Low energy ion scattering is a qualitative and quantitative surface analysis technique. Its supreme surface sensitivity and straightforward quantification (using a well-defined reference) make it a convenient tool for the study of surface composition and a useful method for surface characterization in cooperation with other surface analysis methods such as XPS and SIMS. Silver (100) monocrystal was analyzed by the primary beam of helium ions. The wide energy range from 1.0 to 4.5 keV covers three distinguished regions. On the low energy side, the charge exchange processes are dominated by Auger neutralization (AN), while collision-induced (CI) processes rule a high energy range. Both mechanisms are mixed in the intermediate region between 1.2 and 2.1 keV (for perpendicular incidence and 145° scattering geometry). The results can serve both as a reference and as an insight into neutralization probability changes (as dependence on primary energy). The neutralization strength is reflected by the characteristic velocity. It was evaluated for AN and CI regions to 0.75 × 105 and 0.38 × 105 ms−1, respectively. The CI reionization energy threshold is around 1700 eV for both Ag (100) and polycrystalline Ag. The reference measurement on polycrystalline copper relates the presented data to those received by other Qtac100 instruments with different sensitivities.
INTRODUCTION
Silver's excellent electrical conductivity, antimicrobial effects, and ability to reflect light define a wide range of its applications. The role of silver in human activities has evolved over many centuries, from its use in jewelry, mirrors, and photography to its application in electrical contacts, batteries, integrated circuits, photovoltaic cells, and display screens. Here, we provide an analysis of two Ag surfaces by low energy ion scattering (LEIS).
LEIS spectroscopy is a well-established method for studying the composition of surface layers (Refs. 1–7). The noble gas ions (He+, Ne+, etc.), alkali metal ions (Li+, Na+, etc.), or hydrogen ions are accelerated (primary energy between 0.5 and 10 keV) toward the surface, where they are scattered in all possible directions, and the experimental setup defines the scattering angle around which the signal is collected and analyzed (Ref. 1). Some of the instruments have fixed polar and azimuthal angles; some were designed to vary the scattering angle significantly or detect the projectiles within a wide range of polar and azimuthal angles, providing information about the periodic arrangement of the atoms (Refs. 8 and 9).
The electrostatic energy analyzers (ESA-LEIS) and time-of-flight energy analyzers (TOF-LEIS) are widely used. In the case of TOF-LEIS, both charged and neutral projectiles are analyzed (comparing signals from neutralized and ionized projectiles) (Ref. 4), while in the case of ESA-LEIS, only projectiles that are charged after the scattering event are analyzed and detected, usually providing more surface-specific analysis.
There is a high probability of being neutralized for noble gas ions with primary energies in the LEIS range (0.5–10 keV), which penetrate the surface and scatter on the deeper layers compared to those scattered on the outermost layer. Thus, the information from the outermost atomic layer is highlighted in the form of the so-called binary peaks (sometimes also surface peaks); this fact, combined with a relatively high cross section (due to a low primary energy range), makes the LEIS technique extremely surface-sensitive and it also makes acquired data different from spectra measured in medium energy range (MEIS) or with high energies (RBS). The detected peaks are accompanied by mound-shaped background for elements that are effective in reionization. For illustrative examples, see Ref. 3.
For data presented here, the noble gas ions with the primary energy between 1.0 and 4.5 keV were used, and the energy analysis was done by an ESA double toroidal energy analyzer described in Ref. 10. The primary beam is perpendicular to the surface, the scattering angle is fixed to 145° ± 0.2°, and projectiles are collected over the entire azimuth (360°), which excludes the structural analysis but significantly increases the instrument sensitivity (Qtac100 instrument, IONTOF GmbH). The scattering geometry could be considered typical of many experimental setups like 136° (Ref. 11) and 129° (Ref. 12).
Another advantage of the LEIS is that collision between projectile and target can usually be approximated as binary collision or sequence of binary collisions (Refs. 1, 13, and 14). This is basically an idea behind the computer simulation codes like SRIM or TRIM (Transport of Ions in Matter) for BackScattering being used to describe these collisions (Ref. 15).
