Nanoscale chemical heterogeneities control magma viscosity and failure

doi:10.21203/rs.3.rs-3891365/v1

Download PDF

Article

Nanoscale chemical heterogeneities control magma viscosity and failure

https://doi.org/10.21203/rs.3.rs-3891365/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Explosive volcanic eruptions, resulting from magma fragmentation, pose significant threats to inhabited regions. The challenge of achieving fragmentation conditions in less evolved compositions, such as andesites and basalts, stems from their low viscosities. Recent research highlights the role of Fe-Ti-oxide nanocrystals (nanolites) in increasing melt viscosity, yet the mechanisms behind the impact of nanocrystallization remain a subject of ongoing debate. To assess their effect on melt viscosity, we introduce innovative viscosity models exclusively utilizing nanolite-free viscosity data. Our study unveils the first in-situ imaging of nanolite formation in andesitic melt resulting in a heterogeneous distribution of elements, generating a relatively SiO₂-enriched matrix and Al-enriched shells around nanolites. This phenomenon results in a substantial, up to 30-fold increase in magma viscosity at eruptive temperatures. By incorporating nanoscale observations of fragmented magma from the literature, we deduce that elemental heterogeneities might play a critical role in driving magmas towards failure conditions.

Earth and environmental sciences/Solid Earth sciences/Volcanology

Earth and environmental sciences/Solid Earth sciences/Petrology

andesite

viscosity

nanolite

differential scanning calorimetry

Brillouin spectroscopy

Raman spectroscopy

TEM

Explosive eruptions are recognized as one of the most hazardous natural phenomena on Earth ¹, capable of injecting a large amount of gas and ash into the atmosphere, posing a threat to inhabited regions. These eruptions result from the magma failure and production of a hot gas-pyroclast mixture ^2–5. This phenomenon, known as magma fragmentation, can be triggered by bubble overpressure linked to their limited expansion^6,7 or induced by sufficiently high strain rates relative to the structural relaxation time, primarily dictated by the chemical composition of the melt ^2,8,9. Ocasionally, sub-Plinian and Plinian volcanic activities are fueled by andesite and basalt magmas ^10–13. Nevertheless, these kind of magmas face challenges in attaining the requisite for magma fragmentation due to their relatively low viscosity ^14–19, unless a substantial volume of microlite crystallization occurs²⁰. Recent studies have demonstrated that the formation of Fe-Ti-oxide nanocrystals (referred to as nanolites) can significantly increase the viscosity of andesitic and basaltic melts during laboratory measurements^21–23. Interestingly, these nanolites can be present in natural volcanic products erupted during explosive events ^24–28. It has been argued that such nanocrystals could have a significant impact on magma fragmentation, as they increase magma viscosity and provide sites for bubble nucleation, consequently influencing the eruptive style. However, despite the increasing frequency of discoveries of nanolites in natural products in the literature, the mechanisms and the extent to which they impact magma viscosity continue to be topics of ongoing debate ^{21,22,27,29–35}. In this study, we present the first in-situ high-temperature nanoscale observation of nanolite formation in a volcanic (andesitic) melt. We subsequently correlate the fundamental insights gained from such observations with flow behavior of various andesitic melts in the laboratory. To establish a reliable basis for comparison, we developed new viscosity models exclusively using viscosity data derived from samples devoid of nanolites. This necessity arises from the observed tendency of previous literature, including widespread viscosity models³⁶, to overestimate the nanolite-free melt viscosity of compositions prone to nanocrystallization^22,23,25,37. Ultimately, we demonstrate that nanolite formation results in structural heterogeneities at the nanoscale, creating highly viscous domains that could lead to substantial changes in the physical properties of andesitic plugs and domes. This phenomenon could potentially provide crucial insights into understanding the molecular basis of magma fragmentation.

2.1 Preliminary in-situ observation of nanolite formation during heating

We produced four anhydrous andesitic melts and one transition-metal-free analog (Table 1). The sample AND100 (Fe³⁺/∑Fe_tot = 0.64) was designed to mirror the andesitic chemical composition of the magma erupted at Sakurajima volcano (Okumura et al. 2022). Samples AND100red (Fe³⁺/∑Fe_tot = 0.27) and AND100ox (Fe³⁺/∑Fe_tot = 0.71) represent isochemical analogues of AND100 with lower and higher Fe³⁺/∑Fe_tot ratios, respectively (Table 1). AND65 (Fe³⁺/∑Fe_tot = 0.70) and AND0 were produced based on the composition of AND100, from which 35% and 100% of the total transition metal content (FeO_tot, TiO₂ and MnO) were removed. The pristine glassy nature of all specimens was confirmed through a combination of scanning electron microscopy (SEM) imaging, in backscattered electron (BSE) mode, and Raman spectroscopy analysis; additional details can be found in Supplementary Information (Supp. Inf. Section S1; Table S1).

Initially, we conducted a comprehensive exploration of the thermal response of AND100 composition from a fundamental standpoint, performing in-situ high-temperature measurements. High-temperature Raman spectra (Fig. 1a) reveal that AND100 is unstable against thermal treatments, leading to the formation of titanomagnetite nano-sized crystals already ⁓70 ℃ above the glass transition temperature (T_g), which was determined by differential scanning calorimetry (Table 1; Supp Inf. Section S2.1). This is evidenced by the appearance of vibrational features assignable to Fe-Ti-oxides such as titanomagnetite ^37–40, which became distinctly identifiable upon cooling our samples to room temperature (Supp Inf. Section S3; Fig. S4b). To monitor the nanostructural changes associated with the non-stoichiometric precipitation of titanomagnetite, we also performed the first in-situ nanoscale observation of nanolite formation in an andesitic melt using transmission electron microscopy (TEM) with a heating stage (see videos in the Supp. Material). The experimental procedure, as previously optimized by Zandonà et al. (2023), minimized possible artefacts arising from electron irradiation, which could be limited to a simple shift of thermally activated processes (i.e., phase separation and crystallization) to lower temperatures and a minor loss of alkali during the heating in vacuum. During the heating experiment (Fig. 1b), we observed that the initially homogeneous material underwent amorphous phase separation with increasing temperature, leading to the development of amorphous higher-contrast particles with a diameter that was lower than 5 nm. Notably, these particles exhibited structural ordering at higher temperatures (> 550 ℃), followed by a gradual growth of iron-rich nanosized crystals surrounded by aluminium-rich domains (Supp Inf. Fig. S5). Fast Fourier transforms (FFTs) of the images (Fig. 1b) revealed consistent with the formation of titanomagnetite (Wechsler et al. 1984; Zinin et al. 2011). Energy-dispersive X-ray spectra (EDS) acquired before and after the in-situ experiment (Supp Inf. Fig S5; Table S3) confirmed that the overall bulk composition remained almost constant after the experiment, except from an unavoidable loss of alkalis due to electron irradiation in a vacuum. The observed phenomena (chemical diffusion inducing amorphous phase separation, followed by nanocrystal formation and growth within Fe enriched domains) mirror very closely those inferred from ex-situ experiments on basaltic melts ²³. They therefore provide an accurate overview of processes that should be expected to occur in deeply supercooled melts and reheated glasses.

