Thermoluminescence dating is generally not very accurate.
The accuracy of thermoluminescence dating is only about 15% for a single sample and 7 to 10% for a suite of samples in a single context.
Both assumptions become less realistic with the passage of time.
Another problem with the TMRD is the calibration of the detector, since different crystals in an artifact can contain different amounts and/or types of luminescence material.
Natural crystalline materials contain imperfections: impurity ions, stress dislocations, and other phenomena that disturb the regularity of the electric field that holds the atoms in the crystalline lattice together.
This leads to local humps and dips in its electric potential.
Create fake pottery that will pass the thermoluminescence test One way to pass a fake through a TL test is to expose the newly-made pottery to a high dose of artificial radiation sources, thus fooling the measurement instruments.
However, producing fakes with this method calls for expertise on the subject, as well as expensive instruments.
The thermoluminescence (TL) properties of plagioclase separates from 11 independently dated alkalic basalts 4,500 years to 3.3 million years old and 17 tholeiitic basalts 16 years to 450,000 years old from the Hawaiian Islands were investigated for the purpose of developing a TL dating method for young volcanic rocks.
Ratios of natural to artificial TL intensity, when normalized for natural radiation dose rates, were used to quantify the thermoluminescence response of individual samples for age-determination purposes.
Instead, a less sophisticated method that would deceive TL testing is to reuse original broken and unmarketable pieces.