The Flexibility of β-Content, γ-Confidence Tolerance Intervals to Qualimetry a Simultaneous 22 Aromatic Amines Derived from Azo Dyes in Fabric Using a Sensitive GC-MS Technique.

Background: Azo dyes are among the most widely used dyes in the textile industry, releasing a series of carcinogenic aromatic amines that can be absorbed through the skin.

Objective: This work aims to show that 22 azo dye amines in a textile matrix can be quantified using a GC-MS method.

Methods: Based on the notion of total error and β-content, γ-confidence tolerance intervals (β,γ-CCTI), a chemometric approach known as the "uncertainty profile" has been used to completely validate a GC-MS method for the simultaneous assay of 22 azo amines in fabrics. According to International Organization for Standardization (ISO) in ISO 17025 guidelines, analytical validation and measurement uncertainty estimates have evolved to be two main principles for ensuring the accuracy of analytical results and controlling the risk associated with their use.

Results: The calculated tolerance intervals allowed for the determination of the uncertainty limits at each concentration level. These limits when compared to the acceptable limits show that a significant portion of the expected outcomes is in conformity. Additionally, the relative expanded uncertainty values, calculated with a proportion of 66.7% and a 10% risk, do not exceed 27.7, 12.2, and 10.9% for concentration levels 1, 15, and 30 mg/L, respectively.

Conclusion: The capability and flexibility of the β-content, γ-confidence intervals have been established through the use of this innovative approach to carrying out qualimetry of the GC-MS method depending on the behavior, required conformity proportion, and acceptable tolerance limits of each amine.

Highlights: An efficient GC-MS technique for the simultaneous determination of 22 azo amines in a textile matrix has been developed. Analytical validation using a new strategy based on the uncertainty concept is reported, uncertainty associated to measurement results is estimated, and the applicability of our approach to the GC-MS method is investigated.