Thin Fe-oxide films (15–150 nm thick) have been prepared using Langmuir–Blodgett (LB) films as precursors consisting of a certain number of Fe–arachidate monolayers (ML). The organic component of the LB films was removed by different heating procedures: heating in ambient air at 360 and 900°C, and at 500°C in a H2 gas flow. The remaining Fe-oxide films were analyzed by a number of methods. The most important results are: all Fe-oxide films are smooth and homogeneous (roughnesses between 0.8 and 4 nm), the magnetic properties are, however, not sufficient for an application as magnetic storage medium. Only a film prepared from 35 ML Fe–arachidate by heating at 500°C in H2 had magnetic properties (hysteresis loop of the single domain type, large coercivity and saturation magnetization) suited to use the film as a high data density magnetic storage medium. The films heated at ambient air consist mainly of α-Fe2O3. In addition they contain most probably γ-Fe2O3 and Fe3O4 (films heated at 360°C) or Fe3O4 and FeO (films heated at 900°C). Films heated in H2 consist of Fe3O4, FeO and metallic α-Fe.

A novel molecular ion source based on spontaneous desorption (SD) processes has been developed. The molecular ions from the SD source are used as incident particles for the production of secondary cluster and molecular ions. The SD source presents attractive features for surface analysis and cluster ion beam production. It is shown that analytical applications can be performed with this source with better efficiency, in term of analysis time, than with fission fragment impacts used in plasma desorption mass spectrometry. Low intensity beams of clusters ions could be produced with good stability for a period of several weeks. © 1998 John Wiley & Sons, Ltd.

It is shown by means of a few examples that MeV cluster ions from small accelerators are able to produce very high energy densities in matter, densities which are not accessible with the heaviest and fastest atomic ions from the big heavy ion facilities. The examples given are the pulse-height-defect in Si-detectors, defect production in mica and HTSC thin films, and cluster induced desorption of secondary ions.

MeV sputtering from CsI and organic samples induced by the polyatomic ions Cn+ (n = 1–12), Agn+ (n = 1–7), and Aun+ (n = 1–5) has been investigated. It is found that the desorption yield Yn increases more than linearly with increasing nonber n of projectile constituents. In case of the Cn+ induced desorption from a CsI sample it could be shown that the nonlinear increase of Yn reflects a linear increase of the energy density produced by the cluster projectile. This is a collective effect resulting from the simultaneous action of all n constituents of the polyatomic ion.

Bi2Sr2Ca1Cu2O8 + δ and YBa2Cu3O7 − δ thin films were irradiated by 23 MeV C60 and 24 MeV Au5 cluster ions from a tandem accelerator. The defect production of the C60 cluster ions in Bi2Sr2Ca1Cu2O8 + δ exceeds that of 238U ions with an energy of 2.7 GeV. The analysis of the mechanism underlying the defect production leads to different conclusions: (i) for C60 clusters we find independence of nuclear and electronic contributions to the defect production as in the case of atomic projectiles, (ii) for Au5 cluster ions this independence has lost it's validity. A possible interpretation of the Au5 results is that nuclear defects are annealed by the electronic energy loss. The size of the effective damage radii obtained for the Bi2Sr2Ca1Cu2O8 + δ film may indicate that continuous tracks are formed in this film via the C60 and Au5 irradiation.

It is shown that polyatomic ions at high energy (MeV) are able to emit large size cluster ions with a high yield when they bombard solid surfaces. Results on sputtering of clusters ejected from inorganic and organic targets as well as sputtering of Cx− clusters from carbon foils are presented. Gold cluster Aun, carbon cluster Cn and C60 in the energy range from 2 to 20 MeV were used as projectiles.

The vacuum stability of Langmuir-Blodgett films consisting of fatty acids and Cd salts of fatty acids deposited on various substrates was investigated by means of plasma desorption mass spectrometry. Plain fatty acid films (particularly thin ones) exhibit a remarkable instability. Exceptions occur when the films are deposited on metal substrates covered with the natural oxide. In this case cation exchange occurs which results in the formation of salt (or a mixture of salt and acid) films exhibiting reasonable stability. Thin films consisting of 100% salt are fairly stable in vacuum.

Cation exchange taking place in Langmuir-Blodgett films of fatty acids deposited on hydrophobic Cu substrates was investigated by plasma desorption and spontaneous desorption mass spectrometry. Complete cation exchange was found in plain arachidic acid monolayers. Incomplete exchange was found in bilayer and multilayer films of Cd arachidate with thicknesses up to 10 monolayers. The fraction of exchanged ions is independent of the number of layers.

A novel molecular ion source based on the spontaneous desorption process has been investigated. The results obtained indicate that it is well suited for the analysis of organic molecules with masses up to 10(3) u in a time-of-flight mass spectrometer.

Yttrium arachidate multilayer films deposited on gold substrates by the usual Langmuir—Blodgett dipping technique have been investigated by means of Rutherford backscattering and plasma desorption mass spectrometry. It turns out that the cation of the film-forming molecules is Y3+. This is in contrast to a previous analysis where the cation was determined to be Y(OH)2+.

