# RHX-Dating revisited…

[bibshow file=all.bib]
Our latest publication on RHX-Dating is online!

Rehydroxylation-Dating is based on the observation that fired clay grows heavier and larger over time. This gain in mass is associated with water that is slowly moved very deeply into the material in a process called diffusion. The water is bound not as water molecule but as OH-groups – hence this process is called Hydroxylation. This process seems to follow a strict law over time, the mass gain $m_{OH}$ is [bibcite key=Wilson2003] $m_{OH}(t)=\alpha_m \times t^{1/4}$

This $t^{1/4}$ factor is very interesting, because in principle it should make it possible to date a fired piece of clay with very high precision!

To date the brick or ceramic artefact, we only need to measure its mass at the current time, remove the hydroxilated water and measure the mass again, in effect measuring $m_{OH}$. Afterwards, we could let the sample gain mass again for some time and measure $\alpha_m$.

This would indeed be a beautifully simple measurement to make but there’s always a catch, isn’t there?

In fact, there are two problems, one being that the constant $\alpha_m$ depends on the material and on the average temperature during the mass gain. Effectively this means that we have to measure $\alpha_m$ at at least two different temperatures and then use the Arrhenius equation to calculate $\alpha_m$ at our best estimate of the temperature during storage[bibcite key=Hall2013].

The other problem that we talk about in our article [bibcite key=Numrich2015] is that not only are there numerous ways in which water can stick to the sample (plain “loosely bound” water sticking to the surface and others) but there’s also the chance that there is some mass in the sample that is lost when heating above 500°C that is not hydroxylated water. If this is the case, then the measurement for $m_{OH}$ will be too large, making the sample much older than it really is.

In the publication we try to identify this additional mass (called non-refractory compounds or NRC) and offer a few recipes how to remove any NRCs that might be in the sample.

[/bibshow]