Ionization probability of sputtered particles as a function of their energy: Part II. Positive Si ions

In this paper we represent the experimental ionization probability of sputtered silicon atoms as a function of their energy, which has been obtained for positive Si⁺ ions sputtered from silicon by O₂⁺ ion beam. To explain the experimental data, we have considered ionization of an outgoing atom at a critical distance from the surface, which occurs due to the electron transition between this atom and the surface, and suggested the formation of a local surface charge with the polarity opposite to that of the outgoing ion that has just been formed. Then we have considered the interaction between those two charges, outgoing ion, and surface charge as a process of the particle passage through a spherical potential barrier; as a result, we have obtained the theoretical energy distribution of secondary ions. Together with the well-known Sigmund-Thompson energy distribution of sputtered atoms, the obtained ion energy distribution allowed us to derive the equation for the secondary ion yield versus the sputtered particle energy. Both equations derived have exhibited a quite good correlation with our experimental results and also with a large number of published experimental data.     

Secondary ion emission from Si under nitrogen ion bombardment

The yield and energy distribution of positive secondary ions emitted from Si under N₂⁺ ion bombardment were measured. The obtained mass peaks correspond to three types of secondary ion species, that is, physically sputtered ions (Si⁺, Si₂⁺), chemically sputtered ions (SiN⁺ Si₂N⁺) and doubly charged ions (Si²⁺). The dependence of secondary ion emission on the primary ion energy was studied in a range of 2.0–20.0 keV. The yields of physically and chemically sputtered ions were almost independent of the primary ion energy. The yield of the doubly charged ion strongly depended on the primary ion energy. The energy distribution of secondary ions of the three types showed the same dependence on the primary ion energy. The most probable energy of the distribution increased with the primary ion energy. On the other hand, for the energy distribution curves of sputtered ions, the tail factors N in E^?N were constant and showed a m/e dependence.     

Quantitative fundamental SIMS studies using O implant standards

The use of dilute ‘minor-isotope’ ¹⁸O implant reference standards for quantification of surface oxygen levels during steady-state SIMS depth profiling is demonstrated. Some results of two types of quantitative fundamental SIMS studies with oxygen (¹⁶O) primary ion bombardment and/or oxygen flooding (O₂ gas with natural isotopic abundance) are presented: (1) Determination of elemental useful ion yields, UY(X^±), and sample sputter yields, Y, as a function of the oxygen fraction c_O measured in the total flux emitted from the sputtered surface. Examples include new results for positive secondary ion emission of several elements (X = B, C, O, Al, Si, Cu, Ga, Ge, Cs) from variably oxidized SiC or Ge surfaces. (2) The dependence of exponential decay lengths λ(Au^±) in sputter depth profiles of gold overlayers on silicon on the amount of oxygen present at the sputtered silicon surface. The latter study elucidates the (element-specific) effects of oxygen-induced surface segregation artifacts for sputter depth profiling through metal overlayers into silicon substrates.     

Characterizing SIMS transient effects apparent in matrix secondary ion signals from silicon under 1 keV Cs impact

The low energy Si₂⁻, Si₃⁻ and Si₄⁻ secondary ion signals resulting from Cs⁺ impact on Si appear to scale with the Cs uptake noted over the SIMS transient region in a manner consistent with the electron tunneling model. These populations, particularly Si₃⁻ and Si₄⁻ also exhibit a relative insensitivity to the presence of O (shown once sputter rate variations are accounted for). Profiles that more closely match the expected Si concentration gradient from a native oxide terminated Si wafer present within the SIMS transient region can also be obtained by simply dividing the Si₃⁻ or Si₄⁻ secondary ion intensities by the Si₂⁻ intensities. This suggests a possible alternative route for reducing transient effects present in the negative secondary ion populations from Si wafers.     

Cesium ion sputtering with oxygen flooding: Experimental SIMS study of work function change

We investigated the work function (WF) change of a silicon surface being under cesium ion bombardment and simultaneous oxygen flooding with various oxygen pressures at the sample surface. It was found that WF of Cs⁺ ion sputtered Si decreases under oxygen flooding. This decrease provides an essential grow of secondary ion yields of some negative ions, sputtered from Si. At the same time Si⁻ ion yield decreases approximately in two times. In the paper we have discussed possible explanations of our experimental data: we considered a surface composition change, formation of surface dipoles and work function change caused by oxygen adsorption, and their relationships between each other.     

