Gas-phase Photoelectron Spectroscopy

BY R. G. EGDELL

1 Introduction

The present Report aims to provide a comprehensive review of work in gas-phase photoelectron spectroscopy (p.e.s.) and related areas for the period 1977-early 1980, although overlap with the previous article in this series' is avoided as far as possible. The field has now reached a level of considerable maturity, with relatively few new innovations in experimental or theoretical techniques. But the reviewer is unaware of any decline in the general level of activity in the field.

The usefulness of photoelectron spectroscopy as a means of illustrating a variety of concepts involved in a discussion of electronic structure at an elementary level is reflected in the introduction of p.e.s. in general textbooks on physical and inorganic chemistry. More specialized undergraduate texts on and the theory of chemical bonding and symmetry also give prominence to the technique. A useful monograph by Ballard attempts a development of molecular orbital theory, drawing on the results of p.e. experiments to illustrate basic ideas. A similar approach is adopted in an Open University Course Unit which presents a more limited discussion of molecular valence-level p.e.s. in relation to the theory of electronic structure. The Unit also deals with core-level and solid-state p.e.s. Nonetheless there still remains a clear need for a generally available undergraduate text dealing in a unified way with valence- and core-level p.e.s. of solids and gases. An enigmatically titled volume published in India comes close to satisfying this need but is difficult to obtain in Western Europe or the U.S.A. As regards more advanced books we draw attention to a number of authoritative articles on various aspects of p.e.s. in Volumes 2 and 3 of the series edited by Baker and Brundle. Special mention should also be made of a recent book by Berkowitz. This provides a thorough account of the dynamic aspects of p.e.s., including the theory of photoabsorption, photoelectron angular distributions, and fragmentation of molecular ions. This comprehensive text includes some previously unpublished experimental results. Finally we note accounts of p.e.s. given in Volume 9 of Comprehensive Analytical Chemistry and in a book on instrumental analysis.

The recognition of the importance of p.e.s. in contemporary physics and chemistry is reflected in the appearance of articles concerned with historical aspects of development of the subject. A volume edited by Carlson contains reprints of key papers in X-ray photoelectron spectroscopy (X-p.e s.). Looking to the future, Bock has addressed himself to the question of what will become of p.e.s. He concludes that the technique is of value in analysis and optimization of gas-p hase reactions, although the reviewer can foresee other areas of application.

Turning now to review articles of various sorts, we again commend the terse but comprehensive and timely articles by Baker and co-worker. These provide coverage of both X-p.e.s. and u.v.-p.e.s. for the periods 1976-1977 and 1978-1979. Reviews of a general nature but with much more limited scope have been given by Fellner-Feldegg, Turner! Larkins: Leckey: and Siegbalahn. Instrumentation in molecular p.e.s. has been authoritatively reviewed by Eland, whilst Schwartz has provided a useful survey of techniques used to calculate ionization energies. We also draw attention here to a burgeoning body of literature of an introductory or review nature published in foreign journals without international circulation. The extensive series of articles in Volume 16 of the Japanese journal Kagaku Sosetu deserve special mention, but we have also noted surveys of p.e.s. from Japan, China, Hungary, Poland, Russia, and Uganda. Whilst of limited general interest these articles are clearly of great value to workers in the respective countries.

The largest body of review articles is concerned with application of p.e.s. in restricted areas of chemistry or physics. The impact of electron spectroscopy on the study of the electronic structure of atoms (particularly in its dynamic aspects) is reflected by the appearance of excellent surveys of the area by Wuilleumier and co-workers, Codling, and Samson. The basic theory of atomic photoionization and its application to specific problems has been dealt with by Manson. As regards more chemically oriented articles, Jolly has presented a timely review which deals critically with the use of X-p.e.s. in the study of solid-state and molecular inorganic systems. Elsewhere, Jonathan and co-workers have surveyed the application of p.e.s. to the study of transient species. A number of articles deal with p.e.s. of metal complexes. Furlani and Cauletti presented the first comprehensive review of He-I p.e.s. of transition-metal complexes, dealing for the most part with papers published before 1977. A review by Cowley covers some more recent material but adopts a similar comprehensive approach. By contrast, Hillier has surveyed the area in a more personalized manner, dealing with both core- and valence-level spectra and emphasizing the relationship between molecular orbital calculations and photoelectron spectra. In the published proceedings of a NATO Advanced Study Institute devoted to the organometallic chemistry of the f-block elements, Fragala has discussed p.e. spectra of actinide complexes, including in his paper some previously unpublished results. Other reviews concerned with electron spectra of specific groups of inorganic compounds include those on organotin compounds, organophosphorus compounds, and organosulphur compounds. The u.v.-p.e.s. of organic compounds has been surveyed comprehensively by Ra0, whilst Gleiter deals with the more restricted topic of p.e.s. and bonding in small ring hydrocarbons. A selective and critical article by Sandorfy explores the relationship vacuum U.V. absorption and p.e. spectra of organic systems. The 'impact' of photoelectron spectroscopy on biology has been discussed by McGlynn. Although a number of molecules of biological importance have now been studied by p.e.s., it is unclear whether these measurements have had a significant influence on biological science.