This approximation provides a reasonable agreement with experimental data which means that there is virtually no so-called matrix effect (effect where the vicinity of the target influences the measured data). This allows straightforward quantification of the data using a reference spectrum. However, one has to be aware that for a limited number of cases, it has been shown that the matrix effect (through neutralization and reionization processes) affects the intensity of the ESA LEIS signal (Refs. 4 and 11); hence, the straightforward quantification (comparing the evaluated and reference spectra) is not possible. Therefore, it is crucial to identify specific neutralization and reionization processes; ergo, comparative measurements are needed (Refs. 16 and 17).
When the charge exchange processes are the same for the sample and reference, quantification using the reference is relevant. The inverse velocity plot (see below) is a useful tool for it. Here, we present the selected spectra for He+ scattering on the silver surfaces and the inverse velocity plots measured within the wide energy range. The calibration measurement on the polycrystalline Cu surface links the absolute signals received on instruments with different sensitivities. Thus, the data can be shared between the instruments using identical or similar scattering geometry. It is important to note that in the Qtac100 instrument, the signal is collected over the entire azimuth range, which integrates any azimuth-dependent effects occurring on crystalline or polycrystalline surfaces.
The most straightforward system for studying the charge exchange processes in a low energy range is the scattering of He+ ions on metals with high work function where the neutralization processes are spatially separated (Ref. 14). Helium has the simplest electronic structure compared to other noble gas projectiles; thus, the range of possible combinations of local and collective interactions with the target is limited by a single He 1s level. Nevertheless, even for He+ projectiles, several charge exchange processes and their combinations are possible. The experimental effort investigated He+ scattering on metallic surfaces in the last two decades (Refs. 4, 11, 14, and 18–21). Two charge exchange processes dominate He+ scattering on metals with high work function. Auger neutralization (AN) is a long-range process and acts during all scattering events. Thus, it is the first mechanism that starts to neutralize the incoming He+ beam, acts during the close scattering event, and is also the last mechanism capable of neutralizing the He+ ions on their way from the sample surface. The AN has a nonlocal character, and thus, the Auger transition rate depends on the perpendicular distance of the He+ projectile from the plane defined by the target surface. The AN is possible for any kinetic energy of the projectile. On contrary, the collision-induced (CI) resonant charge exchange processes (sometimes referred to as resonant processes in close proximity or just collision-induced processes, Refs. 1, 4, 11, 14, and 21) appear below a certain distance between the He+ projectile and individual target atoms, where molecular orbitals are formed (Ref. 22). Thus, the CI processes are localized, act only during hard collisions, and require certain projectile energy that exceeds the so-called reionization threshold energy (specific for given scattering geometry) (Refs. 11 and 23).
The quantification in the AN energy range is surface-sensitive. For metallic targets where electrons are supplied from delocalized states, the quantification is independent of scattering geometry but depends on surface crystal faces (Ref. 4). The quantification at the CI energy range is almost independent of the surface orientation but modifies the experimental spectra by adding the signal from deeper atomic layers in the form of a background at the low energy side of the surface peak. This was observed by Puurunen et al. (Ref. 24) on island growth mode of ZrO2 on hydrogen-terminated silicon (the development of Zr background with a number of ALD deposition cycles was shown). Additionally, materials with opened structures enable direct scattering from deeper atomic layers. For example, the triplet layer structure (F––Ca2+–F−) of CaF2 (111) enables direct scattering from both F layers and a Ca layer between them (Ref. 16). The quantification in the AN energy range is simplified on polycrystalline surfaces using the primary beams perpendicular to the surface plane. This is a case of experimental data presented in our contribution. The Qtac100 instrument uses the energy analyzer with full angle (360°) azimuth acceptance, further favoring the quantification in the AN energy range.
