Resonant two-colour four-wave mixing spectroscopy of the E 0g+ and f 0g+ ion-pair states of iodine vapour

Resonant two-colour four-wave mixing spectroscopy of the E 0g+ and f 0g+ ion-pair states of iodine vapour

28 October 1994 ELSEVIER CHEMICAL PHYSICS LETTERS Chemical Physics Letters 229 (1994) 285-290 Resonant two-colour four-wave mixing spectroscopy of...

440KB Sizes 0 Downloads 71 Views

28 October 1994

ELSEVIER

CHEMICAL PHYSICS LETTERS

Chemical Physics Letters 229 (1994) 285-290

Resonant two-colour four-wave mixing spectroscopy of the E 0: and f 0: ion-pair states of iodine vapour M.D. Wheeler, I.R. Lambert, M.N.R. Ashfold ’ School ofChemistry,

Unwersity

ofBristol,

Bristol

BS8 ITS,

C’K

Received 15 August 1994

Abstract

The technique of two-colour four-wave mixing spectroscopy has been used to investigate the E 0: and f 0: ion-pair states of iodine vapour in the two-photon energy range 49 100-52 700 cm- ‘. Spectroscopic analysis indicates the need for a further modest refinement of the Dunham parameters used to define the E-state potential energy function.

1. Introduction Four-wave mixing (FWM) methods traditionally have been employed in the areas of optical-phase conjugation [ 1 ] and ultra-fast relaxation phenomena in liquids and solids [ 2 1. Only in recent years has its potentials as a powerful probe in high-resolution gas-phase spectroscopy [3-61 in combustion and plasma diagnostics [ 7,8], and in molecular photodissociation dynamics [ 91 been realised. Although FWM techniques in the gas phase are still in their infancy considerable theoretical analysis has been carried out [ lo- 12 ] in order to investigate at least some of the many features afforded by such experiments. The most common and conceptually most pleasing visualisation of the FWM process is that of the laserinduced grating [ 21 (LIG) in which two laser beams are crossed in an absorbing non-linear medium to form a spatially modulated intensity pattern and, since the laser is tuned to an absorption resonance, a spatial modulation in the concentrations of ground and excited state molecules. This pattern then acts as

I E-mail address: [email protected]

a transient diffraction grating which may be probed by a third laser to obtain a diffracted signal beam. Amongst the major advantages one can foresee for LIGS methods over many of the more conventional forms of spectroscopy are (i) the highly directional nature of the signal (thus minimising laser scatter and/or source background emission) and (ii) the ability to study non-fluorescing and predissociated systems [ 13 1. Degenerate FWM is a special case, wherein the grating and probe beams all have the same frequency: it has been applied to a wide variety of atomic and molecular systems [ 14 1. Here we provide further illustration of the spectroscopic virtues of the less common two-colour variant, in which the grating is formed by overlapping two beams of frequency og and then probed by a third laser beam of frequency w, [ 15: 161; the benefits so derived, and the information obtained, closely parallel those associated with traditional double-resonance spectroscopies. The present study is concerned with the observation and analysis of transitions involving hitherto unreported rovibrational levels of the E and f ion-pair states [ 17-201 of molecular iodine. Ion-pair states

286

M.D. Wheeler et al.

/Chemical

have attracted considerable attention of late, both experimental and theoretical [ 201, it is now recognised that such excited states should be identifiable for many molecules. The halogens are proving to be especially popular for study, not least because their lowest lying ion-pair states lie at energies amenable to many contemporary spectroscopies. The interactions between zeroth-order ion-pair states, Rydberg states lying at comparable energies and (generally repulsive) valence excited electronic states, and the consequences of these interactions on the observable excited state photophysics, are proving to be a particularly fruitful area of study [ 201. The first observations of ion-pair states in molecular iodine were in emission [ 2 11, following high-energy electrical discharge excitation of iodine vapour. The spectra so obtained were impenetrably complex. This is hardly surprising, given the unselective nature of the excitation scheme and the fact that, because of the much greater equilibrium bond lengths of the ion-pair states? the emission spectrum from each ion-pair state comprises a very extensive Franck-Condon progression. More selective, higher resolution data has since been obtained via a number of double resonance experiments - both in the bulk and in a beam - generally proceeding via the bound B 311,(0:) valence state. This choice of intermediate state automatically limits the number of accessible ion-pair states since u++u transitions are dipole forbidden. This problem may be overcome, however, by double resonance excitation via perturbed levels near the dissociation threshold of the B state [ 221. All 18 ion-pair states arising from the combinations I+(3PZ)+I-(‘S0), I+(3P,,0)+I-(1S0) and I+ ( ‘DZ) +I- (‘So) have now been characterised, at least partially, amongst the cluster of states lying at excitation energies z 4 1400,470OO and 55400 cm- ‘, respectively. Molecular constants for many of these states are now available; the most recent and complete set of spectroscopic parameters [ 191 for the E 0: ( 3PZ) and f 0: ( 3P0) states span a very wide range of vibrational levels (E state 123 < v< 346 and f state 40< v< 2 14)) but are restricted to the non-rotating molecule. The present double resonance experiments involve selective preparation of (five) rovibronic levels of the B 311U(0: ) state and further excitation to levels of the E and f ion-pair states lying in the energy range