2.2 The viscosity of homogeneous and crystal-free andesitic melts

The in-situ high-temperature measurements revealed the high reactivity of andesitic glasses and melts at a temperature above T_g, emphasizing the need to apply a meticulous experimental methodology in deriving accurate pure melt viscosity ^22,23,42. In light of these considerations, we employed direct viscometry techniques, specifically micropenetration and concentric cylinder viscometry (Supp. Inf. Sections S3 and S5). Additionally, we expanded the dataset by incorporating indirect viscosity derivations through conventional and flash differential scanning calorimetry (C-DSC and F-DSC, respectively; Supp. Inf. Sections S2.1 and S2.2). To address potential alterations in the samples during measurements, such as crystallization and/or iron oxidation, we performed Raman and Mössbauer spectroscopy before and after experiments (Supp. Inf. Sections S2.1, S2.2, S3 and S5).

The accurate fitting of melt viscosity requires data derived from samples that have preserved a consistently homogeneous amorphous structure throughout all measurement stages. However, post-experiment Mössbauer and Raman results (Supp. Inf. Sections S2.1, S2.2, S3 and Table S2), as well as micropenetration (Fig. 2b) and post-micropenetration TEM results (Fig. 3, Fig. 4; Supp. Inf. Table S4, Fig S11, Fig. S12), evidenced that a portion of our measured viscosity values was compromised due to iron oxidation and nanocrystallization of titanomagnetite occurring during the measurements. Relying on such comprehensive experimental investigations, we exclusively utilized reliable data acquired from samples that were free of nanolites after the experiments to formulate accurate viscosity parametrizations for our andesitic compositions (further details in Supp. Inf. Sections S2.1, S2.2, S3 and S6). Additionally, we accounted for iron oxidation states, consistently categorizing the obtained datasets based on the iron valence determined after viscosity measurements. We utilized the MYEGA parametrization (Eq. 5), with log₁₀η_∞ fixed at -2.93 ^43,44 to fit the melt fragility index (m). Together with our C-DSC-derived T_g (Table 1), this approach enables us to characterize melt viscosity over a broad range of temperatures. Viscosity results obtained from samples that are devoid of nanolites are represented by coloured symbols in Fig. 2, and the fitting parameters of the corresponding MYEGA parametrizations are summarized in Table 1.

Table 1. Electron microprobe chemical compositions, Brillouin spectroscopy results and MYEGA (Eq. 5) fit parameters.

	AND100	AND100red	AND100ox	AND65	AND0
SiO₂	60.38 (0.36)	60.47 (0.40)	60.56 (0.35)	62.52 (0.19)	65.91 (0.34)
TiO₂	0.79 (0.04)	0.80 (0.07)	0.81 (0.06)	0.56 (0.04)	0.04 (0.03)
Al₂O₃	16.69 (0.18)	16.83 (0.22)	16.79 (0.19)	17.28 (0.22)	18.01 (0.11)
FeO_tot	6.77 (0.12)	6.87 (0.12)	6.76 (0.15)	4.40 (0.11)	0.04 (0.03)
MnO	0.17 (0.03)	0.18 (0.04)	0.18 (0.03)	0.11 (0.03)	0.01 (0.01)
MgO	3.00 (0.07)	2.94 (0.06)	2.95 (0.06)	3.02 (0.07)	3.21 (0.06)
CaO	6.62 (0.09)	6.51 (0.11)	6.49 (0.11)	6.76 (0.13)	7.26 (0.11)
Na₂O	3.50 (0.15)	3.46 (0.10)	3.40 (0.10)	3.49 (0.11)	3.82 (0.14)
K₂O	1.58 (0.07)	1.65 (0.07)	1.66 (0.06)	1.70 (0.05)	1.75 (0.06)
P₂O5	0.18 (0.05)	0.18 (0.05)	0.18 (0.05)	0.16 (0.05)	0.04 (0.03)
Fe³⁺/∑Fe_tot^a	0.64	0.27	0.71	0.70	-
NBO/T^b	0.34	0.34	0.34	0.27	0.16
K/G^c	1.57 (0.02)	1.57 (0.02)	1.58 (0.02)	1.53 (0.02)	1.48 (0.02)
T_g^d (℃)	654 (1)	645 (1)	662 (1)	696 (1)	737 (1)
m^e	30.5 (0.5)	30.5 (0.5)	30.5 (0.5)	31 (0.4)	31.8 (0.2)
^aRatios derived using Mossa software ⁴⁵.
^bNBO/T calculated after Prata et al. (2019).
^cValues derived using Eq. 4 and Brillouin spectroscopy data.
^dDerived via DSC measurements (see Methods or SI).
^eFitted fragility indices (m) using the Mauro–Yue–Elli- son–Gupta–Allan (MYEGA, Eq. 5) parametrization using η_∞ = -2.93.

Our results (Fig. 2a) show that the homogeneous viscosity of andesitic melts increases as their transition metal oxide content is reduced (progressing from AND100 to AND65 and then AND0). A 35% removal of transition metal oxides (FeO_tot, TiO₂ and MnO) from AND100 would result in a viscosity increase of 0.7 ± 0.05 log₁₀ Pa s (⁓ 5 times) at the eruptive temperature for Sakurajima volcano (900–1050 ℃; Araya et al. 2019). In contrast, a complete removal of transition metal oxides (i.e., AND0) would results in an increase of about ⁓1.3 ± 0.1 log₁₀ Pa s (⁓ 20 times), or a rise of ⁓1.5 ± 0.1 log₁₀ Pa s (⁓ 30 times) if starting from AND100red, within the same temperature range. Moreover, the only increase in iron oxidation, from Fe³⁺/∑Fe_tot = 0.27 to 0.71, leads to higher viscosity up to 2 times at the previously mentioned temperature range (as exemplified by the comparison between AND100red and AND100ox in Supp. Inf. Fig. S8), which is consistent with previous literature ^19,47,48. It is crucial to emphasize that these parametrizations pertain specifically to homogeneous melts, free of nanolites or observable nanoscale compositional fluctuations in the amorphous state, as further discussed below. Additionally, the Giordano et al. (2008) model (dashed lines in Supp. Inf. Fig. S8), significantly overestimates the viscosity of AND100 (red dashed lines in Supp. Inf. Fig. S8) by up to 25 and 3 times between 700 and 900°C, respectively. Nevertheless, we do not observe such difference when comparing the AND0 compositions (black dashed lines in Supp. Inf. Fig. S8). This observation is consistent with the results reported by Valdivia et al. (2023) for basaltic compositions, implying that the Giordano et al. (2008) model might have been constructed based on viscosity data derived from melts exhibiting nanoscale heterogeneity.