The detailed study of He+ scattering on silver surfaces was reported by Primetzhofer et al. (Ref. 12) over the energy range covering AN processes, and for 6.0 keV primary energy well in the CI energy range. The work uses the TOF energy analyzer (scattering angle 129°) and compares the signal of ionized projectiles with neutral fraction when varying the angle of incidence. The neutralization by polycrystalline silver and Ag (111) in the AN energy range [about (0.5–0.85) keV for polycrystalline Ag and about (0.6–1.2) keV for Ag (111)] was found to be essentially the same. The characteristic velocities published are 1.39 × 105 and 1.40 × 105 ms−1. Note that the energy threshold for CI processes was reported below 1500 eV (Ref. 25) and around 2000 eV (Ref. 26). The energy range used for AN satisfied both values. The scattering data measured at Ag (111) at energies above the energy threshold for CI processes (data for 6.0 keV primary energy presented) show strong modulation in the detected signal intensity when the angle of incidence is varied. This was related mainly to the increase in the number of scattering centers involving deeper atomic layers into the scattering (effect strong for TOF-LEIS measurements).
Theoretical background
Since dσi/dΩ can be determined for every projectile-target pair (we used scattering cross sections calculated according to the Thomas–Fermi–Molière potential, Ref. 27), roughness factor is R = 1 for reasonable flat surfaces (Ref. 28) and instrumental factor σ is a constant for Qtac100 instrument, the ion fraction is our variable of interest.
The ion fraction characterizes the charge exchange processes between the projectile and the target atoms. They are ruled by the energy states in projectile and target atoms, the distance between them, and the time provided during the collision. The time the projectile spends within the vicinity of the surface atoms or surface as a whole is inversely proportional to its velocity (with regard to the scattering angle). The neutralization probability generally decreases with higher energy (due to less time spent near the target atoms) with documented oscillations in specific cases. According to Refs. 1 and 29 when there is only one of these processes dominantly present, the neutralization probability is changing as an exponential decay (with inverse velocity and characteristic velocity in the exponent). The characteristic velocity vc is a measure of the neutralization strength and is characteristic of every projectile-solid pair. It describes neutralization probability during the whole encounter, and it is derived from the transition rate integrated over the distance from the surface. At energies where a mixture of processes occurs, the neutralization probability oscillates. The strong regular oscillations were reported by Erikson for He scattering on Pb (Ref. 30) or He scattering on gallium (Ref. 1). More complex oscillations were recently reported by Vaníčková et al. for He scattering on bismuth for a wide range of primary energies and a scattering angle of 145° (Ref. 31).
39 spectra of helium projectiles scattered on monocrystal of silver Ag (100). The range of primary energies is (1–4.5) keV. A similar set of spectra was measured for polycrystalline silver foil. Both sets are available in the supplementary material.
39 spectra of helium projectiles scattered on monocrystal of silver Ag (100). The range of primary energies is (1–4.5) keV. A similar set of spectra was measured for polycrystalline silver foil. Both sets are available in the supplementary material.
Dependence of the modified signal (in logarithmic scale) of helium projectiles (scattered on silver) on the inverse perpendicular velocity of these projectiles (primary ion energy is on the top horizontal axis for comparison). Ranges of primary energies are (1–4.5) keV and (1–3) keV for silver Ag (100) (triangles) and polycrystalline silver (circles). Characteristic velocities are also indicated for AN (E<1200 eV) and CI (E>2100 eV) energy regions.
Dependence of the modified signal (in logarithmic scale) of helium projectiles (scattered on silver) on the inverse perpendicular velocity of these projectiles (primary ion energy is on the top horizontal axis for comparison). Ranges of primary energies are (1–4.5) keV and (1–3) keV for silver Ag (100) (triangles) and polycrystalline silver (circles). Characteristic velocities are also indicated for AN (E<1200 eV) and CI (E>2100 eV) energy regions.
The ion fraction P+ is often used to define the axis as well. In this case, the intercept for infinite primary energy can be assumed to be equal to unity. This theoretical point is important and is usually used together with raw experimental data to fit the linear dependencies, significantly influencing the value of characteristic velocity. This approach is relevant only for data received in the energy range where AN dominates (Ref. 23).
