Effect of Laser Parameters on Ultrafast Hydrogen Migration in

ARTICLE pubs.acs.org/JPCA

Effect of Laser Parameters on Ultrafast Hydrogen Migration in Methanol Studied by Coincidence Momentum Imaging Huailiang Xu,†,‡ Tomoya Okino,† Tatsuya Kudou,† Kaoru Yamanouchi,*,† Stefan Roither,§ Markus Kitzler,§ Andrius Baltuska,§ and See-Leang Chin^ †

Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan State Key Laboratory on Integrated Optoelectronics, College of Electronic Science and Engineering, Jilin University, Changchun 130012, China § Photonics Institute, Vienna University of Technology, Gusshausstrasse 27, A-1040 Vienna, Austria ^ Center for Optics, Photonics and Laser (COPL) & Department of Physics, Engineering Physics and Optics, Laval University, Quebec City, QC, Canada G1 V 0A6 ‡

ABSTRACT: The effect of intensity, duration, and polarization of ultrashort laser pulses (795 nm, 40
100 fs, and 0.15
1.5  1015 W/cm2) on the hydrogen migration in methanol is systematically investigated using Coulomb explosion coincidence momentum imaging. The ratio of the ion yield obtained for the migration pathway CH3OH2+ f CH2+ + OH2+ with respect to the sum of the yields obtained for the migration pathway and for the nonmigration pathway CH3OH2+ f CH3+ + OH+ exhibits a small (10
20%) but clear dependence on laser pulse properties, that is, the ratio decreases as the laser peak intensity increases but increases when the pulse duration increases as well as when the laser polarization is changed from linear to circular.

1. INTRODUCTION When molecules are exposed to an intense laser field, they exhibit a variety of characteristic dynamics such as ionization, isomerization, and fragmentation.1
6 In recent years, a so-called hydrogen migration process within a hydrocarbon molecule that leads to large-scale deformation of the molecular skeletal structure and chemical bond rearrangement associated with ultrafast motion of hydrogen atom(s) or proton(s) induced by an intense laser field has become one of the most attractive research themes.7 This ultrafast molecular behavior in intense laser fields has been demonstrated experimentally in a variety of hydrocarbon molecules.8
16 It has been argued that this ultrafast process may not be described well by conventional adiabatic and Born
Oppenheimer (BO) approximations, as was discussed in a recent theoretical study.17 In the series of our recent experimental studies, we demonstrated that the hydrogen migration plays a decisive role in the breaking and rearrangement processes of chemical bonds in hydrocarbon molecules. For example, we showed that when one of the two initially equivalent C
C bonds in allene, CH2dCdCH2, is broken, it is sensitively dependent on the position of a migrating proton within the parent molecule,18 and the breaking of one of the two terminal C
C bonds in 1,3butadiene is triggered by the migration of the hydrogen atom(s) from the central carbon site to the terminal carbon site.19 Therefore, in view of the importance of ultrafast hydrogen migration in controlling chemical bond-breaking processes in hydrocarbon molecules, it is of particular interest to investigate r 2011 American Chemical Society

what are the most decisive parameters of ultrafast intense laser pulses that govern the dynamics of the ultrafast hydrogen migration. It has been reported that parameters of ultrashort laser pulses such as intensity, pulse duration, and wavelength are key factors in controlling photoinduced chemical processes.20
22 It is thus expected that the ultrafast hydrogen migration process induced by an intense laser field is also affected by such laser parameters. In the present study, we chose three primary laser parameters, that is, intensity, pulse duration, and polarization, and examined their effect on the ultrafast hydrogen migration in methanol, CH3OH, using the coincidence momentum imaging (CMI) method. We record CMI maps of the following two types of two-body Coulomb explosion processes in methanol, that is, the pathway in which the C
O bond is broken without the hydrogen migration23 CH3 OH2þ f CH3 þ þ OHþ

ð1Þ

and the pathway in which the C
O bond is broken after the migration of one hydrogen atom from the methyl group to the hydroxyl group CH3 OH2þ f CH2 þ þ OH2 þ

