Search

An in-vacuum position-sensitive micropixelated detector (Timepix) is used to investigate the time-dependent spatial distribution of different charge state (and hence different mass-to-charge (m/z)) ions exiting an electrospray ionization (ESI)-based quadrupole mass spectrometer (QMS) instrument. Ion images obtained from the Timepix detector provide a detailed insight into the positions of stable and unstable ions of the mass peak as they exit the QMS. With the help of image processing algorithms and by selecting areas on the ion images where more stable ions impact the detector, an improvement in mass resolution by a factor of 5 was obtained for certain operating conditions. Moreover, our experimental approach of mass resolution enhancement was confirmed by in-house-developed novel QMS instrument simulation software. Utilizing the imaging-based mass resolution enhancement approach, the software predicts instrument mass resolution of ?1,0000 for a single-filter QMS instrument with a 210-mm long mass filter and a low operating frequency (880?kHz) of the radio frequency (RF) voltage.

In this paper, we have employed an ion imaging approach to investigate the behavior of ions exiting from a quadrupole mass spectrometer (QMS) system that employs a radio frequency octopole ion guide before the QMS. An in-vacuum active pixel detector (Timepix) is employed at the exit of the QMS to image the ion patterns. The detector assembly simultaneously records the ion impact position and number of ions per pixel in every measurement frame. The transmission characteristics of the ion beam exiting the QMS are studied using this imaging detector under different operating conditions. Experimental results confirm that the ion spatial distribution exiting the QMS is heavily influenced by ion injection conditions. Furthermore, ion images from Timepix measurements of protein standards demonstrate the capability to enhance the quality of the mass spectral information and provide a detailed insight in the spatial distribution of different charge states (and hence different m/z) ions exiting the QMS.

The dissociative scattering of CF 3 + ions from a clean Ag(111) and a barium-covered Ag(111) surface was explored in the energy range 100–600 eV. Both negative (CF 3 − , CF 2 − and F − ) and positive (CF + and C + ) ions were observed. The complete dissociation of the CF 3 + ions into C + fragments was found at low (normal) incoming energies. The breaking of three CF bonds is due to concerted electronic and mechanical energy transfer. The azimuthal orientation of the crystal has a large influence on the degree of dissociation, which indicates the importance of direct momentum transfer. For the barium-covered Ag(111) surface the negatively charged scattering products only showed F − and the positives showed CF + and C + fragments. No negatively charged polyatomic ions were observed. The lowering of the work function facilitates resonant neutralization into the excited 3s Rydberg state of the neutral molecule. Consequently, the spectral features associated with Auger neutralization are not observed.

The dissociative scattering of CF3+ ions from a liquid insulating perfluorinated polyether surface was explored in the incident energy range 100–500 eV. We have measured the angular and energy distributions of the positive and negative fragment ions. The impulsive energy transfer to the liquid molecules and to internal energy of the scattered ions has been determined as a function of incoming energy and total scattering angle. The energy transfer in the ion/surface collision indicated scattering off the –CF3 groups of the liquid molecules. The dissociation of the CF3+ ions was due to a single impulsive collision with these groups. The parent ion was observed as well as CF2+, CF+, and C+ ions. The degree of dissociation depends only on the total scattering angle and the incoming energy, as in gas phase scattering. Remarkably, negatively charged F− fragment ions were also observed. These were formed in a process referred to as collisionally-induced “ion-pair” formation from neutrals leaving the surface. The...

The dissociative scattering of CF + 3 ions from a barium-covered Ag(111) surface has been studied for incident energies between 100 and 500 eV. The negatively charged scattering products only showed F − and the positives showed CF + and C + fragments. No negatively charged molecular ions are observed. The results are compared with those on scattering from a clean Ag(111) surface. The lowering of the work function facilitates resonant neutralization into the excited 3s Rydberg state of the neutral molecule. Consequently, the spectral features associated with Auger neutralization are not observed.

