, 2013 Published: January 18, 2013 1393

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

Article

(TAIGA; Thales Laser). The laser beam passed through several materials such as a beam splitter (quartz, t = 1 mm), a focusing lens (quartz, t = 5 mm), and an ionization chamber window (quartz, t = 3 mm). The same materials were placed in front of the autocorrelator, and group-velocity dispersions introduced by these materials were compensated with an acousto-optic programmable dispersive filter (Dazzler; Fastlite) to obtain the minimum pulse width. We did not achieve gain-narrowing compensation using this filter, and the typical pulse width was 40 fs. A linear mode of reflectron-type time-of-flight mass spectrometer (KNTOF-1800; Toyama) was used for the ion analysis. The chamber pressure was monitored 20 cm from the laser focus point with a cold cathode pressure gauge. The base pressure of the ionization chamber and the time-of-flight chamber was below 5 × 10−7 Pa. The sample pressure in the ionization chamber was kept at 5.5 × 10−5 Pa during the experiments to avoid space-charge effects. The pressure of the time-of-flight mass chamber was kept 10 times below the ionization chamber using differential pumping to avoid collision-induced fragmentation. A slit of 500-μm width was located on the extraction plate perpendicular to the laser propagation direction to collect the ions that were generated in the most tightly focused point of the laser beam (achieving ion collection from an axially symmetric parallel-beam geometry). The acceleration voltage was 4000 V, and the extraction field potential of 303 V cm−1 was optimized so that the double peaks of atomic ions originating in the Coulomb explosion would have similar heights. Under these conditions, the yield of ions emitted backward to the detector with sufficient kinetic energy due to the Coulomb explosion was decreased through the narrow slit located on the extraction plate. Thus, the ions emitted forward to the detector were used to evaluate the ion yield and kinetic energy, because ions were extracted and detected efficiently. The resolution (m/Δm, fwhm) was 560 at m/z = 129. The output signal from a microchannel plate (F4655-11X; Hamamatsu Photonics) was averaged by a digital oscilloscope (Wave Runner 6100, 1 GHz; LeCroy Japan) for 1000 shots. The ion yield was obtained by integrating over the appropriate peaks in the time-of-flight spectrum. The error in the evaluation of the ion yield was about ±5%. The direction of laser polarization (800 nm, linear) with respect to the time-of-flight axis was changed by a zero-order half-wave plate and was confirmed using a broadband polarizing cube beam splitter. The laser beam was focused into the ionization chamber with a planoconvex quartz lens of 200-mm focusing length. The position of the lens was adjusted to achieve a maximum signal intensity for the highest observed charge state of xenon (Xe4+ or Xe5+). The laser energy was attenuated by the combination of a half-wave plate and plate polarizers before the pulse compressor. A part of the laser beam was reflected by a beam splitter at a small incident angle, and the laser pulse energy was measured using a Spectralon-coated integrating sphere (Labsphere) and a calibrated Si pin photodiode. The actual laser intensity of the linear polarized pulse at the focus was determined by measuring the saturation intensity, Isat, of xenon (1.1 × 1014 W cm−2 for a 45-fs pulse) by the method of Hankin et al.,25 and the error in the determination of the absolute laser intensity was about ±10%. The DCE ions were measured successively after the measurement of the Isat value of xenon without changing experimental conditions between runs.

