Summary
A critical evaluation and summary of experimental vibrational and electronic energy level data for neutral and ionic transient molecules and high temperature species possessing from three to sixteen atoms is presented. Although the emphasis is on species with lifetimes too short for study using conventional sampling techniques, there has been selective extension of the compilation to include data for isolated molecules of inorganic species such as the heavy-metal oxides, which are important in a wide variety of industrial chemical systems. Observations in the gas phase, in molecular beams, and in rare-gas and diatomic molecule matrices are evaluated. The types of measurement surveyed include conventional and laser-based absorption and emission techniques, laser absorption with mass analysis, and photoelectron spectroscopy.
Outline
Introduction
Earlier Data Compilations
Recent Progress
Scope of Database
Stable Molecules and High-Temperature Species
Types of Measurement
Ground-State Vibrational Spectra
Gas-Phase Measurements
Vibrational Frequencies from Rydberg Transitions
Matrix Isolation Measurements
Matrix Shifts for Uncharged Covalently Bonded Molecules
Matrix Shifts for Molecular Ions
Matrix Shifts for Ionic Bonds
Electronic Spectra
Gas-Phase Electronic Spectra
Flash Photolysis
Laser-Based Techniques
Photoelectron Spectroscopy
Matrix Shifts in Electronic Spectra
Guide to the Compilation
Molecular Formulas
Isotopic Species
Units
Error Estimates
Excited Electronic States
Symmetry
Assignments Based on Photoelectron Spectroscopy
Energy of the Electronic Transition
Wavelength Range of the Electronic Transition
Ground- and Excited-State Vibrations
Vibration Numbering Conventions
Renner-Teller Interaction
Inversion Splitting
Intensities
Abbreviations
Standard Abbreviations for Transition Energies
Type of Measurement
Acknowledgment
General References
Figures
Introduction
Most chemical processes--including not only laboratory and industrial chemical syntheses but also those which occur in flames, propellant systems, the initiation of energetic materials, atmospheric pollution, chemical vapor deposition, and plasma processing--consist of a complicated sequence of interrelated reactions in which neutral and charged molecular fragments play essential roles. Although these fragments are present in only very small concentration, they are highly chemically reactive. If a specific molecular fragment is removed from the system, as by introducing a scavenger molecule, the reactions in which that fragment participates stop. Other parts of the overall process continue, resulting in very significant changes in product distribution and yield.
In the early studies of complex chemical processes, it was necessary to postulate mechanisms involving such transient intermediates, present in concentrations too small for direct detection. Conventional end product analysis aids in the selection of suitable mechanisms, but generally does not yield a complete description of the system. Consequently, the improvement of industrial chemical processes often is achieved by semiempirical experimentation. The determination of the detailed chemical mechanism would, in turn, permit the development of rational strategies for removing undesired products and enhancing the yield of the desired species.
In recent years, there has been great progress in the development of techniques suitable for monitoring chemical reaction intermediates. Molecular spectroscopy is especially well suited to this task. Optical detection can be used not only for gas-phase measurements, but also for studies of processes that occur on surfaces or in the condensed phase. It also permits remote sensing, an important advantage. A wide variety of recently developed laser-based spectroscopic detection schemes are not only highly sensitive but also space and time specific. Although the development of spectroscopy-based diagnostics for chemical reaction systems is in its infancy, already the laboratory application of sophisticated sampling and observation techniques has yielded a wealth of vibrational and electronic spectral data for reaction intermediates.
Earlier Data Compilations
For many years, the most important source of vibrational and electronic energy level data for small polyatomic reaction intermediates was the compilation of spectroscopic data for small polyatomic molecules (3-12 atoms) given by Herzberg [1]. To meet the need for an updated, critically evaluated compilation, a series of publications [2,3,4] have appeared in the Journal of Physical and Chemical Reference Data, culminating in the publication in 1994 of a monograph [5] (Vibrational and Electronic Energy Levels of Polyatomic Transient Molecules) which presented evaluated spectral data for more than 1550 small polyatomic transient molecules, defined as species that have a lifetime of less than a few minutes in the pressure range (typically 0.1 to 1.0 Torr) encountered in their production. Vibrational fundamentals in the ground and excited electronic states and radiative lifetimes were included. To aid in spectral identification, the principal rotational constants were also given to three decimal places. These tables provide the basis for this database, designed to supplement the published compilation by providing a capability for rapid searches by molecule or wavenumber.
