Search

Nous présentons ici une méthode originale et sécurisée pour l'élaboration en une seule étape de films nanostructurés composés de nanoparticules enrobées dans une matrice. Ce procédé polyvalent combine sous vide la technologie des jets de nanoparticules avec la pulvérisation magnétron classique. Le principal avantage de la technique est la possibilité de choisir indépendamment la nature chimique des nanoparticules et celle de la matrice. La capacité à réaliser la synthèse sur de grandes surfaces est fournie par l'utilisation d'une série de plusieurs lentilles aérodynamiques placées en parallèle entre la source des nanoparticules et la chambre de dépôt sur une machine prototype., We present here an original safe-by-design method for the elaboration of nanostructured films composed of nanoparticles embedded in a matrix. This versatile single step process operates under vacuum by combining the jet of nanoparticles technology with classical magnetron sputtering. The main advantage of the technique is based on the possibility to choose independently the chemical nature of nanoparticles and matrix. The ability to achieve synthesis on large surfaces is provided by the use of a series of several aerodynamic lenses implemented on a prototype set up between the source of nanoparticles and the deposition chamber.

Le Groupe des Applications Plasma (GAP) du CEA a Saclay participe au projet national PREUVE du Reseau Micro- et Nano-Technologies. Ce projet a ete lance fin 1999 pour reunir et developper les competences en France sur la lithographie dans l'extreme ultraviolet (LEUV). Au sein de PREUVE, notre objectif a ete le developpement d'une source plasma laser dans l'EUV autour de 13nm afin de contribuer a la realisation d'un premier banc d'essai pour la lithographie (BEL) en Europe. Afin de realiser cette source, nous utilisons un plasma emetteur qui est produit par l'interaction d'un laser de type Nd :YAG sur un jet de gouttelettes de xenon. A la fin du projet PREUVE, cette source satisfait les principales specifications et repond en particulier aux besoins en flux de photons EUV pour realiser des tests d'insolation EUV avec le banc d'essai. Suite a ces resultats prometteurs, nous demarrons actuellement un projet industriel EXULITE avec nos partenaires du CEA, d'Alcatel et de Thales sur le developpement d'une source EUV de puissance pour des machines de lithographie de production. Ce projet se terminera en 2005.

We exposed small size-controlled lead clusters with a few hundreds of atoms to laser pulses with peak intensities up to 1015 W cm-2 and durations between 60 fs to 2.5 ps. We measured kinetic energies and ionic charge of fragments as a function of the laser intensity and pulse duration. Highly charged Pbn+ ions up to n = 26 have been detected presenting kinetic energies up to 15 keV. For comparison with our experimental results, we have performed simulations of the laser coupling with a cluster-sized lead nanoplasma using a qualitative model that was initially proposed by Ditmire and co-workers at LLNL for the case of rare gas clusters. From these simulations we conclude that two mechanisms are responsible for the explosion dynamics of small lead clusters. As already observed for large rare gas clusters (n = 106), fragments with charge states below +10 are driven by Coulomb forces, whereas the higher charged fragments are accelerated by hydrodynamic forces. The latter mechanism is a direct consequence of the strong laser heating of the electron cloud in the nanoplasma arising from a plasmon-like resonance occurring at n e = 3n c. In order to obtain an optimized laser-nanoplasma coupling, our results suggest that the plasma resonance should occur at the peak intensity of the laser pulse. Due to inertial effects, even for such small-sized clusters, the observed optimum pulse duration is in the order of 1 ps which is in good agreement with our theoretical results.

Within PREUVE, our group developed an EUV source that will meet α-tool specifications in 2002. With this aim, a laser-produced plasma source was built that uses a dense and confined xenon jet target. This device assures high conversion efficiency and low debris flux which are necessary for future extreme ultraviolet lithography. Our technical solution is based on a specific target injector design and the use of well adapted nozzle materials. After injection, the xenon is recycled and purified to reach a low cost of operation. Our ELSA source will allow 2 years full operation of the BEL lithography test bench.

A liquid jet of either nitrogen or argon of 20 μm diameter was exposed to intense laser fields with pulse durations between 70 fs and 250 ps, leading to intensities of 1016 W cm-2 and 1013 W cm-2, respectively. The emission of extreme UV light and soft X-rays shows the characteristic lines of hydrogen-like nitrogen and carbon-like argon. For nitrogen the emitted photon flux at 250 ps was about two orders of magnitude higher than for 70 fs pulses. A weak dependence on the laser polarization with respect to the liquid jet axis was found. The kinetic energy of the emitted ions easily exceeded 100 keV for nitrogen and 200 keV for argon for a pulse duration close to 2 ps.