2.3 The viscosity of nanolite-bearing andesitic melts

While conducting micropenetration measurements on AND100 samples, we observed a time-dependent increase in viscosity at constant temperatures (red arrows in Fig. 2b; details in Supp. Inf. Section S3). Subsequent post-experiment analyses (Supp. Inf. Section S3) revealed that the samples underwent nanocrystallization of titanomagnetite and iron oxidation. As such, to further explore the mechanisms behind the increase in viscosity, we employed transmission electron microscopy (TEM) explorations on these samples.

Electron diffraction patterns acquired from post-micropenetration AND100 samples in TEM mode (Supp. Inf. Fig S9) confirmed the presence of nanocrystals (additional images in the Supp. Material). Most of the recorded d-spacings can be associated with the structure of titanomagnetite ^49,50. Specifically, the diffraction patterns evolved from diffused circular halos in the case of AND100_MP₆₆₀ to a more distinct set of well-visible diffraction rings in AND100_MP₇₂₃ and especially AND100_MP₈₀₈. Notably, AND100_MP₈₀₈ displayed more pronounced diffraction features along the rings, providing evidence for the formation of well-ordered and larger nanocrystals. These results are consistent with our high-temperature in-situ experiments (Fig. 1; Supp. Inf. Fig. S4), suggesting that nucleation and growth of nanolite become more pronounced as temperatures increase further above T_g. Additionally, chemical micrographs were collected in scanning transmission electron microscopy (STEM) on FIB-made lamellas of known thickness (Supp. Inf. Table S4) to obtain quantitative data on the nano-structural changes contributing to the increase in viscosity. The analyses of high-angle annular dark-field (STEM-HAADF) images (Supp. Inf. Fig. S10) provided minimum radius, minimum nanolite content (in vol%) and minimum nanolite number density (NND) for the three AND100 samples subjected to micropenetration (Supp. Inf. Table S4). The average minimum radius of nanocrystals increased from 1.4 ± 0.5 to 2.6 ± 0.8 nm as the temperature explored during viscosity measurements increased from 660 to 808 ℃ (Supp. Inf. Table S4), corroborating the results of Raman spectroscopy and electron diffraction (Supp. Inf. Sections S3 and S6). Moreover, compositional elemental maps were extracted by EDS as shown in Fig. 3.

We observe that the Fe-rich regions can be identified with the nanocrystals appearing bright in the STEM-HAADF images (Fig. 3, Supp. Inf. Fig S11 and Fig. S12), as they correspond to the denser phase. Conversely, Ti appears to be distributed around the Fe-rich zones. Similarly, we notice that Al is preferentially distributed around nanolites, leaving Al-depleted regions between them. This unique behaviour becomes more apparent when the Fe, Ti and Al elemental maps are overlapped, as shown in Fig. 3 (and in Supp. Inf. Fig S11 and Fig. S12). To specify these observations, and following our previous image analyses, we computed the average composition of the following sub-regions: 1) the bulk image, 2) the nanolites, 3) the nanolites and the compositionally differentiated halo around them, 4) the residual amorphous matrix (as illustrated in Fig. 3). Due to the thickness of the samples, results obtained for nanolites and halos inevitably include a contribution from the surrounding amorphous matrix. First, we confirmed that our bulk EDS composition closely aligns with the electron microprobe chemical composition of AND100 starting material (Table 1 and Supp. Inf. Table S4), evidencing that only a negligible migration of alkalis occurred during EDS measurements. Figure 4 illustrates the relative compositional differences between the various sub-regions, normalized to the bulk chemistry (Supp. Inf. Table S4). Our measurements reveal that the residual amorphous matrix in all three samples experienced enrichment in SiO₂, with average values of approximately 70 wt% (Supp. Inf. Table S4), which is ca. 10 wt% higher than the original AND100 composition (Table 1). Conversely, a gradual depletion in Al₂O₃ and FeO_tot was observed as the experimental temperature increased from 660 to 808 ℃ during viscosity measurements. In contrast, we observed a progressive increase in concentration of FeO_tot and TiO₂ in nanolite regions, aligning with the nucleation and growth of titanomagnetite nanolites observed in our high-temperature in-situ experiments (Fig. 1). Notably, we observed that nanolites exhibited an initial enrichment in Al₂O₃ during the early stages of formation (660 ℃), while Al seemed to be expelled from the developing nuclei at higher temperatures (808 ℃), when ordered titanomagnetite nanolites began to form (Supp. Inf. Fig. S9c). We infer that this phenomenon produces the preferential distribution of aluminium around nanocrystals noticed in Fig. 3 and observed after high-temperature in-situ TEM experiments (Supp. Inf. Fig. S5), whereas the surrounding matrix is enriched in SiO₂. This behaviour is consistent with well-established observations in glass-ceramic materials, exhibiting Al-enriched shells around TiO₂- and/or ZrO₂-bearing nanocrystals acting (upon further heating) as seeds for the controlled crystallization of the surrounding aluminosilicate matrix ⁵¹. Recent tracer diffusivity data obtained from a TiO₂-containing albite glass further confirmed that the mobility of Al in supercooled aluminosilicate melts is closely related to that of (and strongly enhanced by) transition metals ⁵².