Recent Progress
The rapid growth in the scientific literature concerned with the spectroscopic study of transient molecules and with their detection in chemical reaction systems continues. Since the October 1993 cutoff in the data evaluation for the monograph, substantive new spectroscopic data have been published for many species included in it, and the first data have become available for almost 2000 other molecules. There has been especially great progress in the spectroscopic characterization of transient species produced by the reaction of metal atoms with oxygen and other small molecules. Despite this rapid progress, many gaps remain in our knowledge of the energy levels of the species represented, and many new and potentially important transient molecules are still to be discovered. This compilation attempts to provide a comprehensive, critically evaluated summary of vibrational and electronic energy level data for small polyatomic transient molecules, in order to support further research and new technologies such as those of plasma processing and chemical vapor deposition.
Scope of Database
A critical evaluation and summary of experimental vibrational and electronic energy level data for neutral and ionic transient molecules and high temperature species possessing from three to sixteen atoms is presented. By early 2003, evaluated data were available for approximately 3500 molecules, and the published version encompassed non only the original monograph [5], but also two supplements [6,7]. The Chemistry WebBook database brings these results together in an ongoing effort to provide evaluated data to the scientific community. Although radiative lifetimes and principal rotational constants are not presently included in this version of the database, references to the original literature concerned with then are given.
Stable Molecules and High-Temperature Species
Data have been selectively included for some molecules which are important in environmental and industrial chemical reaction systems but which can be studied only with difficulty using conventional sampling techniques because of the ease with which they decompose, rearrange, or polymerize. Also included are data derived from spectra of many high-temperature species, such as metal oxides, studied in molecular beams and in rare-gas matrices. Unfortunately, with these few exceptions it is not possible to include data for stable molecules. However, the spectra of many of these species are relatively well established, and sources of data such as the tables of Herzberg [1] remain extremely useful. In obtaining spectral identifications with the help of the present data, it is crucial that the possible contribution of the absorptions or emissions by a stable molecule also be considered.
Types of Measurement
Ground-State Vibrational Spectra
Gas-Phase Measurements
Studies in the gas phase offer the potential for the most precise, detailed measurements. Because of the high chemical reactivity of transient molecules, it is difficult to obtain gas-phase survey infrared spectra of them. The well known advantages of Fourier transform infrared measurements, coupled with sophisticated digital data handling procedures, have permitted the acquisition of gas-phase survey spectra for a number of transient molecules.
Vibrational Frequencies from Rydberg Transitions
Although vibrational frequencies of ground-state molecular ions have frequently been estimated from structure in Rydberg transitions of the parent neutral species, in this compilation vibrational frequencies of the ions are not inferred from these Rydberg transitions. Many of these transitions have residual valence character, resulting in significant variations in vibrational frequencies from one Rydberg state to another.
Matrix Isolation Measurements
As in the earlier compilations, spectral data obtained for molecules trapped in dilute solid solution, with the solvent a rare gas or a small covalent molecule, are included. The application of this sampling technique, known as matrix isolation, for the stabilization and spectroscopic study of uncharged reaction intermediates has recently been reviewed [8]. Because nitrogen and the rare gases are transparent through the entire infrared spectral region, matrix isolation measurements provide a potentially valuable survey tool. In these matrices, infrared absorptions are typically sharp, with half band widths between 0.1 and l cm-1. Rotational structure is, with few exceptions, quenched. Multiple trapping sites occur, often resulting in the appearance of several absorption maxima--usually one or two of which predominate--over a range of a few cm-1.
Matrix Shifts for Uncharged Covalently Bonded Molecules
Matrix shifts for covalently bonded molecules trapped in solid neon or argon often are quite small. A comparison [9] of the positions of the ground-state vibrational fundamentals of over two hundred diatomic molecules observed in the gas phase and in nitrogen and rare-gas matrices has shown that, typically, the smallest matrix shift occurs for neon matrix observations, with successively greater matrix shifts for the heavier rare gases and for nitrogen. Except for very weakly bonded molecules and for the alkali metal and Group IIIa halides, matrix shifts of most diatomic molecules isolated in solid argon are smaller than 2%. Similar conclusions resulted from a comparison of neon- and argon-matrix shifts for the ground-state vibrational fundamentals of larger molecules [10]. The generalization that matrix interactions are minimal for neon and that they increase as the mass of the rare gas is increased and become even more important for nitrogen and most other small molecule matrices is supported both by experimental observations on larger molecules and by ab initio calculations [11] for the weakly bonded CaH2 and CaF2 molecules complexed with the rare gases and with nitrogen. Figure 1 compares the observed matrix shifts for the ground-state fundamental vibrations of transient molecules trapped in solid neon and argon. For neon matrices, the maximum in the distribution lies near 0.0%, and for argon matrices, near 0.2%. For both neon and argon matrices, fewer than one-tenth of the matrix shifts are greater than 1%.
