GasPhase Beamline

The Gas Phase Photoemission (GAPH) beamline is the only one at Elettra specifically devoted to research on gaseous systems.
GAPH offers a multi-technique approach for investigation of electronic properties of free atoms, molecules and clusters in the photon energy range 13-900 eV.

The broad energy range, the high resolving power and flux together with the purpose built end-stations, make this facility ideal for investigating the spectroscopy and dynamics of basic processes like inner-shell and multiple excitations and ionisation, as well as for characterising key processes relevant to several areas of science and technology (for example atmospheric chemistry, material science and biomedical sciences).

Gas Phase highlights | Publications

slideshow image

The Fragmentation Dynamics of Simple Organic Molecules of Astrochemical Interest Interacting with VUV Photons

An experimental investigation on the fragmentation dynamics following the double photoionization of simple organic molecules of astrochemical interest, propylene oxide and N-methylformamide molecules, induced by VUV photons has been reported.  Falcinelli  (2019) ACS Earth Space Chem 

Read More
slideshow image

Fragmentation of Small Molecules after Core Excitation and Core Ionization Studied by Negative-Ion/Positive-Ion Coincidence Experiments

We have studied the fragmentation of the methanol molecule after core excitation and core ionization by observing coincidences between negative and positive ions.
Kivimäki et al.; J. Phys. Chem. A  122 (2018) 224

Small molecules usually break into parts after the absorption of an X-ray photon. More than 99% of the states with a hole in a C 1s, N 1s or O 1s shell have been calculated to decay through electron emission. Subsequent dissociation mainly produces positive and neutral fragments owing to the positive charge of the molecular ion after normal and resonant Auger decay. Negative ions (or anions) have also been observed at the core edges, but there is very little information about their formation. We have studied the fragmentation of the methanol molecule after core excitation and core ionization by observing coincidences between negative and positive ions. 
Our experimental setup consists of two time-of-flight (TOF) spectrometers facing each other; one of the TOF spectrometers is used to detect positive ions, the other for negative ions. Electrons, which often interfere with negative ion detection, are deflected by a weak magnetic field created by permanent magnets, placed outside the vacuum chamber. Coincidences between negative particles (anions + residual electrons) and positive ions were recorded using a constant extraction field in the interaction region. The arrival times are analyzed afterwards, searching for arrival-time-differences (ATD) between the particles. This procedure allows for the possibility that a negative ion can arrive before or after the positive ions, depending on their masses. The analysis can thus yield also multiple negative-ion/positive-ion coincidences.
For methanol five different negative ions ― H-, C-, CH-, O-, and OH- - were observed both at the C 1s and O 1s edges. As negative ion formation occurs after resonant and normal Auger decay of core-hole states, it is necessarily linked with the release of positively charged fragments. Our data show that such fragmentation can happen in many different ways: We found approximately 30 negative-ion/positive-ion/positive-ion coincidence (NIPIPICO) channels. All involve only singly charged positive ions. Fragmentation channels leading to atomic ions are the most probable, but positive molecular ions are also frequently found in the context of anion formation. We could also verify the occurrence of four-ion coincidences, which involved one negative ion (H- or O-) and three positive ions.
The coincidence detection of negative and positive ions not only helps us to identify negative ions released by a given sample molecule, but also gives information on dissociation channels that produce these anions. The most intense NIPIPICO channels belong to the series O/H+/CHn+ (n =0−3), H/H+/CHn+ (n = 0−2), and H/H+/COHn+ (n = 0,1), where the intensities typically decrease when n increases. As an exception, the O/H+/CH3+ channel gains much intensity at the excitations of O 1s electrons to high-Rydberg orbitals.

