The challenges of analyzing a large variety of alloying elements with wide ranges of concentrations by XRF are well known to the metals industry. The Metals edition of the Zetium spectrometer with NiFeCo-FP is the solution to these challenges. The power and flexibility of Malvern Panalytical’s FP matrix correction algorithm is used to provide a single program for wide-range alloys analysis.
Special steels, high-temperature alloys and super alloys are used in a wide variety of high-performance components needing characteristics such as durability, hardness, temperature stability and corrosion resistance. They are used extensively in the manufacture of products like aero-engines, turbines, tools and grinding equipment.
The production of high-specification steels requires process and quality control with very tight tolerances. XRF spectrometry, capable of producing fast, accurate and very precise chemical analyses, is one of the principal methods for production control.
Special steels, high-temperature alloys and super alloys are used in a wide variety of high-performance components needing characteristics such as durability, hardness, temperature stability and corrosion resistance. They are used extensively in the manufacture of products like aero-engines, turbines, tools and grinding equipment.
The production of high-specification steels requires process and quality control with very tight tolerances. XRF spectrometry, capable of producing fast, accurate and very precise chemical analyses, is one of the principal methods for production control. The challenges of analyzing a large variety of alloying elements with wide ranges of concentrations by XRF are well known to the metals industry. In particular, the matrix corrections needed and the large number of reference materials required for high-quality results are generally a concern.
The Metals edition of the Zetium spectrometer with NiFeCo-FP is the solution to these challenges. The power and flexibility of Malvern Panalytical’s FP matrix correction algorithm is used to provide a single program for wide-range alloys analysis. In addition, specially formulated reference samples are used to transfer calibrations for up to 19 elements in nickel-, iron- and cobalt- based alloys. The master calibration is based on more than 120 Certified Reference Materials (CRMs) and includes special elements like Ta and Hf, for which very few CRMs are available.
Calibration standards (CRMs), NiFeCo-FP setup samples and the CRMs used for accuracy verification were re-surfaced to obtain a quality surface for analysis. This was done on a fine-surface milling machine using 150 grit Al2O3 paper, which is suitable for a variety of materials of varying hardness. However, for some of the hardest alloys, initial re-surfacing was done with 80 grit ZrO2 paper before finishing with the Al2O3 paper.
The total measurement time of the analytical program was 430 seconds, with the Metals edition of the Zetium spectrometer operated at 2.4 kW. The program includes measurements on both peak and background positions for 19 elements and assures accuracy at low concentration levels (< 0.1 wt%). Measurement times (s) for each element (Table 1) were calculated using the formula:
This formula is a useful ‘rule of thumb’ and is based on the premise that at the 100% concentration level approximately 4 x 106 counts are required to obtain a relative counting statistical error of 0.05%, which is a practical and common measure in the analysis of metals. The elements from vanadium to tungsten were measured using a LiF220 crystal to reduce spectral interferences. In addition, serious line interferences were avoided for copper and tantalum by using the Cu Kβ1 and Ta Lβ1 lines, respectively. To simplify installation, the SuperQ software for the Metals edition of the Zetium spectrometer contains a template for the NiFeCo-FP application and calibration only requires measurement of the NiFeCo-FP setup samples.
Table 1. Callibration data for the NiFeCo-FP master calibration
The six NiFeCo-FP setup samples were calibrated against master calibrations containing up to 126 CRMs. These CRMs were sourced from Analytical Reference Materials International (IARM-series), Bureau of Analysed Samples (BCS-&SS- series), Brammer Standards (BS-series), Iron and Steel Institute of Japan (JSS- series), National Bureau of Standards (NBS-series), Turret Alloys and Willan Metals (WM- and WT-series). The accuracy of the master calibrations are shown in Table 1 and calibration plots for Ni, Co, Si and P (Figures 1 - 4). The ‘root mean square’ (RMS) factor (effectively 1 standard deviation) and the K-factor are an indication of the calibration quality, showing the magnitude of the differences between the certified and measured concentrations. Lower values for RMS and K-factor indicate more accurate calibrations.
Typical compositions for the NiFeCo-FP setup samples are shown in Table 2. In the calibration, the Malvern Panalytical Fundamental Parameter (FP) model was used to calculate the matrix correction term (M) in the calibration formula:
The wide-range calibrations shown in Figures 1 and 2 clearly demonstrate the power of the FP model. Corrections for the inter-element effects are optimized for each of the calibration standards, making it possible to calibrate concentration ranges from 0 – 70 wt% or more.
