摘要:

Predissociation lifetimes in the B-state of I79Br E. A. Volkers,a A. E. Wiskerke,a R. Mooyman,a M. J. J. Vrakkingb and S. Stolte*a a Laser Centrum Vrije Universiteit Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands. E-mail: stolte@chem.vu.nl b FOM Institute for Atomic and Molecular Physics (AMOLF), Kruislaan 407, 1098 SJ Amsterdam, The Netherlands Received 29th August 2000, Accepted 14th September, Published 20th September 2000 The predissociation of the B (3 0+) state of I79Br is studied by measuring its one photon absorption spectrum in a cold molecular beam by REMPI detection of the atomic photo fragments. In this work, we present the absolute frequencies as well as the line shapes of the transitions to the vibrational levels Y = 22 to Y = 34 of the B (3 0+) state.These results provide the lacking information needed to fully interpret the outcome of the femtosecond experiments. They also provide important spectroscopic data required to acquire a more accurate theoretical description of the intersecting 3 0+ surfaces. The predissociative lifetimes of the short lived levels ( < 2 ps) have been determined from the observed linewidth. It is found that the linewidths, and thus the predissociative lifetime, oscillate as a function of the vibrational quantum number. All observed spectral transitions obey a Fanoprofile lineshape from which the Fano-parameters have been extracted. Fig. 1 Potential energy curves for IBr. The ground state, the B (3 0+) state and the 0+ (2341) state are drawn as solid red lines.The dotted black curves are the other states in this region but are not of interest for this study. The numbers behind the symmetry give the electron configuration for the (p), (p), *(p), *(p) levels. (Picture adapted from Jung et al.2) threshold corresponds to diabatic photodissociation,3–5 the rapid onset of predissociation above the crossing between the B (3 0+ DQGWKH% +) states is a clear manifestation of the importance of the adiabatic path.6,7 Furthermore, early work on the room temperature absorption spectrum showed just a few isolated regions where sharp rotational structure could be identified.6 Sharp rotational structure is only obtained when the vibrational levels of the diabatic and adiabatic state coincide.1,8 Child carried out some extensive calculations in both the diabatic and the adiabatic approach in which he discovered that the rotational constant of the B (3 0+) state was intermediate Introduction Our understanding of the nature of chemical reactions is based on the Born–Oppenheimer approximation with an assessment of the adiabaticity of the chemical evolution along Born–Oppenheimer potential curves.The Born– Oppenheimer approximation exploits the fact that in a molecule the electrons move much faster than the nuclei. Electronic wavefunctions therefore adjust instantaneously to the positions of the nuclei, and determine the local potential energy that describes the interactions between the nuclei. For each different electronic configuration a diabatic potential energy curve governing the motion of the nuclei is thus obtained.Crucial to chemistry is the behaviour at crossings where the potential energy in two different electronic configurations becomes equal. Depending on the strength of the coupling between the two electronic configurations, the molecule may proceed diabatically or may cross to the other configuration. In the latter case, the coupling can be taken into account by constructing an adiabatic potential energy curve, which changes from one electronic configuration to the other along the crossing. The propensity for diabatic vs. adiabatic passage of the crossing is determined by the strength of the aforementioned coupling and the rate at which the system goes through the crossing.As the coupling strength is increased, the behaviour at the crossing will change from diabatic to adiabatic. A challenging case is presented by molecular systems where the coupling between two electronic configurations is of intermediate strength, and where the molecule cannot be appropriately understood in either a diabatic or an adiabatic picture. Predissociation of the B (3 