biology, and chemistry; however, it is generally difficult to discriminate geometric isomers by means of their mass spectra alone. The fragmentation pattern in a mass spectrum can help to identify positional isomers such as substituted benzenes to some extent; however, geometric isomers such as cis and trans isomers have generally shown the same fragmentation patterns. Although ion-mobility spectrometry is a promising method for identifying isomers of different sizes,17 a conventional chromatographic technique is typically used for isomer separation prior to mass analysis. Conventional spectroscopic methods such as NMR and IR spectroscopies have a great advantage in isomer identification if a single species is separated in abundance. In this study, we investigated whether Coulomb explosions can be used to identify three dichloroethene isomers having the same mass by observing the angular distribution of atomic ions. Field (tunnel) ionization rather than multiphoton ionization dominates when the electric field of the laser is much larger than that of the valence electrons in atoms or molecules. The Keldysh adiabaticity parameter γ defines the border between the multiphoton-ionization and tunnel-ionization regimes of atoms, and the tunnel-ionization mechanism dominates when γ is much smaller than unity.18 Assuming that the first ionization potential of dichloroethene is 9.8 eV, for the minimum laser intensity used in this work, 1.3 × 1015 W cm−2 (0.8 μm), γ is 0.25, and thus tunnel ionization dominates in this study. Under these conditions, successive ionization proceeds by sequential tunneling processes under the strong alternating electric fields, and charged atoms are prepared within the molecular framework.19 As a result of the strong Coulombic repulsion between atomic ions in close proximity, these ions fly away with a certain kinetic energy. These Coulomb explosion20 dynamics are strongly affected by the charge, the mass,21 the initial geometric configuration of atomic components in molecules, the structural deformation,22 and the migration of atoms23 under a strong electric field. These effects should be strongly related to the molecular properties. Ethene derivatives exposed to intense laser fields provide interesting information because of their planar structures and several possible substitution patterns. We found that the angular distributions of chlorine ions, carbon ions, and protons were similar between 1,1- and cis-1,2-dichloroethene, whereas trans-1,2-dichloroethene showed strikingly different features. We discuss the origin of the distinct anisotropic ion angular distribution due to the geometric selection in the tunnel-ionization process followed by a Coulomb explosion.

■

EXPERIMENTAL SECTION 1,1-Dichloroethene (1,1-DCE, Aldrich, 99%), cis-1,2-dichloroethene (cis-1,2-DCE, Wako Chemicals, 99%), and trans-1,2dichloroethene (trans-1,2-DCE, Nacarai Tesque, 98%) were dried over calcium chloride (1,1-DCE) or diphosphorus pentaoxide (cis- and trans-1,2-DCE) for a few days and then separated and degassed by repeated freeze−thaw cycles before use. Xenon (Japan Air Gases, 99.99%) was used without further purification. The experimental details have been described elsewhere.24 Multiple ionization of DCE and xenon was carried out with a 40-fs pulse centered at 0.8 μm, and the ions were detected by a linear time-of-flight mass spectrometer. An all-diode-pumped Ti:sapphire laser (Alpha 100/1000/XS hybrid, 15 mJ, 800 nm; 2 mJ, 800 nm; Thales Laser) was used in this study. The pulse width was measured by a second-order single-shot autocorrelator 1394

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

■

Article

RESULTS Dissociation of Dichloroethenes with Femtosecond Laser Pulses. The fragmentation patterns of the three dichloroethene isomers in the mass spectrum were identical except the relative abundances, as expected (Figure 1). The

