Mass-energy Analyzer Research Articles

Energy distributions of H and H2 in a Ar–SiH4–H2 plasma during transition from a-Si to nc-Si film deposition

The assessment of H and H2 densities by mass spectrometry methods in silane–hydrogen (SiH4–H2) plasma-assisted CVD processes requires estimation of the depletion factor for the parent molecules (SiH4 and H2). Such an estimation is non-trivial as H2 is both depleted and generated in SiH4–H2 plasmas, and the usual depletion factor becomes a ‘depletion-generation’ factor. A method for calculating the depletion-generation factor in SiH4–H2 plasmas is presented and applied to de-convolute energy distributions of H and H2 from energy-resolved residual gas analysis (ERRGA) performed using a HIDEN EQP300 mass-energy analyser. The ERRGA method is described and the energy distribution of the H radicals and H2 molecules is obtained for an LEPECVD plasma during the transition from a-Si to nc-Si deposition at constant discharge operation parameters for gas input flow rates of 50 sccm Ar and 10 sccm SiH4, with the addition of H2 (0–50 sccm) at reactor pressures of 1–3 Pa. The H and H2 energy distributions show two main groups, one below 0.7 eV and another at 1.2–1.5 eV, with a strong variation in total density with increase in H2 input. Above 30 sccm H2 input, a step increase in the lower energy component is observed. Energy distributions are discussed in relation to H and H2 generation reactions and gas heating in PECVD processes.

Growth stress in tungsten carbide-diamond-like carbon coatings

Growth stress in tungsten carbide-diamond-like carbon coatings, sputter deposited in a reactive argon/acetylene plasma, has been studied as a function of the acetylene partial pressure. Stress and microstructure have been investigated by wafer curvature and transmission electron microscopy (TEM) whereas composition and energy distribution functions of positive ions were obtained by electron probe microanalyzer, elastic recoil detection analysis, and mass-energy analyzer (MEA). It has been observed that the compressive stress decreases with increasing acetylene partial pressure, showing an abrupt change from −5.0 to −1.6 GPa at an acetylene partial pressure of 0.012 Pa. TEM micrographs show that by increasing the acetylene partial pressure in the plasma from 0 to 0.012 Pa, the microstructure of the coating changes from polycrystalline to amorphous. MEA results show that the most probable energy of positive ions bombarding the substrate during deposition in pure argon and argon/acetylene atmosphere is the same. Based on the results, it is concluded that the huge variation in the compressive stress at low acetylene partial pressures is due to a change in the microstructure of the coating from polycrystalline to amorphous and not to the energy of positive ions bombarding the film.

Experiment and simulation of the compositional evolution of Ti–B thin films deposited by sputtering of a compound target

The evolution of the coating stoichiometry with pressure, target-substrate distance, and angle was analyzed for dc sputtering of TixB (x=0.5, 1, 1.6) compound targets by elastic recoil detection analysis. For an investigation of the underlying fundamental processes primarily Ar was used as sputter gas. Additionally, the effect of a reactive gas (N2) as well as bias voltage (floating up to −200 V) was briefly cross-checked. For deposition along the target normal (90°) a pronounced Ti-deficiency of up to 20% is detected. Increasing the pressure or distance from 0.5 to 2 Pa and from 5 to 20 cm, respectively, leads to an almost equivalent linear increase in Ti/B ratio surpassing even the target composition. Off-axis depositions at lower angles (30° and 60°) on the other hand result in a higher Ti/B ratio. This is consistent with results obtained from Monte Carlo simulations combining the respective emission characteristics from the sputter process as well as the gas-phase transport. Hence, the pressure, distance, and sample position induced changes in chemical film composition can be understood by considering gas scattering and the angular distribution of the sputtered flux. The theoretically determined transition from a directional flux to thermal diffusion was experimentally verified by mass-energy analysis of the film-forming atoms.

