Volume 35(2) - Summer 2019

Refractory products are materials that can withstand high temperatures, above 1500°C. They are used in a wide range of applications, including as the lining of furnaces that perform melting and heating processing of materials for metallurgical, chemical, ceramic, machine, glass industries and so on. There are many types of refractories, including shaped refractories that have been molded and fired beforehand in the form of the final products, powder granules, or paste-like monolithic refractories that are formed into a specific shape at a construction site. Furthermore, refractory products can have different chemical properties. For example, there are acidic refractories that mainly consist of acidic oxides such as SiO₂ and ZrO₂, basic refractories that mainly consist of basic oxides such as MgO and CaO, and neutral refractories. The type of refractory is chosen depending on its intended use. In order to maximize the performance of such refractories, it is necessary to precisely control their elemental composition to meet the needs of specific applications.

Analysis of refractories can be performed according to standardized methods prescribed by Japanese Industrial Standards (JIS R 2216) and ISO 12677 (2011), utilizing X-ray fluorescence (XRF) spectrometry, which is known as a rapid and accurate quantitative analysis method for elemental analysis(2). To obtain accurate analysis results, samples are prepared by the fusion bead method to eliminate grain size and mineralogical effects.

In order to meet customer needs, Rigaku was the first company to release quantitative application packages, including for refractories. These application packages have been well received for their ability to easily and accurately perform quantitative analysis without any specialized technical skills.

This paper describes an analysis example using an application package for quartzite refractory products, an acidic refractory. Quartzite refractories are effective for repeated heating and cooling cycles due to their small change in volume above 600°C. In addition, due to their excellent heat properties, they are widely used as construction furnaces for coke ovens, hot stoves, and glass melting chambers. It is necessary to add 4 to 5 mass% of Al₂O₃ or Fe₂O₃ etc. as a sintering aid to quartzite refractories. However, when used in a glass melting chamber, it is necessary to use a low-porosity Al₂O₃ or other low-alkali component to avoid attack by alkaline vapor in the atmosphere, and accurate analysis of coexisting elements is also required.

Raman spectrometry is becoming a common method for identification of hazardous compounds and illicit drugs. Historically, infrared absorption spectroscopy was the common method, but the mainstream has gradually shifted to Raman spectroscopy. In particular, Raman spectrometry has a couple of advantages over infrared spectrometry for onsite rapid analysis. First, handheld Raman spectrometers can analyze contents through transparent or translucent containers, which infrared absorption spectrometers cannot. Second, Raman spectrometer can analyze wet samples, but infrared absorption spectrometer cannot.

In the past, handheld Raman spectrometers had issues with overlap of fluorescence, depending on sample type. The reason behind this was that the wavelength of excitation laser was mostly 785 nm. The novel handheld Raman spectrometers for hazardous materials and drugs with 1064 nm excitation laser described in this article can reduce the overlap of fluorescence drastically. Other advantages of the novel handheld Raman spectrometers are described in the later sections.

X-ray computed tomography (CT) is a nondestructive imaging technique that can be used to examine the internal features of an object in three dimensions (3D). The first commercial X-ray CT scanner was introduced 45 years ago and the technique has been widely used in the medical and industrial fields since. Recently X-ray CT for examining microstructures of size less than 10 μm (X-ray micro-CT) has seen rapid growth in the research and development of materials. This method is mainly used in the automotive industry to analyze defects such as voids or cracks in castings and resin molded products, and in the electronics industry to evaluate bonding defects in electronic printed circuit boards (PCBs).

As the X-ray micro-CT technique has become more popular, the demand for new sample analysis applications is emerging. One of these applications is examination of micrometer-scale structures of electronic components or PCBs. Typical X-ray micro-CT systems can be used to examine such objects, but their spatial resolution is not very high (5 μm). Therefore, it has been difficult for these instruments to render the internal microstructure of those objects in detail needed for this application.

Another application is examination of engineering materials made up of only carbon-containing compounds, that is, polymers and carbon fiber reinforced polymers (CFRP), which are essential to develop lightweight parts, composite materials, etc. In order to meet such demand, we have developed the nano3DX X-ray CT microscope, combined with a chromium (Cr), copper (Cu), or molybdenum (Mo) target, each of which emits highly monochromatic, long wavelength X-rays yielding a large interaction cross section (high sensitivity) for the materials described above.

Using these X-rays, the nano3DX can examine structures nondestructively at the 1 micrometer scale, in objects such as CFRPs, pharmacological agents and chemical products, with a sensitivity one order of magnitude greater than that of conventional X-ray micro-CT systems. As the result, identifying the relationship between internal structure and material properties, which can’t be done by the cross section method (e.g., conventional optical microscope, electron microscope), is now possible.