these differences in the relative values found upon femtosecond laser ionization did not help in the isomer identification. In contrast, the chlorine- and carbon-ion yieldswhich were strongly dependent on the configuration between the direction of laser polarization and the ion flight axiswere found to be strikingly different among the isomers. When the direction of laser polarization was parallel or perpendicular to the ion flight axis, we refer to this situation as “parallel” or “orthogonal” conditions, respectively. Intact molecular ions, C2H2Cl2x+, were emitted isotropically with respect to the laser polarization direction because they had nearly zero kinetic energy. In contrast, Cly+ and Cz+ showed anisotropic emissions, as the detection of fragment ions with certain kinetic energies was limited by a narrow slit located on the extraction plate in the mass spectrometer. The inset in Figure 1 shows the peaks of 35 + Cl . The origin of the peak splitting appeared to be the Coulomb explosion (charge-separation processes) of fragment ions: the emissions of ions forward and backward with respect to the flight axis. A significant suppression of the chlorine-ion yield was found in the case of trans-1,2-DCE under orthogonal conditions (Figure 1c inset). On the other hand, the degrees of suppression under orthogonal conditions seemed to be small and similar in the cases of 1,1- and cis-1,2-DCE. The peak kinetic energies evaluated from the time-of-flight spectra27 of Cl+, Cl2+ and Cl3+ measured under parallel conditions at 1.3 × 1015 W cm−2 were 1.8, 13, and 57 eV, respectively, for 1,1DCE; 6.6, 39, and 41 eV, respectively, for cis-1,2- DCE; and 2.8, 4.8, and 23 eV, respectively, for trans-1,2-DCE. Anisotropic Emission of Atomic Ions Due to Geometric Selection by Tunneling. Figures 2−4 compare the angular distributions of 35Cly+ (y = 1−4), 12Cz+ (z = 2 and 3), and H+ in polar coordinates at different laser intensities. The angular distribution of 12C+ was not obtained because of spectral overlap with 37Cl3+. Here, we define θ as the angle measured with respect to the polarization plane of the laser fields from the ion flight axis to the detector. The data were measured experimentally for the two quadrants in 4° steps, and those data were averaged to improve the ratio of signal intensity to noise intensity because the two quadrants were almost identical. The averaged data were then used to make a polar plot for clearer presentation. The symmetries of 1,1-, cis-1,2-, and trans-1,2-DCE are C2v ′ , C2v ″ , and C2h, respectively. As expected from these symmetries, 1,1- and cis-1,2-DCE showed similar angular distributions of chlorine ions, carbon ions, and protons. Mostly chlorine ions were emitted along the laser polarization direction (θ = 0°, 180°), but both 1,1- and cis-1,2DCE had significant fractions in the orthogonal direction (θ = 90°, 270°). In contrast to the differences between the positional isomers, the chlorine-ion angular distributions between the cis and trans geometric isomers were strikingly different. The chlorine ions emitted from trans-1,2-DCE showed a sharp distribution, as in the cases of diatomic molecules. The effect of laser intensity on the chlorine angular distribution was not significant. Regarding the carbon ions, C2+ ions ejected from 1,1- and cis-1,2-DCE showed a football-like distribution with its long axis parallel to the laser polarization direction. The effect of laser intensity on the carbon angular distribution was insignificant in both cases. We note that 1,1-DCE had a relatively large fraction in the orthogonal direction. In contrast, the angular distribution of carbon ions from trans-1,2-DCE showed quite different features compared to the other isomers. C2+ ions ejected from trans-1,2-DCE measured at 1.3 × 1015 W cm−2 showed a rounded rectangular distribution with its long

Figure 1. Time-of-flight mass spectra of (a) 1,1-dichloroethene, (b) cis-1,2-dichloroethene, and (c) trans-1,2-dichloroethene at 1.3 × 1015 W cm−2. The laser polarization was parallel to the ion flight axis. The ion signal was normalized to unity at m/z = 96. The insets show magnifications of a portion of the same spectra. The laser polarizations were parallel (black) and orthogonal (red) to the ion flight axis.

principal ions were C2H2Cl2x+ (x = 1, 2), C2HkCl+ (k = 1, 2), CHlCl+ (l = 0, 1), C2Hm+ (m = 0−2), CHn+ (n = 1, 2), HCl+, Cly+ (y = 1−4), Cz+ (z = 1−3), H2+, and H+. It should be noted that the electron-impact ionization (70 eV) of the three dichloroethene isomers shows the same fragment-ion distribution, including the relative abundances of those ions. In addition, C2H2Cl22+ was missing in the case of electron-impact ionization. The trends of the ion yields as a function of laser intensity did not show significant differences among the three isomers. The vertical ionization potentials of 1,1-, cis-1,2-, and trans-1,2-DCE were 10.0, 9.80, and 9.80 eV, respectively.26 Thus, the values of the tunnel-ionization probability, which is dominated by the ionization potential, were expected to be similar. However, it should be noted that trans-1,2-DCE showed a relatively high abundance of multiply charged chlorine ions but a lower abundance of multiply charged carbon ions. In addition, the ratio between C2H2Cl22+ and C2H2Cl2+ varied. The values of this ratio were 0.11 for 1,1DCE, 0.95 for cis-1,2-DCE, and 0.56 for trans-1,2-DCE at 1.3 × 1015 W cm−2. The error in these ratios was ±10%. However, 1395

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

Article

Figure 2. Angular distributions of chlorine ions (black, 35Cl+; red, 35Cl2+; green, 35Cl3+; blue, 35Cl4+) ejected from (a,d) 1,1-, (b,e) cis-1,2-, and (c,f) trans-1,2-dichloroethene at different laser intensities: (a−c) 1.3 × 1015 and (d−f) 2.2 × 1015 W cm−2. The radius represents the relative ion intensity.