ð2Þ

Special Issue: Femto10: The Madrid Conference on Femtochemistry Received: August 5, 2011 Revised: September 27, 2011 Published: October 24, 2011 2686

dx.doi.org/10.1021/jp207483y | J. Phys. Chem. A 2012, 116, 2686–2690

The Journal of Physical Chemistry A

ARTICLE

Figure 1. The observed momentum images of CH3+ and CH2+ appearing in coincidence, respectively, with OH+ and OH2+ through the two-body Coulomb explosion pathways of (a) CH3OH2+ f CH3+ + OH+ and (b) CH3OH2+ f CH2+ + OH2+ obtained with laser pulses with a peak intensity of 4.9  1014 W/cm2 and a pulse duration of 40 fs. The laser polarization direction is set to be parallel with the py axis as indicated by the arrow. The total event numbers for (a) and (b) are 9383 and 2989, respectively.

and derive the relative ion yield of the migration pathway (eq 2) with respect to the sum of the yields of the migration pathway and the nonmigration pathway (eq 1) as a function of the three above-mentioned laser parameters.

2. EXPERIMENTS The light source used in the experiments is a Ti:Sapphire femtosecond laser system (Amplitude Technologies, Pulsar 5000), in which output pulses from a Ti:Sapphire oscillator (Femtolasers, Femtosource, Scientific 20) were positively chirped to about 100 ps in an aberration-free stretcher and then amplified using a high-repetition-rate (5 kHz) amplification stage that consisted of a regenerative amplifier, a two-pass preamplifier, and a cryogenically cooled four-pass amplifier. To shorten the pulse duration, an acousto-optic programmable dispersive filter (Fastlite, Dazzler) was placed between the stretcher and the regenerative amplifier to control simultaneously the spectral phase and amplitude of the pulses with a central wavelength at 795 nm. The laser pulses were compressed by a two-grating compressor, by which the pulses can easily be chirped positively and negatively. The pulse energy was adjusted by a half-wave plate and a polarizer, which are located downstream from the compressor. The pulse duration was measured by a SPIDER apparatus, and the radius (∼12 μm) of the focal spot was measured by a CCD camera. The laser pulses were introduced into an ultrahigh vacuum chamber through a quartz lens (f = 15 cm) and a quartz plate window (thickness: 3.5 mm). The sample vapor of methanol (CH3OH) was introduced into the sample vacuum chamber through a microsyringe (0.51 mmϕ) and skimmed by a skimmer (0.48 mmϕ) to form a molecular beam in the differentially pumped ultrahigh vacuum chamber whose base pressure was about 7  10
11 Torr. The molecular beam and the laser beam crossed at right angles, and the ions generated at the laser focal spot in the molecular beam were projected onto a position-sensitive detector (PSD) with delay line anodes (RoentDek DLD 80) by three equally spaced parallel plate electrodes in the velocity mapping configuration.24 The laser polarization direction, electrode plates, and the surface of the detector were all set to be parallel to the plane formed by the molecular beam and laser beam axes. The three-dimensional momentum vector of each fragment ion was determined by its

Figure 2. Laser field intensity dependences of the kinetic energy distributions of the fragment ions released from the two-body Coulomb explosion pathways (a) CH3OH2+ f CH3+ + OH+ and (b) CH3OH2+ f CH2+ + OH2+.

position and arrival time on the detector plane. In order to securely achieve coincidence conditions, the number of generated ions per laser shot was kept at 0.3
0.5 during the experiment. More details on the coincidence measurement method can be found elsewhere.25

3. RESULTS AND DISCUSSION Typical momentum images of CH3+ and CH2+ recorded in coincidence with OH+ and OH2+, respectively, are shown in Figure 1a and b. In this measurement, the laser polarization was linear, the laser peak intensity was I = 4.9  1014 W/cm2, and the pulse duration was Δt = 40 fs. The signals appearing in the central part of the CMI maps of Figure 1a and b are false coincidence events that may originate from the dissociation of singly charged CH3OH+. In Figure 2a and b, we plot the distributions of the kinetic energy Ekin of the fragment ions ejected from the two types of Coulomb explosion pathways, eqs 1 and 2, where Ekin is the sum of the kinetic energy of a pair of fragment ions. The central area of the CMI maps in the kinetic energy range of Ekin < 2 eV, where the false coincidence events also appear, is not shown. The kinetic energy distributions of both the nonmigration and migration pathways are peaked at ∼5.7 eV, and their shapes are very close to each other in a wide range of laser peak intensities, 0.15  1015
1.5  1015 W/ cm2. However, the full width at half-maximum (fwhm) of the kinetic energy distribution for the migration pathway, ∼2.1 eV, is slightly larger than that for the nonmigration pathway, ∼1.8 eV. The released kinetic energy, Ekin, can be converted into the distance between the charge centers of the two fragment moieties, R, by the simple formula of Ekin = q1q2/(4πε0R), with q1 and q2 being the charges of the two fragment ions. For the nonmigration pathway, the distance range of R = 2.2
3.0 Å peaked at ∼2.5 Å is obtained from the kinetic energy distribution 2687

