Electron rescattering and the dissociative ionization of alcohols in intense laser light

JOURNAL OF CHEMICAL PHYSICS VOLUME 119, NUMBER 23 15 DECEMBER 2003

Electron rescattering and the dissociative ionization of alcoholsin intense laser light

F. A. Rajgara, M. Krishnamurthy, and D. MathurTata Institute of Fundamental Research, 1 Homi Bhabha Road, Mumbai 400 005, India

~Received 14 July 2003; accepted 18 September 2003!

The fragmentation dynamics of a series of alcohol molecules, from methanol and ethanol, throughhexanol to dodecanol, has been studied by irradiating these molecules with 100 fs duration pulsesof linearly and circularly polarized, infrared, intensity-selected laser light. At laser intensities of1016 W cm22, the yields of singly and multiply charged atomic fragments from all these moleculesare suppressed when circularly polarized light is used. This dependence of the fragmentationdynamics on polarization is rationalized using a simple electron rescattering model. Circularpolarization switches ‘‘off’’ electron rescattering and leads to suppression of multiple ionization andmolecular fragmentation. The degree of suppression depends upon the amount of energy transferfrom the optical field to the molecule: the larger the energy transfer that is required for a particularfragmentation channel, the more marked is its suppression when circular polarization is used. Themaximum kinetic energy that is released upon fragmentation appears to be more or less independentof the polarization state of the incident light. The observation that the actual values of kinetic energyreleased are less than Coulombic indicates that the enhanced ionization mechanism also holds forcircularly polarized light. ©2003 American Institute of Physics.@DOI: 10.1063/1.1625637#

l-mro

ne

htld

dvever,co

tyucesitatioa

ed

cc

oo

-

-ro-

dy

ing

, itightore

c-

li-

de-ani-es inularer,be

l-he-the

ce;c-larto

ent

I. INTRODUCTION

Ready availability of intense optical radiation from utrafast lasers has opened new vistas for probing the dynaof molecular ionization and dissociation. Such dynamics pceed in nonlinear fashion in the strong fields that are geated at light intensities that exceed;1012 W cm22. Field-induced ionization is a foregone conclusion in such ligmolecule interactions. A special feature of strong-fieionization dynamics is that electrons that are ionized frommolecule continue to ‘‘feel’’ the effect of the optical fieleven after the initial ionization event is over. The wapacket that describes the ejected electron initially moaway from the vicinity of the parent molecule. Howevewhen the incident laser light is linearly polarized, the eletronic wave packet is pulled back towards the parent mecule half a cycle after its initial formation. The probabiliof recollision between the electron and the parent molecdepends on the laser phase as well as on the initial veloand initial position of the electronic wave packet. Such rcattering allows the nuclear wave packet to be probed wtime resolutions that are lower than the pulse durationforded by the laser that is used. For instance, the correlabetween the electronic and nuclear wave packets thatcreated in the ionization event has been effectively utiliz1

to probe the motion of the vibrational wave packet of D21

over several femtoseconds with a remarkable temporal aracy of 200 attoseconds, and spatial accuracy of 0.05 Å.

In the case of field-induced ionization of atoms, electrrescattering also affords other tangible benefits: it is respsible for the emission of high harmonics,2,3 generation ofenergetic electrons,4 multiple ionization,5 and for the generation of extremely short~attosecond-duration! light pulses.6,7

The effect of rescattering on diatomics like H2 and D2 has

12220021-9606/2003/119(23)/12224/7/$20.00

Downloaded 18 Dec 2003 to 158.144.56.13. Redistribution subject to A

ics-r-

-

a

s

-l-

leity-hf-nre

u-

nn-

been probed1,8 but, in general, its effect on molecular ionization dynamics is often masked by another strong-field pcess, enhanced ionization,9,10 that can dominate the multipleionization process. Recently, an intense-field, many-boS-matrix theory has been developed11 that explicitly takescognizance of electron wave packet dynamics in determinionization yields in polyatomic molecules.