Data are beginning to appear for molecules trapped in a hydrogen matrix. Insufficient information is available for generalization on the magnitude of matrix shifts in this medium. For the few species heretofore studied, including several transient molecules present in this compilation, the matrix shifts have been comparable to those in a neon matrix.
Many other matrix materials have also been employed for spectroscopic studies. However, complications due to reaction or to relatively strong interaction (e.g., hydrogen bonding) of the transient molecule with the matrix frequently occur. Therefore, observations in such media as solid hydrocarbons and aqueous solutions and studies of condensed reaction products without an inert carrier have been excluded.
Matrix Shifts for Molecular Ions
For molecular ions, neon is the matrix of choice. Polarization and charge-transfer interactions become successively more important for molecules isolated in the heavier rare gases. Charge delocalization sometimes also occurs for ionic species trapped in the rare gases [12,13,14]. The anomalously large matrix shift for ν3 of ClHCl- may be attributed to this phenomenon. As is shown in Figure 2, data for small cation species trapped in solid neon are consistent with the matrix shift generalizations given above. Only twelve comparisons are available for molecular cations observed both in the gas phase and trapped in solid argon. The absolute values of five of the observed matrix shifts are greater than 1%. For several other vibrations of molecular ions which have been observed in both neon and argon matrices but not in the gas phase, there are deviations greater than 1% between the neon- and argon-matrix frequencies. Very few comparisons are possible for molecular anions. A number of these species have been generated in rare-gas (usually argon) matrices by charge transfer between a precursor molecule and an alkali metal atom. Recent studies of such species as CO2- and SO2- generated instead by photoionization and/or Penning ionization and trapped in solid neon indicate that shifts on the order of 50 cm-1 may be attributed to the relatively strong interaction of the anion with the nearby alkali metal cation. On the other hand, when the uncharged molecule has a relatively large electron affinity, as is true for C2 and for NO2, charge transfer occurs at a relatively great separation, and a substantial fraction of the anion population may be trapped in sites in which interaction with the alkali metal cation is minimal.
Matrix Shifts for Ionic Bonds
Matrix shifts for vibrations associated with ionic bonds are often considerably larger than those associated with uncharged molecules or with intramolecular vibrations of molecular ions. Criteria for inclusion of data for species that include an ionic bond are exemplified by the selection process for the heavy-metal oxides. Often the stable dioxide structures include an M+O2- species with significant covalent bond character for the attachment of M+, evidenced by a substantial shift in the O2- stretching fundamental as M+ is varied. Such species are included in the compilation. On the other hand, there is little evidence for substantial metal-atom participation in the vibrations characteristic of the O3- moiety of M+O3-. Accordingly, spectral data are given for O3-, but not for M+O3-.
Electronic Spectra
Gas-Phase Electronic Spectra
Gas-phase studies of the electronic spectra of transient molecules were for many years much more readily conducted than were studies of ground-state vibrational spectra. The concentration of transient molecules in flames, chemiluminescent reactions, or various types of discharge may be sufficiently high for spectroscopic study. The photographic plate provides a cumulative detector for visible and ultraviolet radiation. Much of the electronic spectral data summarized in these tables was obtained using conventional gas-phase ultraviolet absorption or emission spectroscopy, which affords the potential for both a broad spectral survey and very high resolution.
Flash Photolysis
Flash photolysis permits the observation of relatively high concentrations of transient species at short time intervals after the flash. Because the products are generally formed with much less internal energy than is typical of systems with detectable emission spectra, the absorption spectra obtained in flash photolysis studies are more readily analyzed. Furthermore, the time-resolved detection used in flash photolysis studies provides information on the rates of formation and disappearance of transient molecules in the system.
Laser-Based Techniques
More recently, a wide variety of laser-based techniques have also been used for electronic spectral observations, often with exceptionally high detection sensitivity. In recent years, very high resolutions have been achieved in laser-based spectroscopic measurements. Since a given laser is tunable over a relatively limited spectral region, laser studies of transient molecules are greatly aided by the availability of survey spectra obtained using other techniques. The spatial configuration of the laser beam makes it an extremely powerful tool for studies of the energy levels of molecules in molecular beams. If the molecule of interest is present in a supersonic molecular beam, excited rotational and vibrational energy levels can be very effectively depopulated, and the absorption spectrum of the molecule is greatly simplified. The spatial configuration of the laser beam also makes it amenable to the development of probes for chemical reaction intermediates not only in the laboratory but also in the environment and in industrial processes.
Laser studies may be broadly classified according to whether the interaction of the molecule with the laser beam(s) is followed by photon or mass detection. Photon-based observations are amenable to remote sensing applications. Because pulsed lasers offer an exceptionally wide range of time specificity, they are very useful for determining radiative lifetimes and rates of elementary chemical reactions. The coupling of lasers with mass detection can lead to the identification of transient molecules for which survey spectra are not available. Among the mass-based detection schemes are photofragment spectroscopy and resonance-enhanced multiphoton ionization (REMPI).