An improved method for IEDF determination in pulsed plasmas and its application to the pulsed dc magnetron

In this paper we demonstrate an improved technique for obtaining time-resolved ion energy distribution functions (IEDFs) in pulsed plasma discharges using a commercial quadrupole mass energy analyser. The method involves extracting ions from the plasma at selected times during the pulse cycle through the application of a synchronized electrical bias on a grid assembly built in the barrel of the instrument. Placing the triggered mesh in the barrel of the energy analyser makes the ‘transit’ gap length (between the chopping mesh and neighbour electrodes) ∼2 mm and that results in a significant increase of the time resolution. Due to the high vacuum conditions inside the barrel, no unwanted ‘parasitic’ discharge will exist between the extractor and orifice. The electrostatic chopping by the built-in mesh allows time-resolved IEDFs to be determined without reference to the inherent time delay caused by the ion flight time through the spectrometer. The technique has been applied to the low-pressure pulsed dc magnetron discharge operating in Ar gas and at a pulse frequency of 100 kHz. Time-resolved IEDFs were obtained with a 1 µs time resolution, and these revealed large variations in the form and width of the ion energy spectra at different times during the discharge cycle, governed by the temporal evolution of the plasma potential. A calculation based on ion dynamics in the grid assembly agrees well with experimentally obtained results and demonstrates that this method can provide time resolutions as small as 200 ns.

Low-energy linear oxygen plasma source

A new version of a constricted plasma source is described, characterized by all metal-ceramic construction, a linear slit exit of 180 mm length, and cw operation (typically 50 kHz) at an average power of 1.5 kW. The plasma source is here operated with oxygen gas, producing streaming plasma that contains mainly positive molecular and atomic ions, and to a much lesser degree, negative ions. The maximum total ion current obtained was about 0.5 A. The fraction of atomic ions reached more than 10% of all ions when the flow rate was less then 10 SCCM O(2), corresponding to a chamber pressure of about 0.5 Pa for the selected pumping speed. The energy distribution functions of the different ion species were measured with a combined mass spectrometer and energy analyzer. The time-averaged distribution functions were broad and ranged from about 30 to 90 eV at 200 kHz and higher frequencies, while they were only several eV broad at 50 kHz and lower frequencies, with the maximum located at about 40 eV for the grounded anode case. This maximum was shifted down to about 7 eV when the anode was floating, indicating the important role of the plasma potential for the ion energy for a given substrate potential. The source could be scaled to greater length and may be useful for functionalization of surfaces and plasma-assisted deposition of compound films.

Time-resolved measurements of the ion energy distribution function in a pulsed discharge using a double gating technique

This paper presents an improved technique for obtaining the time-resolved ion energy distribution functions (IEDFs) in a pulsed plasma using a commercial quadrupole mass energy analyser. In this technique, both the extractor and the detector of the analyser are gated at selected times relative to the pulse. The combination of the gating pulses allows the time-resolved IEDFs to be determined without a time delay inherent in the ion transit time through the spectrometer. The technique is applied to a low pressure RF discharge in Ar. The time-resolved IEDFs are obtained for two different modulation frequencies with 10 µs time resolution: (i) 1 kHz of modulation frequency, 50% duty (when the afterglow plasma exists within the entire ‘off’ period) and (ii) at 200 Hz modulation frequency, 10% duty (when the remnant plasma completely collapses before the beginning of the following ‘on’ pulse). The peak values in the ion energy are found to be constant in a steady state of the ‘on’ time (∼42 eV). The highest ion energy at 200 Hz is 90 eV, but only 80 eV at 1 kHz. At the beginning of the ‘off’ time they fall sharply to 10–15 eV and then slowly decrease in time to 4–6 eV. The normalized integral of the counts under the IEDFs and the IEDF peak energies are compared and agree well with the plasma densities and the plasma potential respectively obtained at the same time points by the single Langmuir probe.

