AVS2010 Session AS-TuA: Advances in Surface and Interface Imaging

1. It is difficult to detect any peak shifts between chemical state differences because almost peaks in an AES spectrum are broader than XPS.
2. It is difficult to measure stably a standard AES spectrum for an insulator sample because a focused electron beam irradiation causes stronger surface charging.
3. A focused electron beam often makes some elements change to another chemical state.

The influence of embedded nanoclusters on the optical, magnetic and electrical properties of bulk and surface oxides has been an active area of investigation. The establishment of new atom probe tomography (APT) and related high-resolution chemical imaging facilities at EMSL, the Environmental Molecular Sciences Laboratory, provides a world-class user facility for performing nanoscale microscopy. In this study we report on Au-rich nanoclusters that have been embedded into MgO and TiO₂ substrates. The effect of high temperature annealing on the properties of the matrix and the secondary phase (Au) are studied in detail. Electron microscopy analysis has shown that the embedded metal particles are often associated with various defects, which further contribute to property modification.

We report the first Local Electrode Atom Probe (LEAP^®) analysis of bulk MgO and TiO₂ implanted with 2 MeV Au ions using the accelerator facility at EMSL. Both as-implanted and annealed samples were critically analyzed using a combination of APT and the results are compared with high-angular annular dark-field scanning transmission electron microscopy (HAADF STEM) imaging. High-resolution transmission electron microscopy (HRTEM) clearly resolves the Au-rich nanoclusters and allows observation of the pronounced vacancy clustering associated with these features [1]. These Au-rich nanoclusters were also observed in the atom probe data with the average cluster size (~ 5 nm diameter) in good agreement with those seen using HRTEM. The APT technique, however, due to the high three-dimensional (3D) spatial resolution, is also able to detect the presence of finer-scale Au clusters. It can also directly measure residual Au composition within the MgO matrix and any MgO-Au mixing within the clusters. Besides variations in compositional microstructure, evolution of mass spectrum quality as a function of Au content is also observed. Efforts are ongoing in EMSL to confirm this observation and eliminate the possibility of any contribution from experimental artifacts.

During the course of this presentation we will highlight the advantages of using 3D APT in combination with electron microscopy. Specifically, correlative microscopy provides a means to evaluate the capability of APT to detect the presence of the vacancy clusters.

[1] Wang et al., Applied Physics Letters 87, 153104, 2005

A significant problem in the combustion of coal is the release of toxic elements into the atmosphere in gaseous and solid (particulate) forms. In China, villagers use coal, exposed at the surface by erosion, to heat their homes and cook their meals. It is believed that coal combustion is a potential source of endemic diseases [1, 2].

Coal samples were collected from the province of Guizhou in China. In this study, an ION-TOF V equipped with a bismuth primary ion source was used to analyze coal directly without any chemical treatment. Mechanical polishing was performed to create a flat surface for analysis. Analyses revealed that there are distinctions between the organic and inorganic phases of coal. Fluorine was been determined to come from an inorganic phase rather than an organic phase. Two different forms of elemental sulfur were determined through ion imaging. The first one was determined to be organic in nature, while, the second derives from pyrite (FeS₂).

The high spatial resolution of TOF-SIMS can be used to distinguish between different domains in coal, revealing the relationship between specific components suspected of involvement in toxicity of particulate emissions and the coal components.

Spectral shifts in STXM NEXAFS spectra allow carbonate species to be distinguished from carboxylic species. With the help of reference materials, aromatic species could be distinguished from aliphatic species. However, using TOF-SIMS for distinguishing such species would prove to be very difficult, which is why STXM is a good complementary technique for coal studies.

[1] Finkelman et al. (2004) Int. J Coal Geol. 59, 19–24.

[2] Finkelman et al. (2002) Int. J Coal Geol. 50, 425– 443.

TOF-SIMS characterization of materials in the range of several microns from the sample surface has become somewhat routine. Nevertheless, there are practical limitations to the use of ion beam sputtering for probing both organic and inorganic specimens beyond the surface region. Certain matrix components do not sputter well and are susceptible to ion beam-induced molecular damage. This accumulated beam damage gives rise to incorrect molecular distributions. Some matrix components may sputter at a different rate than others which results in a misrepresentation of the elemental and molecular distributions. Finally, the time requirements to achieve uniform (i.e. representative) depth profile analysis under ideal instrumental conditions can become prohibitive. Even under optimized experimental conditions, the efficacy of sputter depth profiling for 3D TOF-SIMS imaging is limited to < 5 μm in the case of a favorable matrix and to < 300 nm in the case of an unfavorable matrix. An alternative approach for 3D TOF-SIMS imaging the interior of a specimen is to utilize FIB milling and sectioning. With FIB milling, the interior of a specimen is revealed to depths of ~ 50 μm within a reasonable analytical timeframe. Additionally, 3D chemical imaging of ~ 10 μm deep volumes may be achieved in the same time it would take to perform a low voltage sputter depth profile. The advantage of the FIB-TOF approach is that the artifacts caused by sputter depth profiling, i.e. differential sputtering and accumulated ion beam damage to matrix molecules, are avoided. The union of successive FIB sectioning and TOF-SIMS analysis cycles to achieve 3D chemical imaging will be discussed and illustrated using inorganic and organic examples.