To elucidate the impact of nanocrystallization and iron oxidation on the viscosity of andesitic melts, we compare our viscosity micropenetration results for AND100 samples with our novel pure melt viscosity parametrizations (Fig. 2b). For AND100_MP₆₆₀, our initial viscosity measurement (η = 10¹² Pa s) closely aligns to the pure melt viscosity of AND100 at 660 ℃, but we observed a progressive increase in viscosity at isothermal conditions, recording a final viscosity value of η = 10^12.7 Pa s (the overall iron oxidation state slightly increased during the measurement to Fe³⁺/∑Fe_tot = 0.7). This viscosity is roughly twice greater than the viscosity parametrization of AND100 (Fe³⁺/∑Fe_tot = 0.64) at 660 ℃. However, our STEM and EDS results (Fig. 4; Table S4) show that, on average, less than 3% of the FeO_tot precipitated in the form of nanocrystals in this sample. As such, neither the presence of solid crystalline particles (< 1%; Supp. Inf. Table S4), as suggested by classic particle suspension models ²⁰, nor the overall compositional variations (e.g., Fe-Ti removal; Fig. 5) of the surrounding amorphous phase are sufficient to explain such an increase in viscosity. We argue that the heterogeneous distribution of chemical species (Figs. 3 and 4) is responsible for the observed surplus in measured viscosity as illustrated in Fig. 5. Similar conclusions can be drawn from the viscosity measurements performed at higher temperatures. For AND100_MP₇₂₃, the final viscosity value (η = 10^11.9 Pa s; Fe³⁺/∑Fe_tot = 0.79) is almost twice the pure melt parametrization of AND65 (Fe³⁺/∑Fe_tot = 0.70), although the residual matrix of this sample still contains approximately 40% of the initial FeO_tot content (Supp. Inf. Table S4). For AND100_MP₈₀₈, the final viscosity value (η = 10^10.17 Pa s; Fe³⁺/∑Fe_tot = 0.83) slightly surpassed the melt parametrization of AND0 at 808 ℃ (η = 10¹⁰; Fig. 2), although the residual amorphous phase was far from being completely free of transition metal oxides: approximately 25% of the initial FeO_tot content (Fig. 4; Supp. Inf. Table S4) was still contained by the aluminosilicate glass surrounding the nanolites. Even in this comparatively evolved sample, the simple presence of solid crystalline particles (~ 1 vol%) is too negligible to play any relevant rheological role ²⁰, especially due to the absence of any evident physical interaction between them (such as coalescence or aggregation). We stress that our findings refute the assumptions of previous authors, who argued that the increase in viscosity due to titanomagnetite nanocrystallization is solely attributed to the removal of iron from the residual aluminosilicate matrix ^21,35. Moreover, our post-micropenetration Mössbauer spectroscopy results (Supp. Inf. Section S3; Table S2) indicate that the Fe³⁺/∑Fe_tot values of post-micropenetration AND100 samples are significantly higher than the required stoichiometric conditions for titanomagnetite to consume all the iron in the melt. This suggests that in samples AND100_MP₇₂₃ and AND100_MP₈₈₀, which are shown to have well-formed titanomagnetite nanolites, the remaining iron contained in the SiO₂-enriched matrix may be predominantly in the Fe³⁺ state. In this scenario, the remaining iron may acts preferentially as a network former ⁵³, further increasing the nanoscale local viscosity of the remaining matrix.

As such, a notable surplus in viscosity (Fig. 5; between 0.2 and 0.8 log₁₀ units) emerges across all three cases with respect to values expected for homogeneous melts. Considering the very low crystal content of the melts (< 2 vol%; Supp. Inf. Table S4), this phenomenon must arise from the pervasive heterogeneity produced by the concurrent diffusion and segregation of elements (e.g., Al, Ti, Fe; Fig. 3, Fig. 4; Supp. Inf. Table S4) driven by the iron oxidation and nanocrystallization of titanomagnetite. This process leads to the formation of chemically differentiated nanodomains in the initially homogeneous melt, including nano-sized solid particles, Al-enriched shells, and highly SiO₂-enriched regions (up to ⁓70 SiO₂ wt%; Fig. 4 and Supp Inf. Table S4), thereby significantly increasing the overall viscosity (Fig. 5). Looking at this behaviour, one could assert that the viscosity of AND0 (a homogeneous melt that is devoid of transition metals) merely establishes a lower boundary for the nanolite-bearing viscosity of andesitic compositions at eruptive temperatures, since compositional fluctuations at the nanoscale appear to play a substantial role, at least as relevant as that of the overall chemical composition of the melt. It is now evident that the measurement of physical properties of titanomagnetite-bearing silicate melts cannot be comparable with homogeneous materials, as we have shown that the nucleation and growth of nanolites produce chemically differentiated nanodomains. For example, we have observed that the progressive nanocrystallization of nanolites produces a sustained increase of DSC-derived characteristic temperatures (Sup. Inf. S2.1 and S2.2). Furthermore, despite the low titanomagnetite content observed in natural melts, the impact of titanomagnetite crystallization should not be overlooked.

Our results carry direct implications for comprehending the dynamics of natural domes and plugs, as they are known to be exposed to naturally occurring reheating processes ⁵⁴. Additionally, the presence of nanoscale liquid immiscibility (phase separation) under eruptive conditions has already been observed in lavas from the 2018–2021 Fani Maoré eruption ⁵⁵. Moreover, Bamber et al. (under review) reported the presence of a more viscous melt around Fe-rich nanolites in pyroclasts erupted during the basaltic Plinian events at Masaya volcano. As anticipated, nanolite agglomeration is not observed in our samples, as the high-viscosity landscape we investigated does not allow for such agglomeration ^23,33,37. Nevertheless, Bamber et al. (under review) also reported elemental nanoscale heterogeneities around nanolite aggregates in basaltic glasses, suggesting that the implications on magma viscosity might be even more profound. Furthermore, we deduce that the existence of differentiated domains at the nanoscale might play a crucial role in controlling eruptive dynamics, as suggested in previous works ^28,33, indicating that Fe-rich nanolites could serve as proto-fragmentation surface for future ash particles. Indeed, recent investigations ⁵⁶ have revealed nanoscale Al-rich heterogeneities at the surface of andesitic ash particles, suggesting that ash-forming fractures preferentially propagate through boundary layers around nanosized Fe-rich phases. These findings align with the formation of chemically differentiated nanodomains around nanolites, as magma failure should propagate through the most viscous zones (i.e., SiO₂-enriched matrix), resulting in the observed Al-rich surfaces in ash particles. Additionally, we argue that nanoscale elemental heterogeneities may not only increase the magma viscosity but also facilitate bubble nucleation sites ⁵⁷, hinder bubble connectivity and outgassing, promote gas-melt coupling, enhance ascent velocity, and increase strain rates. Collectively, these factors could potentially act as a gateway to magma fragmentation, explaining the occurrence of less evolved explosive eruptions. Indeed, earlier research ^22,33 has demonstrate that water-bearing basaltic melts can display similar viscosity increases during heat treatments, linked to nanolite crystallization and nucleation of a substantial number of bubbles. Consequently, we posit that the heterogeneous distribution of elements induced by nanolite crystallization contributes significantly to the observed viscosity increase in Fe-bearing aluminosilicate melts, influencing the formation and propagation of fractures and potentially controlling degassing dynamics of magmas.