Finally we note that several conferences have been concerned with p.e.s. The meeting held in Uppsala in 1977 was mentioned in the previous report, although the proceedings were not discussed in detail. We remedy this deficiency in the present report. A second major conference on electron spectroscopy was held in Melbourne in 1978; its proceedings are published as a volume of J. Electron Spectrosc. Relat. Phenom. The 5th International Conference on Vacuum Ultraviolet and Radiation Physics was concerned in part with p.e.s. Two meetings in also touched on the subject.

Finally we note that NATO Advanced Study Institutes series on photoinization, electron and ion spectroscopy, and excited states and quantum chemistry dwelt on certain aspects of photoelectron spectroscopy.

Despite the high level of activity in the field the basic question as to whether p.e. spectra should be presented with the ionization energy scale increasing from right to left or vice versa still arouses discussion. The IUPAC recommendation is that the ionization energy scale should increase from right to left. We apologize for failing to adhere to this convention in reproducing certain original published data.

2 Experimental Aspects

There have been few major developments in experimental technique in the period covered by the report, although there have been important new applications of technically demanding but well established techniques in the study of high-temperature and transient species, in the measurement of negative-ion and threshold p.e. spectra, and in the determination of photoelectron branching ratios and asymmetry parameters using synchrotron exciting radiation. One of the few new innovations concerns measurement of p.e. spectra using supersonic beam targets.

Spectrometer Systems. - A conventional instrument incorporating a 127° cylindrical sector analyser has been described by Hegde and basu. Another simple instrument with a slotted grid retarding potential analyser and an electrometer electron detector is recommended by Price for teaching purposes. In general, though, the trend is towards increased sophistication in instrumentation. Ohta and Sergeevs and respective co-workers have discussed automation and computer control of photoelectron spectrometers. Flexible data accumulation and handling facilities become particularly important when large amounts of data are to be collected, as for example in angle-dispersive studies. Both Leng and Nyberg and Kuppermann and associates have described in detail automated instruments in which both the data collection angle and analyser potentials are microprocessor controlled. The former incorporates a simple retarding-field analyser, whereas the latter exploits a more elaborate 180° spherical sector analyser. Systems of this sort facilitate collection of p.e. data at a wide range of acceptance angles which should lead to more accurate β values than those derived from measurements of p.e. intensity patterns at just two angles.

Computer control of data acquisition is also useful in the study of temperature-dependent effects and of gas-phase chemical equilibria. For example Schwe igSg has described a microprocessor-con t rolled Perkin-Elmer PS-16/ 18 spectrometer used to study gas-phase conformational equilibria, whilst Peel and associatesg0 have made extensive use of a computer-controlled spectrometer to accumulate spectra of gas-phase transient species. Where spectra of two or more molecules overlap, an interactive stripping program may be used to obtain spectra of individual components, as illustrated in Figure 1

A novel and elaborate instrument described by Weeks et al. incorporates a vidicon camera detector system mounted behind a retarding-field analyser. It was used to obtain angle-dispersive p.e. spectra of solids, but presumably a machine of this sort could also be applied to gas-phase work. Retarding-field electron energy analysis was also used in a dedicated gas-phase spectrometer described by Lindemans. Photoionization in this instrument is excited by a monochromatized continuum discharge; β-independent branching ratios are easily measured .

Two novel spectrometer designs have been described by Turner. The first is a retarding-field instrument with a channel-plate electron detector. Both neutral target species and charged ions trapped in a strong magnetic field are amenable to study. In principle the integral p.e. spectrum produced by this instrument contains information about photoelectron angular distributions, although it proved impossible in practice to extract meaningful values for the asymmetry parameter for the 3p subshell of argon. A β-independent integral p.e. spectrum can be obtained from the spectrometer provided that electrons are able to exchange momentum in elastic collisions with target gas atoms. Turner's second instrument is based on electron kineticenergy analysis in an axially symmetric inhomogeneous magnetic field. It offers the advantages of high spatial resolution and sensitivity.