A perfluorinated polyether (PFPE) surface is used to enhance the fragment ion yield for surface-induced dissociation (SID) in a tandem time-of-flight mass spectrometer. Selection of the parent ion by separation of M+ from (M — H)+ is obtained with unit resolution by pulsing the voltage on the collision surface. The benzene radical cation is used as a model compound to characterise the instrument. Fragment ions are collected with a mass resolution of 100. The distribution of recoil energies of the SID fragments leaving the collision surface is not sufficient to clarify the limited mass resolution. We estimate an average conversion efficiency of kinetic into internal energy of 30±7% between 20 and 40 eV collision energy by analysing the relative intensities of the SID fragments as a function of collision energy.

To evaluate the necessity for cooling laser-desorbed molecules to obtain molecular ions and minimal fragmentation, mass spectra of a series of methoxybenzenes were measured in two different experimental configurations. In one geometry, laser-desorbed molecules were entrained in a pulsed supersonic jet before ionization. In the other, the molecules were ionized directly after laser desorption. The samples were ionized with laser-generated vacuum ultraviolet radiation in a single-photon process or ultraviolet radiation in a multi-photon process.

In this paper two semi-empirical models are presented. By the one model the electron impact ionization cross sections of polyatomic molecules can be estimated for electron energies in the range from 15 to 150 eV. Using the other model the photoionization cross section of polyatomic molecules is determined at photon energies of 11.8 and 16.7 eV. In the case of electron impact, the model reproduces the molecular ionization cross sections within 20% at electron energies above 50 eV. The photoionization cross sections calculated with the model deviate not more than 20% from the experimentally determined values for all polyatomic molecules tested.

We present absorption spectra of the rotational band contours of the 6 1 0 and 7 1 0 bands in benzene, 12 C 6 H 6 and 13 C 12 C 5 H 6 . These spectra were measured using resonant two-photon ionization in combination with time of flight mass spectrometry. A rotational temperature of 2.0 K was obtained by coexpanding the benzene in a supersonic jet of argon. We have resolved different J levels in the 6 1 0 band. Contours of the bands are simulated using a rigid-rotor model and accounting for Coriolis splitting. For the isotopomer, the vibrational modes 6 and 7 each consist of two nondegenerate modes a and b and the positions of these band origins relative to the band origin of the twofold degenerate state of benzene have been determined. The splitting between bands of mode 6 is in good agreement with values reported in literature which were calculated with force field methods for the ground state. The splitting in band 7 is dominated by either a difference in anharmonicity or a shift due to a Fermi resonance.

We present absorption spectra of the rotational band contours of the 6 1 0 and 7 1 0 bands in benzene, 12 C 6 H 6 and 13 C 12 C 5 H 6 . These spectra were measured using resonant two-photon ionization in combination with time of flight mass spectrometry. A rotational temperature of 2.0 K was obtained by coexpanding the benzene in a supersonic jet of argon. We have resolved different J levels in the 6 1 0 band. Contours of the bands are simulated using a rigid-rotor model and accounting for Coriolis splitting. For the isotopomer, the vibrational modes 6 and 7 each consist of two nondegenerate modes a and b and the positions of these band origins relative to the band origin of the twofold degenerate state of benzene have been determined. The splitting between bands of mode 6 is in good agreement with values reported in literature which were calculated with force field methods for the ground state. The splitting in band 7 is dominated by either a difference in anharmonicity or a shift due to a Fermi resonance.

Butane, methylpropane, 1−,2-butene and isobutene are studied as a complete set of doubley methylated ethane and ethene molecules in order to probe the dissociation dynamics of their excited states. The butane and butene molecules are formed in electronically highly excited states by neutralization of their respective radical cations in collisions with alkali metal atoms. Processes that lead to dissociations into two fragments are measured with differential translational spectroscopy. it is found that the first triplet state is dominantly populated. Butanes exhibit direct dissociation of a carboncarbon (CC) bond via a repulsive state. Intramolecular vibrational energy redistribution (IVR) prior to dissociation is observed for the unsaturated butenes. For the butenes the CC bond cleavage is even preceded by a hydrogen shift, resulting in molecular losses like methane and ethene. In this paper we stress the relations between the present results and optical excitation and electron scattering experiments. For butane the proposed dissociation via a repulsive triplet state is corroborated in ab initio calculations.

