Comparative studies of MCs+-SIMS and e?-beam SNMS for quantitative analysis of bulk materials and layered structures

For the quantification of heterostructure depth profiles the knowledge of relative sensitivity factors (RSF) and the influence of matrix effects on the measured profiles is necessary. Matrix dependencies of the measured ion intensities have been investigated for sputtered neutral mass spectrometry (SNMS) and MCs⁺-SIMS. The use of Cs as primary ions for SNMS is advantageous compared to Ar because the depth resolution is improved without changing RSFs determined under Ar bombardment. No significant amount of molecules has been found in the SNMS spectra under Cs bombardment. Using MCs⁺-SIMS the RSFs are matrix dependent. An improvement of depth resolution can be achieved by biasing the sample against the primary ion beam for SNMS due to a reduction of the net energy of the primary ions and a resulting more gracing impact angle.     

MCs+ secondary ion and sputtering yields of oxygen-exposed semiconductors and glasses

The emission of MCs⁺ molecular ions sputtered by Cs⁺ ion impact from a variety of elemental (Si and Ge) and compound (GaAs, InP, InSb, ZnSe, CdS, CdSe, CdTe and CdZnTe) semiconductors and a selection of glass samples of different composition has been investigated. For the glass samples a set of relative sensitivity factors has been determined which are largely composition-independent and provide the possibility of a reliable quantification of glasses by MCs⁺ SIMS. For the semiconductors fractional ion yields (i.e. the number of detected MCs⁺ ions per sputtered M atom) range from 10^?6 to some 10^?4 and exhibit little variation with the oxygen surface coverage of the specimen. Depending on M, the emission of MOCs⁺ molecular species becomes prominent (or even dominating) at high oxygen concentrations. Furthermore, total sputtering yields for 5.5 keV Cs⁺ impact and different oxygen partial pressures have been determined from sputtered craters.     

MCs+ secondary ion and sputtering yields of oxygen-exposed semiconductors and glasses

The emission of MCs⁺ molecular ions sputtered by Cs⁺ ion impact from a variety of elemental (Si and Ge) and compound (GaAs, InP, InSb, ZnSe, CdS, CdSe, CdTe and CdZnTe) semiconductors and a selection of glass samples of different composition has been investigated. For the glass samples a set of relative sensitivity factors has been determined which are largely composition-independent and provide the possibility of a reliable quantification of glasses by MCs⁺ SIMS. For the semiconductors fractional ion yields (i.e. the number of detected MCs⁺ ions per sputtered M atom) range from 10^–6 to some 10^–4 and exhibit little variation with the oxygen surface coverage of the specimen. Depending on M, the emission of MOCs⁺ molecular species becomes prominent (or even dominating) at high oxygen concentrations. Furthermore, total sputtering yields for 5.5 keV Cs⁺ impact and different oxygen partial pressures have been determined from sputtered craters.     

Relative sensitivity factors for hard X-ray photoelectron spectroscopy with photon energies of 3.0, 5.9, 7.9, and 9.9 keV

Empirical relative sensitivity factors (RSFs) for the 1s, 2p_3/2, and 3d_5/2 levels relative to O1s were derived from the hard X-ray photoelectron spectroscopy measurements with photon energies of 3.00, 5.95, 7.94, and 9.92 keV. The data for 5.95, 7.94, and 9.92 keV were obtained at BL46XU in SPring-8, and those for 3.00 keV were obtained at BL6N1 in AichiSR (note that the measurement conditions, i.e., electron spectrometer and measurement geometry, at the two beamlines were different; for details, see Section >3). It was found that the empirical RSFs showed the linear behaviors as a function of the binding energy on a log–log plot. Also, from the comparison of the empirical RSFs and the theoretical RSFs calculated from Hartree–Slater cross sections reported by Scofield combined with the energy dependence on the spectrometer function and the inelastic mean free path, it is observed that there is an agreement between the empirical and the theoretical RSF values with several exceptions.     

Quantitative determination of element distributions in silicon based thin film solar cells using SNMS

The determination of elemental distributions in thin film solar cells based on amorphous silicon using electron beam SNMS is possible by quantifying the measured ion intensities. The relative sensitivity factors (RSFs) for all elements measured have to be known. The RSFs have been determined experimentally using implantation and bulk standards with known concentrations of the interesting elements. The measured RSFs have been compared with calculated RSFs. The model used for the calculation of the RSFs takes into account the probability for electron impact ionization and the dwell time of the neutrals inside the postionization region. The comparison between measured and calculated RSF shows, that this model is capable to explain the RSFs for most elements. Differences between calculated and measured values can be explained by the formation of hydride and fluoride molecules (in case of H and F) and influences of the angular distribution of the sputtered neutrals in case of Al. The experimentally determined RSFs have been used for a quantification of depth profiles of the i-, buffer-, p- and front contact layers of a-Si solar cells.     