Physics Letters 229 (I 994) 285-290

49 100-52700 cm-‘. Spectroscopic analysis yields refined term values for a range of rotational levels associated with high vibrational levels ( V= 100-l 06, 113-121, 130-140and 150_157)oftheEstate,and for lower-lying (cl== 27-66) levels of the f state.

2. Experimental The TC-LIGS experimental arrangement has been described elsewhere [ 5 1. The present experimental arrangement uses a small portion of the second harmonic output of an etalon-narrowed Nd : YAG laser (Quanta-Ray DCR-2) at 532.05 nm (18790 cm-‘), unfocused and irised to = 1 mm spot size, to form the grating whilst the remainder is employed to pump a tunable dye laser (Quanta-Ray PDL-3, operating with a range of dyes). The output of the laser is itself frequency doubled in a KDP crystal to give an output in the range 290-360 nm, then irised and used (unfocused) to excite the probe transition. The phasematch angle of the crystal was controlled by an autotracker (Inrad). The TC-LIGS excitation scheme adopted was therefore as follows: 290-360

E,fO,+,v’,J’ 532.05

nm

B 0: , v: J

-

nrn



x

0:

, Ll’!.

J”

The two grating forming beams are both temporally and spatially overlapped at a crossing angle of N 3’ in the centre of a glass cell containing iodine vapour at a pressure of z 300 mTorr. The probe beam is then overlapped at an angle determined by the phasematching requirement sin 0 g=_g_ sinf3,

1 ip’

where 0, and ep are the grating and probe crossing half-angles, respectively, and & and d, are the corresponding wavelengths. The signal detection axis is aligned by placing a thin cuvette containing a dilute solution of iodine in cyclohexane in the interaction region and observing the visible signal beam that may be traced along the 2-3 m to the photomultiplier detector via a number of apertures and a band pass filter. The dye laser wavelength was calibrated in the vis-

M.D. Wheeler et al. /Chemical Physics Letters 229 (1994) 285-290

ible by recording, simultaneously, portions of the well documented I2 (B-X) fluorescence excitation spectrum.

3. Results and discussion Fig. 1 shows a portion of the TC-LIGS spectrum involving levels ofboth the E 0: (3P,) and f 0: ( 3P0) ion-pair states of Iz. The linewidth of the grating forming laser is such that we are not truly state specific in the grating forming transition. This increases the spectral complexity, but not to the extent of causing any ambiguity in interpretation. The Iz(B-X) system has been sufficiently well studied that it is possible to calculate the positions of all excitations in the wavelength region of the Nd:Y.4G second harmonic. This simulation, and the accompanying line assignments, are both displayed in Fig. 2. Subsequent analysis of our TC-LIGS spectra indicates that the output of the pump laser used in these experiments had a bandwidth of ~0.4 cm-’ (fwhm) centred at 18790.0 cm-‘; as a result we see EtB and f+B features originating from B state levels with 2;= 32, J= 5 1 and J=56. u=33. J=84, v=39, J=43 and ~~40, J=65. -.7-7-