The performance of a newly developed photoionization source in combination with a high resolution mass spectrometer is tested. The total ion currents for several linear alkanes ( n -pentane to n -decane) and for some thiophenes (2-methylthiophene, 2,5-dimethylthiophene and 2-ethylthiophene) are measured at three fixed wavelengths: the Kr I (10.03 eV and 10.64 eV), Ar I (11.62 eV and 11.83 eV) and Ne I (16.67 eV and 16.85 eV) resonance lines. These total ion currents are compared with the data for the alkanes and thiophenes obtained with low energy (10.6 eV, 11.8 eV and 16.7eV) electron impact ionization on the same mass spectrometer. The loss in ion intensity at several positions throughout this instrument is determined for the photo-ionization source and for the electron impact ionization source. One out of every 1500 ions created in the photoionization source is measured by the detector. For the electron impact ionization source, one out of 180 ions is measured. The introduction of a five-element “Heddle” lens for the transfer of the ions from the photoion source to the mass analyser resulted in an approximately 1.5 fold loss in ion current. From the measured total ion intensities, the photoionization and electron impact ionization cross-sections at energies of 10.6 eV, 11.8 eV and 16.7 eV are calculated. The photoionization cross-section values of the linear alkanes are found to be in the range of 2.5 Mbarn to 355 Mbarn, and the electron impact ionization cross-section values are between 40 Mbarn and 735 Mbarn. The photoionization cross-section values of the thiophenes range from 4 Mbarn to 31 Mbarn, and the electron impact ionization cross-section values from 81 Mbarn to 760 Mbarn.

VUV light of 118.2 nm (10.49 eV) is produced by frequency tripling the third harmonic of a Nd: YAG laser (354.6 nm) in a Xe cell. With these photons time-of-flight mass spectra of n -hexane, 3-methylpentane, 2,2-dimethylbutane, 2,3-dimethylbutane and cyclohexane are obtained by single photon ionization. The simple set-up for production of 118.2 nm photons, the necessity of separating the 354.6 nm photons and the obtained mass spectra are discussed. The fragmentation of hexane as a function of its temperature is compared with the predictions from quasi-equilibrium theory. The model shows that for n -hexane less than 50% of the difference between photon energy and ionization energy is deposited in the molecular ion.

The combination of laser desorption followed by multiphoton ionization (MPI) with time of flight mass spectrometry (TOF/MS) produces a versatile method of high sensitivity and selectivity. The first step is laser desorption in which molecules are desorbed from the sample intact or with little degradation. Laser ionization is employed as the second step to produce the ions and also induce fragmentation. Soft ionization, which yields primarily the parent ion, and hard ionization, which yields primarily the fragments, are both possible. Instrumental details are described and preliminary results from recent laser desorption MPI-TOF/MS experiments on a perbenzoylated sugar and on a porphyrin are presented. Examples from single photon ionization (SPI) experiments on two alkanes are also shown.