dx.doi.org/10.1021/jp207483y |J. Phys. Chem. A 2012, 116, 2686–2690

The Journal of Physical Chemistry A

Figure 3. The ratio γ of the yield obtained for the migration pathway of CH3OH2+ f CH2+ + OH2+ with respect to the sum of the yields of the migration pathway and the nonmigration pathway of CH3OH2+ f CH3+ + OH+ as a function of GDD and laser pulse duration. The error bars attached to the data points originate from the uncertainties in the measured pulse duration. The ratio γ is calculated using the event numbers of these two pathways accumulated during the experiment, and therefore, no error bar is attached to the data points along the vertical direction.

Figure 4. The peak values of the kinetic energy distributions of the fragment ions released from the nonmigration pathway, CH3OH2+ f CH3+ + OH+ (open circles), and the migration pathway, CH3OH2+ f CH2+ + OH2+ (filled squares), as a function of GDD and laser pulse duration.

with fwhm = 1.8 eV peaked at ∼5.7 eV. For the migration pathway, the distance range of R = 2.1
3.1 Å peaked at ∼2.5 Å is obtained from the kinetic energy distribution with fwhm = 2.1 eV peaked at ∼5.7 eV. In order to discuss the dependence of the hydrogen migration on the laser parameters, we introduce a parameter γ defined as γ = ηmig/(ηmig + ηnonmig), where ηmig and ηnonmig are the coincidence event numbers of the migration pathway of eq 2 and the nonmigration pathway of eq 1, respectively. In Figure 3, we plot the ratio γ as a function of the pulse duration, which was adjusted by changing the distance between the two gratings in the compressor. In this measurement, the peak laser intensity is kept constant at ∼6.5  1014 W/cm2. It can be seen in Figure 3 that there is a clear tendency that the ratio increases when the magnitude

ARTICLE

Figure 5. Intensity dependence of the ratio γ of the yield of the migration pathway, CH3OH2+ f CH2+ + OH2+, with respect to the sum of the yields of the migration pathway and the nonmigration pathway, CH3OH2+ f CH3+ + OH+. The ratio γ is calculated using the event numbers of these two pathways accumulated during the experiment, and therefore, no error bar is attached to the data points along the vertical direction.

of the group delay dispersion (GDD) lc, which may also be called the linear chirp rate, increases. The ratio γ increases by ∼20% when the GDD value increases from lc = 0 to |lc| ≈ 1322 fs2, that is, when the laser pulse duration increases from 40 to 100 fs. This shows that the probability of the hydrogen migration from the carbon atom site to the oxygen atom site increases when the temporal duration of the light
molecule interaction becomes longer. The larger relative ratio of the hydrogen migration pathway in the longer pulse duration may be explained in terms of fieldinduced nonadiabatic transitions among the electronic states26 possibly in the singly charged stage, CH3OH+, as in the case of the dissociative ionization of ethanol in intense laser fields where the relative yield of the C
C bond-breaking pathway and that of the C
O bond-breaking pathway were discussed.27 In Figure 4, we plot the peak values of the kinetic energy distributions as a function of the GDD value and laser pulse duration for both the nonmigration pathway (circle) and the migration pathway (square). It can be seen that the kinetic energy decreases as the GDD value increases either positively or negatively. This is consistent with the fact that longer pulses allow for more time for the C
O bond to be stretched, resulting in lower kinetic energy release. In addition, it can be seen in Figure 4 that the kinetic energy in the migration pathway is generally higher than that in the nonmigration pathway. Thus, the mean distance R reaches larger values for the nonmigration pathway as the pulses become more strongly chirped. This observation suggests the existence of a threshold value of R beyond which the probability of the hydrogen migration decreases. If the distance R becomes too large, hydrogen migration from the carbon atom site to the oxygen atom site may be suppressed. In Figure 5, we plot the ratio γ as a function of the laser peak intensity. As the laser peak intensity increases from ∼0.15  1015 to 1.5  1015 W/cm2, γ decreases by ∼17% from γ = 0.24 to 0.20. This behavior may reflect the competition between the hydrogen migration and ionization that occur simultaneously in the singly charged stage of methanol. As previously shown in ref 18, the hydrogen atom (or proton) can migrate from the carbon site to the oxygen site in the singly charged stage of methanol within the laser pulse. As the laser intensity increases, the ionization rate should become larger and compete with the hydrogen migration. This means that CH3OH+ ions prepared on the singly charged stage are 2688