CHARACTERIZATION OFAl+SECONDARY ION EMISSION PRODUCED BYNe+ANDAr+BOMBARDMENT OF ALUMINIUM SURFACE

This paper reports the characterization of the velocity (energy) dependencies of the Al+secondary ion emission produced by 0.5 keV and 5 keV Ne+and Ar+bombardment of polycrystalline pure aluminium. The distributions of secondary Al+ions over their kinetic energy were measured for emission energies of 1–1000 eV without applying electric fields to force the ions into the mass–energy analyzer. To extract the ionization probability, the measured energy distributions of emitted ions were normalized with respect to reference energy distributions of neutral atoms. The reference distributions were obtained by original numerical simulations, as well as analytically, through a sophisticated normalization of the Thompson distribution. It was shown that for both extraction methods, the logarithmic plots of the normalized secondary ion fraction versus the normal component of the reciprocal ion velocity (the reciprocal or inverse velocity plots) are nonmonotonic, with two peaks and two linear portions situated at a low emission energy (Ek=5–25 eV ) and at a high emission energy (Ek=80–280 eV ). The linear portions were fit by exponential dependency P+∝ exp (-v0/vn) with two different values of the characteristic velocity v0. For the low emission energy, the value v01~(3.3±0.2)×106cm / s was independent of the mass and energy of the projectiles. However, for the high emission energy, the characteristic velocity depended on the projectile's mass, M, namely v02~(5.3±0.3)×106cm / s for Ne+and v02~(8.1±0.3)×106cm / s for Ar+; the ratio v02( Ne+)/v02( Ar+) is close to the value [Formula: see text]. This indicates that ballistic mechanisms might contribute to affect the high-energy part of the reciprocal velocity plots along with nonballistic ionization processes, which are generally believed to be the only significant factor for the plots.

Effect of plasma flux composition on the nitriding rate of stainless steel

The total ion flux and nitriding rate for stainless steel specimens exposed to a modulated electron beam generated argon-nitrogen plasma were measured as a function of distance from the electron beam axis. The total ion flux decreased linearly with distance, but the nitriding rate increased under certain conditions, contrary to other ion flux/nitriding rate comparisons published in the literature. Variation in ion flux composition with distance was explored with a mass spectrometer and energy analyzer as a possible explanation for the anomalous nitriding rate response to ion flux magnitude. A transition in ion flux composition from mostly N2+ to predominantly N+ ions with increasing distance was observed. Significant differences in molecular and atomic nitrogen ion energy distributions at a negatively biased electrode were also measured. An explanation for nitriding rate dependence based on flux composition and magnitude is proposed.

Energy distributions of Ga + and In + secondary ions sputtered from A IIIB V compound semiconductors by noble gas ions: Mass-dependence of the high-energy yield on the second component (P, As, Sb) of the compounds

Experimental and simulated energy distributions of Ga + and In + secondary ions produced by 4 keV Ne +, Ar + and Kr + bombardment of the A IIIB V semiconductors (GaP, GaAs, GaSb, InP, InAs and InSb) are reported. The measurements were carried out for a wide range of initial energy (up to 1000 eV) in a small solid angle along the surface normal, without applying electric field to extract the ions into the mass-energy analyser. It is shown that the energy spectra are complex, with evident high-energy hump, whose relative intensity increases with the mass of the second component (P, As, Sb) of the compound. The Sigmund–Thompson distribution cannot fit reliably these data, and a satisfactory approximation of the measured spectra was obtained with a sum of two decaying exponential functions to describe the contribution of both, the isotropic linear collision cascades and the outward knock-on atoms. The experimental results are compared with simulations based on the MARLOWE computer code.

Ion energy distributions versus frequency and ion mass at the rf-biased electrode in an inductively driven discharge

In this article, we present ion energy distributions (IEDs) at a rf-biased surface as a function of driving frequency and ion mass. The experiments were carried out in high-density inductively coupled rare-gas (Ne,Ar,Xe) plasmas. Our quadrupole mass and cylindrical-mirror energy analyzer sampled ions incident on a rf-biased pinhole located in the center of the wafer chuck. The electron density, electron temperature, and plasma and chuck potential oscillations were measured, and they provided inputs to numerical models used to predict IEDs, which were shown to closely match our experimental results under certain conditions. For a given driving frequency, heavier ions showed narrower IEDs and, for a given ion mass, the IED became narrower and shifted to a higher mean energy with increased driving frequency, in agreement with calculations.

