Ceramic Materials Characterization Using X-Ray Diffraction
Successful ceramic manufacturing requires the correct identification of phases and an understanding of microstructure in starting powders and finished products.
Powder X-ray diffraction (XRD) techniques can trace their origin to the pioneering work of Debye and Scherrer in Europe (1916) and Hull in the U.S. (1917).1 Their results dispelled the belief that grinding a single crystal to a powder would destroy crystallinity. In 1938, Hanawalt, Rinn and Frevel published over 1,000 reference XRD patterns tabulated as diffraction spacings (d-spacings) and intensities.2 More importantly, they also described a new method of materials identification that used the tabulated diffraction data and was described by the journal editor as a complete, new workable system of analysis.
Three years later, the American Society of Testing Materials (ASTM) published the first set of Hanawalt et al diffraction patterns on individual 3 x 5 in. cards, with these cards eventually referred to as Powder Diffraction File™ (PDF®) cards. In 1969, this diffraction database activity was separated from ASTM, and the new parent organization was named the Joint Committee on Powder Diffraction Standards (JCPDS). To better reflect the international scope of the organization, the name International Centre for Diffraction Data™ was added in 1978, today referred to as ICDD®.
The first consideration in any analysis of XRD results is to determine if a material is crystalline or amorphous. If amorphous, the diffraction pattern will comprise broad diffuse regions of scatter, rather than well-defined diffraction peaks. Once a crystalline phase(s) has been observed to be present in a sample based on the XRD data, identification of the phase or phases requires a comprehensive database of reference diffraction data. The 2016 ICDD PDF-4+ database3 has 384,600+ entries, including 15,000+ classified as ceramic phases. In 1941, the original PDF cards had limited information primarily consisting of d-spacings and intensities. Today, entries in the PDF also contain crystallographic information that can be used to extend the capabilities of XRD analysis (see Figure 1).
Analysis of XRD patterns often identifies the presence of more than one crystalline phase, including polymorphic forms (same chemistry, different crystal structure). Elemental analysis of a powder sample using the EDAX capability of a scanning electron microscope (SEM) revealed the presence of calcium, carbon and oxygen. An interpretation of this result would be to consider the sample is calcium carbonate (CaCO3). XRD data indicate that CaCO3 was present; however two CaCO3 phases were detected (see Figure 2). Aragonite (PDF 00-005-0453) and calcite (PDF 00-047-1743) polymorphic forms of CaCO3 were identified, demonstrating one of the key capabilities of XRD analysis—phase identification.
For many years, qualitative phase analysis was the end point of most XRD analyses, as preferred orientation of anisotropic crystallites in a powder sample resulted in significant variations in recorded diffraction peak intensities. More recently, whole-pattern-fitting (WPF) techniques such as Rietveld refinement have been used for quantitative XRD. A requirement for WPF methods is the availability of atomic coordinates, allowing for calculation of powder diffraction patterns that can be compared and refined relative to observed XRD raw data patterns.
The PDF, with more than 271,000+ entries that contain atomic coordinates including 11,700+ ceramic phase entries, is an optimal source for data used in WPF analysis. For composite samples, ceramic powders are often compounded with a base matrix such as a polymer. In the manufacturing of ceramic components, it is necessary to confirm that the ceramic phase is the desired phase. In addition, knowing how much of the phase is present is important for product function and cost control. A biomedical film comprising polymer polyethylene terephthalate (PET) and titanium dioxide was analyzed by XRD. The XRD results confirmed the TiO2 phase was the anatase (PDF 00-021-1272) polymorph, with identification possible even in the presence of the biaxially oriented PET polymer matrix (see Figure 3).
Using the atomic coordinates included in the anatase PDF entry, along with the raw data PDF reference pattern for the biaxially oriented PET (PDF 00-061-1413), WPF analysis was performed. The TiO2 content was determined to be 1.3 wt%, in excellent agreement with the aim of 1.5 wt%. With WPF, XRD has become accepted as a standard method for quantitative phase analysis.
Numerous ceramic manufacturing methods require thermal processing to produce the final product. Many of the nonambient steps can be simulated using in situ high-temperature XRD (HTXRD) analysis. An aliquot of sample is placed in a chamber that is equipped with a heating (or cooling) stage, with an ambient environment that represents the conditions used during manufacturing, including ambient air, high (or low) relative humidity air, nitrogen, argon, vacuum, and others.
A process to produce CaO lime powder for application as a desiccant was evaluated using HTXRD with a flowing dry N2 gas as ambient. The HTXRD data are presented in Figure 4. The starting powder was identified as containing Ca(OH)2 (PDF 00-044-1481) portlandite and CaCO3 (PDF 00-047-1743) calcite. Upon heating to 350°C, Ca(OH)2 begins to lose water and converts to CaO (PDF 00-037-1497), with conversion complete at 425°C. At 800°C, CaCO3 loses CO2 and converts to CaO. From the HTXRD data, a process engineer would have knowledge of temperatures required for total phase conversion to CaO, and would likely work with the diffractionist to evaluate reaction kinetics by using isothermal scans in the range of 700-800°C to help optimize oven temperature and powder transport rates.