REMPI provides a powerful tool for mapping the Rydberg transitions of transient molecules. Whereas laser-excited fluorescence measurements depend upon the presence of electronic energy levels which decay by photon emission, all molecules possess Rydberg energy levels. REMPI measurements depend on multiphoton excitation into a suitable electronic energy level, most often one of Rydberg character. The selection rules may permit excitation of levels which are not accessible by one-photon excitation from the ground state. The range of tunability of the laser is multiplied by the number of photons required for the excitation of the Rydberg level, significantly broadening the spectral regin which can be probed with a given laser. When the parent molecule is a free radical, almost all of the mass signal is generally found to arise from the parent cation, with very little fragmentation.
Photoelectron Spectroscopy
Much valuable information on the energy levels of molecular cations has been obtained from photoelectron spectroscopy. These tables include selective coverage of the voluminous literature on photoelectron spectroscopic measurements. The number of stable molecules that possess more than six atoms for which photoelectron spectra have been reported is too great to permit the inclusion of low-to-moderate resolution photoelectron spectral data for molecular cations with more than six atoms. Those who need such data for larger molecules may find the reviews by Turner et al. [15], Rabalais [16], and Kimura et al. [17] helpful. Several criteria are important in determining whether a given reference should be included in the present work. The first of these is resolution. In the few instances in which high resolution photoelectron data are available, these data are heavily weighted. Where direct spectroscopic observation is possible, the measurements generally are of considerably higher precision than are the photoelectron data, which are then omitted from the tables. A second criterion is the availability of adiabatic ionization potentials. In order to obtain information on the positions of electronic transitions from photoelectron spectral data, it is necessary to subtract the first ionization potential from the energy of the photoelectron band. Where there is little change in the molecular geometry in the transition, the difference between the vertical ionization potentials gives a reasonable approximation to the position of the electronic transition. However, this is not the general case. Therefore, priority is given to papers that include adiabatic ionization potentials.
For most photoelectron spectroscopic transitions, structure has not been resolved. Many of these states are dissociative. Further information on the dissociation products can be obtained from values of the appearance potentials for various products in photoionization studies on the parent molecule. Such studies are beyond the scope of this review. The tables of published ionization and appearance potentials by Lias and co-workers [18,19] and the on-line version included in the Chemistry WebBook constitute a valuable source of information on the appearance potentials of photofragments.
Matrix Shifts in Electronic Spectra
The range of tunability of visible and ultraviolet lasers, like that of infrared lasers, is limited. Therefore, a preliminary survey using conventional gas-phase and/or matrix-isolation spectroscopic studies is often desirable. A comparison of the positions of the electronic band origins of diatomic molecules in the gas phase and in rare-gas and nitrogen matrices has been published [20]. As in the determination of ground-state vibrational energy levels, neon is the matrix material of choice, with a sharp maximum at 0.0% in the distribution of matrix deviations for valence transitions of covalently bonded molecules. In argon-matrix observations, most such band origins are shifted by less than 2% from the gas-phase values. At the somewhat higher temperatures often used for electronic spectral observations in matrices of the heavier rare gases or of nitrogen, relatively broad phonon bands become prominent. The blue shift of the phonon maximum from the zero-phonon line in absorption measurements, and the red shift in emission measurements, typically amount to approximately 1 to 1.5%. Rydberg transitions of molecules in matrices often are greatly broadened and experience much larger shifts. Further details of the behavior of electronic transitions of matrix-isolated molecules have previously been discussed [3,8,20].
Guide to the Compilation
Considerable effort has been expended to provide a critical evaluation of the data. However, for many species the available data are meager. The identities of some species have been proposed on the basis of chemical evidence. While such evidence may be quite compelling, it is not definitive. Many examples could be cited in which a spectrum was later reassigned to characteristic impurities in the sample. Where chemical evidence has provided a reasonable basis for the assignment of vibrational or electronic bands to a transient molecule, data have been included in this compilation, in the hope that further testing of the assignment will be facilitated.
While every effort has been made to make these tables as complete as possible, for various reasons omissions do occur. There remains some selectivity in the coverage of electronic spectral data for larger molecules. Where low-resolution photoelectron spectral data have been superseded by spectroscopic observations with appreciably higher resolution and greater precision, often the photoelectron data have not been cited. Candidate molecules or energy levels may also have been inadvertently omitted. Suggestions of additions or needed revisions to the data to be included in subsequent extensions of this database are welcome, as are inquiries regarding new data added after the cutoff date for this compilation.