In considering laser–matter interactions in generalmight be expected that the polarization state of the laser lwould influence the overall dynamics because of one or mof the following factors:~a! The ellipticity of the laser radia-tion and, hence, the direction of the light’s electric field vetor, would affect the trajectory of the electron~or electrons!ejected upon field-induced ionization~or multiple ioniza-tion!; ~b! At the same laser intensity, the electric field amptude would differ for circular and linear polarization; and~c!Angular momentum selection rules are polarization statependent. In the case of atomic ionization, these factors mfest themselves as changes in the ionization rate, changthe energies of the ejected electrons, and on their angdistributions. In the case of molecular ionization, howevadditional facets of the field–molecule interaction need toconsidered, such as:~i! The polarization tensors in the moecule that might lead to spatial alignment, specifically in tcase of linear polarization;~ii ! The dependence of the ionization rate on the angle between the induced dipole inmolecule and the electric field of the incident light;~iii ! Therovibrational couplings in the electronic states that influeninteratomic distances;~iv! The effect of enhanced ionizationand ~v! Differences in the quantal descriptions of the eletronic states that are excited, owing to the different angumomentum selection rules. All these factors contributemaking the physics that underlies polarization-depend

4 © 2003 American Institute of Physics

IP license or copyright, see http://ojps.aip.org/jcpo/jcpcr.jsp

d

ath

on

-er

htpl

i-tey-laprycele

eto

en

n

nda

ouione

ereac

ed

enlaa

traouyrle

Thic

rtc-

ag-thed,ser

ablethel forea-fo-

ro-e

ar-o-

elliestheH

y,nolra-

areas,-tals

ederi-iesea-

d wer-in

sein-ha-ortheero-ticx-

ion-in-thefordy

eenstng.–

12225J. Chem. Phys., Vol. 119, No. 23, 15 December 2003 Electron rescattering and dissociative ionization

molecular dynamics in intense light fields both difficult anat the same time, of interest and importance.

While controversies persist in theoretical formulationsto whether, and to what extent, the polarization state ofincident laser radiation influence atomic ionization,12–15 re-sults of experiments with picosecond pulses offer indicatithat in both the intense (.1013 W cm22) and superintense(.1015 W cm22) regimes, rates for atomic ionization decrease with the ellipticity of the incident light. For instancexperiments on above-threshold-ionization have cleashown that the ionization rate decreases as the laser ligmade more elliptical; this has been rationalized by simsemiclassical formulations.16–18 In the tunnel ionization re-gime, multielectron dissociative ionization of molecular ntrogen has been studied by 100-fs-long laser pulses of insity in the 1015 W cm22 range, using linearly and circularlpolarized infrared light.19,20 Substantial suppression of ionization channels has been observed in the case of circupolarized light, even when laser intensities were approately adjusted to ensure that the laser field experienced b2

was identical in the two cases. Interestingly, the enhanionization mechanism was shown to be valid for multipionization of N2 with circularly polarized light.19 Circularlypolarized laser light, of intensity in the rang1012– 1015 W cm22, has also been recently shown to leada reduced propensity for ionization of a chiral molecule21 inthe picosecond regime. On the other hand, it has also breported that the fragmentation pattern of molecules isaffected by the laser polarization: Talebpouret al.22 haveconducted experiments on benzene with a 100 fs, 800laser at intensities up to 1015 W cm22, and they have re-ported that the fragmentation pattern, and relative ratiosfragment ion yields, are essentially similar for linear acircular polarization. In contrast, results of experiments cried out by Bhardwajet al.23 at a long wavelength~1400 nm!but with similar laser intensities to those used by Talebpet al. indicate that both molecular ionization and dissociatof benzene into molecular fragments exhibit a strong depdence on ellipticity.

We have also recently reported the results of expments on intense-field dissociative ionization of benzen24

with linearly and circularly polarized laser light in whichdistinct lowering was observed of the propensity to produmultiply charged fragment ions when circularly polarizlight is used at laser intensities in excess of 1015 W cm22. Atpeak laser intensities lower than this, light-induced fragmtation appears to be more or less independent of the poization state of the incident intense light. Thus, there appeto be a satisfactory reconciliation with the apparently condictory observations made in the experiments of Talebpet al.22 At higher laser intensities, the lowering of multiplcharged fragment ion yields that is observed with circulapolarized light is attributed to the lowered probability for threscattered electrons to induce dissociative ionization.effect of electron rescattering on the fragmentation dynamof benzene was also noted by Bhardwajet al.,23 although thecontradiction between their observations and those repoby Talebpouret al. were not remarked upon; this contradition remains unresolved.


