Status Of EDXRF Method Validation and Quality Control in the Chemistry Division of Atomic Energy Centre, Dhaka, Bangladesh

X-ray Fluorescence (XRF) Analysis is a non-destructive instrumental method used for the qualitative and quantitative analysis of chemical elements. It is based on the measurement of characteristic fluorescent radiation emitted when inner-shell vacancies, created by an external radiation source, are de-excited. Among the various types of XRF analysis, the two major approaches are Wavelength Dispersive X-ray Fluorescence (WDXRF) and Energy Dispersive X-ray Fluorescence (EDXRF). These techniques are primarily differentiated by the type of detector used to measure the characteristic emission spectra. WDXRF employs a diffracting crystal to separate and determine the characteristic wavelengths of emitted X-rays. EDXRF utilizes detectors that directly measure the energy of the X-rays by collecting the ionization produced in a suitable detecting medium. A wide range of EDXRF analytical systems has been developed using various excitation sources, including radioisotopes, X-ray tubes, charged particle accelerators, microprobe electron beams, and synchrotron radiation sources. Energy Dispersive X-ray Fluorescence (EDXRF) spectroscopy is a widely used analytical technique for non-destructive, multielemental analysis in environmental, industrial, and material science applications. The Chemistry Division of the Atomic Energy Centre, Dhaka (AECD) employs EDXRF for the determination of elemental composition in various samples, including soil, sediments, road dust, and industrial materials. To ensure the reliability of EDXRF results, method validation and quality control (QC) procedures are strictly followed. These steps confirm the accuracy, precision, detection limits, and robustness of the analytical method, ensuring compliance with international standards and national research objectives. This report highlights the current status of EDXRF method validation and quality control at AECD, focusing on calibration procedures, performance evaluation, and quality assurance measures [1-10].

When radiation, such as charged particles or photons, interacts with an atom, it is absorbed, transferring its energy to an orbital electron. This absorption process, known as photoelectric absorption, results in the ejection of the electron, creating a vacancy in the atom’s energy level. The ejected electron, referred to as a photoelectron, carries the acquired kinetic energy away from the atom. To restore equilibrium, an electron from a higher energy level transitions to fill the vacancy. This transition releases a quantum of energy in the form of fluorescence X-rays. Since electron energy levels in an atom are well-defined, the emitted X-rays have characteristic energy values unique to each element. By analyzing the energy of these X-rays, elements can be identified (qualitative analysis). Furthermore, by measuring the intensity of the emitted X-rays, the concentration of the element can be determined (quantitative analysis) [1]. In X-ray spectrometric analysis, the intensity of the secondary spectral lines emitted by the analyte is measured to estimate its concentration.

The incident X-ray beam interacts with the sample, leading to excitation of atoms. The intensity (I0I_0I0 ) and energy distribution of this beam determine how many atoms can be excited. Absorption of Primary X-rays in the Sample The probability of the incident X-rays exciting the analyte depends on how much energy is absorbed by the analyte atoms relative to the matrix material. This is governed by the mass absorption coefficient (μ\muμ) and thickness of the sample.

Not all excited atoms emit characteristic X-rays; some undergo non-radiative transitions. The fluorescence yield (ω\omegaω) describes the probability of X-ray emission after excitation. Once X-rays are emitted, they may be reabsorbed by the sample or matrix before reaching the detector. The self-absorption effect is influenced by the sample composition and thickness. The efficiency of detecting X-ray photons depends on the sample-detector configuration. Factors like take-off angle, detector efficiency, and collimation affect the measured intensity. This theoretical framework is essential for quantitative XRF analysis, where corrections for matrix effects and instrumental factors must be applied to accurately determine analyte concentrations.

Method Validation

Method validation is essential to confirm that the EDXRF technique produces accurate and reproducible results. The Chemistry Division follows rigorous validation protocols, which include: Calibration and Standardization: Calibration is performed using certified reference materials (CRMs) with known elemental concentrations. Calibration curves are constructed by plotting elemental X-ray intensities against their concentrations. Matrixmatched standards are used to minimize interference effects. Energy Dispersive X-ray Fluorescence Spectrometry (EDXRF) was employed to quantify heavy metals (Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Hg, and Pb) in wastewater and cultured fish samples. The analytical method was validated by constructing a calibration curve using the certified reference material (CRM) “Tuna Fish Homogenate/ IAEA-350.” To ensure the accuracy of the calibration curve, another CRM, “TORT-2,” was used as part of the quality control process. For water samples, the calibration curve was developed using a cellulose-based, lab-synthesized multi-elemental standard (Cellu-1), while another lab-synthesized water standard (Cellu-2) was used to verify the quality of the curve. The results, presented in Table 1, indicate that the percentage error remained within ±10%. The complete methodology for sample analysis and method validation for both water and fish is detailed [11]. When radiation, such as charged particles or photons, interacts with an atom, it is absorbed, transferring its energy to an orbital electron. This absorption process, known as photoelectric absorption, results in the ejection of the electron, creating a vacancy in the atom’s energy level. The ejected electron, referred to as a photoelectron, carries the acquired kinetic energy away from the atom. To restore equilibrium, an electron from a higher energy level transitions to fill the vacancy. This transition releases a quantum of energy in the form of fluorescence X-rays. Since electron energy levels in an atom are well-defined, the emitted X-rays have characteristic energy values unique to each element. By analyzing the energy of these X-rays, elements can be identified (qualitative analysis). Furthermore, by measuring the intensity of the emitted X-rays, the concentration of the element can be determined (quantitative analysis). In X-ray spectrometric analysis, the intensity of the secondary spectral lines emitted by the analyte is measured to estimate its concentration.