Figure 3. Angular distributions of carbon ions (black, 12C2+; red, 12C3+) ejected from (a,d) 1,1-, (b,e) cis-1,2-, and (c,f) trans-1,2-dichloroethene at different laser intensities: (a−c) 1.3 × 1015 and (d−f) 2.2 × 1015 W cm−2. The radius represents the relative ion intensity.

(Figure 4). The angular distribution of protons contained a significant fraction in the parallel direction but was less sharp compared to that of acetylene.28 However, only the protons ejected from trans-1,2-DCE contained a significant fraction in the orthogonal direction or an isotropic component in the angular distribution, which decreased at higher laser intensity. These differences in atomic-ion ejection found between cis- and trans-1,2-DCE are helpful in discriminating the geometric isomers.

axis orthogonal to the laser polarization direction (Figure 3c). As the laser intensity increased, the angular distribution of C2+ ions ejected from trans-1,2-DCE became broader and was finally emitted in an isotropic manner (Figure 3f). A similar trend was also found in a rigid linear molecule, diiodoacetylene (I−C≡C−I), in which the heavy iodine-ion obstacles disturbed the direction in which the light carbon ions moved and the structural deformation enabled them to travel in an orthogonal direction.28 The initial atomic configuration of hydrogen in the molecule also reflected the angular distribution of protons 1396

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

Article

Figure 4. Angular distributions of H+ ions ejected from (a,d) 1,1-, (b,e) cis-1,2-, and (c,f) trans-1,2-dichloroethene at different laser intensities: (a−c) 1.3 × 1015 and (d−f) 2.2 × 1015 W cm−2. The radius represents the relative ion intensity.

■

The migration of protons under an intense laser field has been reported,32 but chlorine is too heavy to have migrated within 40 fs to form a different isomer structure before the Coulomb explosion. Figure 5 shows the molecular structures and isocontours of the molecular orbitals (HOMO and HOMO − 1) of the three DCEs. The HOMOs of the three DCEs consist of π orbitals made from the carbon−carbon double bond and the four chlorine lone pairs with nodes between them. The nodes of each lobe have a perpendicular configuration between the HOMO and HOMO − 1. Based on the amplitudes of the

DISCUSSION The origin of the anisotropy in the ion angular distributions measured in this work was the restriction of the detection of energetic ions caused by the narrow slit located on the extraction plate and the geometric selection in the tunnelionization process, even though the molecules were randomly oriented in the gas phase. Molecular alignment with intense nano- and picosecond pulses, which allows molecules to rotate and align along the laser polarization direction within the laser pulse duration (known as dynamic alignment), has been well studied.29 Anisotropic emission of atomic ions by Coulomb explosion is obtained as a result of alignment by such longduration pulses. The alignment of large and heavy molecules was not completed within our laser-pulse duration (40 fs); however, geometry-selected tunnel ionization by intense linearly polarized femtosecond laser pulses occurred (known as geometric alignment). Generally, electrons were stripped from the large-amplitude lobe of molecular orbitals along the laser polarization direction by tunneling.4,5 Assuming that the ionization probability was not saturated, the molecules that had a suitable orientation with respect to the laser polarization direction were ionized. Many molecules having different HOMO structures have been examined, and it has been shown that the molecules having HOMOs with asymmetric structures are selectively ionized using asymmetric electric fields.30 Because we used symmetric electric fields in the present study, we expected that alignedrather than orientedmultiply charged molecular ions would be prepared by tunneling. Those molecular ions were exploded into pieces by a Coulomb explosion, and thus, the direction of atomic-ion ejection was strongly correlated with the relative geometry between the molecules and the laser field. However, it should be noted that the direction of ion emission is also dependent on the rotation of molecules, the structural deformation,22 the existence of heavy-atom obstacles within a molecule,20,28 and the migration of atoms31 within the framework of the molecule.