,

se

s

,lyis

e

n-

rlyi-Nd

enot

m

of

r-

r

n-

i-

e

-r-rs-r

y

es

ed

We have carried out systematic experiments on the frmentation dynamics of a series of polyatomic molecules,linear alcohols, using linearly and circularly polarizefemtosecond-duration, infrared, intensity-selected lapulses of intensities (1016 W cm22) that are large enough togenerate electric fields whose magnitudes are comparwith the interatomic Coulombic fields. We have selectedalcohol series, methanol, ethanol, hexanol, and dodecanoour experiments, as they typify small, medium-sized and rsonably large-sized linear molecules. We have chosen tocus attention mainly on low-yield fragment ions that are pduced in the interaction for the following reason. In thintensity regime that we probe, reliable quantification of pent molecular ion signals, and signals due to high-yield mlecular fragments, is very difficult as ionization occurs win the saturation regime. Determination of relative intensitof saturated ion signals is, obviously, not possible. Incase of methanol, ethanol, and hexanol, ions such as C3

1

and CH1 are examples of low-yield fragments. Similarllong-lived doubly and triply charged methanol and hexamolecules also yield relatively low ion signals. The satution intensities of such ions lie beyond 1016 W cm22. On theother hand, atomic fragment ions, particularly those thatmultiply charged, are convenient probes in this respectfor example, C31 from methanol is observed only for intensities greater that 131015 W cm22, and saturation occurs aintensities that are very much higher. In comparison, signdue to molecular ions like CH3OH1 and C2H5OH1 lie wellin the saturation regime even at 1014 W cm22. It is also im-portant to take cognizance of the fact that highly chargfragment ions possess large values of kinetic energy. Expmental design has to ensure that unit collection efficiencare achieved for detecting such energetic ions. In our msurements we have made special effort to ensure this, anfind that fragment ion yields are significantly lower for cicularly polarized light. We rationalize our observationsterms of a simple electron rescattering model.

The dynamics of fragmentation of a molecule in intenfields can be perceived to occur in two steps. First, thetense laser field ionizes the molecule. The ionization mecnism could be multiphoton ionization, tunnel ionizationover-the-barrier ionization, depending on the intensity ofinteracting laser field. Secondly, the molecular ion, eithsingly or multiply charged, dissociates on the repulsive mlecular ion potential energy surface, giving rise to energefragment ions. Before the molecular ion rolls down the ecited potential surface, rescattering of electrons that areized but undergo oscillation under the influence of thetense laser field significantly affects the fragmentation ofmolecules. Is this rescattering process more significantlarger polyatomics, like the series of alcohols that we stuhere?

II. EXPERIMENTAL METHOD

Our experimental apparatus and methodology have bdescribed recently25 and only those features that are mopertinent to the present study are mentioned in the followiLight pulses~of wavelength 806 nm and pulse duration 90


rain

ghasv-ung–

-v

n

erh

tinmng

OFreti

sroars

ither

fger

aumoura

rion

rosieatyp

innt

inio

ach

thel-hat

oftern

ed

ionhe

itsro-ter.

-r-nd

alsou-

cu-rved

lesond

hat

for

12226 J. Chem. Phys., Vol. 119, No. 23, 15 December 2003 Rajgara, Krishnamurthy, and Mathur

120 fs! were obtained from high-intensity~55 mJ!, chirpedpulse amplification, titanium–sapphire laser system opeing at 10 Hz repetition rate. The laser light was focused usa biconvex lens, of 15 cm focal length, in an ultrahivacuum chamber capable of being pumped down to a bpressure of 2310210 Torr. Our vacuum chamber waflooded with alcohol vapor~after degassing by means of seeral freeze–pump–thaw cycles in a clean, greaseless vacline! such that typical operating pressures were in the raof 631028 Torr, or lower. Ions formed in the lasermolecule interaction were electrostatically extracted intotwo-field, linear, time-of-flight~TOF! spectrometer. The polarization state of the light was varied by use of a half-wa~or quarter-wave! plate. The extent of elliptical polarizatiois defined by the ellipticity parameter,«5(Ex/Ey); in ourexperiments circular polarization implies an«-value of 0.9–1.0.