We present the first in-situ imaging nanoscale observation of nanolite formation in andesitic melt and introduce novel melt viscosity models tailored for various andesitic compositions. Our observations indicate that above the glass transition temperature, iron oxidation and nanocrystallization readily occur. Our study challenges conventional explanations, demonstrating that the increase in viscosity due to titanomagnetite nanocrystallization cannot be solely attributed to the depletion of iron in the remaining matrix or the presence of solid crystal particles. Instead, we propose a nuanced mechanism: the precipitation of nanocrystals induces a heterogenous distribution of elements in the residual melt, generating a relatively SiO₂-enriched matrix and Al-enriched shells around nanolites. This results in a significant, up to 30-fold, surge in magma viscosity at eruptive temperatures. This heightened magma viscosity, coupled with molecular-scale variations in viscosity, may play a key role in fracture formation and propagation within magmas, potentially contributing to conditions leading to explosive eruptions.

4.1 Synthesis of starting glasses

AND100, AND65 and AND0 were synthesized by mixing powder reagents (SiO₂, TiO₂, Al₂O₃, Fe₂O₃, MnO, MgO, CaCO₃, Na₂CO₃, K₂CO₃ and P₂O₅) according to their target compositions. The mass of each oxide and carbonate component was determined through molar mass calculations. All reagents were mixed using an agate mortar and ethanol. The mixture underwent manual grinding for ~ 45 minutes before being dried using an infrared light. Subsequently, the dry mixture was placed in an alumina crucible and subjected to an overnight heat treatment at 900°C to eliminate CO₂ from the carbonate compounds. Following decarbonization, the material was transferred to a Pt crucible and melted for 24 hours at 1400°C. Afterwards, the melt was rapidly quenched in water to prevent crystallization. The resulting quenched glass was crushed to powder using a stainless-steel percussion mortar and then manually mixed before performing a second melting to achieve chemical homogenization. The second round of melting at 1400°C lasted for 4 hours, after which the crucible was swiftly immersed in water for rapid cooling. Subsequently, AND100red was produced by re-melting AND100 sample in a hanging Au₈₀Pd₂₀ open capsule at 1275 ℃ and 1 atm for 24 hours, using a gas mixing furnace with a gas mixture of 95% CO₂ and 5% CO. The resulting melt was rapidly quenched in water by melting the Pt wire.

4.2 Electron microprobe analyses (EMPA)

The major elemental composition (Si, Ti, Al, Fe_tot., Mn, Mg, Ca, Na, K, and P) of samples AND100, AND100ox, AND100red, AND65, and AND0 was determined using a JEOL JXA-8200 electron microprobe at the Bayerisches Geoinstitut (University of Bayreuth, Germany) (Table 1). Glasses were embedded in epoxy, polished, and carbon coated. Measurements were performed using 15 kV voltage, 5 nA current, and 20 seconds of counting time under a defocused 10 µm beam. We collected 20 to 30 points per sample to account for heterogeneities. Synthetic wollastonite (Ca, Si), periclase (Mg), hematite (Fe), spinel (Al), orthoclase (K), albite (Na), manganese titanate (Mn, Ti), and apatite (P) were used as calibration standards. Sodium and potassium were analysed first to prevent alkali migration effects ⁵⁸.

4.3 Micropenetration viscometry

We conducted micropenetration (MP) viscometry measurements on plane-parallel and polished glass chips of 2–3 mm in thickness. These measurements were carried out utilizing a vertical dilatometer (Bähr VIS 404) at the Institute of Non-Metallic Materials, TU Clausthal (Germany). We measured the indentation rate of a sapphire sphere (r = 0.75 mm) during isothermal dwells at temperatures controlled using an S-type thermocouple (Pt-PtRh) placed at ~ 1.5 mm from the sample surface. The temperature error is estimated to be ± 5°C considering the accuracy of the S-type thermocouple and its distance from the sample ⁵⁹. We followed standard procedures ^22,23,37,60 to achieve thermal equilibration at the target temperature. The indentation depth was measured as a function of time and the viscosity curve was determined according to Eq. 1 ⁶¹:

$$\begin{array}{c}\eta =\frac{9F}{32 \sqrt{2r} \sqrt{{L}^{3}}}t\#\left(1\right)\end{array}$$

where η is the Newtonian viscosity (Pa s), F is the applied force (N), t is the time (s), r is the radius of the sphere (m) and L is the indentation depth (m). The dilatometer was previously calibrated using a standard glass DGG-1, reproducing the certified viscosity data ⁶² with a deviation of ± 0.1 in log units. Data points are reported in the text according to the scheme SampleName_MP_Temperature, with temperature representing the final experimental temperature expressed in °C, and the duration in minutes.

4.4 Concentric cylinder (CC) viscometry

High-temperature viscosity measurements were conducted using a Rheotronic II Rotational Viscometer (Theta Instruments) at the Experimental Volcanology and Petrology Laboratory (EVPLab, Roma Tre University, Italy). The experimental apparatus featured an Anton Paar Rheolab Qc viscometer head with a maximum torque capacity of 75 mN m ⁶³. Temperature monitoring was carried out using a factory-calibrated S-type thermocouple, with a precision of ± 2°C. The concentric cylinder was previously calibrated using a standard glass NIST 717a, reproducing the certified viscosity data with a deviation of ± 0.03 in log units ⁶⁴. To ensure thorough thermo-chemical homogenization, the glass materials were loaded into Pt₈₀Rh₂₀ cylindric crucible (62 mm in height, and 32 mm inner diameter) and stirred at γ̇ = 10 s⁻¹ using a Pt₈₀Rh₂₀ spindle (3.2 and 42 mm in diameter and length, respectively) at 1435°C at air oxygen fugacity and ambient pressure for 5 hours. Subsequently, the temperature was lowered by steps of 25–50°C down to 1150 and 1130°C for the samples AND100ox and AND0, respectively. The viscosity was measured at every step, holding the conditions constant until steady viscosity and temperature values had been achieved (~ 45 min). At the end of the experiments, the temperature was quickly raised to 1430 ℃ where a portion of the melt was rapidly quenched in water to determine the iron oxidation state of the high-temperature viscosity measurements.