The Energy Adjustment Procedure corrects energy drift in analytical instruments, ensuring optimal performance, especially after a software upgrade. It realigns energy settings using reference standards, maintains accuracy, and prevents deviations in measurements. The process involves pre-checks, calibration, validation, and saving adjustments to ensure consistent and reliable results. Comparison between Present Results and the Certified Values of Standard (Tort-2) Reference Materials (mg kg^-1).

Table 1:Comparison between Present Results and the Certified Values of Standard (Tort-2) Reference Materials (mg kg^-1).

A comparison between the present results and the certified values of the standard (Tort-2) reference materials can be described as follows: The Tort-2 reference material provides certified values for various elements, typically expressed in mg/ kg. The present results are obtained from the EDXRF analysis of the sample. The comparison evaluates how closely the measured concentrations of elements in the sample match the certified values of Tort-2, which serves as a benchmark. This comparison helps assess the accuracy and reliability of the analytical method used for the sample analysis. The differences between the two values can be used to identify potential biases or errors in the calibration or measurement process. Here is the stacked column graph Figure 1 comparing the standard (0.65 mg/kg) and experimental (0.59 mg/ kg) concentrations of Chromium (Cr) in fish standard. We found the result certificate values Mn 13.60 mg/kg and experimental values 12.11mg/kg. Similarly, we got the others elements Fe, Ni, Cu, Zn, As, Se, Hg and Pb respectively. To validate the method, the following parameters are evaluated.

Accuracy and Precision

Accuracy is verified by analyzing reference materials and comparing results with certified values [12]. Precision analysis is the process of evaluating the consistency or repeatability of results obtained from repeated measurements or experiments. In the context of tuna fish or any scientific measurement, precision refers to how consistently results can be repeated under the same conditions. Precision: Reproducibility in repetitive measurements. Precision is assessed through repeated measurements of the same sample, ensuring low standard deviation and high reproducibility. (I) Repetitive measurements (ii) Duplicates (iii) Reference samples (iv) Control charts. Analyze TORT-2 and compare results with certified values. Calculate percent recovery using:
% Recovery= (Measured Value /Certified Value) ×100

Precision (Repeatability & Reproducibility). Analyze TORT- 2 multiple times on different days. Calculate Relative Standard Deviation (RSD %) using:
RSD= (Standard Deviation/ Mean Value) ×100 Typically, RSD ≤ 5% is acceptable.

The EDXRF calibration results view shows plots with red data points for each element. Two modes are available: Calculated vs. Given: Plots analysis results for standards as unknowns against known concentrations. Intensity vs. Concentration: Compares measured intensities with known concentrations. These plots help assess calibration accuracy. Calibration corrections are used to manually adjust the slope and intercept values of the calibration. A regression value of 0.99965 likely represents the coefficient of determination (R²) or a correlation coefficient (r) in a statistical model. Here’s what it means: If it is an R² value: R2measures the proportion of variance in the dependent variable that is explained by the independent variable(s). An R² of 0.99965 (very close to 1) suggests that the model explains 99.965% of the variance, indicating an excellent fit. If it is a correlation coefficient (r): The correlation coefficient measures the strength and direction of a linear relationship between two variables. An r of 0.99965 suggests an almost perfect positive correlation between the variables. Such a high value is rare in real-world data and may indicate overfitting, measurement bias, or an artificially perfect dataset [13]. The RMS (Root Mean Square) value of 1.55147 typically represents the measure of deviation or error in various contexts, such as: Regression Analysis (RMSE - Root Mean Square Error). If this is the RMSE in a regression model, it indicates the average deviation of predicted values from actual values. A lower RMSE suggests better model performance.

Limit of Detection (LOD) and Limit of Quantification (LOQ)

The LOD and LOQ are determined for each element to establish the lowest measurable concentration. These values are optimized by adjusting instrumental parameters like tube voltage, current, and detector settings.

Table 2:Certified Values of Standard (Tort-2) Reference Materials (mg kg^-1).

Repeatability in the context of TORT-2 (Tissue Oxidation Reduction Test-2) refers to the ability to obtain consistent and reproducible results when the same test is performed under the same conditions, using identical methodologies and samples. The concept of “12 times run time standards” could refer to repeating the TORT-2 test under varying conditions or extended periods. Repeating a test multiple times ensures that results are consistent and reliable, especially when trying to account for variability in environmental conditions, biological response, and equipment performance. This could also be a standard procedure for ensuring that the system or method remains stable and that potential outliers or errors are minimized. In research, having multiple trials can also assist in understanding the variability of the response of organisms to contaminants over time [14-15]. By performing the test 12 times, researchers can identify patterns and potential anomalies, improving the overall robustness of the study.
Determine using:
LOD=3.3× σ/S
LOQ=10×σ /S
where σ = standard deviation of blank and S = slope of calibration curve.
Prepare a calibration curve with metal standards. Ensure R² > 0.999 for good linearity.