Anomalous enhancement of negative sputtered ion emission by oxygen

Enhancement of negative sputtered ion yields by oxygen (either O⁺₂ bombardment or O₂ gas with Ar⁺ bombardment) is demonstrated for Si^?, As^?, P^?, Ga^?, Cu^? and Au^?, sputtered from a variety of matrices. Because oxygen also enhances positive ion yields of the same species, this effect cannot be simply explained on the basis of existing sputtered ion emission models. To rationalize these phenomena, a surface polarization model is developed which invokes localized electron emissive or electron retentive sites associated with differently oriented surface dipoles in the oxygenated surface. Such sites are considered to dominate the emission of negative and positive ions respectively. The model is shown to correctly predict that Au⁺ and Au^? ion yields are much more strongly enhanced by oxygen in dilute Au-Al alloys than in pure gold.     

Work function, valence band and secondary ion intensity variations noted during the initial stages of SIMS depth profiling of Si and SiO2 by Cs

Work function, valence band and ²⁸Si⁻ secondary ion intensity variations from various Si substrates sputtered by 1 keV Cs⁺ at 60° were measured. Oxide free Si wafers and native oxide terminated wafers did not reveal any appreciable valence band variations close to the Fermi edge. Their work functions however, decreased substantially with an exponential trend noted between this and Si⁻ secondary ion intensities from the O free Si wafer. This is consistent with the electron tunneling model which assumes a resonance charge transfer process. Native oxide terminated wafers exhibited deviations from this exponential trend, while Si wafers with thicker oxides revealed the growth of sub-band features in the valence band spectra on sputtering with Cs⁺. These features, may partially, if not fully, explain the Cs⁺ induced enhancement effect noted on SiO₂ substrates where work function based models are not applicable.     

Secondary ion emission from Si bombarded with large Ar cluster ions under UHV conditions

Si was bombarded with size-selected 40 keV Ar cluster ions and positive secondary ions were measured using the time-of-flight technique under high and ultra-high vacuum (HV and UHV respectively) conditions. Si⁺ ions were main species detected under the incidence of 40 keV Ar cluster ions, and the yields of Si cluster ions such as Si₄⁺ were also extremely high under both conditions. On the other hand, oxidized secondary ions such as SiO⁺ were detected with high intensity only under the HV condition. The yield ratios of oxidized ions decreased in UHV to less than 1% of their values in HV. The effect of residual gas pressure on Si cluster ion yields is relatively low compared to oxidized ions, and the UHV condition is required to obtain accurate secondary ion yields.     

Characterization of ion species of silicon oxide films using positive and negative secondary ion mass spectra

Secondary ion species of silicon oxide films have been investigated using time-of-flight secondary ion mass spectrometry (TOF-SIMS). Characterization of thermally grown SiO₂ films on silicon has been performed. A diagram showing secondary ion spectra of SiO₂ films in both positive and negative polarities indicates the pattern of change in polarities and intensities of ion species from SiO⁺ to Si₅O₁₁⁻. The ions mostly change from positive to negative polarity between Si_nO_2n−1 and Si_nO_2n. Ion peaks with the strongest intensities in the respective cluster ions correspond to the Si_nO_2n+1 negative ion. Intensities of ion species of Si_nO_2n+2 appear negligibly small. Ion species of Si₃O⁺, Si₃O₂⁺ and Si₃O₃⁺ have been found at the interface between silicon and SiO₂ films. The intensity patterns of these ion species compared to those of SiO₂ films indicate that most of these species are not emitted from the SiO₂ films, but likely from the SiO structures.     

Effect of residual oxygen on ionization processes of Si and Si sputtered from Si(1 1 1)-7 × 7 surface

The effect of residual oxygen impurity on ionization processes of Si⁺ and Si²⁺ has been studied quantitatively. In this study, ion sputtering experiments were carried out for a Si(1 1 1)-7 × 7 surface, irradiated with 9-11 keV Ar⁰ and Kr⁰ beam. Even if the oxygen concentration is less than the detection limit of Auger electron spectrometry, SiO⁺ and SiO₂⁺ ions have been appreciably observed. Moreover, as the SiO⁺ and SiO₂⁺ yields increases, the Si⁺ yield is slightly enhanced, whereas the Si²⁺ yield is significantly reduced. From the incidence angle dependence of secondary ion yields, it is confirmed that Si⁺* (Si⁺ with a 2p hole) created in the shallow region from the surface exclusively contributes to Si²⁺ formation. By assuming that the SiO⁺ and SiO₂⁺ yields are proportional to the residual oxygen concentration, these observations are reasonably explained: The increase of Si⁺ with the increase of residual oxygen is caused by a similar effect commonly observed for oxidized surfaces. The decrease of Si²⁺ yield can be explained by the inter-atomic Auger transition between the residual oxygen impurity and Si⁺*, which efficiently interferes the Si²⁺ formation process.     