0+) state of the IBr molecule is a benchmark case where this situation is encountered.1 The potential energy curves of IBr are shown in Fig. 1. While the predominance of excited state Br* (2P1/2) photofragments at excitation energies above the I + Br* DOI: 10.1039/b006982p PhysChemComm, 2000, 10between the rotational constants of the adiabatic and diabatic state.1 To explain this behaviour, Child proposed the idea of the intermediate coupling in IBr.The first quantum mechanical confirmation of predissociation at intermediate coupling strength was given by Child and Lefebvre.9 Later calculations of Guo also support this approach.10 For a transition to a state, which is a mixture of a bound state and a continuum, like the B-state of IBr, Fano and coworkers11–13 derived the profile of the peaks. The shape of the peak is asymmetric and is given by the relation: ( )2 (1) b 1+ G/2 and the e2 where q is the asymmetry factor and is a reduced energy. The reduced energy is defined in terms of the resonance width and the resonance energy Er of the predissociative state: E E - r (2) G = h (3) q + e s s = = e The relation between the resonance width lifetime of the resonance is given by: t Recently, the predissociation of I79Br was studied in a femtosecond pump–probe experiment.14,15 In this experiment, a tuneable femtosecond pump laser pulse excited a wavepacket near the inner turning point of the IBr B (3 0+) state and the subsequent wavepacket evolution was monitored by probing the wavepacket at a variable time delay with a second femtosecond laser pulse, which ionised the molecule.The experiments revealed the time scales for vibrational motion in the IBr B (3 0+) state and the wavepacket de-phasing and re-phasing which one expects in time-resolved experiments.Importantly, the experiments indicated that the molecule can predissociate in a few picoseconds. On this timescale, the predissociation rate is a non-trivial function of the central wavelength of the pump laser pulse. For the longer timescales, occurring for central wavelengths between 551 and 572 nm, vibrational wavepackets were observed which predissociated relatively slowly, so that more than 20 revivals could be observed in the pump–probe signal prior to complete dissociation. At other central wavelengths however, the predissociation proceeded significantly faster and only a few vibrational oscillations could be observed before the molecule had dissociated completely. A fruitful way of analysing time-domain pump–probe experiments is frequently an analysis of the Fourier transform power spectrum.When a coherent superposition of a number of discrete states is excited, its power spectrum will show a number of peaks corresponding to contributions to the total signal originating from coherences between pairs of states. The frequencies at which these peaks occur correspond to level spacings between these pairs of states. In the femtosecond IBr experiment, only a small portion of the vibrational level spacings belonging to levels lying within the bandwidth of the excitation laser could be observed. For a central wavelength around 569 nm only one peak was observed at 48.1 cm–1 while for a central wavelength around 551 nm two peaks were observed at 28.65 and 31.25 cm–1.In addition, a few next-nearest neighbour coherences ( v = 2) were also seen. The conclusion was reached that the predissociation lifetime in IBr oscillates as a function of vibrational excitation, and becomes relatively longer ( > 20 ps) in the vicinity of Y = 22-23 and Y = 28–31.† 0+) The femtosecond pump–probe experiments on IBr have been analysed in two subsequent theoretical treatments. Using the split-operator technique, Hussain and Roberts16 carried out wavepacket propagation calculations on the coupled B (3 0+ DQG % +) states, while Shapiro et al. simulated the pump–probe experiments starting from a time-independent formalism.15 In both treatments the variations in the excited lifetimes were qualitatively reproduced as a function of the vibrational quantum number.Quantitatively, the agreement between the femtosecond experiments and the theoretical treatments was limited by uncertainties