Rare gas or lead clusters of several hundreds to many thousands of atoms per cluster were irradiated by intense laser pulses producing peak intensities up to 1017W/cm2. Fast, highly charged fragment ions were observed by means of standard TOF mass spectrometry as well as a magnetic deflection TOF device (MDTOF). Charge states of up to Xe30+ or Pb18+ were detected with maximum kinetic energies reaching far beyond 100 keV for Xe. A strong dependence of the charging and heating process on the laser pulse duration was found. While the resulting ionic fragment spectra are qualitatively similar for rare gas and lead clusters, the kinetic energies for the metal clusters seem to be significantly lower.

Photofragmentation of Fe(H2O)n + clusters (n=1-9) is investigated at three different wavelengths, 532, 355 and 266 nm. Two fragmentation pathways are observed depending essentially on the photon energy, but also on the parent size n. The fragmentation products belong to two ion families, Fe(H2O)m + and FeOH(H2O)m + , which correspond to dehydration and intracluster dehydrogenation reactions respectively. The ion yields are studied as a function of the laser fluence in order to determine the number of photons implied in the photofragmentation process. This allows us to estimate that the D[(H2O)n-1Fe+-(H2O)] bond energy is ranging between 0.44 eV and 0.55 eV for \(\).Photon absorption cross sections are also derived from the fluence experiments, and two different behaviors are observed: i) At 355 nm, far away from any \(\) transition, progressive solvation of the metal ion results in an increasing absorption cross section from n=2 to n=9. This can be attributed to a forbidden transition of bare \(\), which becomes progressively allowed because of the interaction with more and more water ligands. ii) At 266 nm, close to several allowed transitions of bare \(\), a distinct maximum is observed for the absorption of \(\) ion. It may be attributed to a change in the spin multiplicity when switching from \(\) and \(\) on one hand to Fe(H2O)\(\) on the other.

International audience; The study of chemical reactions in clusters is a rapidly growing field that is attractive for investigating medium effects in reaction dynamics. Cluster isolated chemical reaction (CICR) is a new direction developed in our laboratory, which enables quantitative studies to be made in that field. CICR experiments consist in depositing a controlled number of reactants on large van der Waals clusters that play the role of a solvent of fairly well known size, structure and temperature. This offers enormous advantages, both on the experimental side and for further theoretical investigations. The present review intends to draw together informations that are pertinent for developing experiments and concepts towards understanding chemical reactions in clusters, through CICR-type experiments. In particular, it reviews the questions first of generating and controlling the properties of large van der Waals clusters and secondly of attaching reactants to clusters and controlling their number and location in the cluster. Besides reviewing thoroughly the CICR work done so far in our laboratory, the present paper also reviews part of the vast literature pertaining to reaction dynamics in large van der Waals clusters.

Barium dimers have been produced and stabilized by the successive pickup of two barium atoms on large argon clusters (average size of 8000). By adding chlorine and oxygen molecules, chemiluminescent reactions of these dimers have been observed. This technique of dimer formation can be extended to higher order clusters to study reactions of cold metallic clusters.

Nous avons etudie la chimiluminescence produite par la reaction du baryum avec le chlore moleculaire sur un agregat de 8000 atomes d'argon. L'intermediaire reactionnel (BaCl 2 ) est le principal produit chimiluminescent observe lors de la collision binaire (Ba + Cl 2 ) sur l'agregat. Au contraire, en phase gazeuse, en regime de collision unique, la reaction conduit exclusivement a la paire radicalaire BaCl* + Cl. On interprete ce resultat comme un piegeage efficace de l'intermediaire reactionnel (BaCl 2 ) par l'agregat. On en discute les mecanismes

Barium atoms at the surface of large argon clusters are excited into the 6s6p1P1 state leading either to ejection and subsequent fluorescence of the free barium atom or to fluorescence of solvated barium. Relative probabilities for these two processes as a function of the excitation energy and of the cluster size are presented. By comparison with the binding energy from a molecular dynamics (MD) calculation, we find that the barium atom is not excited above the limit for direct dissociation. The nonzero desorption probability is attributed to the gain of surface energy of the argon cluster which is partially transferred to the barium atom. Using linearly polarized laser light for the excitation, a positive polarization of the fluorescence of the ejected barium atoms is observed, whereas the polarization fluorescence of the solvated barium is negative. This indicates that the solvated barium atoms undergo a relaxation process which rotates the orientation of the excited orbital before the fluorescence. The...