A critical evaluation and summary of experimental vibrational and electronic energy level data for neutral and ionic transient molecules and high temperature species possessing from three to sixteen atoms is presented. Although the emphasis is on species with lifetimes too short for study using conventional sampling techniques, there has been selective extension of the compilation to include data for isolated molecules of inorganic species such as the heavy-metal oxides, which are important in a wide variety of industrial chemical systems. Observations in the gas phase, in molecular beams, and in rare-gas and diatomic molecule matrices are evaluated. The types of measurement surveyed include conventional and laser-based absorption and emission techniques, laser absorption with mass analysis, and photoelectron spectroscopy.
Outline
Introduction
Earlier Data Compilations
Recent Progress
Scope of Database
Stable Molecules and High-Temperature Species
Types of Measurement
Ground-State Vibrational Spectra
Gas-Phase Measurements
Vibrational Frequencies from Rydberg Transitions
Matrix Isolation Measurements
Matrix Shifts for Uncharged Covalently Bonded Molecules
Matrix Shifts for Molecular Ions
Matrix Shifts for Ionic Bonds
Electronic Spectra
Gas-Phase Electronic Spectra
Flash Photolysis
Laser-Based Techniques
Photoelectron Spectroscopy
Matrix Shifts in Electronic Spectra
Guide to the Compilation
Molecular Formulas
Isotopic Species
Units
Error Estimates
Excited Electronic States
Symmetry
Assignments Based on Photoelectron Spectroscopy
Energy of the Electronic Transition
Wavelength Range of the Electronic Transition
Ground- and Excited-State Vibrations
Vibration Numbering Conventions
Renner-Teller Interaction
Inversion Splitting
Intensities
Abbreviations
Standard Abbreviations for Transition Energies
Type of Measurement
Acknowledgment
General References
Figures
Introduction
Most chemical processes--including not only laboratory and industrial chemical syntheses but also those which occur in flames, propellant systems, the initiation of energetic materials, atmospheric pollution, chemical vapor deposition, and plasma processing--consist of a complicated sequence of interrelated reactions in which neutral and charged molecular fragments play essential roles. Although these fragments are present in only very small concentration, they are highly chemically reactive. If a specific molecular fragment is removed from the system, as by introducing a scavenger molecule, the reactions in which that fragment participates stop. Other parts of the overall process continue, resulting in very significant changes in product distribution and yield.
In the early studies of complex chemical processes, it was necessary to postulate mechanisms involving such transient intermediates, present in concentrations too small for direct detection. Conventional end product analysis aids in the selection of suitable mechanisms, but generally does not yield a complete description of the system. Consequently, the improvement of industrial chemical processes often is achieved by semiempirical experimentation. The determination of the detailed chemical mechanism would, in turn, permit the development of rational strategies for removing undesired products and enhancing the yield of the desired species.
In recent years, there has been great progress in the development of techniques suitable for monitoring chemical reaction intermediates. Molecular spectroscopy is especially well suited to this task. Optical detection can be used not only for gas-phase measurements, but also for studies of processes that occur on surfaces or in the condensed phase. It also permits remote sensing, an important advantage. A wide variety of recently developed laser-based spectroscopic detection schemes are not only highly sensitive but also space and time specific. Although the development of spectroscopy-based diagnostics for chemical reaction systems is in its infancy, already the laboratory application of sophisticated sampling and observation techniques has yielded a wealth of vibrational and electronic spectral data for reaction intermediates.
Earlier Data Compilations
For many years, the most important source of vibrational and electronic energy level data for small polyatomic reaction intermediates was the compilation of spectroscopic data for small polyatomic molecules (3-12 atoms) given by Herzberg [1]. To meet the need for an updated, critically evaluated compilation, a series of publications [2,3,4] have appeared in the Journal of Physical and Chemical Reference Data, culminating in the publication in 1994 of a monograph [5] (Vibrational and Electronic Energy Levels of Polyatomic Transient Molecules) which presented evaluated spectral data for more than 1550 small polyatomic transient molecules, defined as species that have a lifetime of less than a few minutes in the pressure range (typically 0.1 to 1.0 Torr) encountered in their production. Vibrational fundamentals in the ground and excited electronic states and radiative lifetimes were included. To aid in spectral identification, the principal rotational constants were also given to three decimal places. These tables provide the basis for this database, designed to supplement the published compilation by providing a capability for rapid searches by molecule or wavenumber.