Matrix effects and spectral interferences are evaluated and corrected using empirical and fundamental parameter-based methods. Software-based corrections are applied to improve measurement accuracy.

Quality control: A system of activities whose purpose is to control the quality of need of users. The aim is to provide quality that is adequate satisfactory and economic.

Quality Assurance: A system of activities to provide the assurance that a product or service meets the defined standards of quality with a stated level of confidence.

Quality Assessment: A system of activities whose purpose is to provide assurance that control is being done effectively [16]. It involves a continuing evaluation of performance of a production /service system and quality of products produced /services provided.

To ensure continuous analytical reliability, AECD follows strict quality control protocols: Instrument Performance Monitoring. Regular calibration checks are performed using quality control samples (QCS). Detector efficiency and stability are assessed through periodic tests.

Duplicate and triplicate analyses are conducted to ensure consistency. Standard Deviation in Replicate Measurements.
S_c=S_o+ K*C
S_c= Standard deviation of replicate analysis
C= Concentration
S_o and K are constants which define relative’s errors
Relative standard deviation (RSD) is monitored to maintain data integrity. Figure 3

The laboratory participates in inter-laboratory comparison programs, following international proficiency testing standards. Results from AECD’s EDXRF system are compared with those from other accredited laboratories.

Analytical results are validated through statistical evaluation and uncertainty analysis. Data is reported with appropriate correction factors and confidence intervals to ensure reliability. Compare results with regulatory limits (e.g., WHO, FAO, EU standards). Report validated data with uncertainty estimation.

The Chemistry Division of the Atomic Energy Centre, Dhaka has successfully implemented rigorous method validation and quality control measures for EDXRF analysis. The use of certified standards, calibration techniques, and quality control protocols ensures high accuracy and reliability in elemental analysis. The use of TORT-2 and implementing multiple run time standards (e.g., 12 times) in environmental testing provides a reliable foundation for understanding the effects of contaminants on ecosystems. This approach aligns with rigorous scientific methodologies, ensuring the reliability of data, which is critical for environmental health risk assessments. With repeated trials, researchers can develop a deeper understanding of pollution patterns, enhance prediction models, and contribute to more informed decision-making in pollution control and policy development. The division’s adherence to international best practices, IAEA guidelines, further strengthens its analytical capabilities. Moving forward, continuous improvements in instrument calibration, software-based corrections, and interlaboratory comparisons will enhance the robustness of EDXRF at AECD. These advancements will play a vital role in supporting environmental monitoring, industrial applications, and scientific research in Bangladesh.

The authors acknowledge the assistance of the staff members of the Atmospheric and Environmental Chemistry Laboratory, Nuclear and Advanced Chemistry Division, Atomic Energy Centre, Dhaka-1000, Bangladesh.

No Conflict of Interest.

1. E P Bertin (1975) Principles and Practice of X-ray Spectrometric Analysis, 2^nd Ed, v, Plenum Press, New York, pp. 571-641.
2. R Jenkins (1974) An Introduction to X-Ray Spectrometry, Heyden and Son, New York.
3. HA Leibhatsky, G Pleiffer, EH Winshow, PD zemany, X-Rays (1972) Electrons and Analytical Chemistry, Wiley-Interscience, New York.
4. E Elad, M Nakamura (1966) J Nucl. Instr. Methods 41: 161.
5. HR Bowman (1966) J. Sci 151: 652.
6. DA Landis, FS Goudling, BV Jarrett (1972) J. IEEE Trans. Nucl. Sci 101(1): 127-135.
7. FS Goulding, DA Landis (1974) Semiconductor Detector Electronics in Nuclear Spectroscopy and reactions, Part A., Academic Press, New York.
8. GF Knoll (1971) Radiation Detection and Measurement, John Wiley and Sons. New York.
9. EE Haller, FS Goulding (1981) Handbook on Semiconductors, In C. Hilsson (Ed.), Nuclear Radiation detectors, North Holland, Amsterdam Vol. 4.
10. E Van Grieken Rene, A Markowicz (Eds.) Andrzej, a Handbook of X-Ray Spectrometry, Chapter VIII.
11. Yeasmin N Jolly, Sadia A Surovi, Sheikh M Mizanur Rahman, Jamiul Kabir, Shirin Akter, et al. (2022) Biological Trace Element Research 201(1): 1996-2010.
12. AKM Khan (2008) M.Sc. Thesis, Department of Geology, Dhaka University.
13. A Islam, YN Jolly (2007) J. Bang. Aca. Sci 31(2): 16.
14. G Muller (1969) Geo Journal 2, pp.108.
15. C Lin, M He, Y Zhou, W Guo, Z Yang (2008) Enviro. Monit. Asses. 137(1-3): 329-342.
16. DL Sparks (2002) Environmental Soil Chemistry (Amsterdam: Academic).