The presence of crystallographic texture (preferred orientation) in polycrystalline materials has a significant effect on the anisotropy of the properties of these materials. That means that quantitative description of the orientation distribution of crystallites, or the orientation distribution function (ODF), is an important task for materials characterization and prediction of their properties. Direct measurement of the ODF is not possible; instead, pole figures (PF) can be measured to determination the ODF. Reconstruction of the ODF from measured PFs is a main goal of quantitative texture analysis. Thus, two problems should be solved to obtain an ODF: measurement and processing of experimental PFs and ODF reconstruction from PFs.

In the X-ray diffraction technique, there are two basic modes for PFs measurement: the conventional mode with a 0D detector and a more advanced mode using a 2D detector(1). While measurement of PFs with 2D detectors is more advanced, it requires additional tools for conversion of the detector’s data into PFs.

When the PFs are prepared, it is possible to start the ODF reconstruction process. Currently, three methods are used for ODF reconstruction: the series expansion method(3), the component method, and direct methods like WIMV or ADC. Each method has advantages and disadvantages. The series expansion method is more general, but it requires a large number of measured PFs and has some problems with numerical calculations. The components method represents the ODF as a set of model functions (components) that have clear physical meaning. This method is most convenient for interpretation and representation of the results, but can require a lot of time for selection of components and fitting their parameters. Direct methods use a numerical calculation of the ODF on a discrete grid in rotation space. They are the most simple and convenient to use but do not provide an interpretation of the ODF.

In the next sections we will describe the Texture plugin of SmartLab Studio II, which is intended for data processing and quantitative texture analysis. This plugin implements two of the above-mentioned methods of ODF reconstruction: WIMV and the components methods. Both can be used for all types of crystal systems and two types of sample symmetry—triclinic and orthorhombic. Also, the plugin can use three of the most popular definitions of Eulerian angles in the texture community: Bunge notation (φ1,Φ,φ2), Roe notation (Ψ,Θ,Φ) and Matthiers notation (α, β, γ). Roe and Matthiers notation are physically equivalent, with the only difference being the letters used in the notation.

Quantitative phase analysis (QPA) using the X-ray diffraction technique is routinely employed to find weight ratios of individual component phases in a mixture. Techniques for QPA have been widely used not only in research and development but also routinely deployed for quality control of industrial products. Various techniques have been proposed for QPA in past decades. Some techniques are designed for exclusive use in QPA of specific materials such as zirconia, silicon nitrides etc., while the calibrationcurve method, the reference intensity ratio (RIR) method, and the Rietveld method have been applied to QPA of general materials. The direct derivation method (DDM) is also a QPA technique suitable for use with general materials. The observed diffraction pattern of a mixture is the superposition of component patterns for individual phases. In conducting QPA, the observed diffraction intensities of the mixture must be separated into intensity datasets of the individual components. In deriving the weight ratios from intensity datasets, the Rietveld method uses crystal structure parameters while the RIR method uses experimentally derived or calculated RIR. Unlike other methods, the DDM requires only chemical composition data of the individual components. Chemical composition data are available in almost all cases, since QPA is usually conducted after phase identification or applied to chemically known materials. Therefore, the DDM has no limitation in applying QPA to any materials as long as the intensity datasets for the individual components are available.

Basic principle of the DDM is very simple. Parameters, used for deriving the weight ratios, can definitely be calculated using chemical composition data. Therefore, its accuracy in QPA is mostly dependent on the accuracy in the intensity datasets for individual components. Thus the correct choice of a decomposition tool, for a given situation, will deliver the highest quality result. Therefore, it is important to understand handing of various techniques used for separating the observed pattern into the component patterns. Readers of this article may be engaged in various analytical works, with many of them having experiences with QPA. Some of them will also be interested in reading original articles. So core mathematical formulas are presented at each step of this article. Theoretical and experimental details are located in references 10 to 15.

Wavelength dispersive X-ray fluorescence (XRF) spectrometers have high spectral resolution and can therefore identify peaks with high accuracy. However, if the analysis line overlaps with a higher order line, peak identification and semi-quantitative analysis results may not be reliable. To perform accurate analysis for such cases, measurement conditions that reduce the influence of the higher order lines need to be set up, followed by remeasuring the sample with the optimized conditions. This can require advanced knowledge of XRF principles and a high level of familiarity with the software. This is not in line with the demand for semi-quantitative analysis being easily and quickly able to give accurate and reliable results for unknown samples.

The new software feature described in this article was developed to overcome this obstacle. It performs semi-quantitative analysis by the FP method using a new procedure that accurately calculates results after the software automatically selects data from the optimal measurement conditions when a higher order line interferes with the analysis line. This allows anyone, including users with little experience in X-ray analysis, to obtain analysis results with higher accuracy and reliability.