Threshold-photoelectron-spectroscopy-coincidence study of the double photoionization of HBr

Transcription

1 JOURNAL OF CHEMICAL PHYSICS VOLUME 120, NUMBER APRIL 2004 Threshold-photoelectron-spectroscopy-coincidence study of the double photoionization of HBr Michele Alagia ISMN-CNR Sez.Roma 1, P.le A. Moro 5, 00185, Rome, Italy and Laboratorio TASC-INFM, Area Science Park, Basovizza, Trieste, Italy Brunetto G. Brunetti Dipartimento di Chimica, Università di Perugia, Perugia, Italy Pietro Candori and Stefano Falcinelli Dipartimento d Ingegneria Civile ed Ambientale, Università di Perugia, Perugia, Italy and Unità INFM di Perugia, Università di Perugia, Perugia, Italy Marc Moix Teixidor Dipartimento di Chimica, Università di Perugia, Perugia, Italy Fernando Pirani Dipartimento di Chimica, Università di Perugia, Perugia, Italy and Unità INFM di Perugia, Università di Perugia, Perugia, Italy Robert Richter Sincrotrone Trieste, Area Science Park, Basovizza, Trieste, Italy Stefano Stranges Laboratorio TASC-INFM, Area Science Park, Basovizza, Trieste, Italy and Dipartimento di Chimica, Università di Roma La Sapienza, Roma, Italy Franco Vecchiocattivi Dipartimento d Ingegneria Civile ed Ambientale, Università di Perugia, Perugia, Italy Received 18 December 2003; accepted 21 January 2004 A threshold-photoelectron-coincidence spectrum of HBr has been recorded in the ev photon energy range, with a resolution of 0.01 ev, using a synchrotron radiation source. The X 3 and a 1 2 states of the HBr 2 dication are clearly observed in the spectrum, while there is no clear evidence for the formation of the b 1 electronic state. For the first two states, the vibrational states v 0 3 have been resolved, while for the ground X 3 state also spin orbit splitting has been detected. The results appear in good agreement with previous experimental observations. A comparison with theoretical predictions indicates the role of noncovalent contributions to the interaction between the two atomic partners for the formation of metastable states American Institute of Physics. DOI: / I. INTRODUCTION Doubly charged molecular dications are usually unstable species, because of their high formation energy. In liquid solutions the two charged parts of the molecular dication can undergo pronounced stabilization effects by the interactions with the solvent molecules. However, when such interactions play a minor role or are absent e.g., in the gas phase, the two charged parts tend to repel each other, making the molecular dication metastable, in some particular cases, and more often unstable, determining the fragmentation in two positive ions. 1,2 Theoretical and experimental efforts aimed at the characterization of the energy, structure, and dynamics of these ionic species are largely increasing 2 because of their relevant role in plasmas, ionized gases, excimer lasers, and interstellar clouds. Among the diatomic dications, the doubly charged hydrogen halides HX 2 (X F,Cl,Br,I) gave rise to a particular interest in their role in several processes and in their peculiar features. In particular, it is well known that HF 2 does not show any bound low-energy states unstable species, while bound levels exist for HCl 2, HBr 2, and HI 2 metastable ions with a stability degree increasing going to the heavier ones. 3 Our attention has been addressed towards some topical questions concerning the HX 2 species: the realization of new spectroscopic investigations, carried out under highenergy resolution conditions, the full characterization of the involved interaction components, and the accounting for the observed behaviors. Specifically, a recent paper from our laboratory reports the identification of the nature of the leading interaction components in HX 2, their modeling, and their combination to describe the potential energy in the lowlying states of HCl 2. 3 In this and in the following paper we focus on the HBr 2 species, providing new experimental results and relevant details on the intermolecular interaction energy in both the metastable and unstable states. Such an interaction, espressed in a simple, natural, and analytical form, as obtained from a direct application of our theoretical /2004/120(15)/6980/5/$ American Institute of Physics