Experimental results
The surfaces of silver and copper were measured in this study. Argon (5.0 keV, fluence 2 × 1016 Ar/cm2) and neon (5.0 keV, fluence 2 × 1015 Ne/cm2) beams were used to sputter the copper foil to prepare a clean surface. Spectra for He+ 3.0 keV scattering are used as a reference to evaluate the instrument's sensitivity. The intensity signal of Cu binary peak is (14 470 ± 120) counts per nC for our instrument. The typical signal intensity for different Qtac100 instruments is in the range of tens of thousands of counts per nC [differences between individual instruments are represented in the instrumental factor, see Eq. (1)]. The Ag polycrystalline foil was cleaned using a procedure identical to the Cu reference. The Ag (100) crystals were first sputter cleaned by argon (5.0 keV, fluence 2 × 1017 Ar/cm2) until the LEIS He+ spectra showed a clean surface free of carbon and oxygen signals. After that, the surface reconstruction was done by annealing the sample at T > 500 °C (similar procedure as in Ref. 12), to recrystallize the sputtered surface (Ref. 33). Heavy sputtering of the polycrystalline samples of Cu and Ag significantly converts the polycrystalline arrangement of the surface layer to the amorphous. Nevertheless, we continue to use the designation poly for these surfaces, as it reflects their state before the experiment began (similar to Ref. 12).
The 39 experimental spectra were collected for He+ scattering on a cleaned Ag (100) surface. The primary energy was varied from 1.0 up to 4.5 keV. The spectra are presented together in the plot in Fig. 1. The individual spectra for primary energies 1.0, 3.0, and 4.5 keV are provided separately in figures 01979-01, 01979-02, and 01979-03 in the supplementary material. The strong Gaussian-shaped peak is presented. The background for 1.0 keV primary energy is practically limited to the noise level. The silver effectively neutralizes the He+ projectiles, and their reionization requires a close approach, which is achieved for primary energies higher than the threshold energy, which is about 1.7 keV according to our results presented here. It is important to note that different methods of determining the reionization energy threshold were developed. Regarding the energy threshold for silver, Souda et al. (Ref. 26) found no energy threshold for the ionization of neutral He0 projectiles scattered on Ag below 2000 eV primary energy (within the investigated energy range). By contrast, Thomas et al. (Ref. 25) determined the threshold to be below 1500 eV using a different approach based on ratios between the scattered ion fractions for He0 and He+ projectiles. The reionization (ionization) threshold energies are compared for different elements and different projectiles in Table 6.1 in Ref. 1, referring to individual citations. Our evaluation for He+ scattering on Ag is based on Method 3, where no data are presented in the table, and it falls between the intervals proposed by Souda and Thomas.
The reionization (ionization) process can be affected by the arrangement of Ag atoms, scattering geometry, and surface contamination. For example, carbon decreases the ion fraction, especially in the sp2 hybridization (Ref. 34), while oxygen enhances it in most cases (Ref. 35). In our experiments, we stopped the sputter cleaning of Cu and Ag poly surfaces well after no evidence of C and O was present in the spectra. The value we propose corresponds to scattering on a poly Ag surface amorphized by Ar and Ne sputtering. Our evaluation method has a relatively high uncertainty compared to approaches used by Souda and Thomas; however, their intervals do not coincide. Proper identification of the reionization threshold energy for He scattering on Ag surfaces would require dedicated research.
Quantification in the LEIS energy range is based on comparing measured signals with those received from well-defined references. This approach is possible only when the matrix effect can be neglected or is well known and identical in both materials. Here, we provide the inverse velocities plot [according to Eq. (2)] for He+ scattering on Ag (100) within the wide energy range (1.0–4.5 keV). This plot is in Fig. 2 and shows the transition between two different regions. The Auger processes (AN) dominate at lower primary energies, while at energies higher than 1.2 keV, the CI processes are starting to take the initiative (while AN is, of course, still contributing but for higher energies AN contribution is minor). The transition range between these energy regions is located in between and covers the range from 1.2 to 2.1 keV.