Turning now to commercially marketed instrumentation we note the availability of a spectrometer incorporating many of the successful design features of the Perkin-Elmer PS/18 machine and the Helectros Developments He-I1 lamp. It is produced by Photoelectron Spectrometer Laboratories Limited. An instrument of novel design is offered by Spinlab. It incorporates a pill-box analyser and yields spectra of impressively high resolution (vide infra). Vacuum Generators Limited still market the UVG-3 dedicated gas-phase instrument, but we note also application of the VG-ESCA3 spectrometer to the study of transient species and of the ADES-400 angle-dispersive spectrometer to gas-phase asymmetry parameter determinations. Other commercially available spectrometers are marketed by MacPhersonl and Leybold-Heraues.

Electronenergy Analysers. - The general field of electron-energy analysis has been reviewed by Roy and Carette. Keski-Rahkonen has discussed the influence of relativistic effects on the calibration of energy analysers. Even the simplest of analysers continue to find new applications. Thus Smith has described a simple and lightweight plane-mirror analyser suitable for use in angle-resolved photoemission experiments. It achieves a resolution ΔE/E of 6%. (But we note that compact spherical sector analysers can also be incorporated in angle-dispersive instrument.) Impressive resolving power has been achieved with a retarding-field instrument which serves as an energy-band pass filter with a large solid angle of acceptance. Cylindrical-mirror analysers also remain popular. An instrument described by Baudais and Taylor was used to study autoionization in ethylene excited with monochromatized synchrotron radiation. The resolution ΔE/E achieved was 7%. This compares with ΔE/E< 0.2% in a double-pass cylindrical-mirror analyser described by Hungarian research workers. Elsewhere Hotop and Huebler used a novel differential retarding-field energy analyser to measure photoelectron and Penning ionization spectra of N2, O2, and HCl. The instrument was calibrated for energy-dependent variation in analyser transmittance and operated at sufficiently high resolving power to reveal shifts between photoelectron and Penning ionization energy scales. However, the peak shapes were somewhat asymmetric (Figure 2).

Time-of-flight electron-energy analysis appears to have reached its full potential in an instrument described by Shirley and co-workers (Figure 3).

Designed for use in conjunction with the Stanford synchrotron SPEAR (where of course the photon source is pulsed), ΔE/E< 5% was achieved at high angular resolution (< 30%). The resolving power is somewhat higher than that achieved by Comer and co-workers in an instrument where time-off-light energy analysis was used to detect electrons ejected in coincidence with electrons inelastically scattered by autoionizing resonances in nitrogen.

Turning now to lens systems used in electron spectrometers, Wannberg and Sköllermo have discussed computer optimization of retarding lenses used in X-p.e.s., whilst Nozoye et al. describe an empirical optimization of a lens incorporated into a spectrometer used for gas-phase Auger measurements. The effects of retarding lenses on the transmittance of energy analysers were discussed by Vulli et al., who experimentally determined the analyser brightness of their photoelectron spectrometer.

Deconvolution. - The application of deconvolution methods in p.e.s. has been reviewed recently by Carley and Joyner. They provide a useful coverage of the mathematical background to various deconvolution techniques and also give a comprehensive and critical discussion of various applications of deconvolution procedures in both core- and valence-level p.e.s. In more limited vein, Lloyd has discussed mathematical aspects of Fourier-transform deconvolution procedures, with particular reference to their application in core-level spectroscopy, whilst Gurker has discussed the alternative technique of iterative deconvolution. Allen et al. have discussed the application of iterative deconvolution to gas-phase p.e. spectra. They present impressive spectra of Ar, O2, N2, and H20 obtained with the Spinlab 560 spectrometer (vide supra). These spectra are then deconvoluted to reveal fine structure, e.g. rotational shouldering in the 2Σ+g ([bar.x]) band of N2+. However, the deconvoluted 2Πu (A) band of N2+ fails to reveal the spin-orbit structure found earlier by Potts in a high-resolution study which did not rely on subsequent data processing (Figure 4).

Radiation Sources. - U.v. Discharge Sources. Bruck and co-workers have made a critical comparison of the relative merits of microwave discharges, glow d.c. discharges, duo plasmatrons, and low-pressure arcs as sources of U.V. radiation for electron spectroscopy. They conclude that the duo plasmatron is the most powerful source of vacuum U.V. radiation and in a later publication discuss conditions required to optimize the He-I1 photon flux from a duoplasmatron source.