Evidence of ionized metal clusters in ion plating discharges

A quadropole mass-energy analyzer has been used to detect positively charged metal ions at the substrate (cathode) of an ion plating system. By studying titanium, evaporated into thermionically supported argon glow discharges at pressures of 1–4 Pa, we have detected the presence of ionized clusters of the vapor material up to the analyzer mass limit of 2500 a.m.u. In contrast, a similar analysis from an equivalent titanium-only discharge has revealed a relatively insignificant cluster detection rate. The results support an earlier prognosis that these clusters may nucleate and grow through vapor cooling, caused by collisions with gas atoms in the discharge.

A technique for obtaining time- and energy-resolved mass spectroscopicmeasurements on pulsed plasmas

We present a simple technique for obtaining the time-resolved ion energydistribution function (IEDF) at a boundary in pulsed plasmas using acommercial quadrupole mass energy analyser. In this technique, ions areextracted from the plasma at selected parts of the pulse cycle, through thesynchronized electrical biasing of a grid assembly attached to the barrel of theinstrument, forming an electrostatic shutter. This sampling method has theadvantage over the normal technique of electronically gating the detectedion signal to achieve time resolution, since the IEDFs can be obtainedeven when the ion flight time through the instrument (typically 100 µs) is greater than the pulse period or the characteristic time oftransients in the plasma under investigation. The arrangement allowsus therefore to diagnose plasmas pulsed at high frequencies (≥10 kHz). Presently, atime resolution of 4 µs can be obtained, limited only by the driving electronics design. Thetechnique has been tested on a DC magnetron discharge operated inargon. The plasma was pulsed at a low frequency of 2 kHz, but with adischarge voltage waveform containing fast transients (on the µstime scale or faster). The results show clearly the evolution of the IEDFs duringthe pulse, responding to these fast transients with a significant number of ionscreated at plasma potentials above 140 V. These time-evolved IEDFs cannot beobtained using the conventional, manufacturer’s time-resolved method for thisinstrument. However, the new technique does introduce a small distortion in themeasured IEDFs at energies above 120 eV, which is always observed,irrespective of the position of the shutter window during the pulse. This isdue to the transient nature of the discriminating grid bias used in theelectrostatic shuttering. Its effects and possible elimination are discussed.

Ion energy distributions at rf-biased wafer surfaces

We report the measurement of ion energy distributions at a radio frequency (rf)-biased electrode in inductively driven discharges in argon. We compare measurements made with a gridded energy analyzer and a commercial analyzer that contains a mass spectrometer and energy analyzer in tandem. The inductive drive and the rf bias in our Gaseous Electronics Conference reference cell were both at 13.56 MHz. By varying the plasma density, we were able to examine the transition region between the “low frequency limit” for rf bias and the intermediate frequency region where, at fixed bias frequency, the ion energy distribution width varies with the plasma density. We find that the experimental ion energy distributions become narrower as the time for ion transit through the sheath approaches the rf period, but that the ion distributions still have widths which are ∼90% of their low frequency limit when the ion transit time is 40% of the rf period. Space-charge-induced beam broadening inside our analyzers appears to significantly affect our measurements of ion angular distributions, especially at low ion energies.

Energy distributions of secondary ions sputtered from aluminium and magnesium by Ne +, Ar + and O 2+: a comprehensive study

The energy distributions of positive and negative secondary ion species produced by 5 keV Ne +, Ar + and O 2 + ion bombardment at 30° to polycrystalline aluminium and magnesium targets have been studied. The measurements were carried out in a small solid angle along the surface normal, without extracting ions into the mass–energy analyser. The most probable and average energies along with the widths of the distributions were tabulated; essential features of the spectra were interpreted in the terms of a linear collision cascade model with a knock-on contribution. Three energy sub-ranges with appreciable different energy power dependence E − s , which could approximate the experimental curves, were revealed for both Al + and Mg + atomic ions within a wide energy range E mp< E<1000 eV. Nearly all the energy spectra of negative secondary ions demonstrated complex structure, with two distinct peaks, caused by impurity species with different binding energies.