2 J. Chem. Phys., Vol. 120, No. 15, 15 April 2004 The double photoionization of HBr 6981 approach, 3 permits the calculation of several properties of HBr 2. A comparison between experimental results and functions appears to be crucial to quantitatively analyze the observed features. In a previous experiment carried out at the GASPHASE beamline of the Synchrotron Laboratory ELETTRA in Trieste we have studied the double photoionization of HBr using a time-of-flight TOF mass spectrometer coupled with monochromatic UV synchrotron radiation. 4 This investigation provided the following threshold energies for the double-photoionization processes: HBr h HBr 2 2e, ev, HBr h H Br 2 2e, ev, HBr h H Br 2e, ev. More detailed information about the structure of these dications can be obtained by detecting, in coincidence, the two electrons emitted with nearly zero kinetic energy. This technique, known as TPEsCO threshold photon-electron s coincidence, has been already successfully applied to the study of the double photoionization of several diatomic molecules, including HCl, 5,6 and has provided the vibrational energy spacing for the three low-lying electronic states of the HCl 2 dication. In the present paper, we report the TPEsCO spectrum of HBr measured with a relatively high photon energy resolution 10 mev. This allowed us to obtain the vibrational energy spacing of the lowest-lying 3 and 1 states, and locate the 1 state. For the ground 3 state it has been also possible to resolve the two spin orbit components. These high-resolution data are then compared with previously available theoretical results about this system 7 9 and with experimental results obtained from Auger spectroscopy. 10,11 Very recently, Eland 12 reported an interesting photoelectron spectrum of HBr 2 dication using a time-of-flight-photoelectron-photo-electron-coincidence TOF-PEPECO spectroscopy, where the kinetic energy of the two correlated photoelectrons in double-ionization processes, induced by eV photons, was measured. Also these data are compared with present TPEsCO results. The details of the experimental setup are discussed in the next section, while in the following one an analysis of the experimental data is presented. In the final section the energy levels are discussed in terms of the interaction potential energy curves for the HBr 2 dication, as obtained from theoretical calculations, 13 and the relevant conclusions are drawn. II. EXPERIMENT The measurements were performed using the monochromatic radiation and the ARPES end station of the GASPHASE beamline at ELETTRA. 14 The experimental chamber was equipped with two threshold electron analyzers mounted on the opposite sides of a turntable that can be rotated in the plane perpendicular to the photon beam. A hypodermic needle, mounted on a XYZ translator and perpendicular to both the electron detection axis and the light beam, was used as effusive gas source. The pressure in the chamber was set to mbar during the measurements. The HBr sample was admitted in the chamber together with an argon leak for an accurate energy calibration provided by single-photoionization and autoionization lines in the argon threshold photoelectron spectrum. 15 All the TPEsCO spectra were normalized with respect to the incident photon beam intensity which was continuously monitored during the data acquisition by a silicon photodiode IRD AXUV-100 mounted at the end of the experimental apparatus about 30 cm from the ionization region. The collection of zero-energy electrons is realized using the penetrating field technique developed by Cvejanović and Read. 16 Briefly, an electrostatic field is generated in the interaction region by two extracting electrodes mounted symmetrically on the opposite sides of the ionization region and screened by a grounded diaphragm with a 10-mm aperture. The two threshold electrons are thus extracted and detected separately and simultaneously by the spectrometers that share equally the whole acceptance solid angle. The dispersive electrostatic element was a hemispherical capacitor with the pass energy set to 10 ev. Coincidences between electrons detected by the two analyzers were measured by standard electronics. True and random coincidences and also the photoelectron yield measured by each analyzer were accumulated by a Macintosh computer, which also controls the photon energy scan. The total instrumental resolution achieved in the threshold-electroncoincidence spectra TPEsCO was about 13 mev, full width at half maximum, as measured on the double-photoionization threshold peak of the Ar 2 ( 3 P 0 ) state at ev. The photon resolution used during the measurements was better than 10 mev. The two TPE spectra were recorded obtaining the same peak shape and similar relative intensities from both analyzers. One of the electrons was delayed and used as a STOP pulse, while the other one as a START signal. The total coincidence rate was measured within an acquisition time