The internal energy of protonated leucine enkephalin has been manipulated in electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry with two newly designed pump-probe experiments. Blackbody infrared radiation was applied to pump an ion population into a well-defined internal energy distribution below the dissociation threshold. Following this pumping stage, the internal energy distribution was probed using on-resonance collisional activation to dissociate the ions. These pump-probe experiments were carried out in two different ways: (a) using on-resonance collisional activation with variable kinetic energies to dissociate the ions at a constant initial ion temperature (determining the precursor ion survival percentage as a function of kinetic energy) and (b) using on-resonance collisional activation with a constant kinetic energy to dissociate the ions at variable initial ion temperatures (to investigate the ion survival yield-initial ion temperature dependence). Using this approach, a detailed study of the effects of the initial ion temperature, the probing kinetic energy and the internal energy loss rate on the effective conversion efficiency of (laboratory-frame) kinetic energy to internal energy was conducted. This conversion efficiency was found to be dependent on the initial ion temperature. Depending on the experimental conditions the conversion efficiency (for collisions with argon) was estimated to be about 4.0 +/- 1.7%, which agrees with that obtained from a theoretical modeling. Finally, the reconstructed curves of the ion survival yield versus the mode of the (final) total internal energy distribution of the activated ion population (after pump and probe events) at different pump-probe conditions reveal the internal energy content of the activated ions.

The ease of fragmentation of various charge states of protonated polypropylenamine (POPAM) dendrimers is investigated by surface-induced dissociation. Investigated are the protonated diaminobutane propylenamines [DAB(PA)n] DAB(PA)8 (1+ and 2+), DAB(PA)16 (2+ and 3+), and DAB(PA)32 (3+ and 4+). These ions have been proposed to fragment by charge-directed intramolecular nucleophilic substitution (SNi) reactions. Differences in relative fragment ion abundances between charge states can be related to the occupation of different protonation sites. These positions can be rationalized based on estimates of Coulomb energies and gas-phase basicities of the protonation/fragmentation sites. The laboratory collision energies at which the fragment ion current is approximately 50% of the total ion current were found to increase with the size, but to be independent of charge state of the protonated POPAM dendrimers. It is suggested that intramolecular Coulomb repulsion within the multiply protonated POPAM dendrimers selected for activation does not readily result in easier fragmentation, which is in accordance with the proposed fragmentation mechanism.

Photodissociation of mass-selected ions in a crossed-beam setup with a pulsed excimer laser (248 nm) is described. The mass-to-charge ratios of the photodissociation products are determined using the mass analyzing quadrupole Q of a hybrid mass spectrometer consisting of a double-focusing (BE) part followed by two quadrupoles to give BEqQ geometry. Technical aspects of the experiment, as well as experimental results of the photodissociation of the radical cations of toluene, norbornadiene and cycloheptatriene, are presented.

Laser desorption with either multiphoton or single photon post-ionization is combined with time of flight mass spectrometric detection. In the first step large molecules are desorbed intact from a surface. In the second step laser ionization is employed to produce the ions of interest. Soft ionization, which yields primarily the molecular ion, and hard ionization, which yields primarily fragment ions, are both demonstrated.

The rotational—translational energy in collisions between N 2 and noble-gas atoms has been investigated experimentally as well as theoretically. Experimental relaxation times have been derived from acoustical absorption measurements. A definite influence of the atomic mass on the relaxation of N 2 is observed. The rotational relaxation in N 2  Ne and N 2  Ar systems needs a smaller number of collisions than either N 2 He or N 2  Xe. These experimental observations are confirmed by classical trajectory calculations, using a dumbbell model for the intermolecular potential.

Laser desorption results of sucrose can be explained by a gas phase complexation: Alkali ions from hot parts (T > 700K) of the surface combine in the gas phase with sucrose molecules which desorb from cooler places on the surface. This ionization process seems to be applicable to other “non-volatile” compounds too. Gas phase ionization contributions to other soft ionization techniques are also discussed.

To study the fundamentals of the sputtering process kinetic energy distributions are measured for various molecular and fragment ions sputtered from solid and liquid matrices. These data are compared with theoretical distributions based on a statistical sputter mechanism.

Analytical expressions are presented for the internal and kinetic energy distributions of sputtered molecules and clusters. The molecules and clusters are treated as systems built up of two identical sub-units which are activated independently in the sputter process. First general expressions are obtained from approximate momentum space integrations. Then explicit functions are given for collision cascade and thermal sputtering. The quality of the analytical expressions for the collision cascade model is evaluated by comparison with a numerical calculation of the momentum space integrals.

