Abstract
Matrix-dependent self-absorption constitutes a significant source of uncertainty in gross alpha activity determination. An analytical framework has been developed to integrate the effective alpha-particle mass range into gross-alpha counting efficiency calculations. This approach accounts for variations in alpha-particle stopping power across different residue compositions and offers a physically grounded alternative to the matrix-specific empirical calibration curves commonly employed in gas-proportional counting. The framework was evaluated using independent calibration datasets prepared with CaSO4·2H2O and CaCO3 matrices, and further validated using IAEA proficiency testing samples analyzed between 2021 and 2025. The results closely aligned with experimental efficiency trends for both matrices within the thin-source method (residue mass below 100 mg, corresponding to a mass thickness of ≤ 5.2 mg cm-2 in this study). All testing outcomes met the acceptance criteria of |Z| 4·2H2O calibration approach, with discrepancies ranging from 12.05 to 23.46%. The largest difference occurred in the sample with the highest residue mass thickness (3.99 mg cm-2). The groundwater residues were composed primarily of sodium and chloride, demonstrating that residue composition can significantly affect counting efficiency estimates when a single calibration matrix is used for chemically diverse environmental samples. These findings indicate that incorporating residue composition into efficiency calculations enhances the reliability of gross alpha activity determination. The proposed framework provides a practical approach for matrix-dependent efficiency correction and may facilitate radiological screening and water quality assessments in regions with elevated natural radioactivity.