Interaction of MEV atomic ions with molecular solids: Ion track structure and sputtering phenomena

Recent results obtained in our research groups from studies of the interactions of swift, heavy atomic ions with molecular solids are concisely outlined. The focus is on material ejection (sputtering) and surface track formation. The experimental techniques employed include time-of-flight mass spectrometry, energy analysis, collection and analysis of sputtered material, and scanning force microscopy. Characteristics of the sputtering process probed include the sputtering yield, radial and axial velocity distributions, angular distributions, and surface track morphology. Besides reviewing and correlating experimental results, we also emphasize the common quasi-thermal origin of pressure-pulse/hydrodynamic and evaporative spike sputtering models.

Design of ion energy distributions by a broad beam ion source

We characterize the performance of a built-in hot filament broad beam ion source by mass spectrometry, energy analysis, and beam profile measurements. In the ion energy distribution we detect various peak structures which can be explained by the potential across the ion source and different charge transfer processes. Depending on the typical cross sections for these processes, differences between the energy distributions of the ion species are observed. The total ion current obtained with the source is determined by the ionization rate in the discharge and the current share directed toward the extraction grid system. The performance of the source is strongly dependent on the process gas used. We observe much broader energy distributions in oxygen and nitrogen than in argon. This broadening is explained by spatial inhomogeneities in the discharge region and can be reduced by a suitable setting of the source parameters. The main contribution to the ion flux is caused by species generated directly from the process gas, with important impurities from source materials appearing only with additional chemical reactions in oxygen.

Simultaneous mass and kinetic energy analysis of Ar + and Ar 2+ ions photodesorbed from condensed argon

Mass and kinetic analysis of photodesorbed ions from argon multilayers shows that the kinetic energy distributions depend on both the mass of the ions and the photon excitation energy. Measured kinetic energies are in good agreement with previously proposed models for photodesorption of ions at hv = 40 eV and hv = 100 eV.

Field ionizable cesium metal clusters from a foil diffusion source

Ionized clusters of Cs were studied by time-of-flight mass spectrometry and energy analysis. The typical clusters contained 4000–10000 Cs atoms, assuming singly charged clusters. The clusters were formed in the neutral state by condensation of Cs from a foil diffusion source at 1300 K. Diffusion sources have previously been shown to emit excited states, especially high Rydberg states of Cs (Pettersson and Holmlid, 1989), and the condensation is thought to be driven by the large interaction energies between highly excited states. Ionized clusters were formed by field ionization of the neutral clusters at electric field strengths of 200–400 V/cm, both with pulsed fields and at constant field strengths. The cluster size determined from the neutral transport time to the point of field ionization agrees roughly with the cluster size determined from the ionized cluster flight time.

Time-resolved mass and energy analysis by position-sensitive time-of-flight detection

We describe a method for time-resolved mass and kinetic energy analysis of ionic or neutral species (1–150 amu, 0.5–500 eV) in diagnostic measurements on spacecraft electric thrusters. Time-of-flight mass spectrometry is combined with position-sensitive detection to measure energy spectra for multiple masses at sampling rates as high as 50 kHz. A rectangular microchannel plate detector with a 96-element metal anode array is read out by fast analog-to-digital converters or by discriminators and scalers. The ion drift time varies as the square root of the mass-to-charge ratio, and the displacement along the detector varies as the square root of the energy-to-charge ratio. The energy resolution is enhanced by minimizing the field distortion near grid wires.

Time-resolved mass and energy analysis by position-sensitive time-of-flight detection

We describe a new method for time-resolved mass and kinetic energy analysis of ionic or neutral species in the range of 1–150 amu and 0.5–500 eV. Time-of-flight mass spectrometry is combined with position-sensitive detection to measure energy spectra for multiple masses at burst-mode sampling rates as high as 50 kHz. The detector is a rectangular microchannel plate with a 96-element metal anode array that is read out either by fast analog-to-digital converters or by discriminators and scalers. The apparatus is configured so that the measured ion drift time varies as the square root of the mass-to-charge ratio and the displacement along the detector varies as the square root of the energy-to-charge ratio. Applications are envisioned in plasma analysis, in beam-scattering experiments, and in diagnostic measurements for spacecraft propulsion.