The effects of adding molecules on the LIF at 540 nm of a barium atom at the surface of an argon cluster (average size 420) has been investigated. We showed that molecules like ethanol,n-hexane and O2 from stable complexes with ground state barium. In the case of molecules like N2, CH4 and SF6, the collisional quenching of solvated Ba(1P) is observed. The large quenching rates obtained are interpreted by a surface mobility of the collisional partners. Moreover, we showed that this collisional quenching leads to the ejection of free Ba(3P1).

This work presents a quantitative comparison between experiment and molecular dynamics simulations for the excitation spectra of large van der Waals clusters. The emission and excitation spectra of mixed Ba(Ar)n clusters have been obtained for average cluster sizes ranging between 300 and 4000. The simulation is performed by using classical dynamics and pairwise additive potentials for two cases corresponding to the barium atom at the surface or inside the argon cluster. A very good agreement with the experiment is found when the barium atom is at the surface.

International audience; The present work, performed in the frame of the EXULITE project, was dedicated to the design and characterization of a laser-plasma-produced extreme ultraviolet (EUV) source prototype at 13.5 nm for the next generation lithography. It was conducted in cooperation with two laboratories from CEA, ALCATEL and THALES. One of our approach originalities was the laser scheme modularity. Six Nd:YAG laser beams were focused at the same time on a xenon filament jet to generate the EUV emitting plasma. Multiplexing has important industrial advantages and led to interesting source performances in terms of in-band power, stability and angular emission properties with the filament jet target. A maximum conversion efficiency (CE) value of 0.44% in 2π sr and 2% band-width was measured, which corresponds to a maximum in band EUV mean power of 7.7 W at a repetition rate of 6 kHz. The EUV emission was found to be stable and isotropic in these conditions.

Molecular dynamics calculations have been performed to simulate the low energy collision (0.2 eV) of a rare gas atom (He, Ar, Xe) with a cluster of 125 argon atoms. Depending on its relative mass to argon, the projectile is either deflected (He) or captured (Ar, Xe) by the argon cluster. We have determined the deflection function of the He projectile that is scattered, and for Xe we have determined wether it stays near the surface of the cluster or migrates inside. These results have been discussed in the light of very simple models.

The present paper investigates chemiluminescence in reactions of excited barium atoms (6s6p 1P1) with water and a series of alcohols and ethers. The electronically excited product molecule from the reaction with H2O (and D2O) is BaOH (and BaOD) in the A 2Π, A’ 2Δ, and B 2Σ levels. The product molecule is always Ba–ORx in reactions with alcohols ROH, whatever the size of the alcohol (methanol to butanol), and whatever its class (primary to tertiary). By comparison, no chemiluminescence was observed when the reactant was dimethyl and diallyl ether although allowed energetically. The nature of the product molecules in reaction with alcohols, and the absence of reaction with ethers that were found here are remarkably close to what was found by Davis et al. (submitted to J. Chem. Phys.) for the formation of ground state products in reactions of Ba(6s5d 1D2) with water, methanol, and dimethyl ether. The present work thus allows us to extend the model of Davis et al. for the reactivity of Ba(6s5d 1D2) as a power...

A systematic study of the quenching of Ba(1P1) fluorescence by CH4 in Ar clusters has been performed for different argon cluster sizes and numbers of methane molecules in the clusters. The quenching rate as a function of the number of methane molecules in a given cluster follows a first-order rate law. Quenching rates for different cluster sizes have been deduced from the experimental data. The average time for Ba(1P1) to collide with CH4 and be quenched is 18±5 ns in an Ar1000 cluster. The quenching time varies with the cluster size as n 2 3 .

The spin-changing (6s)1(6p)1, 1P1→(6s)1(6p)1, 3P2 transition of barium induced by single collisions with H2 has been studied using a crossed beam apparatus. An mj-selective preparation and detection of barium was achieved using the pump and probe technique in the presence of a magnetic field that breaks the degeneracy of the barium magnetic sublevels. The probe laser was scanned across the (6s)1(6p)1, 3P2,mj′→(6s)1(7s)1, 3S1, mj″ multiplicity. Analysis of the Doppler profiles of each line reveals the angular distribution of scattered barium atoms, the polarization of which is selected both before and after the collision. This is therefore a four-vector correlation experiment, and it is the first to document spin-changing transitions of alkaline-earth-metal atoms induced by a molecular target. It must be emphasized that all four vectors in the present experiment do not refer to the same quantization axis, whereas in the standard notion of n-vector correlation experiments, the n-vectors refer to the same axis. The results of the present work are discussed in the light shed by other results obtained on the same experiments with only two and three vectors controlled. A collision mechanism is proposed that accounts for all the existing experimental results on this process.