Focal volume effects play a very important role in detmining the ionization pattern observed using time-of-flig~TOF! spectrometers. By using an aperture in the extracplate of the TOF one can choose the extent of focal voluto be sampled, rather than sample the entire Rayleigh raFor example, it has been shown with molecules like N2 andCS2 that intensity-selective and intensity-averaged Tspectra differ from each other, since different intensitygions are ‘‘seen’’ by the TOF spectrometer due to the spavariation in intensity over the focal volume.25 Intensity-selected measurements are very important for the intenregimes that we are probing so that the large ion counts fthe low intensity region that could swamp the detectoravoided. We have conducted the present experimentintensity-selective mode by placing an aperture of 3 mmfront of our TOF spectrometer. However to ensure thatcollection efficiencies are not compromised, we applied vhigh extraction voltages~of the order 1.2 kV! such that theextraction fields were>250 V cm21 and measurements othe fragment ion yield as a function of the extraction voltawere made to ensure unit collection efficiency, even for venergetic atomic fragments like C31.

Pulse-to-pulse intensity fluctuations in the laser beama feature that can never be totally eliminated. Such flucttions may play a significant role in determining signal aplitudes. We have paid particular attention to this facet inexperimental design. A photodiode was used to monitopart of the incident laser beam. The output of this diode wtransferred, via a fast data bus, to the computer that carout on-line, multichannel data acquisition such that a shby-shot record of laser intensity was maintained. A differedata acquisition channel was used to record ion signals fthe time-of-flight spectrometer, also on a shot-by-shot baSuch an arrangement enabled an intensity window to beily set up. In the series of measurements reported here,cal intensity fluctuations of 5% were tolerated. Ion signalsthose laser shots in which the intensity was outside this rawere discarded. Fragmentation patterns of the type showthe following section~Figs. 1–4! were the result of abou5000 laser shots with the intensity window enabled.

Fragmentation spectra that are presented in the followsection have all been normalized such that the relative


t-g

se

ume

a

e

-tgee.

-al

itymeinney

ey

rea--ras

edt-tms.s-i-

ngein

gn

yields correspond to the same number of laser shots in ecase.

III. RESULTS AND DISCUSSION

A. Fragmentation

From an extensive set of mass spectrometric data oninteraction of intense light with a series of alcohol moecules, we present in the following that subset of data tpertains to the question: How does the polarization statethe intense laser radiation affect the fragmentation patwhen femtosecond pulses are used?

Figure 1 shows fragment ion yields that are obtainupon irradiation of CH3OH by linearly- and circularly-polarized, 806 nm, light at an intensity of 1016 W cm22. Wenote that circular polarization results in a distinct suppressof fragment ion yields. This observation also holds for tdoubly charged methanol ion, CH3OH21. The observation ofthis dication in our time-of-flight spectrum indicates thatlifetime against unimolecular dissociation exceeds a micsecond, which is the flight time through our spectromeFigure 2 shows corresponding data for Cq1 (q51 – 3), O21

atomic fragments as well as CH1 and CH31 molecular frag-

ment ions obtained from C2H5OH. As in the case of methanol, suppression of fragment ion yields with circular polaization is clearly observed. The spectra for hexanol adodecanol that are shown in Figs. 3 and 4, respectively,confirm the ion yield suppression that is obtained with circlar polarization. We note that in the case of hexanol~Fig. 3!,unambiguous data was also obtained for long-lived molelar dication and trications, and the same trend was obsefor yields of these ions as well.

In earlier, long-pulse studies on polyatomic moleculike benzene, carried out using nanosecond and picosecpulses,~see Refs. 26, 27, and references therein!, the ‘‘ladderswitching’’ mechanism, together with its modifications,26,27

was invoked to account for the fragmentation pattern t

FIG. 1. Polarization dependence in the relative fragment ion yieldsmethanol at 1016 W cm22. The M21 data refer to the long-lived CH3OH21

dication.


ethled

nirggiatdataatnaghtht

onale ofse-sor

tota-

ngthe

-e,o-

eethelart

owecu-sesblethe

ri-in

eda-cu-edingo-ap-on-

eth

e

de-


was observed. However, in the present experiments, sinclaser pulses are of only 100 fs duration, we expect thatladder switching mechanism is not likely to be applicabHere, the fragmentation of each of the alcohols we stuhere is most probably induced by population of electroexcited states of the molecular ion whose potential enesurfaces are repulsive, at least in the Franck–Condon rethat is vertically accessible from the ground electronic stof the neutral precursor. The potential energy surface anenergy will, of course, be distorted in some indeterminfashion by the intense laser field. The fragment ions thatformed will depend on the nature of the field-distorted stand on the minimum energy path in the multidimensiopotential energy surface. Gaining proper theoretical insialong these lines remains a challenge for theorists. Neverless, it is possible to make some general observations on

FIG. 2. Polarization dependence in the relative fragment ion yields foranol at 1016 W cm22.