Primetzhofer et al. (Ref. 12) studied the characteristic velocities of different crystal orientations and polycrystalline Cu and Ag. The polycrystalline Cu [assumed to consist of (111) facets] exhibits a neutralization behavior similar to Cu (111). This assumption was corroborated by measurements on Ag (111) and polycrystalline Ag, where the characteristic velocities were found to be essentially the same at 1.4 × 105 m/s. In our investigation, we found a value of 0.75 × 105 m/s for Ag (100), which is about half of the values proposed by Primetzhofer for Ag (111) and polycrystalline Ag. This difference could be associated with surface contamination as described above, but more likely due to differences in surface orientation, similar to what Primetzhofer found between Cu (111) and Cu (100) [an even higher decrease in the characteristic velocity was found for Cu (110)]. All values presented here were evaluated from experimental data measured in the AN energy range but using a slightly different scattering angle (145° for Qtac100, 129° for TOF LEIS ACOLISSA).
In the region where the CI process is dominant, the characteristic velocity reaches values of 0.38 × 105 m/s for Ag (100) and 0.22 × 105 m/s for polycrystalline Ag. The CI neutralization and reionization significantly influence the charge exchange at energies higher than Eth and lower values of characteristic velocities as pointed out by Draxler et al. (Ref. 23).
For Ag (100), the concentration is nAg(100) = 12.1 × 1014atoms/cm2.
SPECIMEN DESCRIPTION (ACCESSION # 01979)
Specimen: Ag (100)
CAS Registry #: 7440-22-4
Specimen Characteristics: Homogeneous; solid; single crystal; conductor; metal; other
Chemical Name: Silver
Source: MTI Corporation
Composition: Silver
Form: Single crystal substrate
Structure: CCP
History and Significance: Silver is an important material for electronics known for its high conductivity.
As Received Condition: High-purity (99.999%) substrate, 0.5 mm thick
Analyzed Region: Same as specimen
Ex Situ Preparation/Mounting: As received, the sample was mounted on a conductive sample holder.
In Situ Preparation: Sputtering by 5 keV Ar+, fluence of 2 × 1017 ions cm−2 (2.5 × 2.5 mm), annealing up to 770 K for recrystallization
Charge Control: None
Temp. During Analysis: 300 K
Pressure During Analysis: 5 × 10−7 Pa
Partial Pressure of Reactive Gases During Analysis: N/A
Preanalysis Beam Exposure: None
SPECIMEN DESCRIPTION (ACCESSION # 01980)
Specimen: Ag poly
CAS Registry #: 7440-22-4
Specimen Characteristics: Homogeneous; solid; polycrystalline; conductor; metal; other
Chemical Name: Silver
Source: Alfa Aesar Chemicals, USA
Composition: Silver
Form: Polycrystalline foil
History and Significance: Silver is an important material for electronics known for its high conductivity.
As Received Condition: High-purity foil (99.9985%), 0.5 mm thick
Analyzed Region: Same as specimen
Ex Situ Preparation/Mounting: As received, the sample was mounted on a conductive sample holder.
In Situ Preparation: Sputtering by 5 keV Ar+, fluence of 2 × 1017 ions cm−2 (2.3 × 2.3 mm), and by 5 keV Ne+, fluence of 1 × 1016 ions cm−2 (2.0 × 2.0 mm)
Charge Control: None
Temp. During Analysis: 300 K
Pressure During Analysis: 5 × 10−7 Pa
Partial Pressure of Reactive Gases During Analysis: N/A
Preanalysis Beam Exposure: None
REFERENCE SAMPLE FOR QUANTIFICATION (ACCESSION #01981)
Chemical Name: Copper (Cu poly)
Source: Alfa Aesar Chemicals, USA
Homogeneity: Homogeneous
Form: Solid
Specific Surface Area: N/A
Comment: High-purity foil (99.999%), 0.025 mm thick
Ex Situ Preparation/mounting: As received, the sample was mounted on a conductive sample holder.