Quantitative determination of element distributions in silicon based thin film solar cells using SNMS

The determination of elemental distributions in thin film solar cells based on amorphous silicon using electron beam SNMS is possible by quantifying the measured ion intensities. The relative sensitivity factors (RSFs) for all elements measured have to be known. The RSFs have been determined experimentally using implantation and bulk standards with known concentrations of the interesting elements. The measured RSFs have been compared with calculated RSFs. The model used for the calculation of the RSFs takes into account the probability for electron impact ionization and the dwell time of the neutrals inside the postionization region. The comparison between measured and calculated RSF shows, that this model is capable to explain the RSFs for most elements. Differences between calculated and measured values can be explained by the formation of hydride and fluoride molecules (in case of H and F) and influences of the angular distribution of the sputtered neutrals in case of Al. The experimentally determined RSFs have been used for a quantification of depth profiles of the i-, buffer-, p- and front contact layers of a-Si solar cells.     

Removal of Ar+ beam‐induced damaged layers from polyimide surfaces with argon gas cluster ion beams

An Ar Gas Cluster Ion Beam (GCIB) has been shown to remove previous Ar⁺ ion beam‐induced surface damage to a bulk polyimide (PI) film. After removal of the damaged layer with a GCIB sputter source, XPS measurements show minor changes to the carbon, nitrogen and oxygen atomic concentrations relative to the original elemental bulk concentrations. The GCIB sputter depth profiles showed that there is a linear relationship between the Ar⁺ ion beam voltage within the range from 0.5 to 4.0 keV and the dose of argon cluster ions required to remove the damaged layer. The rate of recovery of the original PI atomic composition as a function of GCIB sputtering is similar for carbon, nitrogen and oxygen, indicating that there was no preferential sputtering for these elements. The XPS chemical state analysis of the N 1s spectra after GCIB sputtering revealed a 17% damage ratio of altered nitrogen chemical state species. Further optimization of the GCIB sputtering conditions should lead to lower nitrogen damage ratios with the elemental concentrations closer to those of bulk PI. Copyright © 2010 John Wiley & Sons, Ltd.     

Argon cluster cleaning of Ga+ FIB-milled sections of organic and hybrid materials

Secondary ion mass spectrometry studies have been made of the removal of the degraded layer formed on polymeric materials when cleaning focused ion beam (FIB)-sectioned samples comprising both organic and inorganic materials with a 30-keV Ga⁺ FIB. The degraded layer requires a higher-than-expected Ar gas cluster ion beam (GCIB) dose for its removal, and it is shown that this arises from a significant reduction in the layer sputtering yield compared with that for the undamaged polymer. Stopping and Range of Ions in Matter calculations for many FIB angles of incidence on flat polymer surfaces show the depth of the damage and of the implantation of the Ga⁺ ions, and these are compared with the measured depth profiles for Ga⁺-implanted flat polymer surfaces at several angles of incidence using an Ar⁺ GCIB. The Stopping and Range of Ions in Matter depth and the measured dose give the sputtering yield volume for this damaged and Ga⁺-implanted layer. These, and literature yield values for Ga⁺ damaged layers, are combined on a plot showing how the changing sputtering yield is related to the implanted Ga density for several polymer materials. This plot contains data from both the model flat poly(styrene) surfaces and FIB-milled sections showing that these 2 surfaces have the same yield reduction. The results show that the damaged and Ga⁺-implanted layer's sputtering rate, after FIB sectioning, is 50 to 100 times lower than for undamaged polymers and that it is this reduction in sputtering rate, rather than any development of microtopography, that causes the high Ar⁺ GCIB dose required for cleaning these organic surfaces.     

VAMAS interlaboratory study on organic depth profiling

We present the results of a VAMAS (Versailles project on Advanced Materials and Standards) interlaboratory study on organic depth profiling, in which twenty laboratories submitted data from a multilayer organic reference material. Individual layers were identified using a range of different sputtering species (C₆₀ⁿ⁺, Cs⁺, SF₅⁺ and Xe⁺), but in this study only the C₆₀ⁿ⁺ ions were able to provide truly ‘molecular’ depth profiles from the reference samples. The repeatability of profiles carried out on three separate days by participants was shown to be excellent, with a number of laboratories obtaining better than 5% RSD (relative standard deviation) in depth resolution and sputtering yield, and better than 10% RSD in relative secondary ion intensities. Comparability between laboratories was also good in terms of depth resolution and sputtering yield, allowing useful relationships to be found between ion energy, sputtering yield and depth resolution. The study has shown that organic depth profiling results can, with care, be compared on a day‐to‐day basis and between laboratories. The study has also validated three approaches that significantly improve the quality of organic depth profiling: sample cooling, sample rotation and grazing angles of ion incidence. © Crown copyright 2010.     