4.5 Differential scanning calorimetry

Conventional differential scanning calorimetry (C-DSC) measurements were performed at the Institute of Non-Metallic Materials, TU Clausthal (Germany). Around 15 mg (± 5) of glass was placed in a Pt₈₀Rh₂₀ crucible under a constant N₂ (5.0) flow rate of 20 ml min^− 1. We used two conventional differential scanning calorimeters (C-DSC, 404 F3 Pegasus and 404 cell, Netzsch) to measure the heat flow at a heating rate (q_h) of 10 and 20 ℃ min^− 1. Additionally, we used ~ 50 ng of glass to perform flash differential scanning calorimetry (F-DSC) analyses, using a Flash DSC 2+ (Mettler Toledo) equipped with UFH 1 sensors, under constant Ar 5.0 flow (40 ml min^− 1).

The C-DSC was calibrated using melting temperatures and enthalpy of fusion of reference materials (pure metals: In, Sn, Bi, Zn, Al, Ag, and Au), and the F-DSC was calibrated using the melting temperature of aluminium (660.3 ℃) and indium (156.6°C). In our C-DSC measurements, we employed the methodology outlined by Stabile et al. (2021). Initially, we erased the thermal history of the glass by subjecting the sample to a two-step thermal treatment. This involved a first upscan at a rate of q_h = 20 ℃ min^− 1 until it reached a temperature slightly above T_peak, namely T_max. Subsequently, we cooled the melt to 100°C at rates of q_c = 10 or 20 ℃ min^− 1. The actual C-DSC measurements were then conducted using the rate-matching method, which entailed an additional upscan (matching heating segment) with a rate matching that of the preceding downscan (i.e., q_h = q_c). From the measured heat flow during the matching upscan, we extracted the characteristic temperatures T_onset and T_peak. For further details see the methodology presented in Valdivia et al. (2023).

For F-DSC experiments, we followed the methodology described above employing a q_h = q_c = 1,000 ℃ s^− 1 (60,000 ℃ min^− 1). Subsequently, we conducted a series of measurements on the same sample, using the same chip, at 10,000 ℃ s^− 1 (600,000 ℃ min^− 1) to investigate the impact of nanocrystallization on the characteristic temperatures T_onset and T_peak due to consecutive thermal treatments.

Following the theoretical background discussed elsewhere ^37,65–67, viscosity values were derived from C- and F-DSC data using the relationship between the matching heating rate (q_h) of the measurement and the shift factors K_onset and K_peak ^37,38 expressed in Eq. 2:

$$\begin{array}{c}{log}_{10}\eta \left({T}_{onset,peak}\right)={K}_{onset,peak}-{log}_{10}\left({q}_{h}\right)\#\left(2\right)\end{array}$$

where K_onset = 11.20 ± 0.15 and K_peak = 9.84 ± 0.20 ^37,38. It is important to mention that when q_h is 10 ℃ min^− 1, η(T_onset) ≈ 10¹² Pa s, and therefore, T_{onset ≈} T_g.

4.6 Brillouin Spectroscopy

Brillouin spectroscopy (BLS) measurements were performed at the Bayerisches Geoinstitut (University of Bayreuth, Germany). Plane-parallel glass plates with a thickness of ~ 50 µm were analysed using a solid-state Nd:YVO4 laser source operating at a wavelength of 532 nm and 50 mW power. The Brillouin frequency shift was quantified utilizing a six-pass Fabry–Perot interferometer ⁶⁸ coupled with a single-pixel photon counter detector. Measurements were conducted using a symmetric forward scattering configuration ^68,69 with a scattering angle of θ = 79.8°. The accuracy of the scattering angle was established through calibration with a reference silica glass. Conversion of frequency shifts (Δω) to longitudinal (v_p) and shear (v_s) sound velocities was carried out using Eq. 3:

$$\begin{array}{c}v=\frac{\varDelta \omega \lambda }{2sin\left(\theta /2\right)}\#\left(3\right)\end{array}$$

where λ is the laser wavelength and θ is the angle between the incident and scattered beams ^68,70. We collected 8 spectra for each sample at different rotation angles (from − 180° to + 180°) to factor for uncertainties. Finally, we calculated the K/G factor using the ratio between v_p and v_s (Eq. 4):

$$\begin{array}{c}\frac{K}{G}={\left(\frac{{v}_{p}}{{v}_{s}}\right)}^{2}-\frac{4}{3}\#\left(4\right)\end{array}$$

4.7 (High-temperature) Raman spectroscopy

Glasses subjected to micropenetration, concentric cylinder viscometry, C-DSC and F-DSC were analysed before and after measurements to account for potential modifications (i.e., crystallization and/or iron oxidation). For this, we used a confocal Raman imaging microscope at the Institute of Non-Metallic Materials, TU Clausthal (alpha300R, WITec GmbH), where spectra were acquired using a 100x objective in the ranges between 10–1300 cm^− 1. Acquisition parameters included an integration time of 10 seconds, an accumulation count of 5, and a laser power of 10 mW. Spectra were smoothed to enhance the signal-to-noise ratio.

Additionally, we performed in-situ high-temperature Raman analyses on AND100 sample. We targeted the same temperatures and heating treatments as those used for micropenetration experiments. We used a Renishaw InVia Qontor Raman spectrometer at the CEMHTI, Orleans (France). Spectra were acquired using a 20x NA 0.35 objective in the ranges between 150–2000 cm^− 1. Acquisition parameters included an integration time of 60 to 120 seconds, and a laser power of 20 mW.

4.8 Mössbauer spectroscopy

Mössbauer measurements were performed at the Bayerisches Geoinstitut (University of Bayreuth, Germany). Glass samples of ~ 4 mm diameter and ~ 600 µm thickness were measured before and after experiments at room temperature (293 K) using a constant acceleration Mössbauer spectrometer equipped with a high specific activity (370 MBq) ⁵⁷Co point source within a 12 µm thick Rh matrix. Calibration of the velocity scale was relative to a 25 µm thick α-Fe foil, and data were gathered within the range of ± 5 mm s^− 1, with acquisition durations of 2 to 3 days each. The recorded spectra were fitted with the full transmission integral using the MossA software ⁴⁵. Finally, the resulting Fe³⁺/Fe²⁺ ratios were calculated using the relative area associated to each iron species.

4.9 (High-temperature) Transmission electron microscopy (TEM) analyses

We performed TEM analyses on AND100 samples post-micropenetration experiments and in-situ heating TEM observations on the AND100 starting glass. TEM explorations after micropenetration were performed at the Bayerisches Geoinstitut (University of Bayreuth, Germany) using a FEI Titan G2 80-200S/TEM. TEM imaging was acquired from lamellas made using a focused ion beam (FIB) with thicknesses ranging from 25 to 50 nm. These lamellas were extracted from the same samples that were used for micropenetration experiments, utilizing a SCIOS Dual Beam system from FEI Company. We used a Gallium ion beam with variable current, from 7.8 pA to 300 nA depending on the precision requirements. Analytical scanning transmission electron microscopy (STEM) micrographs were collected at 200 kV using an energy-dispersive X-ray spectrometer (EDS) system consisting of four silicon drift detectors (Bruker, QUANTAX EDS). Geometrical analyses of STEM-HAADF images were subjected to pixel segmentation and classification using the ilastik software, version 1.4.0 ⁷¹.