Kinetic energy distributions are measured for mass-selected ions sputtered from solids and liquids with an energy resolution of 0.24 eV (FWHM). The energy distributions of organic ions sputtered from a glycerol matrix have FWHM values in the range 0.3–0.7 eV. Larger values are observed in direct sputtering from solid samples. Sample charging is always observed but can be limited to about 0.2 eV by using a glycerol matrix spiked with HCl or NaCl. The residual charging is independent of the use of neutral or charged primary particles. Molecular-ion and fragment-ion energy distributions are in first order highly similar. In second order for most systems analyzed, a weak increase of the width of the energy distributions is observed with increasing mass of the sputtered ions. Collision cascade contributions are visible from the high-energy tail of the distributions. A thermal spike model applied to the low-energy part of the distribution points to “temperatures” of about 2000–4000 K

Quasimolecular ion-formation in LDMS may be explained by gas phase ion-molecule reactions. Molecular evaporation and ionic emission have to be correlated in time but require different substrate temperatures. The contribution of these processes in two LDMS-experiments is investigated.

Orientational effects in collisions between a diatomic molecule and a noble gas atom are studied by classical trajectory calculations. A preferable ”perpendicular orientation“ of the molecule is found at the moment of closest approach of the atom. This orientational effect is more pronounced in collisions with heavy atoms.

The mass spectrum of thermally evaporated glycerol is measured as a function of temperature by photoionization mass spectrometry (PIMS) at two wavelengths, the Ar i (11.8 eV) and Kr i (10.6 eV) resonance lines. Also the PIMS of glycerol molecules sputtered from the liquid at 267 K is measured. By comparison with the vapor phase measurements, the internal temperature of the sputtered molecules is determined as 190±40 K. Then the assumed ejection by a ‘‘thermal spike’’ must be followed by an expansion which cools the sputtered molecules to an internal temperature which is even lower than the temperature of the liquid. Applying an analytical deconvolution, the breakdown curve of the glycerol molecules is calculated. Comparison between the Ar i and Kr i vapor phase data shows that only 0.3 eV of the excess energy of the argon photon is converted into internal energy.

Nonvolatile biological materials, such as biopolymers or even whole cells, can be made accessible to mass spectrometric analysis by pyrolysis techniques. In principle, this enables differentiation, classification and identification of these materials. Moreover, pyrolysis mass spectrometry can provide a general picture of the chemical composition and structure of the sample analysed. During the past few years we have investigated several pyrolysis techniques in combination with different mass spectrometric methods. These investigations included Curie point pyrolysis, infrared laser pyrolysis, low voltage electron impact ionization, field ionization1 and pyrolysis field desorption.1 Pyrolysis reactions were performed either at the gas inlet or directly at the ion source. The relative merits of these different experimental approaches are discussed. Additionally, examples of various applications, such as computer identification of bacteria and qualitative analyses of biopolymers, are given.

The qualitative changes in pyrolysis mass spectra obtained under various experimental conditions were studied by factor analysis. The parameters involved are the pyrolysis method (oven vs. filament), the method of cleaning the pyrolysis wire, the choice of the suspending liquid, the sample size, the filament position, the temperature-time profile and the temperature of the expansion chamber. For a number of these parameters, factor analysis of the data enabled a physico-chemical interpretation of the variations in the pyrolysis mass spectra to be made. An increased formation of dehydrogenated fragments was observed at higher equilibrium temperatures and for oven pyrolysis. Condensation of less volatile pyrolysis products at relatively cold parts of the inlet system (reaction tube and expansion chamber) was clearly observed.

The desorption of neutrals, alkali ions and quasimolecular ions of sucrose was studied as function of substrate temperature in laser desorption mass spectrometry. These phenomena were also investigated in thermal desorption experiments. It was concluded that in these experiments gas phase cationization is the major ionization process.