dx.doi.org/10.1021/jp207483y |J. Phys. Chem. A 2012, 116, 2686–2690

The Journal of Physical Chemistry A

ARTICLE

’ ACKNOWLEDGMENT The authors thank Dr. Katsunori Nakai for fruitful discussion. The present research was supported by two grants from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan (the Grant-in-Aid for Specially Promoted Research on Ultrafast Hydrogen Migration (#19002006) and the Grant-in-Aid for Global COE Program for Chemistry Innovation) and by a joint international grant of the Austrian Science Fund (FWF) under Contract I274-N16 and the Japanese Society for the Promotion of Science (JSPS) (#09035011-000061). H.X. would also like to acknowledge financial support from NSFC 11074098. Figure 6. The ratio of γ obtained with the circularly polarized light (γcircular) with respect to γ obtained with linearly polarized light (γlinear) at four different laser peak intensities of the linearly polarized pulses. No error bar is attached to the data points along the vertical direction because both γcircular and γlinear carry no error bar.

ionized more efficiently prior to the hydrogen migration, resulting in the decrease in the ratio γ, as seen in Figure 5. We also examine the effect of laser polarization on the hydrogen migration. In Figure 6, we plot χ as a function of laser intensity, where χ is defined as χ = γcircular/γlinear, with γcircular and γlinear being the ratios of the ion yield of CH2+ with respect to that of CH3+ obtained, respectively, from the interaction of the linearly and circularly polarized light having the same peak electric field amplitude. It can be seen in Figure 6 that all of the χ values are in the range of 1.10
1.18 in the laser peak intensity range of 0.65
1.02 PW/cm2, showing that circularly polarized light is slightly more efficient than linearly polarized light for inducing the hydrogen migration. The circularly polarized light is known to suppress the ionization induced by the so-called electron rescattering process, and therefore, it may indicate that the suppression of ionization occurs to a lesser extent in the migration pathway than in the nonmigration pathway. Only from these data, it is difficult to discuss the mechanism of this laser polarization effect, but it can be said that the laser polarization is also one of the factors governing the ultrafast hydrogen migration in methanol in an intense laser field.

4. SUMMARY The dependences of hydrogen migration in methanol induced by an ultrafast intense laser field on the laser parameter's intensity, pulse duration, and polarization state have been systematically investigated using ion
ion coincidence momentum imaging of two-body Coulomb explosion processes. It was found that the yield ratio of the hydrogen migration with respect to the nonmigration pathway becomes larger by a factor of ∼20% when the laser pulse duration increases from 40 to 100 fs but becomes smaller by ∼17% when the laser intensity increases from 0.15  1015 to 1.5  1015 W/cm2. It was also found that the ratio of the hydrogen migration is enhanced by 10
18% in a circularly polarized light field as compared to that in a linearly polarized light field with the same laser peak intensity. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Telephone: +81 3 5841 4334. Fax: +81 3 5689 7347.