287

Valence to ion-pair state transitions in Iz involve significant charge transfer along the internuclear axis, and hold with the A.Q= 0 propensity rule (Hunds case (c) ) expected for such a parallel transition. As a consequence, our TC-LIGS spectra are dominated by progressions of P and R doublets. This may be seen more clearly in Fig. 3, which shows a portion of the survey spectrum of Fig. 1 with much expanded frequency resolution. Analysis of this spectrum is comparatively straightforward given accurate knowledge of the X [ 23 ] and B [ 241 state term values, and the available vibrational Dunham parameters of Wilson et al. [ 19 ] and rotational parameters of Brand et al. [ 171 and Ishiwata et al. [ 181 for the E and f states, respectively. The dominance of the f-B features serves to highlight the very different strengths of the f-B and E-B transition moments in this frequency range. In passing we would also highlight the way in which the f-B and E-B transition probabilities vary with vibrational quantum number (Fig. 1). This marked oscillation is a well-documented feature of such ionpair-valence excitations in the halogen molecules, and is a natural consequence of the Franck-Condon principle when applied to transitions involving highly excited vibrational levels and substantial changes in equilibrium bond length. Note also the TC-LIGS

7

31500

32000

32ioo

Wavenumberd

Fig. 1. Long-range two-colour LIGS spectrum of I2 obtained by tuning the probe laser wavenumber in the range 30900-32500 cm-‘, whilst the grating forming laser is held fixed at 18790 cm- ’ (532.05 nm). Vibronic assignments are indicated by the combs superimposed above the spectrum.

288

M.D. Wheeler et al. /Chemical Physics Letters 229 (1994) X35-290

(32’of Rw) ------I

(32,O)PC52

(3370) P(85L__

(40.2) ~(66)

1 *

I

,

t

I

I

,

I

I

t

I

I

I8791

18790

18789

\Gwxumber!cm’i

Fig. 2. Calculated stick spectrum indicating the rovibronic transitions contributing to the room temperature I1 (B-X) spectrum in the region of 532.05 nm ( 18790 cm-‘). Individual lines are labelled (u,v”) AJ(J” ).

I,

j’=

55

5750

E,v’=l15

/

52 --VW3 64

-_____

66

.~.

jv’=39 85

87

i

I

iv’=38

31760

3&o

31780

31790

31800

Wavenumberlcm-

Fig. 3. Detail of the TC-LIGS spectrum shown in Fig. 1, plotted on a greatly expanded wavenumber scale. The combs superimposed above indicate the vibrational (t“ ) and rotation (J’ ) levels of the ion-pair states involved in this region of the double resonance spectrum.

technique greatiy emphasises this feature, due the intensity of each spectral line being proportional to the square of the one-photon intensity. Detailed analysis of the high-resolution TC-LIGS

such as that shown in Fig. 3 yields f state term values in reasonable agreement with those derived using the available literature parameters [ 181 (maximum deviation AE(obs.-talc.) = -0.6 cm-‘); this, spectra

289

M.D. Wheeler et al. /Chemical Physics Letters 229 (1994) 285-290

0.0

-1.0

-2.0 4 100

140

120

i 160

Vibrational Qua~tim Number

Fig. 4. Plot showing the way in which hE(obs.-talc.), the difference between the observed and calculated [ 17,191 term values for two different J’ levels of the E state of Iz, varies with LJ’quantum number. The full line is to guide the eye through our data points, whilst the dashed line represents data from Wilson et al. [ 19 1. ( A ) J= 50, (

however, is still somewhat large given the narrow ( z 0.1 cm-’ ) bandwidth of the probe dye laser. The situation regarding the E state is far worse: the rovibrational term values determined herein show large deviations from the values predicted using the vibrational data of Donovan and co-workers [ 19 ] and the rotational Dunham parameters of Brand and Hoy et al. [ 17 1. This is illustrated in Fig. 4, which shows the vibrational dependence of this discrepancy for two different levels of rotational excitation within the E state (J’ = 50 and 87). The striking similarity between these two data sets strongly suggests that the error is in the vibrational part of the Dunham expansion, as might be expected since the reported parameters [ 191 were derived from fitting data involving vibrational levels with v’ > 123 only. The present experiments involve much shorter probe wavelengths than in our earlier two-colour LIGS study of high vibrational levels of the ground state of I*. Theory [ 4 ] would predict substantially relaxed phase matching constraints for the present combination of pump and probe wavelengths (A~~750 cm-’ (fwhm) at A,=320 nm, using Eq. (7) in Ref. [ 41) - a prediction fully borne out by the fact that we are able to tune over the complete frequency-doubled DCM dye range ( z 2000 cm-’ ) without the need to readjust the phase-matching of the beams.