FIG. 3. Polarization dependence in the relative fragment ion yields for hanol at 1016 W cm22. The M21 and M31 data refer, respectively, to thelong-lived CH3(CH2)5OH21 dication and CH3(CH2)5OH31 trication.


thee.ycy

oneitsereelte-he

basis of data that are presented in Figs. 1–4. Conventiladder switching mechanisms demand a large increasunimolecular dissociation rates with internal energy. Conquently, in the multiphoton ionization scenario, the precurion dissociates before there is time for additional photonsbe absorbed. This is the rationale for not observing mestable multiply charged precursor molecular ions in lopulse experiments. The ion pairs that were observed inpicosecond experiments of Bhardwajet al.27 on benzene in-variably had atomic ions, C1 or C21, as one of the constituents. In contrast, femtosecond experiments on benzen24

have yielded strong signals corresponding to long-lived mlecular ions C6H6

21 and C6H631 in addition to atomic ions

like C1 and C21. Similarly, we have found ample evidencfor long-lived molecular dications and trications in thpresent series of experiments. Moreover, we found thatpropensity for producing these multiply charged molecuions was also distinctly lower with circularly polarized lighas compared to that with linearly polarized light~see Figs. 1and 3!.

B. Electron rescattering

So, in the absence of appropriate theoretical input, hdoes one account for the polarization dependence of mollar fragmentation? As noted earlier, for the very short pulused in these experiments, ladder switching is not applicaas one can safely assume that the nuclear motion in allalcohol molecules would be negligibly small over time peods of the order of 100 fs. We invoke electron rescatteringthe laser field in order to qualitatively explain the observsuppression in ion yields. As in the case of multiple ioniztion in atoms, we postulate that fragmentation of the molelar ion is dominantly due to the rescattering of the ionizelectrons in the presence of the laser field. The followchronology of events is invoked: Upon irradiation, the alchol precursor first undergoes tunnel ionization when theplied field intensity reaches a large enough value. The i

-

x-

FIG. 4. Polarization dependence in the relative fragment ion yields for docanol at 1016 W cm22.


leb

eha

oo

thvro

hene

e-he-omoheextia

d

e

ofionkehea-heisuldring

bleeri-ngensano-

ng

orthee en-heergy

ctsergycergy

es-. 6t iscir-e

nce

t re-up-

romsetro

dnce


ized electron does not totally ‘‘leave’’ the alcohol molecubut interacts with it under the influence of both the Coulomforce and the laser field. At low values of laser field~corre-sponding toI 51014 W cm22) the Coulomb field has a larginfluence in determining the motion of the wave packet tdescribes the ejected electron. However, at large fields~cor-responding toI 51016 W cm22), the electric field of the in-teracting laser becomes comparable in magnitude to the Clomb field and, therefore, exerts a much larger influencethe electron trajectories. To determine the influence ofinteracting field on the motion of the ejected electron wapacket, we have carried out a model calculation for a hydgen atom that serves to illustrate the essential physics.24 Wecompute the electron trajectory by numerically solving tclassical equation of motions of the electron that is tunionized.

The equation of motion along thex-axis is

m]2x

]t2 x5eq

r 2 x1eEx,

wheree, m are the charge and mass of the electron,q is thecharge on the molecular ion,r denotes the distance of thelectron from the ion, andE is the laser field. We numerically solve the differential equations of motion along all tx,y,z directions iteratively, with a time grid of 0.01 a.u. Figure 5 depicts classical electron trajectories that we have cputed for linear and circular polarization at laser intensity1016 W cm22. We note that at these large intensities, telectron trajectories for linear polarized light depend onactly when the electron wave packet is created. If the iniposition of the ejected electron lies on they50 line ~whenEis parallel to thex-axis!, then the electron would be expecteto take part in rescattering. As the initial value ofy deviatesfrom zero, the electron rescattering probability becomsmall.

FIG. 5. Electron trajectories for ionization of a prototype atom~atomichydrogen, at position marked X! by linearly ~thick solid line! and circularly~thin line! polarized light of intensity 131016 W cm22. All distances andthe time axis are indicated in atomic units~a.u.!. Note that the electrontrajectory in the case of linearly polarized light is such that rescattering fthe H-atom occurs at least once; the corresponding trajectory in the cacircularly polarized light follows a helical path such that the ejected elecnever revisits the vicinity of the H-atom.