Figure 5. (a) HOMO and (b) HOMO − 1 of (left) 1,1-DCE, (center) cis-1,2-DCE, and (right) trans-1,2-DCE. The HOMOs are drawn in two different views. The arrows are drawn to connect the centers of mass of each lobe, which are thought to be the most probable electronextraction axes. 1397

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

Article

than cis-1,2-DCE, reflecting the difference in the relative configuration of the carbon atoms within the molecule. cis-1,2DCE has equivalent carbons (C2v ″ symmetry), whereas the carbons in 1,1-DCE (C2v ′ symmetry) are not equivalent: one is bonded to chlorine, and the other is far from chlorine. The ejection of the latter carbon was not disturbed by the chlorineion obstacles but was affected by the neighboring carbon, presumably resulting in the different angular distribution. One proton was ejected along the laser polarization direction (electron-extraction axis), whereas the other proton was expected to fly in the orthogonal direction in the cases of 1,1- and cis-1,2-DCE. However, the light proton would have easily been driven by the electric field, and the angular distribution of protons contained a significant fraction in the parallel direction but less sharp. It has been clearly shown that geometry-selective ionization can be well explained by the structure of the HOMO because the first electron is considered to be ejected from HOMO.36 However, the source of the second electron remains to be clearly determined. The contribution of multiple electrons even in the single ionization process has become an important subject in tunneling ionization under intense laser fields.6,7,37 In the case of CO, the contribution of HOMO − 1 to the single ionization process was estimated to be 30%.7 However, O+ and O2+ were ejected in the same direction from CO,38 indicating that the doubly and triply charged CO was generated in the same orientation at 5 × 1013 W cm−2. HCl has a perpendicular configuration between the HOMO and HOMO − 1, and the contribution of the latter in the single ionization process is estimated to be 0.3%.6 It should first be noted that the formation of singly charged ions was saturated at the peak intensity used in the present study (>1 PW cm−2), and thus molecules in any orientation were ionized. The observed lowercharge ions such as C2H2Cl2+ and C2H2Cl22+ at 1 PW cm−2 were generated at the wings of the laser focus because of the spatial distribution of the laser intensity; thus, the lower-charge ions observed at 1015 W cm−2 (peak intensity) are not important and did not contribute to the ion angular distributions. However, the singly charged ions generated at the leading edge of the laser pulse, should be further ionized and finally exploded into pieces at higher-charge states at 1 PW cm−2. In addition, enhanced ionization39 due to bond elongation during higher-order ionization, which takes at least few femtoseconds, would accelerate the sequential ionization of the singly ionized molecules generated at the leading edge of the laser pulse. Thus, the singly charged ions generated at the leading edge of the laser pulse, where laser intensity was fairly low, determine the anisotropic emission of atomic ions even at 1 PW cm−2. Nevertheless, electrons belonging to low-lying orbitals should be emitted during the multiple ionization because at most two electrons can occupy a given orbital. One should consider the contributions not only of the HOMO but also of low-lying orbitals. A similar angular distribution would be expected even if multiple electrons were removed from trans-1,2-DCE under the assumption that the shapes and amplitudes of the lobes of MOs dominate the overall ionization efficiency, even though the nodal patterns and directions of lobes with respect to the molecular long axis are perpendicular between HOMO and HOMO − 1. The sharp angular distribution of chlorine ions ejected from trans-1,2-DCE indicates that higher ionization processes including the formation of Cl4+ occurred along the same electron-extraction axis even at 2.2 × 1015 W cm−2. Because the angular

