“…This indicates that the phenomenon is not merely a result of the hydration of a pre-existing oxide layer, occurring post-failure of the joint. Thus, the available evidence suggests that failure occurred within a predominantly hydroxide layer, which is known to be weaker than the initial oxide [20,21]. Therefore, oxide weakening arising from hydration of the oxide in-situ by the ingressing moisture appears to have initiated joint failure, as has previously [11] been demonstrated to occur via transmission electron microscopy studies.…”