Sputtering and secondary ion yields of metals subjected to oxygen ion bombardment

The yields of sputtering and secondary ion emission of metals under oxygen ion bombardment were simultaneously measured for multi-layer targets with known layer thickness. The sputtering yield of each layered metal was determined by the time to sputter away the thin film. The experimental results revealed that the sputtering yields of metals were reversely proportional to the energy transfer factor in the classical head-on collision model and a linear relationship was found between the ionization potential and a modified degree of ionization which can be expressed by the ionization potential and the secondary O⁺₂ current.     

Laser induced ion emission from wide bandgap materials

At fluences well below the threshold for plasma formation, we have characterized the direct desorption of atomic ions from fused silica surfaces during 157 nm irradiation by time-resolved mass spectroscopy. The principal ions are Si⁺ and O⁺. The emission intensities are dramatically increased by treatments that increase the density of surface defects. Molecular dynamics simulations of the silica surface suggest that silicon ions bound at surface oxygen vacancies (analogous to E′ centers) provide suitable configurations for the emission of Si⁺. We propose that emission is best understood in terms of a hybrid mechanism involving both antibonding chemical forces (Menzel-Gomer-Redhead model) and repulsive electrostatic forces on the adsorbed ion after laser excitation of the underlying defect.     

Secondary ion emission from silicon and silicon oxide

The emission of Si⁺, Si²⁺, Si³⁺, Si₂⁺, SiO⁺ and B⁺ from boron doped silicon has been studied at oxygen partial pressures between 2 × 10^?10 and 2 × 10^?5 Torr. Sputtering was done with 2 to 15 keV argon ions at current densities between 3 and $\text{40}\text{μ}\text{A}\text{cm}{}^2$. The relative importance of the different ionization processes could be deduced from a detailed study of the yield variation at varying bombardment conditions. Comparison with secondary ion emission from silicon dioxide allows a rough determination of the composition of oxygen saturated silicon surfaces.     

Formation of binary clusters by molecular ion irradiation

In the present study, we have observed silicon–carbon cluster ions (Si_nC_m⁺) emitted from a Si(1 0 0) surface under irradiation of reactive molecular ions, such as C₆F₅⁺, at 4 keV, 1 μA/cm². The cluster Si_n up to n=8 and “binary” cluster Si_nC up to n=6 are clearly detected for the C₆F₅⁺ irradiation. Stoichiometric clusters (Si_nC_m n=m) except SiC⁺ and other binary clusters which contain more than two carbon atoms (m≥2) were scarcely observed. The observed clusters show a yield alternation between odd and even n. The intensities of Si₄, Si₆ and Si₅C clusters are relatively higher than those of the neighboring clusters. In the case of Si₅C, it is considered that doped carbon atom acts as silicon atom. These results imply that the recombination through the nascent cluster emission and subsequent decomposition takes place during the cluster formation.     

Formation of Sim and SimCn clusters by C60 sputtering of Si

The secondary ion mass spectrum of silicon sputtered by high energy C₆₀⁺ ions in sputter equilibrium is found to be dominated by Si clusters and we report the relative yields of Si_m⁺ (1 ≤ m ≤ 15) and various Si_mC_n⁺ clusters (1 ≤ m ≤ 11 for n = 1; 1 ≤ m ≤ 6 for n = 2; 1 ≤ m ≤ 4 for n = 3). The yields of Si_m⁺ clusters up to Si₇⁺ are significant (between 0.1 and 0.6 of the Si⁺ yield) with even numbered clusters Si₄⁺ and Si₆⁺ having the highest probability of formation. The abundances of cluster ions between Si₈⁺ and Si₁₁⁺ are still significant (>1% relative to Si⁺) but drop by a factor of ∼100 between Si₁₁⁺ and Si₁₃⁺. The probability of formation of clusters Si₁₃⁺-Si₁₅⁺ is approximately constant at ∼5 × 10⁻⁴ relative to Si⁺ and rising a little for Si₁₅⁺, but clusters beyond Si₁₅ are not detected (Si_m≥16⁺/Si⁺ < 1 × 10⁻⁴). The probability of formation of Si_m⁺ and Si_mC_n⁺ clusters depends only very weakly on the C₆₀⁺ primary ion energy between 13.5 keV and 37.5 keV. The behaviour of Si_m⁺ and Si_mC_n⁺ cluster ions was also investigated for impacts onto a fresh Si surface to study the effects that saturation of the surface with C₆₀⁺ in reaching sputter equilibrium may have had on the measured abundances. By comparison, there are very minor amounts of pure Si_m⁺ clusters produced during C₆₀⁺ sputtering of silica (SiO₂) and various silicate minerals. The abundances for clusters heavier than Si₂⁺ are very small compared to the case where Si is the target.The data reported here suggest that Si_m⁺ and Si_mC_n⁺ cluster abundances may be consistent in a qualitative way with theoretical modelling by others which predicts each carbon atom to bind with 3-4 Si atoms in the sample. This experimental data may now be used to improve theoretical modelling.     