lobes, it is reasonable to consider not only lobes located on the CC bond but also those on the chlorine atoms. Although the geometry selection in complex molecules is not easily determined, we were able to draw the most probable electron-extraction axis that connects the lobes of large amplitude in the HOMO, as shown in Figure 5. The ionextraction axis is rather simple when the lobes have a linear configuration, as in the case of C6H13I.33 The lobes of the HOMO have a simple pseudolinear configuration along the long axis of trans-1,2-DCE, so the geometric selection of ionization would be straightforward. Because chlorine ions are ejected along the C−Cl bond axis in both directions, the direction of chlorine-ion ejection is not disturbed by another chlorine ion or adjacent carbon ions, and if the C−Cl bond were nearly parallel to the electron-extraction axis, a sharp angular distribution of chlorine ions would be expected, as was, in fact, observed (Figure 2c,f). In addition, the large abundance of multiply charged chlorine ions in the case of trans-1,2-DCE was explained in terms of the generation of multiply charged chlorines on both edges of the molecule.34 The different emission behavior of carbon ions from trans-1,2-DCE (Figure 3c,f) compared to the other isomers can also be explained by the relative configuration of carbons with respect to the electron-extraction axis. Because the CC bond is at an angle of 53° from the electron-extraction axis in the neutral ground state, trans-1,2-DCE does not need to undergo significant structural deformation to emit carbon ions in the orthogonal direction, as is assumed to be the case for diiodoacetylene.28 As expected from Figure 5, the angular distribution of the protons ejected from trans-1,2-DCE contained a significant fraction in the orthogonal direction or an isotropic component because protons were ejected from both C−H bonds that were almost orthogonal to the electron-extraction axis. Protons would be driven in the parallel direction by the electric field, and the orthogonal component in the angular distribution decreased at high laser intensity. The distinct atomic-ion emissions from trans-1,2-DCE can be well explained in terms of the relative configurations of atoms with respect to the electron-extraction axis, the repulsion with adjacent ions, and the persistence of a molecular frame at highly charged states. In contrast to trans-1,2-DCE, the lobes of the HOMOs of 1,1- and cis-1,2-DCE have symmetric but nonlinear configurations. It has been shown that the symmetry of molecular orbitals is strongly correlated to ionization efficiency.9 Nevertheless, electron extraction would be expected to occur along the axis connecting the lobes of large amplitude, judging from previously examined molecules.35 We could draw the two possible electron-extraction axes on the HOMOs of 1,1- and cis-1,2-DCE as shown in Figure 5. If one of those axes happened to be aligned to the laser polarization direction, one chlorine ion would be ejected along the laser polarization direction, whereas the other chlorine ion would be expected to fly in the orthogonal direction for the cases of 1,1- and cis-1,2DCE. However, the fraction of chlorine ions was larger in the parallel direction than in the orthogonal direction (Figure 2). These findings led us to assume that multiply charged chlorine ions were generated34 in one C−Cl bond that was aligned in the laser polarization direction (electron-extraction axis), leading to the Coulomb explosion, whereas the charge in the other C−Cl bond was relatively small, and the corresponding chlorine did not contribute much to the angular distribution in the cases of 1,1- and cis-1,2-DCE. As for the carbon ions, 1,1DCE had a slightly higher abundance of orthogonal component 1398

