Processes and kinetics of mass gain in archeological brick following drying and reheating

Abstract

The mass gain behavior of archeological bricks was examined following drying (130°C)/reheating (500°C) and aging at a range of temperatures (25°C, 35°C, 45°C). For drying or reheating, samples exhibit a two‐stage mass gain behavior, the second stage, Stage 2, continuing indefinitely and better described by a t1/n model (1/n=1/6‐1/2); a correlation between the 1/n value and the specific surface area/pore volume demonstrates diffusion mechanisms with some pore geometry/morphology dependence. Stage 2 is shown to have an Arrhenius temperature dependence with activation energies of similar orders of magnitude following both drying and reheating. Supported by thermogravimetric‐mass spectrometry (TG‐MS), Stage 2 is demonstrated as likely due to the recombination of chemisorbed water, previously removed, whereas following reheating due to two components, a chemisorbed component associated with drying and a component associated with rehydroxyls removed at higher temperatures during reheating. Differences between activation energies of chemisorption and rehydroxylation components support this. Evidence for a fundamental compositional relationship between these processes is presented by a strong linear relationship between the drying and reheating mass gain rates. Stage 1, following drying or reheating, is shown to be likely associated with physisorption processes alone.

Rehydroxylation (RHX) dating: Issues due to short term elevated temperature events

“…Stage 1) and has generally reached completion within 24 hours (Barrett 2017c). The prolonged and indefinite second stage, Stage 2, is likely due to both chemisorption and rehydroxylation processes Le Goff 2015, Barrett 2017c).…”

“…The rate at which chemisorption/rehydroxylation takes place in Stage 2 is well described by a time 1/n (e.g. Bowen et al 2011;Le Goff and Gallet 2014;Barrett 2015;2017c) power law of the general form:…”

“…A fixed value of 1/n = 4 was originally proposed and applied with some success (Wilson et al 2009; but this has been increasingly shown to be unsatisfactory (e.g. Bowen et al 2011;Le Goff and Gallet 2014;2015a;2015b;Barrett 2015;2017c) with 1/n instead observed to be sample dependent and varying between 1/6-1/2 (e.g. Le Goff and Gallet 2014; Barrett 2017c).…”

“…It has previously been shown that the mass gain behaviour is better described by a t 1/n based equation, with 1/n potentially varying from 1/6-1/2 dependent on the sample (e.g. Barrett 2017aBarrett , 2017cGallet 2014, 2015b). For the present exploratory set of experiments 1/n was generally restricted to a value of 1/4 predominantly because it sits in the middle of the typical range of values expected and to a lesser extent because it is the original value used by Wilson et al (2009Wilson et al ( , 2013; this also permits findings from the brick cooling experiment to be discussed more directly in relation to the positive results of Wilson et al (2009) in which four of five sample materials were brick.…”

“…Despite the promising application of these principles by Wilson et al (2009Wilson et al ( , 2012 to date bricks, tiles and pottery up to 2000 years old, there is a growing body of work that demonstrates significant issues with many aspects of the method: aside from the 11 samples dated by Wilson et al (2009Wilson et al ( , 2012 age estimations from other works have been unsatisfactory (for example, Burakov and Nachasova 2013;Le Goff and Gallet 2015a;Numrich et al, 2015;Barrett 2017a); the time 1/4 model is inadequate (for example, Bowen et al, 2011Bowen et al, , 2013Le Goff and Gallet 2014, 2015a, 2015bBarrett 2017aBarrett , 2017c with the use of a time 1/n model considered an improvement (for example, Bowen et al, 2011;Gallet and Le Goff, 2015; Le Goff and Gallet, 2015b;Barrett 2017aBarrett , 2017c; the presence of contaminants (notably organics and gypsum) that interfere with rehydroxyl mass estimation are also an issue (Numrich et al, 2015;Barrett 2017b). Along with other experimental difficulties (for example Bowen et al, 2011;Zhao et al, 2015), questions surrounding the validity of successful dating trials have also arisen (Le Gallet, 2015a, 2015b; see also the review of Barrett 2015).…”