Positive secondary Ion emission from Si1−xGex bombarded by O2

The positive secondary ion yields of B⁺ (dopant), Si⁺ and Ge⁺ were measured for Si_1−xGe_x (0 ≤ x ≤ 1) sputtered by 5.5 keV ¹⁶O₂⁺ and ¹⁸O₂⁺. It is found that the useful yields of Ge⁺ and B⁺ suddenly drop by one order of magnitude by varying the elemental composition x from 0.9 to 1 (pure Ge). In order to clarify the role of oxygen located near surface regions, we determined the depth profiles of ¹⁸O by nuclear resonant reaction analysis (NRA: ¹⁸O(p,α)¹⁵N) and medium energy ion scattering (MEIS) spectrometry. Based on the useful yields of B⁺, Si⁺ and Ge⁺ dependent on x together with the elemental depth profiles determined by NRA and MEIS, we propose a probable surface structure formed by 5.5 keV O₂⁺ irradiation.     

Secondary ion emission under the bombardment of Si by multiply charged Si q+ ions

The spectra of secondary ion emission under the bombardment of a B-doped Si target by multiply charged Si ^q+ ions (q = 1?C5) have been studied in the energy range of 1 to 10 keV per unit of charge. A multifold increase in the yield of secondary cluster Sk _n ⁺ ions, multiply charged Si ^q/+ ion (q = 1?C3), and H⁺, C⁺, B⁺, Si₂N⁺, Si₂O⁺ is observed as the charge of the multiply charged ions increases. The increase in the yield of secondary ions with increasing charge of the multiply charged-ion charge is most significant for ions with relatively high ionization potentials.     

Sputter depth profiling by secondary ion mass spectrometry coupled with sample current measurements

Ion-induced secondary electron emission determined via sample current measurements (SCM) was employed as a useful adjunct to conventional secondary ion mass spectrometry (SIMS). This paper reports on the results of 3-6 keV O₂⁺ SIMS-SCM sputter depth profiling through CrN/AlN multilayer coatings on nickel alloy, titanium dioxide films deposited on stainless steel, and corrosion layers formed onto surface of magnesium alloy after long-term interaction with an ionic liquid. For Au/AlNO/Ta films on silicon, in addition to SIMS-SCM profiles, the signal of mass-energy separated backscattered Ne⁺ ions was monitored as a function of the depth sputtered as well. The results presented here indicate that secondary electron yields are less affected by “matrix effect” than secondary ion yields, and at the same time, more sensitive to work function variations and surface charging effects. SCM depth profiling, with suppression of “the crater effect” by electronic gating of the registration system is capable of monitoring interfaces in the multilayer structure, particularly, metal-dielectric boundaries. In contrast to SIMS, SCM data are not influenced by the angle and energy windows of an analyser. However, the sample current measurements provide lower dynamic range of the signal registration than SIMS, and SCM is applicable only to the structures with different secondary electron emission properties and/or different conductivity of the layers. To increase the efficiency, SCM should be accompanied by SIMS measurements or predetermined by proper calibration using other elemental-sensitive techniques.     

Ion-enhanced gas-surface chemistry: The influence of the mass of the incident ion

There are many examples of situations in which a gas-surface reaction rate is increased when the surface is simultaneously subjected to energetic particle bombardment. There are several possible mechanisms which could be involved in this radiation-enhanced gas-surface chemistry. In this study, the reaction rate of silicon, as determined from the etch yield, is measured during irradiation of the Si surface with 1 keV He⁺, Ne⁺, and Ar⁺ ions while the surface is simultaneously subjected to fluxes of XeF₂ or Cl₂ molecules. Etch yields as high as 25 Si atoms/ion are observed for XeF₂ and Ar⁺ on Si. A discussion is presented of the extent to which these results clarify the mechanisms responsible for ion-enhanced gas-surface chemistry.     

Critical distance for secondary ion formation: Experimental SIMS measurements

We have performed direct experimental measurements of the critical distance for the secondary ion formation process. To this end, we compared the experimentally measured energy distribution of secondary Si⁻ ions with the theoretical energy distribution (Sigmund-Thompson relation) of secondary Si atoms. Our model states that the maxima positions of these two energy distributions differ by the Coulomb interaction potential between the outgoing ion (Si⁻ in our case) and a charge with the opposite polarity formed at the surface after electron transition between the outgoing Si atom and the surface. Quite a reasonable value was obtained for the critical distance, but with a large scatter in experimental data. The conclusion has been made that the experimental technique should be improved to get more precise values of the critical distance, which is of high importance for practical purposes.