window referred to the START. The same time window delayed by 864 ns, the storage ring revolution time, was used for measuring the false coincidence rate, thus removing possible effects due to the time structure of the photon source. Total and false coincidence signals were acquired in separate channels using logic port processing. Coincidence spectra were obtained by the removal of false coincidences from the total coincidence signal and by normalization with respect to photon flux changes monitored simultaneously. III. RESULTS The TPEsCO spectra of HBr 2 reported in Figs. 1 3 have been obtained as the average of more than 70 partial accumulations, normalized to the photon flux and gas density. No smoothing procedures were used in the data analysis. Figure 1 shows the TPEsCO spectrum for the total photon energy range investigated ev, while Figs. 2 and 3 show an expanded view of the photon energy ranges where the two electronic states of the molecular dication, X 3 and a 1 2, are clearly observed in our experiment,

3 6982 J. Chem. Phys., Vol. 120, No. 15, 15 April 2004 Alagia et al. FIG. 1. Threshold-photoelectron-coincidence spectrum of HBr in the photon energy range between 32.2 and 35.8 ev. FIG. 3. Expanded view of the band system in the threshold-photoelectroncoincidence spectrum of HBr for the formation of the HBr 2 dication in the first excited a 1 2 state. The lines are the best-fit curves see text. while there is no clear evidence for the formation of the b 1 electronic state. These spectra indicate that, in the double photoionization of HBr molecules, excited HBr 2 (a 1 2 ) ions are produced with a higher intensity than HBr 2 (X 3 ) in the ground state. The ground state of the dication is obtained by the removal of one electron from each of the two nonbonding orbitals of the neutral HBr ground-state molecule. The first a 1 2 excited state is produced by the removal of two electrons, both from the same nonbonding orbital, while the less stable b 1 involves a hole in the bonding orbital. The analysis of the spectra clearly shows that at least four vibrational levels are supported by both electronic states. However, from the experimental data we cannot exclude formation of the v 4 vibrational level for both observed electronic states of HBr 2 molecular ions. The signalto-noise ratio is rather small, so the assignment cannot be done with certainty. The same considerations can be applied also to the weak structure recorded in the energy range from to ev. This signal could be experimental evidence of the formation of the b 1 electronic state of the dication with an adiabatic ionization potential (v 0) of ev. This value is in reasonable agreement with the value of ev previously reported by Auger spectroscopy 17 and with the value of ev obtained by Eland. 12 Finally, a detailed analysis of the spectrum of Fig. 2, in the energy range from to ev, highlights the separation of the two spin orbit components of the X 3 1,0 fundamental electronic state of HBr 2. The analysis of the spectra has been performed by a best-fit procedure. The solid lines in Figs. 2 and 3 are the best-fit curves obtained by using an asymmetric function, with a tail on the high-energy side and the analytical expression y y 0 A exp exp x x m w x x m w 1, where y 0 is the offset of the function, and x m and A are, respectively, the abscissa and height of the peak s maximum, while w is the width at half height. In the fitting procedure all these parameters have been left to vary free up to the convergence to the minimum of the standard deviation, with the only restriction that the two spin orbit states of the X 3 must be in the 2:1 intensity ratio that is, the statistical one. The energy levels obtained from the analysis of the spectrum are listed in Table I, where they are also compared with other experimental results. 9,18,12 In Table II the relative intensities of the peaks are listed and compared with those reported by Eland 12 and with theoretical Franck Condon factors. 13 IV. DISCUSSION FIG. 2. Expanded view of the band system in the threshold-photoelectroncoincidence spectrum of HBr for the formation of the HBr 2 dication in the ground X 3 1,0 state. The lines are the best-fit curves see text. The results obtained from the analysis of the present TPEsCO spectrum provide interesting information on the HBr 2 dication, especially when they are compared with results from other experimental techniques and from theoretical calculations. In this section we discuss such comparisons in order to assess information on the energy, structure, and dynamics of this diatomic dictation.