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399

The Journal of Physical Chemistry A

Article

(16) Dela Cruz, J. M.; Lozovoy, V. L.; Dantus, M. J. Phys. Chem. A 2005, 109, 8447. (17) McLean, J. A.; Ruotolo, B. T.; Gillig, K. J.; Russell, D. H. Int. J. Mass Spectrom. 2005, 240, 301. (18) Keldysh, L. V. Sov. Phys. JETP 1965, 20, 1307. (19) Nakashima, N.; Shimizu, S.; Yatsuhashi, T.; Sakabe, S.; Izawa, Y. J. Photochem. Photobiol. C 2000, 1, 131. (20) Shimizu, S.; Zhakhovskii, V.; Murakami, M.; Tanaka, M.; Yatsuhashi, T.; Okihara, S.; Nishihara, K.; Sakabe, S.; Izawa, Y.; Nakashima, N. Chem. Phys. Lett. 2005, 404, 379. (21) Gibson, G. N.; Li, M.; Guo, C.; Nibarger, J. P. Phys. Rev. A 1998, 58, 4723. (22) Hasegawa, H.; Hishikawa, A.; Yamanouchi, K. Chem. Phys. Lett. 2001, 349, 57. (23) Tilborg, J.; van; Allison, T. K.; Wright, T. W.; Hertlein, M. P.; Falcone, R. W.; Liu, Y.; Merdji, H.; Belkacem. J. Phys. B 2009, 42, 081002. (24) Yatsuhashi, T.; Obayashi, T.; Tanaka, M.; Murakami, M.; Nakashima, N. J. Phys. Chem. A 2006, 110, 7763. (25) Hankin, S. M.; Villeneuve, D. M.; Corkum, P. B.; Rayner, D. M. Phys. Rev. A 2001, 64, 013405. (26) Wittel, K.; Bock, H. Chem. Ber. 1974, 107, 317. (27) Yatsuhashi, T.; Nakashima, N. J. Phys. Chem. A 2010, 114, 11890. (28) Yatsuhashi, T.; Mitsubayashi, N.; Itsukashi, M.; Kozaki, M.; Okada, K.; Nakashima, N. ChemPhysChem. 2011, 12, 122. (29) Bhardwaj, V. R.; Safvan, C. P.; Vijayalakshmi, K.; Mathur, D. J. Phys. B 1997, 30, 3821. (30) Ohmura, H.; Saito, N.; Tachiya, M. Phys. Rev. Lett. 2006, 96, 173001. (31) Hishikawa, A.; Matsuda, A.; Fushitani, M.; Takahashi, E. J. Phs. Rev. Lett. 2007, 99, 258302. (32) Hishikawa, A.; Hasegawa, H.; Yamanouchi, K. J. Electron Spectrosc. Relat. Phenom. 2004, 141, 195. (33) Ohmura, H.; Saito, N.; Nonaka, H.; Ichimura, S. Phys. Rev. A 2008, 77, 053405. (34) Shimizu, S.; Zhakhovskii, V.; Sato, F.; Okihara, S.; Sakabe, S.; Nishihara, K.; Izawa, Y.; Yatsuhashi, T.; Nakashima, N. J. Chem. Phys. 2002, 117, 3180. Markevitch, A. N.; Romanov, D. A.; Smith, S. M.; Levis, R. J. Phys. Rev. Lett. 2004, 92, 063001. (35) Ohmura, H.; Nakanaga, T. J. Chem. Phys. 2004, 120, 5176. (36) Ohmura, H.; Ito, F.; Tachiya, M. Phys. Rev. A 2006, 74, 043410. (37) Boguslavskiy, A. E.; Mikosch, J.; Gijsbertsen, A.; Spanner, M.; Patchkovskii, S.; Gador, N.; Vrakking, M. J. J.; Stolow, A. Science 2012, 335, 1336. (38) Ohmura, H.; Saito, N.; Morishita, T. Phys. Rev. A 2011, 83, 063407. (39) Seideman, T.; Ivanov, M. Yu.; Corkum, P. B. Phys. Rev. Lett. 1995, 75, 2819. (40) Fuß, W.; Schmid, W. E.; Trushin, S. A. J. Chem. Phys. 2000, 112, 8347. (41) Harada, H.; Shimizu, S.; Yatsuhashi, T.; Sakabe, S.; Izawa, Y.; Nakashima, N. Chem. Phys. Lett. 2001, 342, 563. (42) Yatsuhashi, T.; Nakashima, N. J. Phys. Chem. A 2005, 109, 9414. (43) Murakami, M.; Tanaka, M.; Yatsuhashi, T.; Nakashima, N. J. Chem. Phys. 2007, 126, 104304. (44) Tanaka, M.; Panja, S.; Murakami, M.; Yatsuhashi, T.; Nakashima, N. Chem. Phys. Lett. 2006, 427, 255. (45) Murakami, M.; Mizoguchi, R.; Shimada, Y.; Yatsuhashi, T.; Nakashima, N. Chem. Phys. Lett. 2005, 403, 238. (46) Krishnamurthy, M.; Mathur, D. Phys. Rev. A 2000, 61, 063404.