Nondestructive analysis of single crystals of selenide spinels by X-ray spectrometry techniques

The paper presents possibilities and difficulties in nondestructive analysis of small multielement single crystals performed by means of X-ray spectrometry techniques: micro-X-ray fluorescence spectrometry (μ-XRF), energy-dispersive electron probe microanalysis (ED-EPMA), and X-ray photoelectron spectroscopy (XPS). The capability of the X-ray spectroscopy techniques in elemental analysis is demonstrated with the single crystals of selenide spinels of the general formula M _x N _y Cr _z Se₄ (M⁺² and N⁺³ are, for example, Zn⁺², V⁺³, Ga⁺³, Cd⁺², In⁺³, and Sb⁺³). The results of the nondestructive analyses (μ-XRF, ED-EPMA, and XPS) are compared with those obtained by inductively coupled plasma optical emission spectrometry (ICP-OES) and wavelength-dispersive X-ray spectrometry (WDXRF) following sample digestion. The present study shows satisfactory agreement between the results of μ-XRF analysis performed using the standardless fundamental parameter method and the results obtained with the WDXRF and ICP-OES analyses. If the measured single crystal is precisely positioned, the difference between μ-XRF and wet analysis (WDXRF and ICP-OES) does not exceed 5% rel. The reliable results of ED-EPMA can be obtained only if the measured area is sufficiently large, i.e., of 200 × 300 μm. Even if this condition is fulfilled, the relative difference between the ED-EPMA and the wet analysis may reach 10% rel. In case of the XPS analysis, the accuracy of results depends on the proper preparation of the sample surface. It should be free of contamination that can be obtained by scraping in situ in ultrahigh vacuum. The ion etching, commonly used for cleaning the surface, leads to preferential sputtering; therefore, the reliable results cannot be obtained.     

Optimization of the depth resolution for profiling SiO2/SiC interfaces by dual-beam TOF-SIMS combined with etching

To examine precise depth profiles at the interface of SiO₂/SiC, a high resolution that can detect slight discrepancies in the distribution is needed. In this study, an experimental method to achieve a high resolution of less than 1 nm was developed by using dual-beam time-of-flight secondary ion mass spectrometry (TOF-SIMS). The analysis was preceded by the following three steps: (1) determination of the optimal analytical conditions of the analysis beam (Bi⁺) and sputtering beam (Cs⁺), (2) verification of the etching methods to thin the SiO₂ layer, and (3) confirmation of the benefits of the low-energy sputtering beam directed toward SiO₂/SiC samples. By using the secondary ion intensity peak-to-valley ratio of BN⁻ and BO⁻ of a sample with delta-doped boron multilayers, the appropriate Bi⁺/Cs⁺ condition for a high depth resolution was determined for each energy level of the sputtering beam. Upon verification of the etching methods to thin the SiO₂ layer, slight discrepancies were found between samples that were obtained with different etching methods. The difference in the roughness values of the etched surfaces was proactively utilized for the performance confirmation of the low-energy sputtering beam by means of precise observation of the profiles at the SiO₂/SiC interface. The use of a Cs beam with a low energy between 0.25 and 0.5 keV enabled the detection of slight discrepancies in the roughness of less than 1 nm between samples. The aforementioned method has the potential to accurately detect discrepancies in the intrinsic distribution at the SiO₂/SiC interface among samples.     

Diagnostic and analytical study of post ablation ionization of neutral atoms of major and minor constituents from a low alloy steel

Post ablation ionization (PAI) of neutral atoms from a low alloy steel has been investigated using non-resonant laser ionization in a time-of-flight mass spectrometer. By varying the delay between the ablation and ionization lasers, the velocity distributions of the Ti, V, Cr, Mn and Fe atoms have been determined simultaneously. These distributions have been recorded as a function of ablation laser fluence. The half-range Maxwell-Boltzmann velocity distribution has been used to fit the data and different characteristic temperatures have been determined for the various elements in the sample. The quantitative capability of this method for bulk and surface analysis has been evaluated by calculating the relative sensitivity factors (RSFs) for the various constituent elements. The RSFs for all of the elements are seen to be highly dependent on the delay between the ablating and ionizing lasers. This dependence was reduced by integrating the temporal dependent ion yield, leading to a significant improvement in the calculated RSF values. It was also found that the RSFs were not highly dependent on the power density of the ablation laser beam.     

