I am working on the extraction of tartaric acid from ancient pottery as a biomarker for wine presence using a negative mode LC-MS method. However, I am observing significant variations in the retention time of tartaric acid across different extractions from model pottery samples.

To establish a limit of detection, I prepared model pottery samples by soaking them in modern wine at different dilutions. I also analyzed pure tartaric acid standards and wine dilutions (without extraction) to confirm the identity of the tartaric acid peak. In these cases, the retention time remains stable and within an acceptable range. However, when analyzing extracts from the model pottery samples, the retention time fluctuates despite having the same m/z.

I suspect that matrix effects are causing this variability, but I am unsure how to conclusively demonstrate that these retention time shifts still correspond to tartaric acid. What approaches or additional experiments could I use to verify that these peaks belong to my analyte despite the retention time variation? Any insights or suggestions would be greatly appreciated!