P - Technische Universiteit Eindhoven

1. Introduction 11
This phenomenon was studied by Charles Tomlinson for sodium sulfate solutions.
As Kurzer reported in his overview of Tomlinson’s work [26]: ’... Tomlinson endeavored
to prove that supersaturation was maintained simply by the absence of nuclei that, when
introduced, initiate the separation of the solute by the exertion of adhesive forces. The
extensive and painstaking although simple experimental work that he undertook to this
end can here be summarized only briefly (but is substantiated by complete references). He
confirmed that supersaturated sodium sulphate solutions, prepared under strictly ’clean’
conditions (that is, with the exclusion of all potential crystallization nuclei) when exposed
to an ordinary town atmosphere, precipitated crystals more or less immediately. In the
pure air of the country no such separation occurred for many hours or days. Similarly,
contact of the supersaturated solution with a glass rod, thermometer or spatula resulted
in rapid crystallization, but when such instruments were previously cleansed with strong
acid or alkali, or drawn through a flame, no such effect was produced.’
Among others, Hartley et al. in 1908 measured the supersolubility curve of heptahydrate
[22], which is presented in figure 1.1. This curve indicates the concentration
required for the spontaneous cooling-induced crystallization at a given temperature. They
also measured the cryohydrate solubility line. The transition temperature of heptahydrate
to anhydrous sodium sulfate plus liquid was found in 1938 by Washburn and Clem [27]
to be 23.47 ◦ C.
As shown in figure 1.1, three main crystalline forms of sodium sulfate can be distinguished:
thenardite (anhydrous, Na 2 SO 4 ) and metastable thenardite (being unstable
with respect to common thenardite), mirabilite (decahydrate, Na 2 SO 4·10H 2 O) and the
thermodynamically metastable heptahydrate (Na 2 SO 4·7H 2 O) [22, 28]. As result of about
800 measurements, it was found that the heptahydrate has a well-defined supersolubility
region limited by the so-called heptahydrate supersolubility line, which indicates the maximum
supersaturation that can be reached [28]. An interesting feature of heptahydrate is
that its metastability is related only to the coexistence with mirabilite and some chemical
compounds (like sodium tetraborate), whereas it is stable with respect to many external
physical effects (like storage time, shaking, heating-cooling, dropping).
Despite the widespread use of sodium sulfate in the chemical industry (glass production
or medicine) and in heat storage, the scientific interest in sodium sulfate decreased
in the middle of the 20th century. Heptahydrate seems to have been ’forgotten’ by researchers
of the 20th century, as illustrated by the citation diagram presented in figure 1.2.
However, with the renewed interest in understanding the damage caused by salts in porous
media, the interest in crystallization behavior of sodium sulfate started again during the
last few years [29–32].
1.3 Accelerated weathering
In order to mimic and to model natural salt weathering of porous materials, accelerated
weathering tests are commonly used for evaluation [33–37]. These types of tests usually
represent cycles, consisting of periodic saturation of a porous material with a salt solution
followed by drying at high temperatures. The systematic study of salt weathering started
during the second half of the 19th century. Routine weathering tests were implemented
already in 1828 by de Thury [38]. A first general assessment, review, and systematization