concerning the IBr B (3 DQG % +) potential curves (and the coupling strength between them). Our findings suggest that the better potential energy curves are those obtained by Child1 in his pioneering paper on the origin of the few isolated sharp rotational lines in the diffuse room temperature absorption spectrum by Selin6 and Eberhardt and Sullivan.17 Child showed that sharp lines occurred in the absorption spectrum whenever the energies of ro-vibrational eigenstates in the diabatic potential well coincided with ro-vibrational energies (at the same J) in the upper adiabatic potential well.In order to facilitate more precise theoretical work and to resolve uncertainties in the position of the origin of the various vibrational IBr B states, we report in this communication the first observation of the absorption spectrum of IBr in a cold molecular beam. Our experiment in the frequency domain of 17 800 to 18 200 cm–1, aims at a direct assessment of individual vibrational origins and line shapes. As discussed above [eqn. (1)–(3)] this will be done by applying a Fano profile. It is interesting to note that, complementary to the recent femtosecond studies, our experiment is most sensitive in determining short lifetimes ( < 2 ps) because these correspond to large linewidth .The transitions with large linewidth remain insensitive to the limited spectral resolution and inhomogeneous linewidth due to low J-distribution of IBr in our supersonically cooled beam. Experimental The experimental apparatus consists of a differentially pumped molecular beam machine with a Wiley–McLaren time-of-flight mass spectrometer.18 IBr was held in a darkened glass bulb and co-expanded through a pulsed solenoid valve (General Valve model 91-284-900) employing Ar as carrier gas. Special care was taken to keep the gas inlet system free from moisture and to shield the sample from ambient room light. Inert materials like Teflon and glass are used as much as possible in the inlet system. Under these conditions it was found that an IBr sample can be used for about one day, before deterioration occurs.Such deterioration manifests itself in an increase of the presence of I2. IBr was photo-dissociated with radiation from a nanosecond dye laser operated with Rhodamine 575. The absorption spectrum was determined by monitoring the yield of Br (2P3/2) atoms using a (2 + 1) REMPI schemethrough the 2D5/2 state. Ion signals for the 79Br isotope were detected in a time-of-flight mass spectrometer in combination with a box-car averager (note that the femtosecond experiments were also carried out at the I79Br mass). The dissociation laser (Quanta Ray PDL-2) produces 15 mJ in 10 ns with a bandwidth of 0.2 cm–1, and was defocused to a spot of 3 mm to suppress saturation.The REMPI transition is driven by a second dye laser (Quanta Ray PDL-3) yielding 4 mJ pulses near 250 nm after frequency doubling in a BBO ( -BaB2O4) crystal. This laser was fired 40 ns after the first laser and is focused on a diameter of approximately 0.1 mm. The measured spectra were corrected for the varying fluences of both lasers. cm-1 r /cm–1 2.6 (2) 1.8 (1.1) 17827.4 (1.6) 17875.8 (0.6) 7.5 (2.8) 11.6 (4.6) 17921.0 (1.8) 17961.9 (2.3) 9.8 (1.8) 6.5 (1.0) 17997.8 (1.7) 18035.2 (1.0) 3.1 (0.5) 1.7 (0.3) 18071.1 (0.5) 18105.0 (0.3) 1.4 (0.5) 2.5 (1.1) 18136.6 (0.4) 18165.2 (0.6) 3.1 (0.4) 3.7 (1.4) 18191.3 (0.7) 18214.1 (1.6) Results and discussion Y2 An overview of the measured I79Br absorption spectrum is shown in Fig.2. Unlike the room temperature absorption spectrum of Selin,6 the spectrum is dominated by transitions originating from the ground vibrational state. The spectrum shows a succession of vibrational transitions which tend to be very broad, except for the excitation to = 22–23 and excitation to Y = 29–30, where through the D5/2 state. Ion signals for the 79Br isotope were detected in a time-of-flight mass spectrometer in combination with a through the 2D5/2 state. Ion signals for the 79Br isotope were Table 1 Fano parameters and lifetimes for I79Br, resulting from fits using eqn. (1). Due to inhomogeneous linebroading the lifetimes marked with (*) have to be taken as