Recent Progress
The rapid growth in the scientific literature concerned with the spectroscopic study of transient molecules and with their detection in chemical reaction systems continues. Since the October 1993 cutoff in the data evaluation for the monograph, substantive new spectroscopic data have been published for many species included in it, and the first data have become available for almost 2000 other molecules. There has been especially great progress in the spectroscopic characterization of transient species produced by the reaction of metal atoms with oxygen and other small molecules. Despite this rapid progress, many gaps remain in our knowledge of the energy levels of the species represented, and many new and potentially important transient molecules are still to be discovered. This compilation attempts to provide a comprehensive, critically evaluated summary of vibrational and electronic energy level data for small polyatomic transient molecules, in order to support further research and new technologies such as those of plasma processing and chemical vapor deposition.
Scope of Database
A critical evaluation and summary of experimental vibrational and electronic energy level data for neutral and ionic transient molecules and high temperature species possessing from three to sixteen atoms is presented. By early 2003, evaluated data were available for approximately 3500 molecules, and the published version encompassed non only the original monograph [5], but also two supplements [6,7]. The Chemistry WebBook database brings these results together in an ongoing effort to provide evaluated data to the scientific community. Although radiative lifetimes and principal rotational constants are not presently included in this version of the database, references to the original literature concerned with then are given.
Stable Molecules and High-Temperature Species
Data have been selectively included for some molecules which are important in environmental and industrial chemical reaction systems but which can be studied only with difficulty using conventional sampling techniques because of the ease with which they decompose, rearrange, or polymerize. Also included are data derived from spectra of many high-temperature species, such as metal oxides, studied in molecular beams and in rare-gas matrices. Unfortunately, with these few exceptions it is not possible to include data for stable molecules. However, the spectra of many of these species are relatively well established, and sources of data such as the tables of Herzberg [1] remain extremely useful. In obtaining spectral identifications with the help of the present data, it is crucial that the possible contribution of the absorptions or emissions by a stable molecule also be considered.
Types of Measurement
Ground-State Vibrational Spectra
Gas-Phase Measurements
Studies in the gas phase offer the potential for the most precise, detailed measurements. Because of the high chemical reactivity of transient molecules, it is difficult to obtain gas-phase survey infrared spectra of them. The well known advantages of Fourier transform infrared measurements, coupled with sophisticated digital data handling procedures, have permitted the acquisition of gas-phase survey spectra for a number of transient molecules.
Vibrational Frequencies from Rydberg Transitions
Although vibrational frequencies of ground-state molecular ions have frequently been estimated from structure in Rydberg transitions of the parent neutral species, in this compilation vibrational frequencies of the ions are not inferred from these Rydberg transitions. Many of these transitions have residual valence character, resulting in significant variations in vibrational frequencies from one Rydberg state to another.
Matrix Isolation Measurements
As in the earlier compilations, spectral data obtained for molecules trapped in dilute solid solution, with the solvent a rare gas or a small covalent molecule, are included. The application of this sampling technique, known as matrix isolation, for the stabilization and spectroscopic study of uncharged reaction intermediates has recently been reviewed [8]. Because nitrogen and the rare gases are transparent through the entire infrared spectral region, matrix isolation measurements provide a potentially valuable survey tool. In these matrices, infrared absorptions are typically sharp, with half band widths between 0.1 and l cm-1. Rotational structure is, with few exceptions, quenched. Multiple trapping sites occur, often resulting in the appearance of several absorption maxima--usually one or two of which predominate--over a range of a few cm-1.
Matrix Shifts for Uncharged Covalently Bonded Molecules
Matrix shifts for covalently bonded molecules trapped in solid neon or argon often are quite small. A comparison [9] of the positions of the ground-state vibrational fundamentals of over two hundred diatomic molecules observed in the gas phase and in nitrogen and rare-gas matrices has shown that, typically, the smallest matrix shift occurs for neon matrix observations, with successively greater matrix shifts for the heavier rare gases and for nitrogen. Except for very weakly bonded molecules and for the alkali metal and Group IIIa halides, matrix shifts of most diatomic molecules isolated in solid argon are smaller than 2%. Similar conclusions resulted from a comparison of neon- and argon-matrix shifts for the ground-state vibrational fundamentals of larger molecules [10]. The generalization that matrix interactions are minimal for neon and that they increase as the mass of the rare gas is increased and become even more important for nitrogen and most other small molecule matrices is supported both by experimental observations on larger molecules and by ab initio calculations [11] for the weakly bonded CaH2 and CaF2 molecules complexed with the rare gases and with nitrogen. Figure 1 compares the observed matrix shifts for the ground-state fundamental vibrations of transient molecules trapped in solid neon and argon. For neon matrices, the maximum in the distribution lies near 0.0%, and for argon matrices, near 0.2%. For both neon and argon matrices, fewer than one-tenth of the matrix shifts are greater than 1%.