Langmuir-Blodgett (LB) films of fatty acid salts (Y-stearate, Y-arachidate, Cu-arachidate, Ba-stearate) were thermally oxidized. As a result one obtains ultrathin (a few Å thick) metal-oxide films at the substrate. The surface of the metal-oxide films was found to be rather inhomogeneous. Y and Cu ions remain quantitatively at the substrate despite the heating procedure. A linear dependence between area density of the metal ions in the oxidized films and the number of monolayers of the LB films was observed. The preparation of a mixed metal-oxide film containing Y, Ba, Cu with a given stoichiometry was found to be difficult due to the effect of counter ion exchange. The samples were investigated by means of plasma-desorption and spontaneous-desorption mass spectrometry, by Rutherford back-scattering and electron microscopy.

Ultrathin ferric oxide films (10(15) Fe ions cm(-2)) were prepared from Langmuir-Blodgett (LB) films of Fe arachidate. The LB films were heated for this purpose to a temperature of 300 degrees C. It turns out that the Fe ions remain quantitatively at the substrate during the heating procedure and that a linear dependence exists between the Fe area density in the ferric oxide film and the number of Fe arachidate monolayers. The homogeneity of the ferric oxide films was found not to be satisfactory if LB films with a small number of monolayers are used. The homogeneity increases with increasing number of monolayers. The analysis of the films was performed with mass spectrometric methods, Rutherford backscattering spectroscopy and electron microscopy.

Gold and C60 cluster ions have been accelerated to MeV energies with the EN tandem accelerator at Erlangen. Negatively charged cluster ions were produced in the sputter source and accelerated through both stages of the accelerator. The identification of the MeV cluster ions was performed by time-of-flight and energy measurements.

The conditions for complete salt formation of an arachidic acid monolayer spread onto an aqueous subphase were investigated by means of plasma desorption mass spectrometry in order to demonstrate that this method is very well suited to characterize Langmuir-Blodgett films. The combinations of subphase pH and metal ion concentration necessary for complete salt formation of arachidic acid monolayers were deduced from this analysis.

It is shown that the grid temperature determines the desorption rate of the Spontaneous Desorption for a given sample voltage. The sample temperature has only a minor effect on the rates. From these observations it is concluded that SD is mainly a two-step process: ions are sputtered from the sample surface by keV ions which originate at the grid via the temperature dependent field enhanced desorption of adsorbates.

The desorption rate for 'spontaneous desorption' (SD) has been investigated using various grids with different geometrical dimensions (line width, line distance). The experimental results can be understood if one assumes that the keV ions which sputter ions from the sample in an SD process originate exclusively at the grid via field-enhanced desorption of adsorbates (FEDA). This means that SD is a two-step process. The first step is FEDA, the second step is a sputter process.

The observation recently made by Zubarev [R.A. Zubarev, P.V. Bonderanko, A.N. Knysh and B.V. Rozynov, Rapid Commun. Mass Spectrom., 5 (1991) 32; 278] that kilovolt electrons cause desorption of thermally labile organic molecules has been confirmed by the present investigations. Time-of-flight mass spectra were obtained for several organic samples and CsI. A comparison with mass spectra obtained from the same samples by means of the plasma desorption mass spectrometry is made. Ion yields were measured as a function of the electron energy. A possible desorption mechanism is proposed.

Gold clusters Aun with n ⩽ 5 and masses m ⩽ 985 u have been accelerated to MeV energies with the tandem accelerators at Orsay and Erlangen. Singly charged negative cluster ions were obtained from a conventional sputter ion source and accelerated between ground potential at the entrance of the accelerator and the terminal (in the middle) at several MV. In a gas channel positioned in the terminal they lose one or several electrons and become neutral or positively charged clusters, respectively. The ions are then accelerated towards the exit of the tandem accelerator. Identification of Au+n clusters is achieved by measuring the time-of-flight of the clusters.

Spontaneous desorption (SD) denotes a process in which atoms and molecules are transferred as singly charged ions from the solid phase into the gas phase. For this purpose the solid sample does not need to be bombarded by particles (keV or MeV ions, electrons) from an external source. An electric field between the sample and an acceleration grid with a field strength at the sample surface below the threshold value for field desorption is the only prerequisite to initiate the SD process. The SD mechanism is the subject of the present paper. A number of experimental results obtained recently are summarized. It follows that SD is a two-step process. Primary ions are field desorbed in a first step from the edges of the grid where an enhanced electric field exists. These ions are mainly molecular ions and stem from adsorbates at the grid. They are accelerated in the electric field between the grid and sample to keV energies and sputter, in a second step, secondary ions from the sample with a sputter yield that exceeds the yield of atomic ions considerably.

It is shown that kiloelectronvolt ion bombardment of a sample frequently causes electrons and/or H− ions to be sputtered simultaneously with a massive negative ion. This result is in accordance with an interpretation of the spontaneous desorption (SD) of negative ions in terms of a multistep process, the last step of which being a sputter process. It is also shown that the correlated emission of a low-mass and a massive positive ion is rather unlikely. This explains why SD spectra of positive ions are difficult to obtain with a time-of-flight mass spectrometer compared with spectra for negative ions.