Retrive article

      Fragmentation of Methanol Molecules after Core Excitation and Core Ionization Studied by Negative-Ion/Positive-Ion Coincidence Experiments
Antti Kivimäki, Christian Stråhlman, Robert Richter, and Rami Sankari
J. Phys. Chem. A  122 (2018) 224  

Read More
slideshow image

Highly efficient double ionization of mixed alkali dimers by intermolecular decay

 Here, we report on a new decay mechanism leading to double ionization by intermolecular energy transfer between an electronically excited helium atom and alkali metal dimers.    A. C. LaForge et al., Nature Physics 15, 247 (2019)

Read More
slideshow image

Triple metal-metal bonds of M2(CH2CMe3)6 (M = Mo, W) investigated by photoelectron spectroscopy and density functional theory

The variable PES of a pair of group 6 triply metal-metal bonded alkyl compounds (M2(CH2CMe3)6 (M=Mo, W) were obtained for the first time and benchmarked againest state-of-art DFT calculations.  M. de Simone et al. Organom. 2022 

Photoelectron spectroscopy (PES) is a powerful technique for measuring the quantized energies of electrons in molecules, and the ability to vary the incident photon energy enables determination of their role in chemical bonding. Such data has proved vital in the development, validation, and benchmarking of theoretical methods for the study of electronic structure. Investigations of the nature of chemical bonding between transition metal atoms (M) in molecular compounds began in the 1960s and have increased dramatically over the past half century, with a significant emphasis placed on the synthesis, structure, spectroscopy, and chemical reactivity of compounds containing metal-metal multiple bonds. A remarkable subclass of the latter are dinuclear molybdenum(III) and tungsten(III) compounds of the type M2X6 (M = Mo, W; X = bulky anionic ligand), which exhibit staggered, ethane-like geometries and short metal-metal bonds that are the embodiment of 24 metal-metal triple bonds (Fig 1a-c). In this study, the electronic structure of a pair of M2X6 compounds where the bulky anionic ligand is a sigma-bonded neopentyl group (X = CH2CMe3) were compared and contrasted.

PE spectra were measured for both compounds over a photon energy range from 20 – 70 eV. Significant variations in the band intensities were observed with the ionizations associated with metal-based orbitals increasing in relative intensity with increasing photon energy (Fig. 1d-e). Ionizations from the M-M π- and s-bonding orbitals showed the largest intensity increase and occurred at the lowest ionization energy. M-C ionizations lie between 8 and 10 eV and they also increase in intensity relative to the broad bands between 10 and 16 eV that are associated with C-C and C-H ionizations from the neopentyl ligands.   More detailed assignments were enabled by density functional theory (DFT) calculations using the Amsterdam Density Functional (ADF) modelling suite. Whereas in the case of Mo­2­(CH2CMe3)6 the first ionization band is associated with both metal-metal π- and s-bonding orbitals, for W­2­(CH2CMe3)6 the first band is assigned to a spin-orbit split π-ionization (Fig. 1f-g). Agreement between calculated and experimentally measured ionization energies was excellent. A marked difference between the bonding patterns of orbitals of Mo2(CH2CMe3)6 and those of W2(CH2CMe3)6 is the stabilization of those orbitals containing 6s character for W and is attributable to relativistic effects. The calculations also gave very good agreement with the structural parameters determined by x-ray diffraction and the electronic absorption spectra of the two compounds measured in alkane solution.
The variable photon energy PE spectra of a pair of Group 6 triply metal-metal bonded alkyl compounds, viz., M2(CH2CMe3)6 (M = Mo, W), were obtained for the first time and benchmarked against state-of-the-art DFT calculations. The Amsterdam Modelling Suite (ADF 2019.306) was used in conjunction with PBEO-dDsC functionals, which include dispersion forces, to calculate ground state structures, energy levels, isosurfaces, PES and electronic absorption spectra for a number of M2X6 compounds. The details of the specific metal-metal triple bonds are discussed, as are the similarities and differences in the energy levels and spectroscopic features of the molybdenum(III) and tungsten(III) neopentyl dimers. The agreement between the experimental data (x-ray, PES and UV-Vis) and theory is remarkable. Spin-orbit splitting (0.33 eV) is observed in the π-ionization of the tungsten neopentyl dimer and was successfully modelled with a relativistic calculation on the cation [W2(CH2CMe3)6]+.  