distribution of chlorine ions ejected from 1,1-DCE was broader than that from cis-1,2-DCE, HOMO − 1 of 1,1-DCE, which has a different structure from the HOMO, might contribute to multiple ionization followed by Coulomb explosion. We conclude that only trans-1,2-DCE showed a distinct angular distribution because its HOMO and underlying orbitals have similar geometries. In conclusion, we have demonstrated that the geometric isomers of dichloroethene show quite different angular distributions of atomic ions ejected by a Coulomb explosion. The possibility of isomer discrimination was demonstrated by the photoabsorption of singly charged molecular ions that determine whether intact ions or fragment ions appear under intense femtosecond laser fields.40−44 The wavelength-dependent fragmentation of molecular ions will also help in the identification isomers.45 Shaped laser pulses have the potential to discriminate positional and geometric isomers when the signal ratio between specific ions is observed. 16 The identification of enantiomers using a strong laser field46 is the next challenge; such discrimination might be accomplished by the combination of an asymmetrically shaped electric field and a pump−probe scheme.

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: +81-6-6605-2554. Fax: +81-6-6605-2522. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The present research was supported in part by the JST PRESTO program. REFERENCES

(1) Augst, S.; Meyerhofer, D. D.; Strickland, D.; Chin, S. L. J. Opt. Soc. Am. B 1991, 8, 858. (2) Munakata, T.; Harada, Y.; Ohno, K. Chem. Phys. Lett. 1981, 84, 6. (3) Itatani, J.; Levesque, J.; Zeidler, D.; Niikura, H.; Pépin, H.; Kieffer, J. C.; Corkum, P. B.; Villeneuve, D. M. Nature 2004, 432, 867. (4) Alnaser, A. S.; Voss, S.; Tong, X.-M. ; Maharjan, C. M.; Ranitovic, P.; Ulrich, B.; Osipov, T.; Shan, B.; Chang, Z.; Cocke, C. L. Phys. Rev. Lett. 2004, 93, 113003. (5) Tong, X. -M.; Zhao, Z. X.; Lin, C. D. Phys. Rev. A 2002, 66, 033402. (6) Akagi, H.; Otobe, T.; Staudte, A.; Shiner, A.; Turner, F.; Dörner, R.; Villeneuve, D. M.; Corkum, P. B. Science 2009, 325, 1364. (7) Wu, J.; Schmidt, L. Ph. H.; Kunitski, M.; Meckel, M.; Voss, S.; Sann, H.; Kim, H.; Jahnke, T.; Czasch, A.; Dörner, R. Phys. Rev. Lett. 2012, 108, 183001. (8) Talebpour, A.; Chien, C.-Y.; Chin, S. L. J. Phys. B 1996, 29, L677. (9) Muth-Böhm, J.; Becker, A.; Faisal, F. H. M. Phys. Rev. Lett. 2000, 85, 2280. (10) Okunishi, M.; Shimada, K.; Prümper, G.; Mathur, D.; Ueda, K. J. Chem. Phys. 2007, 127, 064310. (11) Mathur, D.; Dharmadhikari, A. K.; Rajgara, F. A.; Dharmadhikari, J. A. Phys. Rev. A 2008, 78, 013405. (12) Niikura, H.; Dudovich, N.; Villeneuve, D. M.; Corkum, P. B. Phys. Rev. Lett. 2010, 105, 053003. (13) Niikura, H.; Wörner, H. J.; Villeneuve, D. M.; Corkum, P. B. Phys. Rev. Lett. 2011, 107, 093004. (14) Frasinski, L. J.; Codling, K.; Hatherly, P. A. Science 1989, 246, 1029. (15) Hishikawa, A.; Iwamae, A.; Yamanouchi, K. Phys. Rev. Lett. 1999, 83, 1127. 1399

dx.doi.org/10.1021/jp310361x | J. Phys. Chem. A 2013, 117, 1393−1399