,

t

u-nee-

l

-f

-l

s

Our calculations confirm that for relatively low valueslaser intensity, the electron trajectories for both polarizatstates are similar. At higher laser intensities, li1016 W cm22, the optical field becomes dominant, and telectron trajectories are very different for the two polariztion states. As indicated in Fig. 5, while rescattering of tejected electron is possible with linearly polarized light, itabsent in the case of circularly polarized light. So, one woexpect the fragmentation channels that are due to rescatteto be switched ‘‘off’’ in the latter case. It appears reasonato attribute the differences in fragmentation that are expmentally observed in data shown in Figs. 1–4 to the chain electron rescattering probability. Our model calculatioare simple but demonstrative. However, they pertain toatomic target. This simplicity begs the question: does mlecular structure play a role in determining the overall strofield fragmentation dynamics?

C. Molecular structure

In order to probe this, and to lay the groundwork fproper theoretical treatment, we consider in Fig. 6 howsuppression of fragmentation depends on the appearancergy of fragment ions from specific alcohol precursors. Tion appearance energy is a measure of the ionization enof the given fragment, say C21, plus the bond dissociationenergy. The latter accounts for molecular structure effeand, hence, results in different values of appearance enfor C21 from different alcohol precursors. The appearanenergy is, therefore, a measure of the amount of enetransfer from the optical field to the molecule that is necsary in order to produce a given fragment ion. Data in Figare depicted in terms of the ratio of fragment ion yield thameasured using linear polarization to that obtained usingcular polarization, with the light intensities being kept thsame in both cases. Variation of this ratio with appearaenergy clearly demonstrates that circular polarization~theswitching ‘‘off’’ of electron rescattering! results in distinctlymore marked suppression of fragmentation channels thaquire the largest energy transfer. For instance, a sixfold s

ofn

FIG. 6. Ratio of ion yields obtained with linearly and circularly polarizelight for different fragment ions as a function of fragment ion appearaenergy. The laser intensity was 1016 W cm22. The solid line is a guide to theeye.


izath

-r

heit

ce.e

pnlclyfohorxiin

lybiinon

ion

in-ons,

gethe

als.theels

,id-

erarsd inse

theedhatthe

s-

en-hol

ing

.lesssers

on.

he

so-dr-ntht

of

nea


pression of C31 fragments from various alcohols~appear-ance energy in the region of 88 eV! is obtained with circularpolarization.

Fragment ions that are produced by dissociative iontion possess kinetic energies whose distributions reflectbond length (r ) at which dissociation of the multiplycharged molecular precursor occurs. The kinetic energylease~KER! distributions can, in principle, be mapped to tinternuclear separation if the potential energy functionknown. Such an exercise has, indeed, been carried out incase of light-induced multiple ionization of N2 and peaks inthe ionization rate at;2.2 Å for both linear and circularpolarization have been found that are indicative of enhanionization at internuclear distances larger than equilibrium19

A similar exercise is not possible for the polyatomics wprobe in these experiments. Nevertheless, it has beensible to deduce from our time-of-flight data the kinetic eergy release spectra for various fragment ions from the ahols upon their irradiation by linearly and circularpolarized light. The ratio of KER values that are obtainedthe two polarization states are shown in Fig. 7; we note tthis ratio is more or less unity for the fragment ions fwhich we could make reliable determinations of the mamum KER value. The maximum KER values are shownTable I. The actual values of KER for doubly and tripcharged carbon ions are significantly lower than Coulomvalues, and are indicative of dissociative ionization occurrat internuclear distances that are larger than equilibrium b

FIG. 7. Ratio of the maximum values of kinetic energy release obtaiwith linear and circular polarization for different fragment ions from methnol, ethanol, hexanol, and dodecanol.

TABLE I. Maximum kinetic energy released~eV!, in the center-of-massframe, following formation of C1, C21, and C31 ions from different alco-hol precursors at a laser intensity of 1016 W cm22.

Ion Methanol Ethanol Hexanol Dodecanol

C1 3.4 4.6 4.7 8.2C21 7.5 9.0 9.7 18.6C31 9.4 14.0 15.1 31.6


-e

e-

she

d

os--o-

rat

-

cgd

lengths. This is a manifestation of the enhanced ionizat~EI! process that was referred to in the Introduction.