4 J. Chem. Phys., Vol. 120, No. 15, 15 April 2004 The double photoionization of HBr 6983 TABLE I. Energy levels in ev of the lowest lying HBr 2 dication states as obtained from the present TPEsCO spectrum and compared with previous Auger spectroscopy results. State v Present experiment Wannberg et al. a Matila et al. b Eland c X 3 1, d d d d a b ? a Reference 18. b Reference 9. c Reference 12. d The spin orbit splitting is not explicitly given. The vibrational energy spacings obtained in the present experiment are in good agreement with those previously reported in the Auger spectra of HBr Refs. 9 and 18 and with the TOF-PEPECO results by Eland. 12 The spin orbit splitting for the X 3 vibrational states, well evident in our spectrum, is also in good agreement with the experimental and theoretical values reported by Matila et al. 9 In the best-fit procedure adopted in the data analysis, the intensity ratio between the two spin orbit states has been held fixed at its statistical value 2:1. We have also tried to leave this ratio as variable during the fitting procedure, but the obtained final results were very similar. This is an indication that double photoionization leads to the two spin orbit states just in a statistical ratio, at least within the experimental uncertainty. Figure 4 shows a comparison between the present experimental TPEsCO spectrum and a simulation obtained on the basis of the calculation described in the following paper. For the X 3 state the spin orbit splitting has been artificially introduced by the use of the value of 0.05 ev obtained from Auger spectroscopy 9 and a statistical 2:1 ratio between the two substates. It is interesting to observe a good agreement for the ground-electronic-state levels and a systematic energy shift for those of the first excited one. A comparison between the relative peak intensities in our TPEsCO spectrum, the one in the TOF-PEPECO experiment by Eland, 12 and the Franck Condon factors, as described in the following paper, 13 also provides interesting indications see Table II. Both TPEsCO and TOF-PEPECO results are not largely different from the Franck Condon expectation, although not in perfect agreement among them. Actually some intensity ratios appear slightly different, outside the experimental uncertainty e.g., the ratio between the first two vibrational levels of the X 3 and a 1 states. It is important to point out the different characteristics of the two experiments: in our spectrum we are detecting pairs of near-zero-kinetic-energy electrons, while in the experiment by Eland the peaks correspond to couples of correlated electrons with a fixed total nonzero kinetic energy. Obviously, Auger spectra are also dealing with a complete different observation. This implies that each experiment is sensitive to different features related to a different double-ionization dynamics. Consequently, the three types of experiments can provide complementary information which should deserve a theoretical analysis, not yet available. As an example, the lack of a sharp peak for the second excited state b 1 in the TPEsCO spectrum is an indication that the mechanism for TABLE II. Relative intensities of vibrational levels of the first HBr 2 dication states as obtained from the present TPEsCO spectrum and compared with previous experimental results and with theoretical Franck Condon relative intensities. State v Present experiment Eland a Calculated Franck Condon relative intensities Alagia et al. b Banichevich et al. c X 3 1, a a Reference 12. b Reference 13. c Reference 7.