In-depth compositional analysis of ceramic (Bi2O3)0.75(Er2O3)0.25 by AES and XPS

The chemical composition of dense ceramics of erbia-stabilized -Bi₂O₃ was analyzed by Auger electron spectroscopy (AES) depth profiling using Ar⁺ ion sputtering. The relative sensitivity factors (rsf) and sputter rates of bismuth and erbium in this material have been determined by electron probe microanalysis (EPMA) and chemical analysis. These results, supplemented by data from angle resolved X-ray photoelectron spectroscopy (ARXPS), shows a bismuth enrichment at the surface. Evidence has been found for reduction of the bismuth-oxide at the outermost part of the surface layer.Dedicated to Professor Günther Tölg on the occasion of his 60th birthday     

Al and Ti secondary neutral and secondary ion emission from oxide samples in the high-frequency sputtering mode of HF-plasma SNMS

A strong Al⁺ and a minor Ti⁺ peak without a proportional increase of the O⁺ signal in SNMS high-frequency sputtering mode (HFM) time profiles of an insulating μm-thick oxide layer on Ti-48Al-2Cr-2Nb led us to check for a possible contribution of positive secondary ions (SI⁺). SI⁺ and SI^– (negative secondary ions) can be detected in ion energy spectra. This is shown using Al⁺, O^–, AlO^–, and AlO₂ ^– ions sputtered from massive Al₂O₃. Similarly, and depending on stoichiometry, also Ti⁺ from mixed sintered, microscopically inhomogeneous Al₂O₃-TiO₂-SiO₂ pellets has been identified to be partly SI⁺. The subtraction of an assumed contribution of ionized secondary neutrals (SN⁺) suggests that SI⁺ may form several 10% of the detected ions obtained in the HFM sputtering and plasma processes. However, the positive surface potential of some 10 V being necessary to cause detectable SI⁺ contributions does not build up on μm-thin insulating layers. Therefore, we have to conclude that the Al⁺ and Ti⁺ peaks in the sputter time profiles of the μm-thick oxide layer on Ti-48Al-2Cr-2Nb which are accompanied by an O⁺ deficiency cannot have been caused by SI⁺. Instead, their more probable origin is the inhomogeneous Al₂O₃ interlayer itself. Together with the residues of a topmost TiO₂ layer which has strongly been depleted in O by preferential sputtering, the relative O⁺ deficiency may be explained without assuming SI⁺ contributions. Received: 22 February 1999 / Revised: 1 July 1999 / Accepted: 6 July 1999     

Determination of the relative sputtering yield of carbon to tantalum by means of Auger electron spectroscopy depth profiling

The relative sputtering yield of carbon with respect to tantalum was determined for 1 keV Ar⁺ ion bombardment in the angular range of 70°–82° (measured from surface normal) by means of Auger electron spectroscopy depth profiling of C/Ta and Ta/C bilayers. The ion bombardment‐induced interface broadening was strongly different for the C/Ta and Ta/C, whereas the C/Ta interface was found to be rather sharp, the Ta/C interface was unusually broad. Still the relative sputtering yields (Y_C/Y_Ta) derived from the Auger electron spectroscopy depth profiles of the two specimens agreed well. The relative sputtering yields obtained were different from those determined earlier on thick layers, calculated by simulation of SRIM2006 and by the fitting equation of Eckstein. The difference increases with increase of angle of incidence. Copyright © 2009 John Wiley & Sons, Ltd.     

Al and Ti secondary neutral and secondary ion emission from oxide samples in the high-frequency sputtering mode of HF-plasma SNMS

A strong Al⁺ and a minor Ti⁺ peak without a proportional increase of the O⁺ signal in SNMS high-frequency sputtering mode (HFM) time profiles of an insulating μm-thick oxide layer on Ti-48Al-2Cr-2Nb led us to check for a possible contribution of positive secondary ions (SI⁺). SI⁺ and SI^– (negative secondary ions) can be detected in ion energy spectra. This is shown using Al⁺, O^–, AlO^–, and AlO₂ ^– ions sputtered from massive Al₂O₃. Similarly, and depending on stoichiometry, also Ti⁺ from mixed sintered, microscopically inhomogeneous Al₂O₃-TiO₂-SiO₂ pellets has been identified to be partly SI⁺. The subtraction of an assumed contribution of ionized secondary neutrals (SN⁺) suggests that SI⁺ may form several 10% of the detected ions obtained in the HFM sputtering and plasma processes. However, the positive surface potential of some 10 V being necessary to cause detectable SI⁺ contributions does not build up on μm-thin insulating layers. Therefore, we have to conclude that the Al⁺ and Ti⁺ peaks in the sputter time profiles of the μm-thick oxide layer on Ti-48Al-2Cr-2Nb which are accompanied by an O⁺ deficiency cannot have been caused by SI⁺. Instead, their more probable origin is the inhomogeneous Al₂O₃ interlayer itself. Together with the residues of a topmost TiO₂ layer which has strongly been depleted in O by preferential sputtering, the relative O⁺ deficiency may be explained without assuming SI⁺ contributions. Received: 22 February 1999 / Revised: 1 July 1999 / Accepted: 6 July 1999     