Enhanced ionization of diatomic molecules in short,tense laser pulses at large critical internuclear separatiRc , was initially discovered in theoretical simulations28 ofthe ionization rates of H2

1 ; ionization maxima observed inthe range ofRc-values 2–5 Å were initially interpreted interms of electron localization effects brought on by larcharge exchange radiative resonance couplings betweenhighest-occupied and lowest-unoccupied molecular orbitTheoretical interpretations of EI in terms of extensions offield-ionization and barrier-suppression ionization modhave also been successful.29 Experimental confirmation of EIhas since been obtained in a number of studies~see, forinstance, Refs. 30, 31! but almost all hitherto-reported workboth theoretical and experimental, has exclusively consered linearly-polarized light. Taken together with earliresults19 on N2, the present data set indicate that EI appeto be a universal phenomenon that is also encountereinteraction of molecules with circularly polarized, intenlight fields.

Our data on the alcohol series enable us to estimatetimescales involved in the dissociation of multiply chargprecursor ions if we assume, in the first approximation, tthe alcohol di- and tri-cation potentials are dominated byCoulomb term. Consider two C-atoms of reduced massM atan initial internuclear separation of 2ba0 . Following suddenionization by our femtosecond laser pulse to charge stateZ1

andZ2 , the time,t, taken for the products of Coulomb explosion to develop a separation ofx is32,33

t'lc

2pca2F b3M

2meZ1Z2G1/2Fx~122ba0 /x!1/2

2ba0

11

2lnF11~122ba0 /x!1/2

12~122ba0 /x!1/2G G ,where a is the fine-structure constant,me is the electronmass,lc is the Compton wavelength whilea0 andc are theBohr radius and speed of light, respectively. The time depdences for dissociation of various charge states of alcomolecular ions can be estimated. For instance, followCoulomb explosion of the ethanol dication, it takes;30 fsfor the distance between C1 – C1 ion-pairs to become 10 ÅThe fact that this time dependence appears to be more orindependent of the polarization state of the incident lalight implies that enhanced ionization in the alcohols followessentially the same route for linear and circular polarizatiSpecifically, the internuclear distance,Rc , at which dissocia-tion occurs with maximum probability, is independent of tpolarization state of the incident radiation.

IV. SUMMARY AND CONCLUDING REMARKS

We have conducted experiments on intense-field disciative ionization of a series of alcohols with linearly ancircularly polarized laser light. We observe a distinct loweing of the propensity to produce multiply charged fragmeions from all these molecules when circularly polarized ligis used at laser intensities of 1016 W cm22. The lowering offragment ion yields is attributed to the lowered probability

d-


tiol-

imneniothds

slundarep

ula

redph

tsfpy

-og

nd

ys

, L

ic

ture

G.

g,

s.

ys.

ys.

s.

ass

ys.


the rescattered electrons inducing dissociative ionizawhen circularly polarized light is used for irradiating the acohols.

Our data indicate that molecular structure effects areportant in determining the degree of suppression that caachieved by changing the polarization state of the incidlaser radiation from linear to circular. Those fragmentatchannels that require the largest transfer of energy fromoptical field to the molecule are suppressed most markeby using circularly polarized light; the suppression is lemarked for those channels that require smaller amountenergy transfer. Moreover, the kinetic energy release vathat are measured for different fragmentation channels icate that bond breakage occurs at distances that are lthan the equilibrium bond length. This is indicative of thenhanced ionization mechanism, and the present resultsvide evidence that this mechanism also holds in the circpolarization regime.

The present set of experiments have probed electroncattering from molecules that are more complex thanatomic species and have revealed a facet of strong-fieldnomena that has hitherto not been considered, namelydependence of the ionization suppression on energy tranThis facet will have to be accounted for in future develoment of theoretical insights into molecular fragmentation dnamics in strong optical fields.

ACKNOWLEDGMENT

Our T4 ~TIFR Table-top Terawatt! laser system was partially financed by the Department of Science and Technolfor which we are grateful.

1H. Niikura, F. Legare, R. Hasbani, M. Yu. Ivanov, D. M. Villeneuve, aP. B. Corkum, Nature~London! 421, 826 ~2003!.

2P. B. Corkum, Phys. Rev. Lett.71, 1994~1993!.3J. L. Krause, K. J. Schafer, and K. C. Kulander, Phys. Rev. Lett.68, 3535~1992!.