Retrive article


Triple metal-metal bonds of M2(CH2CMe3)6, (M=Mo, W) investigated by photoelectron spectroscopy and density functional theory. Simone, R.Totani, M. Coreno, N. E. capria, L. Messerle, J. C. Green, A. P. Sattelberger Organometallics 41 (2022) 29; doi:10.1021/acs.organomet.1c00586






Read More
slideshow image

Photoionization dynamics of the tetraoxo complexes OsO4 and RuO4

The photoionization dynamics of OsO4 and RuO4, chosen as model systems of small-size mononuclear heavy-metal complexes, has been theoretically studied by the time-dependent density functional theory (TDDFT) and compared to accurate  experimental measurements. Schio et al. Inorg.Chem. (2020)


Mononuclear and polynuclear organometallic clusters are a class of interesting and technologically relevant complexes due to their widespread use in catalysis. They display a complex electronic structure, bearing the signature of strong correlation and relativistic effects. Moreover, their nontrivial nuclear dynamics leads to a high degree of fluxionality and complex fragmentation dynamics. Photoelectron spectroscopy has established itself as one of the principal tools for the investigation of molecular electronic structure, providing both ionization energies and dynamical observables such as partial ionization cross sections and asymmetry parameters β. In particular, the angular distribution of photoelectrons has scarcely been investigated for organometallic systems. Computational methods to describe photoionization observables, based on ground-state density functional theory (DFT) and linear response time-dependent DFT (TDDFT), are well established for relatively simple organic molecules. To benchmark and assess the quality of the theoretical approach in the case of heavy metal atom containing systems, we have here investigated the tetraoxo complexes OsO4  and RuO4  as model cases. Accurate experimental measurements of photoionization dynamics as test for the theory are reported for the photoelectron asymmetry parameters of outer valence ionizations of OsO4, measured in the 17–90 eV photon energy range. Major upgrades of the ARPES-TPES end station, namely a fast 2D-PositionSensitive electron detector and a novel asymmetric fringing field corrector, made the spectrometer, which was specifically designed to study highly reactive and chemically aggressive gaseous species, very efficient to investigate photoionization observables of OsO4.
The valence electronic configuration of OsO4  and RuO4 is as follows: (1a1 )2(1t2 )6(1 e)4(2t2 )6(2a1 )2(3t2 )6(1t1 )6. The valence PE spectrum of OsO4  recorded in this work at hν = 40 eV and θ = 54.7, the “magic angle”, is shown in Fig 1. The first band, A, arises from the ionization from the 1t1 molecular orbitals (MOs). Bands B and C, separated by 0.4 eV in OsO4, are assigned to the spin–orbit components associated with ionization from the 3t2  MOs with strong contribution of O 2p atomic orbitals (AOs), namely the 2T2  ion state. The fourth band, D, is assigned to the 2A1  ion state, while the fifth band, E, encompasses the remaining outer valence MO ionizations (2E and 2T2  ion states). A main contribution to the broadening of band D is associated to the breakdown of the one-particle picture.       Measured photoionization cross-section branching ratios for the first five PE bands of OsO4 are reported in the upper panel of Fig. 2, where DFT and TDDFT results are shown together with the experimental data. Apart from the description of the autoionization resonances, which is given only by the TDDFT method, whose predicted magnitudes appear strongly damped in the experiment, the most clear case where TDDFT is needed for a quantitative agreement with the experimental cross sections is for the ionizations of the d metal-based MOs 1e and 2t2  in the 50–100 eV photon energy range. Compared to the other outer valence ion states, the partial cross section for the ionization of the 2a1  state shows a distinctly different behavior for both the magnitude as well as for the presence of a maximum at around 45 eV in the DFT profile. We ascribe this resonant enhancement to the occurrence of a shape resonance in the t2  continua. The experimental β parameters for the outer valence ionizations of OsO4 obtained in this work are shown in the lower panel of Fig. 2 together with the calculated theoretical values. The agreement between the DFT/TDDFT estimates and the experimental data is overall very satisfactory for all PE bands. Comparison of our theoretical results with those based on a plane waves representation of the outgoing electron, also reported in Fig. 2, demonstrate that the plane waves method provides a poor description of the photoionization dynamics of OsO4. Cooper minima, which are present in the ionization from Os 5d and Ru 4d AOs, do not affect ionizations from MOs that are involved in the formation of the covalent M–O bonds and have substantial metal d orbital character. A Cooper minimum is a purely atomic effect and occurs in the ionization from AOs with radial nodes. Due to the presence of such minimum in the atomic d orbital ionizations, the cross section of this AO exhibits in the low energy region a steeper decrease tha nthat shown by AOs without radial nodes, e.g., the O 2p orbitals. The computed cross section of atomic d orbitals shows, in the low energy region, a decrease larger than that of the TDDFT cross sections of MOs with pronounced d metal character, namely 1e and 2t2, caused by the significant O 2p AO’s contributions  to the MO character (Gelius’ model). The cross section for the 1e orbital ionizations is characterized, in both OsO4  and RuO4, by a somewhat larger decrease than that corresponding to the 2t2  MOs. This is due to the larger metal 5d (4d) contribution to the 1e MOs. The absence of a clear minimum in the TDDFT cross sections is, however, a clear indication of the metal– oxygen mixed character of the 1e and 2t2  MOs due to the covalent M–O bonds formation. Overall, the photoionization dynamics of the two complexes displays signatures of both single particle and manybody effects. Purely single particle effects include the presence of shape resonances in selected ionization channels (2a1  orbital ionization). Many body effects include the super Coster– Kronig decay of np → nd giant resonances that profoundly affect the ionization dynamics around 60 eV photon energy and whose effects are predicted to be stronger in RuO4  compared to OsO4. The occurrence of a Cooper minimum in ionization from MOs with d metal character has been ruled out by our theoretical predictions, highlighting the strong hybridization of the d metal orbitals with ligand orbitals  

