D.E. Huber

Development and application of MIPAR™: a novel software package for two- and three-dimensional microstructural characterization

Three-dimensional microscopy has become an increasingly popular materials characterization technique. This has resulted in a standardized processing scheme for most datasets. Such a scheme has motivated the development of a robust software package capable of performing each stage of post-acquisition processing and analysis. This software has been termed Materials Image Processing and Automated Reconstruction (MIPAR™). Developed in MATLAB™, but deployable as a standalone cross-platform executable, MIPAR™ leverages the power of MATLAB’s matrix processing algorithms and offers a comprehensive graphical software solution to the multitude of 3D characterization problems. MIPAR™ consists of five modules, three of which (Image Processor, Batch Processor, and 3D Toolbox) are required for full 3D characterization. Each module is dedicated to different stages of 3D data processing: alignment, pre-processing, segmentation, visualization, and quantification.With regard to pre-processing, i.e., the raw-intensity-enhancement steps that aid subsequent segmentation, MIPAR’s Image Processor module includes a host of contrast enhancement and noise reduction filters, one of which offers a unique solution to ion-milling-artifact reduction. In the area of segmentation, a methodology has been developed for the optimization of segmentation algorithm parameters, and graphically integrated into the Image Processor. Additionally, a 3D data structure and complementary user interface has been developed which permits the binary segmentation of complex, multi-phase microstructures. This structure has also permitted the integration of 3D EBSD data processing and visualization tools, along with support of additional algorithms for the fusion of multi-modal datasets. Finally, in the important field of quantification, MIPAR™ offers several direct 3D quantification tools across the global, feature-by-feature, and localized classes.

Three-dimensional characterisation of the microstructure of an high entropy alloy using STEM/HAADF tomography

Abstract The microstructure of a high entropy alloy with composition of Mo0.5Al1Nb1Ta0.5Ti1Zr1 (the digits refer to molar volumes) has been characterised directly in three dimensions using TEM dark field (DF) imaging and by recording tilt pair micrographs using STEM high angle annular DF (HAADF) imaging. The microstructure contains disordered bcc precipitates that appeared as orthogonal stacks of plate-like features. A tapered needle sample was prepared in a focused ion beam/SEM and was used to acquire a 180° tomographic dataset of STEM/HAADF images in 2° increments. The tilt series images were registered, and the algebraic reconstruction technique was used to reconstruct the three-dimensional microstructure. The bcc precipitates were segmented using a combinative approach involving two threshold techniques. The precipitates were then visualised using commercial software, which revealed surprisingly the existence of both cuboidal and plate-like morphologies. Colouring each precipitate according to its morphology (determined using the omega-2 moment invariant) revealed a precipitate arrangement where plate-like features appeared parallel to each cuboid face.

MIPAR™: 2D and 3D Image Analysis Software Designed by Materials Scientists, for All Scientists

Many software programs have been developed independently for 2D and 3D image analysis, both commercial and open source. However, few, if any, can equip users with extensive toolsets for 2D and 3D materials characterization in a single package. To this end, Materials Image Processing and Automated Reconstruction (MIPAR) has been developed. MIPAR is based on an app-suite construction. The apps were designed as standalone programs, each suited for different tasks, but each capable of communication with the others. This paper will discuss the capabilities and purpose of each of these salient applications, as well as describe MIPAR’s typical 2D and 3D characterization workflows.

In-situ STEM-EELS observation of ferroelectric switching of BaTiO3 film on GaAs

Ferroelectric metal-oxide thin films grown on semiconductor substrates are being studied due to their potential applications in non-volatile single transistor memory elements [1]. Epitaxial single-crystalline BaTiO3 (BTO) thin film was successfully grown on polar GaAs substrate with a SrTiO3 (STO) interlayer using molecular beam epitaxy method [2]. The spontaneous polarization of the BTO film was characterized using scanning transmission electron microscope (STEM) along with electron energy-loss spectroscopy (EELS) [3], and the macroscopic ferroelectric switching under coercive electrical biases was measured using piezoresponse force microscopy [4]. However, the microscopic ferroelectric behavior and the switching mechanism at the interfaces have yet to be demonstrated. In this work, we performed an atomic-scale STEM-EELS study of the ferroelectric switching of BTO film on STO-buffered GaAs substrate with in-situ electrical biasing.

Composition of Epitaxial ZrO 2 :Y2O3/SrTiO 3 Heterostructures

Thermally activated oxygen ion conductivity in the electrolyte of a solid oxide fuel cell (SOFC) device is critical to its successful operation. The current generation of bulk solid oxide electrolytes requires temperatures in the range of 700 to 1000 ̊C in order to achieve the required level of oxygen conductivity, which leads to longevity issues due to thermal stress and fatigue and the requirement for more costly materials [1]. The search for SOFC materials with enhanced ion conductivity at low temperatures has lead researchers to the arena of strained epitaxial heterostructures. Reports of ion conductivity increases of several orders of magnitude in highly strained multilayers consisting of alternating layers of yttria stabilized zirconia (YSZ) and SrTiO3 (STO) [2] have created a great deal of excitement in the field SOFC development.