In-situ heating experiments were performed at the CNRS CEMHTI in Orléans (France) using a JEOL ARM200F (JEOL Ltd.) Cold FEG microscope operating at 80 kV, mounting a double spherical aberration corrector, a Gatan Imaging Filter (GIF, Gatan Ltd.) and a OneView camera. The experimental procedure was optimized in a previous work ⁴¹ to minimize artifacts and sample damages due to highly energetic electron irradiation. The AND100 glass was crushed and grinded in an agate mortar, adding ethanol to obtain a diluted suspension; one drop of the liquid was then loaded onto an MEMS grid specifically adapted for a Protochips Fusion double-tilt heating holder and dried in air overnight. Plasma cleaning was avoided to prevent major changes in the redox state of iron in the sample. After introducing the sample holder into the TEM column, it was pre-emptively treated at 200°C for 1 h to remove possible volatile contaminants. The subsequent in-situ experiment involved manual heating at 1 ℃ s^− 1 to various temperatures between 200°C and 750°C, where isothermal dwells of 30–60 s were applied to facilitate nanoscale observation in TEM mode (the time-temperature curve is presented in Supp. Inf. Fig S8). Sample drift was manually compensated during the heating ramps. After the experiment, the acquired data was manually resampled (3 images for each isothermal dwell) and re-aligned using the software DigitalMicrograph GMS.3 (Gatan).

4.10 Viscosity parametrization

The combination of C-DSC, F-DSC and CC viscosity data enabled the parametrization of the melt viscosity of our samples as a function of temperature η(T), via the Mauro–Yue–Ellison–Gupta–Allan (MYEGA) equation (Eq. 5) ⁴⁴:

$$\begin{array}{c}{log}_{10}\eta \left(T\right)={log}_{10}{\eta }_{\infty }+\left(12-{log}_{10}{\eta }_{\infty }\right)\frac{{T}_{g}}{T}exp\left[\left(\frac{m}{12-{log}_{10}{\eta }_{\infty }}-1\right)\left(\frac{{T}_{g}}{T}-1\right)\right]\#\left(5\right)\end{array}$$

where ${log}_{10}{\eta }_{\infty }=-2.93\pm 0.3$ is the logarithmic viscosity at infinite temperature ^43,44, ${T}_{g}$ is the glass transition temperature determined by C-DSC (T_onset at q_h,c = 10 ℃ min⁻¹) and m is the melt fragility defined in Eq. 6 ⁷² as the slope of viscosity curve evaluated at T_g:

$$\begin{array}{c}m={\left.\frac{\partial {log}_{10}\eta }{\partial {T}_{g}/T}\right|}_{T={T}_{g}}\#\left(6\right)\end{array}$$

The melt fragility parameter, m, can be determined by fitting Eq. 5 to our viscosity datasets and the measured ${T}_{g}$ via C-DSC. Additionally, m also can be inferred from BLS measurements using the empirical relationship introduced by Cassetta et al. (2021) (Eq. 7) and recently employed by Di Genova et al. (2023) for the viscosity of peridotitic melts,

$$\begin{array}{c}m=43\cdot \frac{K}{G}-31\#\left(7\right)\end{array}$$

Acknowledgements

PV and DDG acknowledge the funding by Deutsche Forschungsgemeinschaft (DFG) project DI 2751/2–1. DDG acknowledges the funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (NANOVOLC, ERC Consolidator Grant – No. 101044772). This project has benefited from the expertise and facilities of the Platform MACLE-CVL, which has been co-funded by the European Union and the Centre-Val de Loire Region (FEDER). JD acknowledges DFG for financial support via the grant DE 598/33-1. The Scios FIB and the Titan G2 STEM at Bayerisches Geoinstitut were financed by DFG Grants INST 91/315-1 FUGG and INST 91/251-1 FUGG, respectively. MA and CG acknowledge funding from the Agence Nationale de la Recherche (ANR) through project ANR-23-CE08-0013-01. AV and CR acknowledge funding by MUR-PRIN Project P20222BP7J. We thank Alexander Rother and Raphael Njul for sample preparation, and Catherine McCammon for facilitating the Mössbauer facilities at the Bayerisches Geoinstitut.

All authors declare that they have no conflicts of interest.

Author contributions

P.V. drafted the original manuscript, synthesised the starting materials, processed the experimental data, performed the analyses, derived the viscosity models, performed Mössbauer experiments and constructed the figures and tables. P.V., A.Z. and D.D.G. conceptualized the original idea. A.Z. and C.G. performed the in-situ high temperature STEM. D.B. and A.C. performed the high-temperature Raman measurements. P.V., J.L. and D.D.G. performed micropenetration, room temperature Raman and calorimetry analyses. P.V. and N.M. performed the room temperature STEM. P.V. and F.D.F performed the concentric cylinder measurements. P.V., A.K. and T.B.B performed the Brillouin analyses. J.D., C.R., A.V. and M.A facilitated laboratories and instruments. All coauthors provided feedback and contributed to the final version of the manuscript.