Retrive article


  Photoionization dynamics of tetraoxo complexes OsO4 and RuO4.
Luca Schio, Michele Alagia, Daniele Toffoli*, Piero Decleva, Robert Richter, Oliver Schalk, Richard D. Thomas, Melanie Mucke, Federico Salvador, Paolo Bertoch, Davide Benedetti, Carlo Dri, Giuseppe Cautero, Rudi Sergo, Luigi Stebel, Davide Vivoda, and Stefano Stranges Inorg. Chem. 2020, 59, 10, 7274–7282  doi: 10.1021/acs.inorgchem.0c00683






Read More
slideshow image

Quantum effects in hydrogen bonding: Core level photoemission spectroscopy of acetylacetone

The core level photoemission spectroscopy of gaseous acetylacetone, its fully deuterated form, and two derivatives, benzoylacetone and dibenzoylmethane has allowed us to examine the effect of the double-well potential on photoemission spectra. V. Feyer  et al. J. Phys. Chem. Lett. (2018)


The molecule acetylacetone, and two of its derivatives, benzoylacetone and dibenzoylmethane show intramolecular hydrogen bonds, with a proton located in a double well potential, whose barrier height is different for the three compounds. Two distinct O 1s core hole peaks were observed, previously assigned to two chemical states of oxygen. Our alternative assignment of the double peaks takes account of the extended nature of the proton wave function, and the shape of the neutral and ionic potentials in which the proton is located. The peaks are explaineed in terms of an unusual Franck-Condon factor distribution.   Siegbahn won the Nobel prize for Electron Spectroscopy for Chemical Analysis (ESCA), in which the shifts in the binding energies of core electrons measured by photoemission are associated with differences in the chemical environment of the neutral ground state elements. Photoemission is a powerful spectroscopy in wide use at synchrotrons. For many years it has been known that there are a few exceptions to this interpretation, such as satellite peaks, where the shifts have other origins. Are there any other exceptions? Experiments performed at the Gas Phase beamline at Elettra have shown that there are indeed some other effects to be accounted for. The molecule acetylacetone is generally believed to have the structure A in the inset of Fig. 1, in which a hydrogen atom is bonded mostly to oxygen atom 1 or 2. This model pre-supposes that the nucleus of the hydrogen atom, a proton, is a point like object. A few “heretics” have suggested that structure B is more appropriate, with the hydrogen atom located more or less equidistant from the two oxygen atoms. The oxygen core level spectrum, Fig. 1, appears to support structure A: there are two distinct peaks, interpreted as due to oxygen bonded, or not bonded, to hydrogen. The spectrum is slightly sharper when deuterium is substituted for hydrogen as expected: the vibrations of deuterium are at lower energy than those of hydrogen, and the width of the peak is due to vibrations. Incidentally, there is a weak satellite peak at higher binding energy.       Is this the end of the story? Our calculations show that it is not, and an alternative interpretation of the spectrum is possible. Protons are not classical point-like charges, but are quantum mechanical objects with an extended wave function, Fig. 2. The green curve is the potential energy in which the proton moves. Its wave function is delocalised, with two extrema, showing the proton is not located mostly close to one oxygen atom or the other: it has two “most likely” locations, near to one or other oxygen atom, but it is also likely to be found in the middle. Our new interpretation of the O 1s spectrum is that the proton is delocalised in the ground state, see the lower dotted lines in Fig. 2. The wave function of the proton extends over nearly an Ångström. Upon ionization, it localises close to or far from the ionized oxygen atom, for example in Fig. 2 the proton is localised far from the ionized atom. The upper blue curve shows the potential, and the energy is lower near the neutral oxygen atom on the left. The corresponding wave functions of the vibrational states, upper dotted curves, are now localised to one or two tenths of an Ångström.
  This situation corresponds to the lower energy photoemission peak, whereas if the proton is localised close to the positively charged oxygen ion, the energy is higher and this corresponds to the high energy photoemission peak. This interpretation of core level spectra may apply to all systems with hydrogen bonding in a strong, double-well potential. The present case is for an intramolecular double well, but in nature there are numerous situations with both intra- and intermolecular hydrogen bonding, especially in biology and wet chemistry. Our studies provide a fundamental insight into the nature of this bonding, and the interpretation of spectra.  

