Glass stability (GS) indicates the glass reluctance or ability to crystallise upon heating; it can be characterised by several methods and parameters and is frequently used to retrieve glass-forming ability (GFA) of corresponding liquids as the case with which such liquids can be made crystal free via melt-quenching. Here, GS has been determined for the first time on six sub-alkaline glasses having complex (natural) compositions, the most widespread and abundant on Earth. KT, KH, KW, KLL and w2 GS parameters increase linearly and monotonically as a function of SiO2, with very high correlations. Moreover, Tx values and GS parameters highly correlate with GFA via Rc (critical cooling rate), previously determined with ex-situ cooling-induced experiments. Therefore, GS scales with GFA for natural silicate compositions. In addition, the in-situ Rc value of B100 measured with DSC results > 45 {\deg}C/min (> 2700 {\deg}C/h), broadly corroborating the Rc of about 150 {\deg}C/min (9000 {\deg}C/h) determined ex-situ. In turn, relevant solidification parameters on heating or cooling can be obtained by DSC investigations also for chemically complex (natural) systems, similar to simple silicate systems. These outcomes are relevant for lavas or magmas that re-heat glass-bearing volcanic rocks, as well as for fabricate glass-ceramic materials with desirable texture and composition of phases starting from abundant and very cheap raw volcanic rocks.
In Earth Sciences, dynamic solidification processes have been studied mainly by ex-situ cooling-induced ( [1][2][3][4][5][6][7][8][9][10][11] and reference therein) and to a lesser extent by decompression-induced degassing experiments [4,[12][13][14][15], starting from a dry or volatile-bearing silicate liquid ± crystals.
The major part of these studies focus on SiO 2 -poor systems, relevant for basaltic lavas and magmas, whereas more SiO 2 -rich systems are by far lesser investigated [4,6,[10][11]. Even fewer investigations are available for solidification behaviours induced by heating or re-heating on glasses, except for a few studies on basaltic to basaltic-andesite systems [16-22. In parallel, differential scanning calorimetry (DSC) and/or differential thermal analysis (DTA) techniques are rarely applied on crystallization of liquids and glasses relevant to geological systems, although these in-situ methods are useful and rapid to quantify kinetic parameters [23][24]. This contrasts with common using of such in-situ approaches in Glass Sciences, especially for characterizing solidification of non-crystalline solids upon heating ( [25][26][27][28][29][30][31][32][33][34][35][36][37] and references therein). DSC is commonly used to retrieve the melting (Tm) and glass transitions (Tg) temperatures, plus key Tx (or Tc) values, i.e. the temperature of onset of crystallization. To the best of our knowledge, only [24,38] provided crucial Tx determinations for basaltic glasses or, in general, for non-crystalline solids with chemically complex compositions as the natural ones. Most systems studied in glass sciences focus on simple compounds with only few (mainly two or three, rarely four) major components; typically, these systems solidify phases on heating or cooling, with compositions identical to those of parent glasses and liquids, respectively [23]. Instead, in natural silicate systems the number of major components are normally seven to eight and solidified phases which have stoichiometry very different from the parent liquid on cooling and/or glass during heating.
From the above literature studies is clear that there is a need to further study nucleation and growth kinetic in order to shed new light on one of the most important phenomenon linked not only to magmatic/volcanic events but also to industry. Therefore, we present new DSC investigations upon heating on six chemical complex systems. Experiments were performed by using a ramp rate of 10 °C/min from room to melting conditions. These six sub-alkaline natural compounds have chemistries ranging from SiO 2 -poor basalt (B 100 ) to SiO 2 -rich rhyolite (R 100 ), for which GFA is already known and determined by cooling-induced experiments via the so called critical cooling rate (Rc) method (for details refer to [10][11]). Experiments on re-heated and then quenched (from Tm) charges have been also structurally and chemically characterized by SEM and EPMA, to corroborate DSC data and to quantify textures and compositions of all produced phases. The attained outcomes can be useful in scenarios where glass-bearing rocks are re-heated, i.e. flowage of lavas on volcanic rocks [16][17][18][19].
In parallel, the huge abundance of these compounds, their very cheap costs and the less effort for preparation could represent valuable alternatives to design new glass-ceramics with desired macroscopic properties for encapsulation of toxic substances.
Starting glasses. The six different silicate starting glass compositions are the same used in [10][11]. Briefly, they were prepared by using two natural rocks: a basalt from Iceland (B 100 ) and rhyolite from the Lipari Island in the Aeolian arc (R 100 ). About 100 g of these two rocks were powdered and melted in Pt crucibles at temperature of 1600 °C for 4 h in air. After quenching on metal plates, they were crushed and re-melted two times at the same conditions in order to improve were then lodged in the DSC (Netzsch STA 449 C) and run with a heating rate of 10 °C/min, from room to melting conditions (1280 °C for of B 100 to 1130 °C for R 100 ), previously estimated by thermodynamic models [10]. When the liquidus region for each system was achieved, the samplecharges were rapidly quenched to ambient conditions. A duplicate DSC experiment was run for B 100 using the same thermal program, but instead a Pt-crucible was used in order to evaluate the possible effect that sample holder could have on phase nucleation and growth. The same apparatus was also used to bracket in-situ the Rc (GFA) of B 100 [10,11] following the method of [38].
SEM and EPMA analyses. The quenched run products were mounted in epoxy and polished for textural and chemical analysis. Images of phases and their textures were first identified and collected by back-scattered electrons using a field-emission JEOL 6500 F scanning electron microscopy, equipped with an energy-dispersive spectrometer (EDS). Their chemical attributes were then accurately det
This content is AI-processed based on open access ArXiv data.