4. Conclusions We have used the technique of two-colour laser-induced grating spectroscopy to record high-resolution double-resonance-like excitation spectra of the E and f ion-pair states of molecular iodine. The deduced term values of these levels highlight limitations in the available Dunham parameters, especially for the intermediate vibrational levels of the E state. The present work indicates that TC-LIGS is one technique by which these deficiencies could be rectified, given the availability of a more widely tunable excitation source for the grating forming step.

Acknowledgement

We thank Dr. CM. Western for the use of his I2 spectral simulation programs and K.N. Rosser for experimental help and advice. We are also grateful to the Science and Engineering Research Council for a postdoctoral fellowship (IRL) and the University of Bristol for the award of a postgraduate scholarship (to MDW).

290

M.D. Wheeler et al. /Chemcal Physm Letters 229 (1994) 285-290

References

[ 1] R.L. Abrams, J.F. Lam, R.L. Lind, D.G. Steel and P.F. Liao, in: Optical phase conjugation. Phase conjugation and high resolution spectroscopy by DFWM, ed. R.L. Fischer (Academic Press, New York, 1983) ch. 8, and references therein. [2] H.J. Eichler, P. Gilnter and D.W. Pohl, Laser induced dynamic gratings (Springer, Berlin, 1986). [ 31R.L. Farrow and D.J. Rakestraw, Science 257 ( 1992) 1894, and references therein. [4] T.J. Butenhoff and E.A. Rohllin. J. Chem. Phys. 98 (1993) 5460.

[ 51 M.D. Wheeler, I.R. Lambert and M.N.R. Ashfold, Chem. Phys. Letters 2 11 ( 1993 ) 39 I. [6] P. DeRose, H.L. Dai and P.Y. Cheng, Chem. Phys. Letters 220 (1994) 207. [ 71 T. Dreier and D.J. Rakestraw, Opt. Letters 15 (1990) 72. [S] D.S. Green,T.G.Owano, S.Williams,D.G. Goodwin,R.N. Zare and C.H. Kruger, Science 259 (1993) 1726. [9] T.J. Butenhoff and E.A. Rohlling, J. Chem. Phys. 98 (1993) 5469. [ lo] R.L. Farrow, D.J. Rakestraw and T. Dreier, J. Opt. Sot. Am. B9 (1992) 1770.

[ 111 E.J. Friedman-Hill, L.A. Rahn and R.L. Farrow, J. Chem. Phys. 100 (1994) 4065. [ 121 S. Williamsand R.N. Zare, J. Chem. Phys. 101 (1994) 1072, 1093. [ 131 G. Hall, A.G. Suits and B.J. Whitaker, Chem. Phys. Letters 203 (1993) 277. [ 141 G. Hall and B.J. Whitaker, J. Chem. Sot. Faraday Trans. 90 (1994) 1. [ 151 E.F. McCormack, S.T. Pratt, P.M. Dehmerand J.L. Dehmer, Chem. Phys. Letters 2 11 ( 1993) 147. [ 161 T.J. Butenhoff and E.A. Rohlfing, J. Chem. Phys. 97 (1992) 1595. [ 171 J.C.D. Brand, A.R. Hoy, A.K. Kalkar and A.B. Yamashita, J. Mol. Spectry. 95 (1982) 350. [ 181 T. Ishiwata, J. Yamada and K. Obi, J. Mol. Spectry. 158 (1993) 237. [ 191 P.J. Wilson, T. Ridley, K.P. Lawley and R.J. Donovan, Chem. Phys. 182 ( 1994) 325. [20] K.P. Lawley and R.J. Donovan, J. Chem. Sot. Faraday Trans. 89 (1993) 1885. [21] R.S. Mulliken, J. Chem. Phys. 55 (1971) 288. [22] M.D. Danyluck and G.W. King, Chem. Phys. 22 (1977) 59. [23] R.F. Barrow and K.K. Yee, J. Chem. Sot. Faraday Trans II 69 (1973) 684. [ 241 P. Luc? J. Mol. Spectry. 80 ( 1980) 41.