’ REFERENCES (1) Yamanouchi, K. Science 2002, 295, 1659–1660. (2) Krishnamurthy, M.; Rajgara, F. A.; Mathur, D. J. Chem. Phys. 2004, 121, 9765–9768. (3) Baldit, E.; Saugout, S.; Cornaggia, C. Phys. Rev. A 2005, 71, 021403/1–021403/4. (4) Saugout, S.; Charron, E.; Cornaggia, C. Phys. Rev. A 2008, 77, 023404/1–023404/10. (5) Rajgara, F. A.; Krishnamurthy, M.; Mathur, D.; Safvan, C. P. J. Chem. Phys. 2009, 130, 231104/1–231104/4. (6) Kato, T.; Kono, H.; Kanno, M.; Fujimura, Y.; Yamanouchi, K. Laser Phys. 2009, 19, 1712–1722. (7) Xu, H. L.; Okino, T.; Nakai, K.; Yamanouchi, K. In Progress in Ultrafast Intense Laser Field; Yamanouchi, K., Charalambidis, D., Normand, D., Eds.; Springer-Verlag: Berlin, Germany, 2011; Vol. 7, pp 35
52. (8) Hishikawa, A.; Hasegawa, H.; Yamanouchi, K. J. Electron Spectrosc. Relat. Phenom. 2004, 141, 195–200. (9) Alnaser, A. S.; Litvinyuk, I.; Osipov, T.; Ulrich, B.; Landers, A.; Wells, E.; Maharjan, C. M.; Ranitovic, P.; Bochareva, I.; Ray, D.; Cocke, C. L. J. Phys. B 2006, 39, S485–S492. (10) Okino, T.; Furukawa, Y.; Liu, P.; Ichikawa, T.; Itakura, R.; Hoshina, K.; Yamanouchi, K.; Nakano, H. Chem. Phys. Lett. 2006, 419, 223–227. (11) Okino, T.; Furukawa, Y.; Liu, P.; Ichikawa, T.; Itakura, R.; Hoshina, K.; Yamanouchi, K.; Nakano, H. J. Phys. B 2006, 39, S515–S521. (12) Itakura, R.; Liu, P.; Furukawa, Y.; Okino, T.; Yamanouchi, K. J. Chem. Phys. 2007, 127, 104306/1–104306/5. (13) Hishikawa, A.; Matsuda, A.; Takahashi, E. J.; Fushitani, M. J. Chem. Phys. 2008, 128, 084302/1–084302/5. (14) Hoshina, K.; Furukawa, Y.; Okino, T.; Yamanouchi, K. J. Chem. Phys. 2008, 129, 104302/1–104302/6. (15) Xu, H. L.; Okino, T.; Yamanouchi, K. Chem. Phys. Lett. 2009, 469, 255–260. (16) Xu, H. L.; Maceau, C.; Nakai, K.; Okino, T.; Chin, S. L.; Yamanouchi, K. J. Chem. Phys. 2010, 133, 071103/1–071103/4. (17) Kato, T.; Yamanouchi, K. J. Chem. Phys. 2009, 131, 164118/ 1–164118/14. (18) Xu, H. L.; Okino, T.; Yamanouchi, K. J. Chem. Phys. 2009, 131, 151102/1–151102/4. (19) Xu, H. L.; Okino, T.; Nakai, K.; Yamanouchi, K.; Roither, S.; Xie, X.; Kartashov, D.; Sch€offler, M.; Baltuska, A.; Kitzler, M. Chem. Phys. Lett. 2010, 484, 119–123. (20) Assion, A.; Baumert, T.; Bergt, M.; Brixner, T.; Kiefer, B.; Seyfried, V.; Strehle, M.; Gerber, G. Science 1998, 282, 919–922. (21) Levis, R. J.; Menkir, G. M.; Rabitz, H. Science 2001, 292, 709–713. (22) Yazawa, H.; Shioyama, T.; Suda, Y.; Yamanaka, M.; Kannari, F.; Itakura, R.; Yamanouchi, K. J. Chem. Phys. 2007, 127, 124312/ 1–124312/5. (23) Okino, T.; Furukawa, Y.; Liu, P.; Ichikawa, T.; Itakura, R.; Hoshina, K.; Yamanouchi, K.; Nakano, H. Chem. Phys. Lett. 2006, 423, 220–224. (24) Eppink, A. T. J. B.; Parker, D. H. Rev. Sci. Instrum. 1997, 68, 3477–3484. 2689

dx.doi.org/10.1021/jp207483y |J. Phys. Chem. A 2012, 116, 2686–2690

The Journal of Physical Chemistry A

ARTICLE

(25) Hasegawa, H.; Hishikawa, A.; Yamanouchi, K. Chem. Phys. Lett. 2001, 349, 57–63. (26) Kono, H.; Sato, Y.; Tanaka, N.; Kato, T.; Nakai, K.; Koseki, S.; Fujimura, Y. Chem. Phys. 2004, 304, 203–226. (27) Itakura, R.; Yamanouchi, K.; Tanabe, T.; Okamoto, T.; Kannari, F. J. Chem. Phys. 2003, 119, 4179–4186.

2690

dx.doi.org/10.1021/jp207483y |J. Phys. Chem. A 2012, 116, 2686–2690