4G. G. Paulus, W. Nicklich, H. Xu, P. Lambropoulos, and H. Walther, PhRev. Lett.72, 2851~1994!.

5Th. Weber, H. Giessen, M. Weckenbrock, G. Urbasch, A. StaudteSpielberger, O. Jagutzki, V. Mergel, M. Vollmer, and R. Do¨rner, Nature~London! 405, 658 ~2000!.

6M. Hentschel, R. Kienberger, Ch. Spielmann, G. A. Reider, N. Milosev


n

-bet

nelysofesi-ger

ro-r

s-i-e-heer.--

y

.

.

,

T. Brabec, P. Corkum, U. Heinzmann, M. Drescher, and F. Krausz, Na~London! 414, 509 ~2001!.

7P. M. Paul, E. S. Toma, P. Breger, G. Mullot, F. Auge, Ph. Balcou, H.Muller, and P. Agostini, Science292, 1689~2001!.

8A. S. Alnaser, T. Osipov, E. P. Benis, A. Wech, C. L. Cocke, X. M. Tonand C. D. Lin~private communication!.

9A. D. Bandrauk, Comments Mod. Phys., Part D1, 97 ~1999!.10K. Codling and L. J. Frasinski, J. Phys. B26, 783 ~1993!.11J. Muth-Bohm, A. Becker, S. L. Chin, and F. H. M. Faisal, Chem. Phy

Lett. 337, 313 ~2001!.12A. M. Perelomov, V. S. Popov, and M. V. Terentev, Sov. Phys. JETP23,

924 ~1966!.13S. Klarsfeld and A. Maquet, Phys. Rev. Lett.29, 79 ~1972!.14M. Protopapas, D. G. Lappas, and P. L. Knight, Phys. Rev. Lett.79, 4550

~1997!.15A. Patel, M. Protopapas, D. G. Lappas, and P. L. Knight, Phys. Rev. A58,

R2652~1998!.16P. B. Corkum, N. H. Burnett, and F. Brunel, Phys. Rev. Lett.62, 1259

~1989!.17P. Dietrich, N. H. Burnett, M. Ivanov, and P. B. Corkum, Phys. Rev. A50,

R3585~1994!.18Y. Liang, A. Augst, M. V. Ammosov, S. Lazarescu, and S. L. Chin, J. Ph

B 28, 2757~1995!.19S. Banerjee, G. R. Kumar, and D. Mathur, Phys. Rev. A60, R25 ~1999!.20Ph. Hering and C. Cornaggia, Phys. Rev. A59, 2836~1999!.21M. Krishnamurthy and D. Mathur, Phys. Rev. A61, 063404~2000!.22A. Talebpour, A. D. Bandrauk, K. Vijayalakshmi, and S. L. Chin, J. Ph

B 33, 4615~2000!.23V. R. Bhardwaj, D. M. Rayner, D. M. Villeneuve, and P. B. Corkum, Phy

Rev. Lett.87, 253003~2001!.24F. A. Rajgara, M. Krishnamurthy, and D. Mathur, Phys. Rev. A68, 023407

~2003!.25S. Banerjee, G. R. Kumar, and D. Mathur, J. Phys. B32, L305 ~1999!; 32,

4277 ~1999!.26H. J. Neusser, U. Boesl, R. Weinkauf, and E. W. Schlag, Int. J. M

Spectrom. Ion Processes60, 147 ~1984!.27V. R. Bhardwaj, K. Vijayalakshmi, and D. Mathur, Phys. Rev. A59, 1392

~1999!.28T. Zuo, S. Chelkowski, and A. D. Bandrauk, Phys. Rev. A48, 3837

~1993!.29T. Seideman, M. Y. Ivanov, and P. B. Corkum, Phys. Rev. Lett.75, 2819

~1995!.30G. N. Gibson, M. Li, C. Guo, and J. Neva, Phys. Rev. Lett.79, 202

~1997!.31J. H. Posthumus, L. Frasinski, A. J. Giles, and K. Codling, J. Phys. B28,

L349 ~1995!.32W. Eberhardt, J. Sto¨hr, J. Feldhaus, E. W. Plummer, and F. Sette, Ph

Rev. Lett.51, 2370~1983!.33K. Boyer, T. S. Luk, J. C. Solem, and C. K. Rhodes, Phys. Rev. A39, 1186

~1989!.


Electron rescattering and the dissociative ionization of alcohols in intense laser light

Documents