Retrive article


Quantum effects in hydrogen bonding: Core level photoemission spectroscopy of acetylacetone   Quantum effects in hydrogen bonding: Core level photoemissione spectroscopy of acetylacetone V. Feyer, K.C. Prince, M. Coreno, S. Melandri, A. Maris, L. Evangelisti, W. Caminati, B.M. Giuliano, H.G. Kjaergaard , V. Carravetta   V. Feyer et al., J. Phys. Chem. Lett. 9, 521 (2018). DOI: 10.1021/acs.jpclett.7b03175  






Read More
slideshow image

The electronic structure and the nature of the P-O bond in PPT

The near-edge X-ray absorption fine structure (NEXAFS) and X-ray photoelectron (XP) spectra of gas-phase 2,8-bis- (diphenylphosphoryl)dibenzo[b,d]thiophene (PPT) and triphenylphosphine oxide (TPPO) have been measured at the S and P LII,III-edge regions. E. Bernes et al. J. Phys. Chem. C(2020) 

During the past decade, numerous advances have characterized the field of organic semiconductor devices mainly related to the synthetic versatility of organic materials, which can be designed with tuned properties, including emission energy, charge transport, and morphological stability. In particular, significant efforts have been made on solid-state lighting applications involving phosphorescent OLEDs (PhOLEDs) for their potential as full-color displays. Taking into account the OLED architectures, homojunction devices represent the commercially most attractive alternative to the traditional heterojunction systems due to their simplicity and ease of processing. In contrast to the heterojunction structures, where different layers of materials with specific properties allow the charge transport and recombination, the homojunction organic devices (Fig. 1a) are based on organic molecules with the multiple roles of hole/electron transport and light emission. This usually requires an ambipolar molecular film, allowing high and balanced mobility of both holes and electrons. To accomplish this, the ambipolar organic material is typically composed of three basic building blocks within the same system: a hole transporting molecule (donor), an electron transporting system (acceptor), and a polycyclic aromatic moiety used as spacer (Fig. 1b). A promising class of ambipolar materials in blue PhOLEDs is represented by derivatives of dibenzothiophene substituted with diphenylphosphine oxide, such as PPT (2,8-bis(diphenylphosphoryl)- dibenzo[b,d]thiophene). PPT is an ambipolar phosphorescent electrontransporting material, with sky-blue emission, high emission efficiency, and characterized by a wide bandgap with high triplet energy level. The PPT structure (Fig. 1c) can be rationalized by the presence of an electron-rich dibenzothiophene core (DBT, Fig. 1d), functionalized by two electronwithdrawing phosphine oxide groups, modeled in the study by the triphenylphosphine oxide (TPPO, Fig. 1e) building block. The coexistence of these two counterparts ensures good electronand hole-transporting properties of the PPT, thus maintaining charge balance in the emissive layer of PhOLEDs.
  Despite these recent developments toward applications, the detailed understanding of the complex electronic processes involved is still lacking. To fill this gap, advantage can be taken from the characterization of the electronic structure of the material, as provided by core−electron spectroscopies such as XPS (X-ray Photoelectron Spectroscopy) and NEXAFS (Near-Edge X-ray Absorption Fine Structure).