Figure 1 - 4: NiFeCo-FP master calibration plots for Ni, Co, Si and P
Table 2. Typical concentration values for NiFeCo-FP setup samples. Each set has a unique identification number because the concentration values may vary from one set to another
The lower limit of detection for each element is shown in Table 3. These values are calculated from:
1 Where separate peak and background measurements are made, the error attached to these measurements means that the LLD calculated with this equation represents the concentration equivalent of 2 standard deviations of the background intensity.
Table 3. Lower limits of detection
The accuracy of the Metals edition of the Zetium spectrometer using the NiFeCo-FP application for major and minor element analyses in a wide range of alloy compo- sitions is excellent.
This is illustrated in accuracy overview plots for 11 elements (Figures 5 and 6), which compare certified and measured values for 12 CRMs of widely varying composition (Table 4). These CRMs were measured as routine samples against a NiFeCo-FP calibration defined by the 6 setup samples. The tungsten and nickel data for CRM SS 484/1 (~20 wt% W) and WT3151 (~ 78 wt% Ni) (Table 4), illustrate that the Malvern Panalytical FP calibration model is capable of accurate extrapolation and analysis outside the range of concentrations defined by the setup samples (~15 wt% W, ~64 wt% Ni, Table 2). The NiFeCo-FP calibration is also flexible and allows easily for extensions to the calibration with in-type standards if required.
The NiFeCo-FP setup samples are an effective means of transferring the NiFeCo-FP master calibration and by using the master calibration K-factor, we can determine an upper limit to the absolute accuracy (1σ) achievable for a given elemental concentration (Table 5; also see side bar on page 4 for details).
Table 4. Analytical accuracy: comparison of certified and measured values for 18 elements in 12 CRMs of various types. The sample, IARM 100B, is supplied with the NiFeCo-FP package for use as a drift monitor and/or a quality sample. Data for Y, Hf, are not included in this table because no reference values are available for these CRMs. For the certified concentrations, values in ( ) are for information only and those in [ ] are determined by difference from 100 wt%
Figure 5. Accuracy overview from 0 – 100 wt%: comparison of certified and measured values for 8 elements in the wide variety of CRMs listed in Table 4
Figure 6. Accuracy overview from 0 – 1.0 wt%: comparison of certified and measured values for 8 elements in the wide variety of CRMs listed in Table 4
Table 5. Absolute error estimation: accuracies achievable with NiFeCo-FP for low and high concentrations (for example, 0.1 wt% and the highest concentrations in the setup samples). Estimates calculated using the K-factors for the master calibration (Table 1)
The precision of the Metals edition of the Zetium spectrometer and the NiFeCo-FP package was determined by repeated measurement of a CRM against the NiFeCo-FP calibration. Nimonic alloy, SS 310/1, was analyzed 20 times consecutively for a short-term precision test and 15 times over a period of 3 weeks for a long-term precision test (Table 6 and Figure 7). The results were obtained without any drift correction.
Comparison of the RMS value (1 σ standard deviation) with the counting statistical error (CSE) shows the excellent precision of the method and the outstanding stability of the Metals edition of the Zetium spectrometer . Theoretically, the CSE is the minimum possible error and the similar magnitude of the RMS and CSE values shows that errors originating from the instrument are negligible.
Table 6. Analytical precision derived from repeated measurement of a Nimonic alloy (SS 310/1)
Figure 7. Short-term and long-term repeatability results for Al in a Nimonic alloy CRM (SS 310/1) show the excellent stability of the instrument. The dashed lines indicate the 3-σ levels
Components typeset in bold were present in the spectrometer used to obtain the data in this note.
Elemental analysis by XRF is one of the principal methods of process and quality control in the metals industry. The Metals edition of the Zetium spectrometer with NiFeCo-FP application setup is an excellent solution for the analysis of Ni-, Co- and Fe-based alloys, super alloys and steels in a single wide-range analytical program. It has been demonstrated that the analyses are accurate and precise and that the stability of the Metals edition of the Zetium spectrometer is excellent, resulting in highly consistent data over time.
The Metals edition of the Zetium spectrometer has been configured to meet the basic needs of the industry, however, it can be further upgraded to enhance performance in terms of speed and precision. Upgrade features include increased power, continuous loading and direct loading for extra speed of analysis, a sealed Xe counter for lower limits of detection, and Omnian for standardless analysis. Integration of the Metals edition of the Zetium spectrometer into laboratory automation setups is supported with the universal automation interface software and optional sample transfer features.