Poly(L-lactic acid) (PLLA) is a synthetic, bioresorbable polyester that is extensively used and studied for many [FDA-approved] commercial applications – such as therapeutic drug delivery and tissue engineering scaffolds.¹ It is generally accepted that the degradation process of bioresorbable polyesters: a) is diffusion-based,² b) occurs in a region of finite thickness, forming an erosion front that moves towards the center of a polymer structure,² and c) its rate increases with higher polymer surface area.³ Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS) has been previously used to determine initial (before the onset of weight loss) degradation kinetics at bioresorbable polymer surfaces.⁴ This analytical method, combined with the high lateral resolution capabilities of imaging with a bismuth ion source, and image processing algorithms, allowed us to determine the spatially-resolved initial degradation kinetics of micropatterned PLLA membranes at several pH levels. The results show that the degradation reaction occurs at different rates, and that these rates depend on the area of the feature. The findings of this study imply that polymer degradation can be controlled not only in a temporal manner, but also in a microspatial manner, by altering micropattern geometry and size distribution across the polymer membrane.

1. Lee, J, Gardella, Jr., J. Analytical and Bioanalytical Chemistry. 2002. 373. 526.

2. Mathiowitz, E, Jacob, J, Pekarek, K, Chickering III, D. Macromolecules. 1993. 26. 6756.

3. Buchanan, FJ, 2008. Degradation Rate of Bioresorbable Materials: Prediction and Evaluation. Cambridge, UK. Woodhead Publishing Limited.

4. Lee , J, Gardella Jr., J. Analytical Chemistry. 2003. 75. 2950.

Applications of multivariate analysis (MVA) methods to surface analysis imaging datasets have increased quite significantly in recent years. Multivariate analysis of 3D imaging TOF-SIMS and XPS data is now quite widely applied (even routinely used by some research groups). Data preprocessing, scaling and selection of appropriate multivariate analysis method have been discussed in the literature quite extensively in a couple of recent years.

What has been addressed in a lesser extent is an important problem of combining quantitative analysis of imaging datasets from the same analytical method from various samples (multi-sample 3D imaging) or combining various analytical data (spectroscopic, imaging, or scalar) obtained for the same sample (multi-modal analysis).

Approaches to quantitatively combine imaging datasets from multiple samples will be discussed on example of multispectral XPS imaging data sets acquired from various paper samples.

Multi-modal analysis will be discussed on example of combining spectroscopic (XPS, XANES), microscopic (SEM) and macroscopic (BET surface area and pore size distribution) data from set of non-Pt group metal electrocatalysts for oxygen reduction reaction treated at different temperatures.

Fundamentally, photovoltaic and fuel cell devices rely upon interfaces for electrical power generation. However, the undesirable formation of poor quality interfaces can also serve to decrease power efficiency. In the case of photovoltaics, interfaces control the generation and extraction of photogenerated charge carriers; however, the formation of dislocations and dopant clustering can result in recombination centers, thus reducing the ability to extract charge carriers. In fuel cells, the three phase boundary between the electrode, gas, and electrolyte controls the cell power output; however, surface contamination at this interface can reduce the electrochemical reaction rate, and thus the power output of the cell. Understanding both of these interfaces at the atomic structural and chemical level allows for a greater understanding of the formation of interface degradation. In order to fully understand the atomic scale chemistry and structure of interfaces in photovoltaics and fuel cells, we have applied a combination of in-situ FIB / SEM electrical probing using EBIC and ex-situ impedance spectroscopy with high resolution analytical STEM imaging and laser pulsed atom probe tomography. These techniques have been applied to III-V based photovoltaics to gain an understanding of dopant profiling across quantum structures and tunnel junctions, and to probe the initial stages of phase separation in multicomponent epilayers. Similarly, EBIC has been used to identify dislocations and grain boundaries in polycrystalline Si photovoltaics, and to determine the atomic level chemistry and structure at these interfaces that leads to an interface acting as a recombination center. Finally, a combination of STEM and atom probe tomography have illustrated 10-17 / cm^3 changes in local Pt, C, and O chemistry and structure around Pt catalysts for use in polymer electrolyte fuel cells. In order to enable atom probe analysis on materials with widely varying field evaporation characteristics, new FIB specimen preparation techniques were required. Details on the complex experimental methods and instrumentation developed in order to enable all of these investigations are illustrated.