Within this context, we performed a joint experimental and theoretical investigation of the electronic structure of gas phase PPT and TPPO through XPS and NEXAFS at the S and P LII, LIII- edge. The experimental results have been rationalized by relativistic time dependent density functional theory (TDDFT), which allows the inclusion of the coupling between 1h-1p excited configurations from the 2p degenerate core holes and gives a good account of the relativistic effects (mainly spin− orbit coupling) which are necessary to describe the transitions converging to the LII and LIII-edges. The calculation of the S 2p and P 2p XP spectra allowed us to analyze the binding energies (BEs) both in terms of the spin−orbit splitting of the 2p coreholes and of the molecular-field splitting of the 2p3/2 levels. A comparison of S 2p and P 2p XP spectra with the two building blocks of PPT (respectively shown in Fig. 2a, b), reveals that both splittings are substantially conserved. The small increase of the S 2p experimental BEs going from DBT to PPT (Fig. 2a) is a consequence of the decreased shielding of the electronic charge density on sulfur due to the addition of two electron-withdrawing phosphine oxide moieties in PPT. The small decrease of the P LII and LIII experimental BEs observed from TPPO to PPT (Fig. 2b) can be instead rationalized by the replacement of a single phenyl ring of TPPO with one condensed ring of the DBT moiety in PPT. Furthermore, the TDDFT results are accurate enough to provide an unambiguous assignment of all absorption bands that characterize the below threshold region of the S 2p and P 2p NEXAFS spectra (Fig. 2c, d). They display strong similarities with those of the PPT building blocks, DBT and TPPO, in line with the similar local environment of the S and P atoms, being little affected by the increased molecular complexity of PPT. This work sheds light into the structure of the P-O bond in PPT, a highly debated topic in the chemical literature and brings ahead our investigation of the electronic structure of ambipolar molecules and their building blocks by means of photoionization techniques. In this respect, while in our previous works the analysis of the O K-edge regions allows a more straightforward mapping of the O 2p molecular orbitals (MOs), the assignment of the NEXAFS P LII, III- edge features requires a higher level of computation: by including the coupling between different excitation channels arising from the 2p degenerate coreholes and relativistic spin-orbit coupling effects, we obtain a quantitative description on the higher-lying localized σ*(P−O) virtual MOs. The results here presented indicate: (a) that P 3d atomic orbitals (AOs) are not involved in the formation of the P−O bond, and (b) the energy ordering of P 2p transitions to σ*(P−O) and π*(P−O) virtual states are compatible with the traditional view of the P−O bond formation through a mechanism of negative hyperconjugation. A further study at the NEXAFS P K-edge would be useful to evaluate in detail the weight of the P 2p AO contributions to the P−O bond and will be the subject of future works.

Retrive article

S 2p and P 2p Core Level Spectroscopy of PPT Ambipolar Material and Its Building Block MoietiesE. Bernes, G. Fronzoni, M. Stener, A. Guarnaccio,* T. Zhang, C. Grazioli, F. O. L. Johansson, M. Coreno, M. de Simone, C. Puglia, and D. Toffoli J. Phys. Chem. C 2020, 124, 27, 14510–14520; doi: 10.1021/acs.jpcc.0c03973

Read More


GAPH is one of the possible CERIC instruments:
CERIC-ERIC, the European Distributed Research Infrastructure, offers more than 40 different and complementary state of art techniques, distributed in 7 different coutries. A detailed description of also other facilities available in CERIC can be found here.
CERIC allows a wide range of methods for sample preparation and analysis in a flexible and project oriented way, through the submission of multitechnique proposals through a single entry point.


User Area

Proposal Sumission

We invite users and collaborators to discuss their proposals with the beamline local contacts well in advance before the submission deadline. This is crucial for a careful assesment of the experiment feasibility and may lead to improvements in the proposed experimental plan. Our website provides a wealth of informaiton on experiment feasibilty and proposal submission. For more info, please vist the user info section.

Call for proposals

Next deadline for proposal submission at link:  Elettra_DL  and : CERIC_DL 

 All proposals requiring use of the laser will be performed at the branch line.

Last Updated on Wednesday, 07 September 2022 15:41