Zeeman cross sections [sigma][sub [ital j],][ital m[ital j]][r arrow][ital j][prime],[ital m][sub [ital j]][sup [prime]] for the spin-changing 6[ital s]6[ital p] [sup 1][ital P][sub 1]--6[ital s]6[ital p] [sup 3][ital P][sub 2] transition of barium induced by single collisions with N[sub 2], O[sub 2], and H[sub 2] have been measured using a crossed-beam apparatus. State-selective preparation of 6[ital s]6[ital p] [sup 1][ital P][sub 1,][ital m[ital j]] and detection of 6[ital s]6[ital p] [sup 3][ital P][sub 2,][ital m[ital j]][sup [prime]] was achieved using the laser excitation technique in the presence of a magnetic field that breaks the degeneracy of the magnetic sublevels. The results showed that the Zeeman cross sections [sigma]([sup 1][ital P][sub 1,][ital m[ital j]][r arrow][sup 3][ital P][sub 2,][ital m[ital j]][sup [prime]]) are fairly dependent upon [ital m][sub [ital j]] and [ital m][sub [ital j]][sup [prime]] for N[sub 2] and H[sub 2], whereas the Zeeman cross sections all have the same value with O[sub 2]. Additionally the flux-velocity contour map of Ba(6[ital s]6[ital p] [sup 3][ital P][sub 2]) for scattering with H[sub 2] depends significantly upon the magnetic sublevel of barium in the exit channel. The results for Ba-(N[sub 2],H[sub 2]) scattering were accounted for semiquantitatively by a very simple model. Themore » model was then used to stress which physical phenomena are at the origin of the preservation of the orbital alignment and orientation during collisions. The present work also compared the kind of information obtained by the direct measurement of Zeeman cross sections to that found in other experiments that deal with the effect of polarizing the collision partners.« less

International audience; Luminescent silicon nanoparticles particles produced by laser pyrolysis are considered as possible alternative to replace toxic Quantum Dot in bioimaging applications. However, these nanoparticles are fully oxidized when kept in water, therefore, the luminescent silicon core must be be protected from oxidation. The Si nanoparticles were embedded in monodisperse silica beads (~50 nm) produced in microemulsion. The silica beads provide protection of the silicon core and allow stability of the photoluminescence over time. They are well dispersed in water and biological medium with a colloidal stability of several days.

Doppler measurements taken over a range of probe‐laser angles in a crossed‐beam experiment were used, in conjunction with forward convolution analysis, to obtain flux–velocity contour maps for Ba(3P2) produced in a collision of Ba(1P1) with H2, N2, O2, and NO. The contour maps suggest a general model for the dynamics of this process in which large impact parameter collisions result in a near‐resonant transfer of initial electronic energy into final vibrational energy, while close collisions produce sideways scattering and effectively couple electronic energy to translation. The molecular collision partners fall into two categories: for one group, comprising O2 and NO, the existence of a well‐defined molecular anion with favorable Franck–Condon factors linking excited vibrational levels to the ground vibrational state of the neutral results in greatly enhanced coupling for the near‐resonant process. Molecules for which there exist no stable anions, such as N2 and H2, represent a second category. The electronically inelastic collision for this group is instead dominated by the nonresonant process yielding the ground vibrational state and large translational energy release.

The absorption and emission of barium atoms trapped onto large argon clusters (about 2000 atoms) have been observed. The emission consists in a single peak to the red of the barium resonance line and the absorption of two bands, one on each side of the resonance line. The existence of two absorptions seems characteristic of a configuration where the barium atom remains at the surface of the argon cluster.