Data are beginning to appear for molecules trapped in a hydrogen matrix. Insufficient information is available for generalization on the magnitude of matrix shifts in this medium. For the few species heretofore studied, including several transient molecules present in this compilation, the matrix shifts have been comparable to those in a neon matrix.
Many other matrix materials have also been employed for spectroscopic studies. However, complications due to reaction or to relatively strong interaction (e.g., hydrogen bonding) of the transient molecule with the matrix frequently occur. Therefore, observations in such media as solid hydrocarbons and aqueous solutions and studies of condensed reaction products without an inert carrier have been excluded.
Matrix Shifts for Molecular Ions
For molecular ions, neon is the matrix of choice. Polarization and charge-transfer interactions become successively more important for molecules isolated in the heavier rare gases. Charge delocalization sometimes also occurs for ionic species trapped in the rare gases [12,13,14]. The anomalously large matrix shift for ν3 of ClHCl- may be attributed to this phenomenon. As is shown in Figure 2, data for small cation species trapped in solid neon are consistent with the matrix shift generalizations given above. Only twelve comparisons are available for molecular cations observed both in the gas phase and trapped in solid argon. The absolute values of five of the observed matrix shifts are greater than 1%. For several other vibrations of molecular ions which have been observed in both neon and argon matrices but not in the gas phase, there are deviations greater than 1% between the neon- and argon-matrix frequencies. Very few comparisons are possible for molecular anions. A number of these species have been generated in rare-gas (usually argon) matrices by charge transfer between a precursor molecule and an alkali metal atom. Recent studies of such species as CO2- and SO2- generated instead by photoionization and/or Penning ionization and trapped in solid neon indicate that shifts on the order of 50 cm-1 may be attributed to the relatively strong interaction of the anion with the nearby alkali metal cation. On the other hand, when the uncharged molecule has a relatively large electron affinity, as is true for C2 and for NO2, charge transfer occurs at a relatively great separation, and a substantial fraction of the anion population may be trapped in sites in which interaction with the alkali metal cation is minimal.
Matrix Shifts for Ionic Bonds
Matrix shifts for vibrations associated with ionic bonds are often considerably larger than those associated with uncharged molecules or with intramolecular vibrations of molecular ions. Criteria for inclusion of data for species that include an ionic bond are exemplified by the selection process for the heavy-metal oxides. Often the stable dioxide structures include an M+O2- species with significant covalent bond character for the attachment of M+, evidenced by a substantial shift in the O2- stretching fundamental as M+ is varied. Such species are included in the compilation. On the other hand, there is little evidence for substantial metal-atom participation in the vibrations characteristic of the O3- moiety of M+O3-. Accordingly, spectral data are given for O3-, but not for M+O3-.
Electronic Spectra
Gas-Phase Electronic Spectra
Gas-phase studies of the electronic spectra of transient molecules were for many years much more readily conducted than were studies of ground-state vibrational spectra. The concentration of transient molecules in flames, chemiluminescent reactions, or various types of discharge may be sufficiently high for spectroscopic study. The photographic plate provides a cumulative detector for visible and ultraviolet radiation. Much of the electronic spectral data summarized in these tables was obtained using conventional gas-phase ultraviolet absorption or emission spectroscopy, which affords the potential for both a broad spectral survey and very high resolution.
Flash Photolysis
Flash photolysis permits the observation of relatively high concentrations of transient species at short time intervals after the flash. Because the products are generally formed with much less internal energy than is typical of systems with detectable emission spectra, the absorption spectra obtained in flash photolysis studies are more readily analyzed. Furthermore, the time-resolved detection used in flash photolysis studies provides information on the rates of formation and disappearance of transient molecules in the system.
Laser-Based Techniques
More recently, a wide variety of laser-based techniques have also been used for electronic spectral observations, often with exceptionally high detection sensitivity. In recent years, very high resolutions have been achieved in laser-based spectroscopic measurements. Since a given laser is tunable over a relatively limited spectral region, laser studies of transient molecules are greatly aided by the availability of survey spectra obtained using other techniques. The spatial configuration of the laser beam makes it an extremely powerful tool for studies of the energy levels of molecules in molecular beams. If the molecule of interest is present in a supersonic molecular beam, excited rotational and vibrational energy levels can be very effectively depopulated, and the absorption spectrum of the molecule is greatly simplified. The spatial configuration of the laser beam also makes it amenable to the development of probes for chemical reaction intermediates not only in the laboratory but also in the environment and in industrial processes.
Laser studies may be broadly classified according to whether the interaction of the molecule with the laser beam(s) is followed by photon or mass detection. Photon-based observations are amenable to remote sensing applications. Because pulsed lasers offer an exceptionally wide range of time specificity, they are very useful for determining radiative lifetimes and rates of elementary chemical reactions. The coupling of lasers with mass detection can lead to the identification of transient molecules for which survey spectra are not available. Among the mass-based detection schemes are photofragment spectroscopy and resonance-enhanced multiphoton ionization (REMPI).