The asymptotic behaviour of the kinetic energy distributions of various ions sputtered from liquid glycerol was measured to study the sputtering mechanism. Evidence for the operation of a collision cascade mechanism was found for Na+ ions sputtered from glycerol by the observed asymptotic E−2 behaviour. Other ions sputtered from glycerol also show a power dependence E−x. Values found for x are GlyH+, x = 5.3 ± 0.5; GlyNa+, x = 4.7 ± 0.5; (Gly)2H+, x = 4.8 ± 0.5; (NaCl)Na+, x = 7.0 ± 1.0; and (NaCl)2Na+, x = 7.8 ± 1.0. Glycerol fragments show dependences varying from x = 2.2 ± 0.3 to x = 4.7 ± 0.5. These results are interpreted using the linear collision cascade model.

Two different biopolymers, bovine serum albumin and glycogen, were examined by pyrolysis-mass spectrometry under various experimental conditions. The parameters tested include the method of cleaning the pyrolysis wire, the choice of the sample solvent, the sample size, the temperature-time profile and the temperature of the mass spectrometric inlet system. In order to determine quantitatively the effects of changes in experimental conditions on reproducibility of the spectra, the differences observed between the pyrolysis mass spectra represented as numerical dissimilarity values. It is concluded that the long-term differences observed are only partly explained by fluctuations in the conditions studied. However, it appears feasible to define a set of standard conditions with respect to the production of the pyrolyzate which should afford a reasonable degree of interlaboratory reproducibility.

Methods currently in use for pyrolysis-mass spectrometry (Py-MS) of macromolecular and complex organic samples are discussed. Selected examples covering various fields of application, viz. technical polymers, geochemical, biological, medical,environmental and forensic samples are given.

Sound-absorption measurements have been made in mixtures of CD4 with He, Ne, Ar and Xe to investigate the effect of the noble gases on the rotational relaxation of CD4. The measurements were carried out in the pressure range of 100 to 150 torr, at 308 K and a frequency of 3 MHz. The relaxation times and rotational collision numbers obtained show the same tendency as was found in CH4—noble-gas mixtures. The collision numbers Zrot, defined as the ratio of the rotational relaxation time and the mean time between two elastic collisions, increase with increasing mass of the noble gas. We found the following numbers: Zrot (CD4CD4) = 10.0 ± 0.3, Zrot (CD4He) = 2.6±0.7, Zrot (CD4Ne) = 4.1±0.6, Zrot (CD4Ar) = 7.3±0.6, Zrot (CD4Xe) = 14±3. A comparison between these experimental results and theoretical predictions by the rough-sphere model is given.

Mass spectra resulting from superthermal collisions between polyatomic ions and atomic or diatomic targets have been acquired as a function of the collision energy over a range of 10–6000 eV for the t-butyl cation and the molecular ions of methane, propane and n-butylbenzene. Changes in relative fragment ion abundances with systematic variation of the collision energy have been monitored in order to assess the effect of this experimental variable on the degree to which the polyatomic ion is excited by collision. The data for the t-butyl cation and the methane molecular ion show an apparent monotonic increase in energy deposition into the ion with increasing collision energy over the entire range investigated. The data for the molecular ions of propane and n-butylbenzene, on the other hand, indicate that maxima in energy deposition into these ions occur at laboratory collision energies of 140 and 70 eV, respectively. The results are interpreted as being due to changes in the relative importance of direct vibrational excitation and electronic excitation with increasing collision energy.

A method is described for the separation of a set of pyrolysis-mass spectra of simulated mixtures into subspectra for each of the components and for calculation of their relative concentrations. The technique is based on factor analysis and graphical rotation and is evaluated on a data set, constructed by linear combination of pyrolysis-mass spectra of three biopolymers.