Chemiluminescence is observed when barium atoms collide with N2O(Ar)n clusters in a crossed-beam experiment (200

Flux-velocity contour maps were obtained for the inelastic collision process Ba({sup 1}{ital P}{sub 1})+O{sub 2}N{sub 2}{r arrow}Ba({sup 3}{ital P}{sub 2})+O{sub 2}N{sub 2} from Doppler scans of scattered Ba({sup 3}{ital P}{sub 2}) taken over a range of probe laser directions in a crossed-beam experiment. Collision with O{sub 2} resulted in sharply forward scattered Ba({sup 3}{ital P}{sub 2}), with efficient conversion of inital electronic energy into O{sub 2} internal energy and little momentum transfer. Collision with N{sub 2} was dominated by wide-angle scattering with most of the available electronic energy appearing in product translation. The results suggest the importance of large-impact-parameter collisions and a near-resonant energy transfer in the case of O{sub 2}, while for N{sub 2} close collisions dominate despite the presence of an analogous near-resonant channel. The results represent the first direct experimental demonstration of a near-resonant quenching process.

A method based upon the tandem use of the Time-Of-Flight and Surface-Induced-Dissociation techniques is proposed for estimating the average cluster size in a neutral molecular beam. It consists of sending the beam through a buffer gas and measuring the variations of the average beam velocity as a function of the buffer gas pressure. The clusters are detected at the mass of the monomer by surface induced dissociation in the ionization source. This method has been applied to an argon cluster beam and the results are in good agreement with determinations using high energy electron diffraction. This technique appears to be a simple alternative for estimating mean cluster sizes in the range of 100 to a few 1000 monomers.

The reaction of Ba(6s 6p1P1) with O2 has been investigated in a crossed beam experiment as a function of the collision energy. The electronically excited product BaO is not statistically populated and the difference from statistics increases with the collision energy. Such a result is markedly in contrast with what was observed in the reactions of Ba(6s21S0) and Ba(6s5d1D2) with O2 where the existence of a long‐lived complex results in a statistical distribution of internal energy of the BaO product. The present result is interpreted as due to the collisional system following a different channel involving excited electronic states of the intermediate, BaO2 above the ionization limit, instead of the strongly bound states of barium peroxide which are involved in reactions of less excited Ba states.

A laser-spectroscopic investigation of the (000-000) band of the B 2 Σ + - X 2 Σ + transition of the BaOH molecule has been performed. The analysis shows that the upper level is badly perturbed, presumably by a Σ vibronic level of the A 2 Π 3 2 state. A deperturbation calculation is presented, giving molecular constants for both the interacting levels. However, due to the probable existence of additional perturbations, as yet unobserved, it is found that different sets of rotational constants are needed for the e and f levels of the B 2 Σ + (000) state. Furthermore, studies of A 2 Π - X 2 Σ + emission, following B state excitation, reveal that the dominating B - A relaxation mechanism is perturbative and collisional rather than radiative, as had been proposed earlier.

The laser system is by far the most expensive part of a laser-produced plasma (LPP) EUV source. Thus LPP source designers have not only to optimize the technical performances of the lasers but also cost, efficiency, reliability, redundancy and industrial availability of their approach. Based on the typical conversion efficiencies of present LPP source targets (CE ~ 1- 3%), the laser system will have to deliver an average power of at least 20 kW. This can hardly be obtained with only few laser chains. Starting from this consideration, the EXULITE consortium has first proposed the concept of a spatially multiplexed and modular LPP source using 10 to 20 identical, moderate power laser units which are simultaneously focused on the same spot. This approach increases system redundancy and reduces the development cost of the laser. We will present an original technical approach which allows both to manage the multi-beam attack of the target using a closed-loop control and to collect a maximum of EUV power. The grazing incidence EUV collector is designed such that optimized vacuum pumping and debris mitigation can be obtained. Finally, we can show that modular LPP sources can satisfy the severe EUV dose stability that is commonly requested.

It was shown in 1999 that D2 cluster explosion under ultrashort intense laser irradiation can lead to ion energies sufficient to drive nuclear fusion reactions [Ditmire et al. Nature 398, 491]. We show how the use of molecular clusters allows to further enhance the ionic acceleration, allowing one to reach optimal fusion cross sections. A new low density high energy regime is described, in which fusion occurs via a spallation-like process. The process increases with laser intensity up to the relativistic threshold, at which the neutron production is overwhelmed by gamma production resulting from electron acceleration.

Laser light scattering, associated with molecular beam flux measurements, provide an easy to perform, sensitive, non-destructive technique for the determination of relative mean sizes of molecular clusters in molecular beams formed with different expansion conditions. The results are compared to an absolute method used in the authors' laboratory and show good agreement in the range investigated (n approximately=150-4000).