5 6984 J. Chem. Phys., Vol. 120, No. 15, 15 April 2004 Alagia et al. we also know that usually covalent bonds decrease in energy with the size of the involved atoms. Therefore, in these diatomic bonds other noncovalent contributions to the bond become more important, as has been already discussed in our previous paper on HCl 2. 3 It becomes therefore interesting to investigate such contributions also for the present diatomic dictation in order to assess the relative importance of the different components to the interatomic bonds. ACKNOWLEDGMENTS The present work was part of a European project within the MCInet Research Training Network Contract No. HPRN-CT Financial contributions from ASI Agenzia Spaziale Italiana and from the MIUR Ministero dell Istruzione, dell Università e della Ricerca are gratefully acknowledged. The authors acknowledge interesting and useful correspondence with I. R. McNab. S.S. gratefully acknowledges financial support by MIUR Cofin98 for development of the TPESCO analyzer. FIG. 4. Experimental TPEsCO for the double photoionization of HBr compared with the simulation of the spectrum by using theoretically calculated energy levels. the production of such a state at the double-ionization threshold is not efficient. This is likely due to a rather low doubleionization cross section at the threshold and to a nonefficient autoionization involving the underlying manifold of HBr continua decaying into HBr 2 states. Further comparisons between features predicted by our theoretical analysis of the interaction, such as thresholds, energy spacing between bound states, present observations, and data available in the literature, are described in the following paper. 13 Moreover, it is interesting to observe that the present results on HBr 2, together with previous experimental data, 5 12,19,20 experimentally confirm the trend of the bond energy of the ground-state dication of hydrogen halides: the potential well of the metastable state increases in depth on going from HF to HI. It is well assessed that the HF 2 dication does not show any bound low-energy state, while metastable states exist for HCl 2, HBr 2, and HI 2, with a stability increasing for the heavier ions. On the other hand, 1 L. Pauling, J. Chem. Phys. 1, S. G. Cox, A. D. J. Critchley, P. S. Kreynin, I. R. McNab, R. C. Shiell, and F. E. Smith, Phys. Chem. Chem. Phys. 5, M. Moix-Teixidor, F. Pirani, P. Candori, S. Falcinelli, and F. Vecchiocattivi, Chem. Phys. Lett. 379, M. Alagia, M. Boustimi, B. G. Brunetti, P. Candori, S. Falcinelli, R. Richter, S. Stranges, and F. Vecchiocattivi, J. Chem. Phys. 117, ; a recalibration of the monochromator photon energy scale of that experiment has indicated that the energy values given in that paper were systematically underestimated by a value of ev. 5 A. G. McConkey, G. Dawber, L. Avaldi, M. A. MacDonald, G. C. King, and R. I. Hall, J. Phys. B 27, A. D. J. Critchley, G. C. King, P. Kreynin, M. C. A. Lopes, I. R. McNab, and A. J. Yencha, Chem. Phys. Lett. 349, A. Banichevich, S. D. Peyerimhoff, B. A. Hess, and M. C. van Hemert, Chem. Phys. 154, Z. Huang and Z. H. Zhu, J. Mol. Struct. 525, T. Matila, K. Ellingsen, T. Saue, H. Aksela, and O. Gropen, Phys. Rev. A 61, Y. F. Hu, G. M. Bancroft, J. Karvonen, E. Nommiste, A. Kivimaki, H. Aksela, S. Aksela, and Z. F. Liu, Phys. Rev. A 56, R R. Püttner, Y. F. Hu, E. Nõmmiste, G. M. Bancroft, and S. Aksela, Phys. Rev. A 65, J. H. D. Eland, Chem. Phys. 294, M. Alagia, B. G. Brunetti, P. Candori, S. Falcinelli, M. Moix Teixidor, F. Pirani, R. Richter, S. Stranges, and F. Vecchiocattivi, J. Chem. Phys. 120, , following paper. 14 R. R. Blyth, R. Delaunay, M. Zitnik, et al., J. Electron Spectrosc. Relat. Phenom , S. Cvejanović, G. W. Bagley, and T. J. Reddish, J. Phys. B 27, S. Cvejanović and F. H. Read, J. Phys. B 7, R. Püttner, Y. F. Hu, G. M. Bancroft, H. Aksela, E. Nõmmiste, J. Karvonen, A. Kivimäki, and S. Aksela, Phys. Rev. A 59, B. Wannberg, S. Svensson, M. P. Keane, L. Karlsson, and P. Baltzer, Chem. Phys. 133, J. N. Cutler, G. M. Bancroft, and K. H. Tan, J. Phys. B 24, L. Karlsson, S. Svensson, P. Baltzer, M. Carlsson-Gothe, M. P. Keane, A. Naves de Brito, N. Correia, and B. Wannberg, J. Phys. B 22, ; L. Karlsson, S. Svensson, P. Baltzer, A. Naves de Brito, N. Correia, and B. Wannberg, ibid. 24, L

6

7 The Journal of Chemical Physics is copyrighted by the American Institute of Physics (AIP). Redistribution of journal material is subject to the AIP online journal license and/or AIP copyright. For more information, see