a lower limit of the real ones. The accuracy is indicated in brackets v–v' transition E 0–22 0–23 0–24 0–25 0–26 0–27 0–28 0–29 0–30 0–31 0–32 0–33 0–34 2.4 (5) 18236.6 (3.0) Fig.2 The measured one photon absorption spectrum of IBr showing oscillating linewidths. detected in a time-of-flight mass spectrometer in combination with a box-car averager (note that the box-car averager (note that the The spectrum supports the earlier prediction of Child1 that the lifetimes of low-J vibrational levels in the B (3 0+) state of IBr oscillate as a function of the vibrational quantum number, with minima in the decay rates occurring for the vibrational levels mentioned above. Between the minimum linewidths occurring near Y = 22– 23 and Y = 29–30, the linewidth goes—near Y = 25— through a maximum, which is almost 10 times larger than the aforementioned minima.Hence, the lifetimes vary by an order of magnitude. Through the relation between the Fano-parameter and the lifetime of the vibrational level [eqn. (3)], it is possible to calculate the lifetime. The resulting Fano parameters, obtained by using a simplex fitting procedure and the extracted lifetimes of the vibrational bands, are collected in Table 1. Because of the low rotational temperature in the beam and the small values of the rotational constants involved (B = 0.0323 cm–1; % § 0.028 cm–1), corrections on Er due to rotational contributions have been suppressed. The rotational temperature is estimated from the linewidth of the narrowest lines at a maximum of 6 K.q /ps 10.6 (1.3) –8.5 (1.9) 2.0* 3.0* 0.70 (0.26) 0.46 (0.18) –5.2 (0.7) –11.8 (1.6) 0.54 (0.10) 0.82 (0.13) 10.8 (1.3) 14.0 (1.4) 47.3 (5.5) –36.9 (5.8) 1.7 (0.3) 3.2* 3.8* –4.2 (0.7) –3.3 (0.5) 2.2* 1.7 (0.2) –3.5 (0.5) –8.5 (2.2) –3.8 (1.4) 1.4 (0.5) 2.2*Table 2 Comparison of our results with the results of Child, Shapiro and Hussain and Roberts E E v–Y r /cm–1 transition according to Child1 this work r /cm–1 r /cm–1 according to according to Shapiro et al.15 —17876 17827 17876 0–22 0–23 17921 17964 17921 17962 0–24 0–25 —— 17998 18035 0–26 0–27 —— 18071 18105 0–28 0–29 18136 18164 18137 18165 0–30 0–31 18192 18216 18191 18214 0–32 0–33 18240 18237 0–34 For the comparison of our results for Er with the calculations of Shapiro and Hussain and Roberts, we read the peak positions from Fig.3 of ref. 15 and from Fig. 7 of ref. 16. The comparison with the results by Child1 and Selin6 was simpler. The data in Table 2 of ref. 1 only had to be corrected for rotational energy and the zero-point energy in the ground state. The comparison of the obtained data is given in Table 2. Our results are close to the results by Child and Selin and show stronger deviations with Shapiro and Hussain and Roberts. The comparison with the calculated spectra of Shapiro et al.15 and Hussain and Roberts16 also shows several transitions with a wrong sign in the Fano asymmetry parameter .As is to be expected, due to inhomogeneous line broadening and limited spectral resolution, our largest lifetimes ( between 2.0 and 3.8 ps) are much smaller than the expectations ( > 20 ps) extracted from the femtosecond experiments. = 22 and Y = 23 would be present, whereas a = 23 and Y = 24 would be suppressed due (3 Our results can be used to rationalise the coherences observed in the femtosecond time-resolved experiments. As explained in the introduction, structure in the Fourier transform power spectrum of a time-domain experiment is expected if two levels can beat against each other during a large enough number of oscillation periods (i.e., a small value of ). Therefore on the basis of Fig. 2, we might anticipate that in the femtosecond experiments a beat between Y beat between Y to the large decay rate (high ) of Y = 24.These results are summarised in Table 3, where beat frequencies (level spacings) for pairs of narrow vibrational levels in the IBr B 0+) state are given, with the restriction that only neighbouring states and next-nearest neighbours are included. Co-excitation of vibrational states that are further apart (for example Y = 23 and Y = 28) becomes less efficient due to the 200 cm–1 bandwidth of the femtosecond laser. The