REMPI provides a powerful tool for mapping the Rydberg transitions of transient molecules. Whereas laser-excited fluorescence measurements depend upon the presence of electronic energy levels which decay by photon emission, all molecules possess Rydberg energy levels. REMPI measurements depend on multiphoton excitation into a suitable electronic energy level, most often one of Rydberg character. The selection rules may permit excitation of levels which are not accessible by one-photon excitation from the ground state. The range of tunability of the laser is multiplied by the number of photons required for the excitation of the Rydberg level, significantly broadening the spectral regin which can be probed with a given laser. When the parent molecule is a free radical, almost all of the mass signal is generally found to arise from the parent cation, with very little fragmentation.
Photoelectron Spectroscopy
Much valuable information on the energy levels of molecular cations has been obtained from photoelectron spectroscopy. These tables include selective coverage of the voluminous literature on photoelectron spectroscopic measurements. The number of stable molecules that possess more than six atoms for which photoelectron spectra have been reported is too great to permit the inclusion of low-to-moderate resolution photoelectron spectral data for molecular cations with more than six atoms. Those who need such data for larger molecules may find the reviews by Turner et al. [15], Rabalais [16], and Kimura et al. [17] helpful. Several criteria are important in determining whether a given reference should be included in the present work. The first of these is resolution. In the few instances in which high resolution photoelectron data are available, these data are heavily weighted. Where direct spectroscopic observation is possible, the measurements generally are of considerably higher precision than are the photoelectron data, which are then omitted from the tables. A second criterion is the availability of adiabatic ionization potentials. In order to obtain information on the positions of electronic transitions from photoelectron spectral data, it is necessary to subtract the first ionization potential from the energy of the photoelectron band. Where there is little change in the molecular geometry in the transition, the difference between the vertical ionization potentials gives a reasonable approximation to the position of the electronic transition. However, this is not the general case. Therefore, priority is given to papers that include adiabatic ionization potentials.
For most photoelectron spectroscopic transitions, structure has not been resolved. Many of these states are dissociative. Further information on the dissociation products can be obtained from values of the appearance potentials for various products in photoionization studies on the parent molecule. Such studies are beyond the scope of this review. The tables of published ionization and appearance potentials by Lias and co-workers [18,19] and the on-line version included in the Chemistry WebBook constitute a valuable source of information on the appearance potentials of photofragments.
Matrix Shifts in Electronic Spectra
The range of tunability of visible and ultraviolet lasers, like that of infrared lasers, is limited. Therefore, a preliminary survey using conventional gas-phase and/or matrix-isolation spectroscopic studies is often desirable. A comparison of the positions of the electronic band origins of diatomic molecules in the gas phase and in rare-gas and nitrogen matrices has been published [20]. As in the determination of ground-state vibrational energy levels, neon is the matrix material of choice, with a sharp maximum at 0.0% in the distribution of matrix deviations for valence transitions of covalently bonded molecules. In argon-matrix observations, most such band origins are shifted by less than 2% from the gas-phase values. At the somewhat higher temperatures often used for electronic spectral observations in matrices of the heavier rare gases or of nitrogen, relatively broad phonon bands become prominent. The blue shift of the phonon maximum from the zero-phonon line in absorption measurements, and the red shift in emission measurements, typically amount to approximately 1 to 1.5%. Rydberg transitions of molecules in matrices often are greatly broadened and experience much larger shifts. Further details of the behavior of electronic transitions of matrix-isolated molecules have previously been discussed [3,8,20].
Guide to the Compilation
Considerable effort has been expended to provide a critical evaluation of the data. However, for many species the available data are meager. The identities of some species have been proposed on the basis of chemical evidence. While such evidence may be quite compelling, it is not definitive. Many examples could be cited in which a spectrum was later reassigned to characteristic impurities in the sample. Where chemical evidence has provided a reasonable basis for the assignment of vibrational or electronic bands to a transient molecule, data have been included in this compilation, in the hope that further testing of the assignment will be facilitated.
While every effort has been made to make these tables as complete as possible, for various reasons omissions do occur. There remains some selectivity in the coverage of electronic spectral data for larger molecules. Where low-resolution photoelectron spectral data have been superseded by spectroscopic observations with appreciably higher resolution and greater precision, often the photoelectron data have not been cited. Candidate molecules or energy levels may also have been inadvertently omitted. Suggestions of additions or needed revisions to the data to be included in subsequent extensions of this database are welcome, as are inquiries regarding new data added after the cutoff date for this compilation.