Novel Investigative Preparation of Human Hair

Preparing hair samples for electron microscopy has been problematic for various reasons. Keratinized hair is densely packed and inherently dry with the proteins heavily cross-linked[1]. This eliminates the need for primary and secondary fixation. However, the inability to uniformly stain through the cuticle layers and throughout the central cortex has shown varying results[1]. The hair samples in this investigation were treated with an oxidative permanent colorant and washed in tap water containing low levels of copper (Cu). The amount of Cu in the hair was confirmed using inductively coupled plasma optical emission spectroscopy (ICP-OES). In order to optimize and better understand the effect of sample preparation on maintaining the native hair structure as well as internal chemical composition, analytical electron microscopy (AEM) characterization was performed [2,3]. In particular, techniques such as S/TEM-HAADF imaging and Super-X XEDS were used to investigate the ultrastructure and chemistry of the cross sectional surface of several hair fibers.

MIPAR™: 2D and 3D Image Analysis Software Designed for Materials Scientists, by Materials Scientists

Many software programs have been developed independently for 2D and 3D image analysis, both commercial and open source. However, few, if any, can equip users with extensive toolsets for 2D and 3D materials characterization in a single package. To this end, Materials Image Processing and Automated Reconstruction (MIPAR) has been developed. MIPAR is based upon a modular (i.e. appsuite) construction. The apps were designed as standalone programs, each suited for different tasks, but each capable of communication with the others. This paper will discuss the capabilities and purpose of each of these salient applications, as well as describe MIPAR’s typical 2D and 3D characterization workflows.

Characterization of Stannous Fluoride Uptake in Human Dentine by Super-X XEDS and Dual-EELS analysis

Stannous fluoride (SnF2) is a common additive to dental products and has been shown to reduce the dental hyper-sensitivity in patients. In order to elucidate and better understand the permeability and mass transport mechanisms, analytical electron microscopy (AEM) characterization was performed on human dentin exposed to SnF2 [1]. In particular, techniques such as S/TEM-HAADF imaging, Super-X XEDS and dual-EELS have been used to investigate the ultrastructure and chemistry of the inner dentine tubule surface. Results on the characterization of the Sn-reacted product on the inner surface of dentine microtubules, as well as the role of dentine “nano-tubules” that branch off from the primary microtubule will be discussed.

Considerations for Physical Facility Design and Management of a State-of-the-Art Electron Microscopy and Analysis Laboratory

Modern electron microscopes demand a physical environment to enable them to achieve or even surpass the manufacturers performance specifications. This can prove to be formidable challenge to facility managers and researchers. Factors that commonly influence the performance of electron microscopes include, but are not limited to temperature fluctuations, air currents and pressure changes, electromagnetic fields, vibrational and acoustic disturbances. However, with proper considerations given to design and implementation of the physical microscopy facility, the influence of all of these factors can be mitigated. The Center for Electron Microscopy and Analysis (CEMAS), at The Ohio State University, opened in September 2013, was designed with the goal of optimizing the environment for state of the art microscopy.

Super-X XEDS STEM Tomography of y' Precipitates in the LSHR Nickel Superalloy

Electron tomography has been proposed to be advantageous for reconstructing sub-micron precipitates in nickel superalloys. However, not every type of electron image provides the necessary contrast for high fidelity tomographic reconstruction. The γʹ′ precipitates in Low Solvus High Refractory (LSHR) form coherently with the matrix and exhibit limited atomic number contrast relative to the matrix when imaged with traditional transmission (TEM) and scanning transmission electron microscopy (STEM) bright field/dark field imaging. Previous research has shown that energy filtered TEM (EFTEM) has displayed successfully γʹ′ morphologies based off the Cr-L3,2 edge[1,2]. However, EFTEM is hindered by poor signal to noise during image collection requiring prohibitive acquisition times. X-ray energy dispersive spectroscopy (XEDS) in STEM offers a robust method to collect a wide range of compositional information and enables spectral image (SI) collection that permits post-processing analysis of multiple elemental species. The primary objective of the work to be presented was to use Super-XTM XEDS in STEM to acquire compositional maps of LSHR to produce tomographic reconstructions that accurately depict γʹ′ precipitates within a dual-microstructure heat treatment (DMHT) gradient [3] as shown in Figure 1(a). Chromium is a significant component of the LSHR alloy at 12.3 wt% and segregates to the γ matrix, therefore chromium XEDS SI can provide strong contrast between the γʹ′ precipitates and the chromium-rich matrix. The XEDS SI shown in Figure 1(b) was used for the full precipitate reconstruction as displayed in Figure 1(c). The DMHT gradient contains a range of intricate γʹ′ morphologies that have proven difficult to characterize with two-dimensional methods. Three-dimensional characterization not only provides the morphology of the γʹ′, but will also permit more accurate microstructural metrics for inputs for integrated computational materials (ICME) models.