results shown in Table 3 demonstrate that on the basis of the energy differences between ‘sharp’ lines in Fig. 2, all coherences observed in the femtosecond experiment can be assigned. Conclusion The absorption spectrum of rotationally cold I79Br has been recorded in the range of 17 800 to 18 200 cm–1.The width E Er /cm–1 according to Hussain and Roberts16 17812 17870 17808 17865 17968 18017 17923 17968 18066 18087 18015 18056 18120 18158 18097 18131 18184 18207 18159 18187 18237 — 18211 18229 — 18242 Table 3 Comparison of measured energy differences between sharp lines in the absorption spectrum and the measured coherences in the femtosecond pump–probe experiments of ref. 15 0+) pump–probe Energy difference Energy difference in I79Br (this work)/cm–1 Pair of B (3 state vibrational levels experiment (ref. 15)/cm–1 48.15 33.9 48.4 33.9 31.25 28.65 31.6 28.6 60.55 55.15 60.2 54.7 Y = 22 – Y = 23 Y = 28 – Y = 29 Y = 29 – Y = 30 Y = 30 – Y = 31 Y = 29 – Y = 31 Y = 30 – Y = 32 of the observed peaks oscillates strongly with the excited state vibrational quantum number. This is in accord with femtosecond experiments15 where only beat frequencies were observed when the femtosecond laser was tuned to a number of specific central frequencies.On the basis of the absorption spectrum we were able to identify the vibrational transitions that are responsible for the beat frequencies observed in the femtosecond experiment. Moreover, the fast decay of the femtosecond pump–probe experiments at the central frequencies where no beat frequencies were observed, corresponds to the bands in the absorption spectrum where only broad lines are present.Fano profile analysis of the various observed peaks provides improved experimental values of the B(Y = 22– 34) X(v = 0) ‘rotationless’ transitions frequencies as present in a cold supersonic beam. These results are rather close to the early results by Child1 based on room temperature experiments.6 The lifetime has been extracted from the lineshape for a number of vibrational levels (Table 1). In the future, the data might be used to establish better potential energy curves for the femtosecond computer simulations.Acknowledgements Dr G. Bazalgette is thanked for enlightening discussions. This research has been financially supported by the Council for Chemical Sciences (CW-NWO) and by Stichting voor Fundamenteel Onderzoek der Materie (FOM), which are both part of Nederlandse organisatie voor Wetenschappelijk Onderzoek (NWO).The work of M. J. J. Vrakking is part of the FOM research program. References 1 M. S. Child, Mol. Phys., 1976, 32, 1495. 2 K. W. Jung, J. A. Griffiths and M. A. El-Sayed, J. Chem. Phys., 1995, 103, 6999. 3 G. E. Busch, R. T. Mahoney, R. I. Morse and K. R. Wilson, J. Chem. Phys., 1969, 51, 837. 4 M. S. de Vries, N. J. A. van Veen and A. E. de Vries, Chem. Phys. Lett., 1978, 56, 15. 5 M. S. de Vries, N. J. A. van Veen, M. Hutchinson and A. E. de Vries, Chem. Phys., 1980, 51, 159. 6 L. A. Selin, Ark. Fys., 1962, 21, 529. 7 R. B. Bernstein, in Chemical dynamics via molecular beam and laser techniques, Oxford University Press, London, 1982, p. 168. 8 A. D. Bandrauk and M. S. Child, Mol. Phys., 1970, 19, 95. 9 M. S. Child and R. Lefebvre, Chem. Phys. Lett., 1978, 55, 213. 10 H. Guo, J. Chem. Phys., 1993, 99, 1685. 11 U. Fano, Phys. Rev., 1961, 124, 1866. 12 U. Fano and J. W. Cooper, Phys. Rev. A, 1965, 137, 1364. 13 E. F. van Dishoeck, M. C. van Hemert, A. C. Allison and A. Dalgarno, J. Chem. Phys., 1984, 81, 5709. 14 M. J. J. Vrakking, D. M. Villeneuve and A. Stolow, J. Chem. Phys., 1996, 105, 5647. 15 M. Shapiro, M. J. J. Vrakking and A. Stolow, J. Chem. Phys., 1999, 110, 2465. 16 A. N. Hussain and G. Roberts, J. Chem. Phys., 1999, 110, 2474. 17 W. H. Eberhardt and W. Sullivan, J. Mol. Spectrosc., 1978, 70, 270. 18 W. C. Wiley and I. H. McLaren, Rev. Sci. Instrum., 1955, 26, 1150. † Note that we employ the numbering of the vibrational states as introduced by Child,1 which differs from the numbering by Selin6 and Vrakking et al.14 by 3 quanta. PhysChemComm © The Royal Society of Chemistry 2000

ISSN:1460-2733    

DOI:10.1039/b006982p

出版商:RSC    

年代:2000

数据来源: RSC