An approach to describing the decomposition patterns of polyatomic ions as a function of target gas pressure is presented. The kinetics of a decomposition scheme can be described employing a set of simultaneous first-order differential equations. From the pressure resolved data, the different contributions to the fragment ion intensities arising from single or multiple collision events can be calculated. Provided certain conditions hold, a high target pressure MS/MS spectrum can be predicted from data acquired at low target gas pressure. The decomposition pattern of 6 keV CH4+ ions has been measured as a function of pressure for the three target gases He, Ar and Xe in order to illustrate the utility and limitations of the approach. The target gas pressure was varied from 0 to 13 Pa. An estimate of the absolute cross-section for collision-induced dissociation of the methane molecular ion is given. Furthermore, we conclude that, for the targets investigated, the internal energy deposition per collision follows the order He &>; Ar &>; Xe.

Kinetic energy distributions of laser-desorbed Na + ions and quasimolecular ions are presented for a pulsed CO 2 laser (τ = 200 ns, 2 × 10 7 W cm −2 ) irradiating a stainless steel surface coated with sucrose. From these distributions, information is extracted on the processes contributing to the formation of quasimolecular ions. The energy distributions are broader than thermal distributions and display a tail at higher energies extending to 1 eV. The energy distributions are interpreted as a transformation of initially thermal distributions ( T 3 K) by ion/molecule interactions in a dense plasma cloud over the surface. Calculations supporting this interpretation are presented.

The temperature profiles throughout a rapidly heated sample layer were calculated as functions of time and sample thickness using the one-dimensional heat equation. The rates of decomposition throughout the layer were then calculated using the Arrhenius equation. The decomposition temperature, T1/e, for a layer is defined as the temperature at which a fraction of 37% of the material is not yet decomposed. The theoretical results are used to predict T1/e in cellulose for different Curie-point temperatures and heating rates. In general, T1/e increases with increasing heating rates. In vacuum Curie-point pyrolysis (e.g., pyrolysis-mass spectrometry) of thin samples (thickness

Translational spectroscopy is applied to study the dissociation of 5 keV acetone ions in collisions with helium atoms. In these experiments, the scattering of the acetone ion in the collisional excitation step is studied. Semi-quantitative scattering angle distributions of the parent ion are obtained from the measurement of the scattering of the ionic and neutral fragments in coincidence. These distributions show a maximum at laboratory scattering angles of 0.15°. The scattering of the parent ion is of a size comparable with the scattering of the fragments in the dissociation process. From these first results, it follows that the excitation proceeds preferentially in close collisions with some exchange of momentum.

The pure tandem mass spectra of different isobaric ions are reconstructed from series of mixture spectra in which the relative contributions of the ions are varied. Principal component analysis is applied in combination with two natural assumptions about the spectral properties: no negative mass peak intensites and at least one “zero intensity mass peak” occurring in a pure tandem mass spectrum. A comparison is made with other methods for spectrum reconstruction. The signal-to-noise ratio, the number of unresolved isobaric precursor ions and the success of the spectrum reconstruction can be estimated from the eigenvalues of the dispersion matrix. The linearity of the data is shown to be of crucial importance. An increase in effective precursor mass resolution from 500 to 5000 is achieved.

In order to establish the influence of forbidden transitions, rotational relaxation numbers Zrot of the gases 14N2, 14N15N, 15N2, 12C16O and 12C18O were determined from ultrasonic absorption measurements. The absorption coefficient was measured in a single-crystal interferometer at a frequency of 3.007 MHz and a temperature of 307.46 K in the pressure range from 75–200 torr. The ratio of the experimental and the classical absorption coefficients determines Zrot. The following results were obtained: Zrot(14N2) = 3.9,Zrot(14N15N) = 4.1 Zrot(15N2) = 4.5, Zrot(12C16O) = 2.8 and Zrot(12C18O) = 2.7. No influence on the Z value due to the symmetry of the homonuclear N2 molecules is observed. A short discussion is given.