Application of a low impedance contactless conductometric detector for the determination of inorganic cations in capillary monolithic column chromatography

Poly(glycidyl methacrylate) cation exchange monolithic column was prepared in fused-silica capillaries of 320 μm i.d. by thermally initiated radical polymerization and utilized in capillary ion chromatography. With 15 mM methanesulfonic acid as the mobile phase, the separations of a mixture of inorganic cations (Li⁺, Na⁺, NH₄⁺, K⁺) was tested by using a capacitively coupled contactless conductivity detector (C⁴D) and a low impedance C⁴D (LIC⁴D). The LIC⁴D is the series combination of a C⁴D and a quartz crystal resonator. At the resonant frequency of the series combination, the capacitor impedance from capillary wall was offset by the inductance impedance from the quartz crystal resonator. A minimum impedance was obtained in the impedance-frequency curve of the combination. The responses of the C⁴D and LIC⁴D were analyzed based on an equivalent circuit model. It was shown that the sensitivity of the C⁴D to the change in analyte concentration is rather poor due to the high ratio of the impedance from the capillary wall capacitor to the solution impedance. The LIC⁴D has the similar sensitivity as a contact conductivity detector but a much smaller cell volume. The on-column detection model was realized by LiC⁴D without preparation of optical detection window in monolithic column.     

Graphene-Enabled Electric-Field Regulation and Ionic Redistribution Around Lithiophilic Aurum Nanoparticles Toward a Dendrite-Free and 2000-Cycle-Life Lithium Metal Battery

Lithium metal batteries (LMBs) have attracted extensive attention owing to their high energy density. However, the uncontrolled volume changes and serious dendrite growth of the Li metal anode have hindered their commercialization. Herein, a three-dimensional Cu foam decorated with Au nanoparticles and conformal graphene layer was designed to tune the Li plating/stripping behaviors. The 3D−Cu conductive host anchored by lithiophilic Au nanoparticles can effectively alleviate the volume expansion caused by the continuous plating/stripping of Li and reduce the nucleation energy barrier. Notably, the conductive graphene not only facilitates the transfer of electrons, but also acts as an ionic rectifier, thereby avoiding the aggregation of local current density and Li⁺ ions around Au nanoparticles and enabling the uniform Li⁺ flux. As a result, the G−Au@3D−Cu/Li anode ensures the non-dendritic and homogeneous Li⁺ plating/stripping. Electrochemical results show that the symmetric G−Au@3D−Cu/Li cell delivers a low voltage hysteresis of 110 mV after 1000 h at 1 mA cm⁻². Matched with a layered LiNi_0.6Co_0.2Mn_0.2O₂ cathode, the NCM622||G−Au@3D−Cu/Li full cell exhibits a long cycle life of 2000 cycles and an ultra-low capacity decay rate (0.01 % per cycle).     

Formation of small cluster ions from alkali halides in SIMS

An experimental investigation of M₂X⁺ cluster ion formation by alkali halides in secondary ion mass spectrometry is presented as evidence of a gas phase cluster ion formation mechanism. Experimental M₂X⁺ ion yields are compared with calculated heats of formation for M₂X⁺ ions formed from binary mixtures of alkali halides, MX (M = Li, Na; X = F, Cl). The cluster ion intensity order was found to be Li₂F⁺ > Li₂Cl⁺ > Na₂F⁺ > Na₂Cl⁺, consistent with gas phase mechanism predictions. This ordering may be affected, however, by experimental artifacts such as mass discrimination by the quadrupole mass spectrometer, and variations in composition caused by sample preparation or preferential sputtering. Steps are taken to limit quadrupole mass discrimination, and X-ray photoelectron spectroscopy (XPS) is used to examine surface stoichiometry changes induced by sputtering. The XPS study suggests that, although preferential sputtering did occur, it did not influence the relative ordering of cluster ion intensities.