In Situ Preparation: Sputtering by 5 keV Ar+, fluence of 2 × 1017 ions cm−2 (2.0 × 2.0 mm), sputtering by 5 keV Ne+, fluence of 2 × 1016 ions cm−2 (2.0 × 2.0 mm)
Charge Control: None
Temp. During Analysis: 300 K
Pressure During Analysis: 5 × 10−7 Pa
Partial Pressure of Reactive Gases During Analysis: N/A
INSTRUMENT DESCRIPTION
Manufacturer and Model: IONTOF GmbH (Münster, Germany)
Qtac100
Analyzer Type: Double Toroidal
Detector: Other
Number of Detector Elements: Continuous
INSTRUMENT PARAMETERS COMMON TO ALL SPECTRA
Spectrometer
Analyzer Mode: Constant pass energy
Energy Dependence of Detection: Constant
Charge Compensation Energy: Not used
Time of Flight Filter Used: No
Time of Flight Filter Comment: N/A
Ion sources
Ion source 1 of 2
Purpose of this Ion Source: Analysis beam
Manufacturer and Model: IONTOF GmbH (Münster, Germany), electron impact ion source
Energy: 1000–4500 eV
Current: (2.1–5.3) × 10−6 mA
Current Measurement Method: Faraday cup
Species: 4He
Spot Size (unrastered): 50 μm
Raster Size: 1500 × 1500 μm
Incident Angle: 0°
Polar Angle: N/A
Azimuthal Angle: 0°–360°
Scattering Angle: 145°
Comment: Qtac100 instrument is equipped with a single ion source used either for sputtering or for analysis. For Ag analysis, the lowest and the highest value of the primary ion beam energy and its current are stated.
Ion source 2 of 2
Purpose of this Ion Source: Sputtering beam
Manufacturer and Model: IONTOF GmbH (Münster, Germany), electron impact ion source
Energy: 5000 eV
Current: 0.3 × 10−3 mA
Current Measurement Method: Faraday cup
Species: 40Ar, 20Ne
Spot Size (unrastered): 50 μm
Raster Size: 2500 × 2500 (2300 × 2300; 2000 × 2000) μm
Incident Angle: 0°
Polar Angle: N/A
Azimuthal Angle: 0°–360°
Scattering Angle: 145°
Comment: Ion beam settings for scattering and sputtering. Qtac100 instrument is equipped with a single ion source used either for sputtering or for analysis.
DATA ANALYSIS METHOD
Energy Scale Correction: No correction
Peak Shape and Background Method: Peak energies, widths (FWHMs), and areas were obtained after subtraction of LEIS background by an error function and a Gaussian function fit as reported in Cushman/Brongersma/Linford.
Quantitation Method: None
Sensitivity Factor (source): Relative sensitivity factor of Ag (100) and polycrystalline Ag to polycrystalline Cu is 2.14 and 1.39, respectively (Ei = 3000 eV).
SUPPLEMENTARY MATERIAL
See the supplementary material for spectral data files 01979-01 through 01979-39, 01980-01 through 01980-08 and 01981-01 as well as the supplemental Spectral Features Table.
ACKNOWLEDGMENTS
Fund of Science (FV22-15) project funded by Faculty of Mechanical Engineering, Brno University of Technology is gratefully acknowledged for the financial support of this work, and CzechNanoLab project LM2023051 funded by MEYS CR is gratefully acknowledged for the financial support of the measurements/sample fabrication at CEITEC Nano Research Infrastructure.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Jan Staněk: Data curation (equal); Investigation (lead); Writing – original draft (lead). Stanislav Průša: Supervision (lead); Writing – review & editing (lead). Tomáš Strapko: Data curation (equal). Tomáš Šikola: Funding acquisition (lead).
DATA AVAILABILITY
The data that support the findings of this study are available within the article and its supplementary material.