International audience; Following the interaction of superintense, short pulse lasers and plasmas, ions can be accelerated to velocities sufficient to drive nuclear fusion reactions, in particular, by the process of Coulomb explosion of clusters [T. Ditmire et al., Nature (London) 398, 491 (1999)]. We show here how short bursts of neutrons can be produced using a jet of low-density deuterated methane clusters. Ion velocity distributions were simultaneously measured by a Thomson parabola mass spectrometer, demonstrating deuteron energies up to 120 keV. We show that, in such conditions, nuclear fusion will occur not only in the hot plasma core, but also in the cold outer region by collision processes.

International audience; The Doppler profiles of Ba($^3P_2$) atoms desorbed from the surface of argon clusters following the deactivation of Ba($^1P_1$) have been measured. These measurements have been performed for desorption from pure Ar$_N$ clusters and as a function of a known average number of CH$_4$ molecules deposited on the cluster. Analysis of the profile widths with respect to the kinetic energy release from deactivation indicates that desorption occurs along a single Ba–Ar and Ba–CH$_4$ coordinate in the former and latter cases, respectively. By comparing the kinetic energy distributions in the desorbed barium with the relative kinetic energy available at the temperature of the cluster it is found that the collisions leading to deactivation in both cases are gas kinetic at the temperature of the cluster (35 K). The residual anisotropies in the Doppler profiles reveal the Ba–Ar deactivation to be a relatively inefficient process allowing the barium to undertake a full migration on the cluster surface before desorbing. This results in an essentially isotropic distribution of recoil velocities. In contrast Ba–CH$_4$ deactivation is sufficiently fast to preserve some degree of anisotropy in the desorbed barium velocity distribution. The anisotropy results from the depolarization of the barium orbital due to both the migration of the barium on the cluster surface and axial relaxation of the orbital by collisions with neighboring argon atoms. Calculations of the anisotropies resulting from both reorientating mechanisms show a significant degree of relaxation and migration to occur before the barium is desorbed.

Large van der Waals clusters are used as microreactors to study collision processes at the contact of a solvent. A barium atom and different molecules are deposited on the surface of an argon cluster through the pick‐up technique. The location of the barium with respect to the cluster is determined through the spectroscopy. By analyzing the changes in the barium fluorescence when adding molecules on the cluster, the collision rate can be determined and consequently the relative mobility of the barium‐molecule pair at the cluster surface.

For the experimental study of solution phase chemistry one would wish for a “perfect” system to be able to investigate. That is, a solution of known size, structure, and internal energy, with a knowledge of interaction energies of the solute and solvent. The need for such “ideal experiments” is made very acute by recent molecular dynamics calculations that open the route for modelling fundamental aspects of reaction dynamics in solution [1, 2].

The filling of solvent shells around the proton was investigated by the method of cluster ionization and fragmentation. Hydrogen-bonded clusters of H{sub 2}O, CH{sub 3}CN, C{sub 2}H{sub 5}OH, and mixtures of these molecules were studied. Well-defined first shells corresponding to (CH{sub 3}CN){sub 2}H{sup +}, (CH{sub 3}CN){sub 3}(H{sub 2}O)H{sup +}, and (H{sub 2}O){sub 4}H{sup +} were observed. Most significantly, only one shell was found in clusters of pure CH{sub 3}CN, whereas a structure corresponding to (CH{sub 3}CN){sub n}(H{sub 2}O){sub n-2}H{sup +} was observed up to n = 7 in multicomponent clusters of water and acetonitrile. These results illustrate how results derived from small clusters in high-pressure mass spectrometric experiments may be extended to large clusters.

We present an experiment adapted to collisional studies of cluster ions based on a laser vaporization setup coupled to a supersonic expansion. The ions are selected in a first time-of-flight, slowed down to the desired energy, and collided in an octopolar guide. The parent and fragment ions are then reaccelerated in order to be mass analyzed in a reflectron time-of-flight. An original method for the extraction of the ion that uses a double voltage pulse, is proposed. The experiment has been applied to collisions of hydrated cobalt ions. An absolute cross section of 17 A2 for the loss of one water molecule by Co(H2O) 2 + in collision with neon at a center-of-mass energy of 10 eV, has been determined, with an accuracy of 10%. The threshold for this reaction has been measured at 1.5 eV and is in good agreement with the existing literature (Dalleska et al. J. Am. Chem. Soc. 1994, 116, 3519). Ions that cannot be formed by conventional ligand exchange methods, can also be studied. As an example, the threshold for dehydration of the Co2(H2O)+ ion has been measured at 1.5 ± 0.2 eV.