Nanomaterials are an active area of research and process control studies. Particles with a size of less than 100 nanometers (nm) are referred to as nano, often exhibiting physical properties that differ from same-phase microcrystalline powders. Powder X-ray diffraction is well-suited for the analysis of small particle powders. As the crystallite size decreases, diffraction peaks broaden since there are fewer lattice planes to maintain destructive interference of non-Bragg condition scatter.
In an attempt to increase the refractive index of a plastic lense, nano-zirconia was to be added to a polymer plastic. A small reaction vessel from Anton-Paar was used to make the zirconia so that the process could be developed using temperatures lower than typical refractory processing by performing the reaction at elevated pressures. Diffraction data for powder samples made using four conditions were analyzed (see Figure 5). At lower temperature (LT) and lower pressure (LP), the sample is amorphous. At LT and higher pressure (HP), the onset of a nanocrystalline phase is observed, with XRD results indicating the average crystallite size is 1.8 nm. Processing at higher temperature (HT) and LP results in an increase in crystallite size to 3.8 nm, and HT and HP processing produces powders with an average crystallite size of 11.0 nm. Phase identification analysis confirmed the nanopowders are ZrO2 (PDF 00-037-1484) baddeleyite, the monoclinic polymorph.
By itself, one must be careful using only XRD to determine if a material is truly nanocrystalline. Transmission electron microscopy data are complementary to XRD, with the TEM image in Figure 5, for the HT and HP sample showing individual particles on the order of 10 nm in diameter. Using the XRD results, the temperature and pressure used for nanoparticle generation can be controlled to produce ZrO2 small enough to not scatter visible light but large enough to increase the refractive index of a composite plastic lense.
As mentioned previously, preferred orientation can bias X-ray diffraction peak intensities relative to what would be expected for a randomly oriented powder. An extreme example would be the formation of large single crystals rather than fine powder during ceramic material processing. The likelihood of producing large crystals can be enhanced, especially when a melt flux is involved with the manufacturing method.
A mixed powder comprised of ZrO2 (baddeleyite type) and laponite clay (Na0.7(Si8Mg5.5Li0.3O20(OH)4) was placed on a magnesia (MgO) plate and sintered. After cooling to room temperature, the sample was removed from the oven. It was noticed that, in addition to the sintered article, there were regions where a hard continuous coating covered portions of the MgO plate. Small pieces of the coating were removed by hammer and chisel. In cases where a small amount of sample is available for testing, it is sometimes advantageous to use a microdiffractometer4 for XRD analysis. A microdiffractometer equipped with a two-dimensional electronic detector will also enhance the ability to observe if the sample is a powder, polycrystalline (diffraction rings), or a single crystal (diffraction spots). In all cases where the hard coating was observed, microXRD identified the crystalline phase as MgSiO3 (PDF 00-034-0189), forsterite (see Figure 6).
When the starting ZrO2/laponite clay powder had greater than 65 wt% laponite, the two-dimensional diffraction pattern of the coating sample showed a spot pattern. In contrast, a starting powder with less than 65 wt% laponite had a two-dimensional diffraction pattern of the coating sample that showed diffraction rings after sintering. In both composition situations, silicon (Si) from the clay reacted with the MgO plate. High levels of laponite clay provided enough of a melt flux that large crystals were able to grow in the flux. At lower clay levels, Si from the clay and MgO still reacted; however, there was not enough flux to grow large single crystals. Combining XRD phase analysis with a two-dimensional detector gives a chemist insight into adjusting the powder mixture composition for control of phase and particle morphology.
Beyond Laboratory X-Ray Diffraction
Laboratory instrumentation results highlight some of the applications of XRD analysis for research, development and manufacturing process control applications related to ceramic materials. Government laboratory facilities around the world that provide high-intensity synchrotron radiation, constant wavelength neutron, and time-of-flight neutron diffraction capabilities are allowing scientists and engineers to achieve lower limits of phase and quantitative analysis detection, as well as provide a better understanding of phase-property behavior in ceramic materials.
1. Blanton, T., “Two Dimensional X-Ray Diffraction Detectors,” Spectroscopy, 31(7), 2016, pp. 42-43.
2. Hanawalt, J.D.; Rinn, H.W.; and Frevel, L.K., “Chemical Analysis by X-Ray Diffraction,” Industrial & Engineering Chemistry Analytical Edition, 10(9), 1938, pp. 457-512.
3. Kabekkodu, S., ICDD PDF-4+ 2016 (Database), International Centre for Diffraction Data, Newtown Square, Pa., 2016.
4. Blanton, T.N., “Applications of X-Ray Microdiffraction in the Imaging Industry,” Powder Diffraction, 21(2), 2006, pp. 91-96.