Loughlin, S. C., Sparks, S., Brown, S. K., Jenkins, S. F. & Vye-Brown, C. Global volcanic hazards and risk. Global Volcanic Hazards and Risk (Cambridge University Press, 2015). doi:10.1017/CBO9781316276273.
Papale, P. Strain-induced magma fragmentation in explosive eruptions. Nature 397, 425–428 (1999).
Gonnermann, H. M. Magma Fragmentation. Annu. Rev. Earth Planet. Sci. 43, 431–458 (2015).
Cashman, K. V. & Scheu, B. Magmatic Fragmentation. The Encyclopedia of Volcanoes (Elsevier Inc., 2015). doi:10.1016/b978-0-12-385938-9.00025-0.
Cassidy, M., Manga, M., Cashman, K. V. & Bachmann, O. Controls on explosive-effusive volcanic eruption styles. Nat. Commun. 9, 2839 (2018).
Kueppers, U., Scheu, B., Spieler, O. & Dingwell, D. B. Fragmentation efficiency of explosive volcanic eruptions: A study of experimentally generated pyroclasts. J. Volcanol. Geotherm. Res. 153, 125–135 (2006).
Mueller, S., Scheu, B., Spieler, O. & Dingwell, D. B. Permeability control on magma fragmentation. Geology 36, 399–402 (2008).
Dingwell, D. B. Volcanic Dilemma: Flow or Blow? Science (80-. ). 273, 1054–1055 (1996).
Okumura, S., Nakamura, M., Nakano, T., Uesugi, K. & Tsuchiyama, A. Shear deformation experiments on vesicular rhyolite: Implications for brittle fracturing, degassing, and compaction of magmas in volcanic conduits. J. Geophys. Res. Solid Earth 115, 1–10 (2010).
Valdivia, P., Marshall, A. A., Brand, B. D., Manga, M. & Huber, C. Mafic explosive volcanism at Llaima Volcano: 3D x-ray microtomography reconstruction of pyroclasts to constrain shallow conduit processes. Bull. Volcanol. 84, 2 (2022).
Marshall, A. A. et al. The mafic Curacautín ignimbrite of Llaima volcano, Chile. J. Volcanol. Geotherm. Res. 421, 107418 (2021).
Bamber, E. C. et al. Pre- and syn-eruptive conditions of a basaltic Plinian eruption at Masaya Volcano, Nicaragua: The Masaya Triple Layer (2.1 ka). J. Volcanol. Geotherm. Res. 392, 106761 (2020).
Parfitt, E. A. A discussion of the mechanisms of explosive basaltic eruptions. J. Volcanol. Geotherm. Res. 134, 77–107 (2004).
Arzilli, F. et al. Magma fragmentation in highly explosive basaltic eruptions induced by rapid crystallization. Nat. Geosci. 12, 1023–1028 (2019).
La Spina, G., Arzilli, F., Burton, M. R., Polacci, M. & Clarke, A. B. Role of volatiles in highly explosive basaltic eruptions. Commun. Earth Environ. 3, (2022).
Vetere, F. et al. Viscosity of andesite melts and its implication for magma mixing prior to Unzen 1991-1995 eruption. J. Volcanol. Geotherm. Res. 175, 208–217 (2008).
Robert, G., Whittington, A. G., Stechern, A. & Behrens, H. The effect of water on the viscosity of a synthetic calc-alkaline basaltic andesite. Chem. Geol. 346, 135–148 (2013).
Richet, P., Lejeune, A. M. M., Holtz, F. & Roux, J. Water and the viscosity of andesite melts. Chem. Geol. 128, 185–197 (1996).
Liebske, C., Behrens, H., Holtz, F. & Lange, R. A. The influence of pressure and composition on the viscosity of andesitic melts. Geochim. Cosmochim. Acta 67, 473–485 (2003).
Mader, H. M., Llewellin, E. W. & Mueller, S. P. The rheology of two-phase magmas: A review and analysis. J. Volcanol. Geotherm. Res. 257, 135–158 (2013).
Okumura, S. et al. Rheology of nanocrystal-bearing andesite magma and its roles in explosive volcanism. Commun. Earth Environ. 3, 241 (2022).
Valdivia, P. et al. Are volcanic melts less viscous than we thought? The case of Stromboli basalt. Contrib. to Mineral. Petrol. 178, 45 (2023).
Scarani, A. et al. A chemical threshold controls nanocrystallization and degassing behaviour in basalt magmas. Commun. Earth Environ. 3, 284 (2022).
Zellmer, G. F. et al. Inferring the effects of compositional boundary layers on crystal nucleation, growth textures, and mineral chemistry in natural volcanic tephras through submicron-resolution imaging. Front. Earth Sci. 4, 1–7 (2016).
Di Genova, D. et al. A compositional tipping point governing the mobilization and eruption style of rhyolitic magma. Nature 552, 235–238 (2017).
Eichheimer, P. et al. Combined numerical and experimental study of microstructure and permeability in porous granular media. Solid Earth Discuss. 1–25 (2020) doi:10.5194/se-2019-199.
Knafelc, J. et al. Havre 2012 pink pumice is evidence of a short-lived, deep-sea, magnetite nanolite-driven explosive eruption. Commun. Earth Environ. 3, 1–11 (2022).
Barone, G. et al. Nanoscale surface modification of Mt. Etna volcanic ashes. Geochim. Cosmochim. Acta 174, 70–84 (2016).
Mujin, M. & Nakamura, M. A nanolite record of eruption style transition. Geology 42, 611–614 (2014).
Mujin, M., Nakamura, M. & Miyake, A. Eruption style and crystal size distributions: Crystallization of groundmass nanolites in the 2011 Shinmoedake eruption. Am. Mineral. 102, 2367–2380 (2017).
Mujin, M. & Nakamura, M. Late-stage groundmass differentiation as a record of magma stagnation, fragmentation, and rewelding. Bull. Volcanol. 82, (2020).
Di Genova, D., Caracciolo, A. & Kolzenburg, S. Measuring the degree of “nanotilization” of volcanic glasses: Understanding syn-eruptive processes recorded in melt inclusions. Lithos 318–319, 209–218 (2018).
Di Genova, D. et al. In situ observation of nanolite growth in volcanic melt: A driving force for explosive eruptions. Sci. Adv. 6, (2020).
Cáceres, F. et al. Can nanolites enhance eruption explosivity? Geology 48, 997–1001 (2020).
Cáceres, F. et al. Oxide nanolitisation-induced melt iron extraction causes viscosity jumps and enhanced explosivity in silicic magma. (2024) doi:10.1038/s41467-024-44850-x.
Giordano, D., Russell, J. K. & Dingwell, D. B. Viscosity of magmatic liquids: A model. Earth Planet. Sci. Lett. 271, 123–134 (2008).
Di Genova, D., Zandona, A. & Deubener, J. Unravelling the effect of nano-heterogeneity on the viscosity of silicate melts: Implications for glass manufacturing and volcanic eruptions. J. Non. Cryst. Solids 545, 120248 (2020).
Stabile, P. et al. The effect of iron and alkali on the nanocrystal-free viscosity of volcanic melts: A combined Raman spectroscopy and DSC study. Chem. Geol. 559, 119991 (2021).
Shebanova, O. N. & Lazor, P. Raman spectroscopic study of magnetite (FeFe2O4): a new assignment for the vibrational spectrum. J. Solid State Chem. 174, 424–430 (2003).
Jubb, A. M. & Allen, H. C. Vibrational Spectroscopic Characterization of Hematite, Maghemite, and Magnetite Thin Films Produced by Vapor Deposition. ACS Appl. Mater. Interfaces 2, 2804–2812 (2010).
Zandonà, A. et al. Crystallization Mechanism of Gel-Derived SiO2-TiO2 Amorphous Nanobeads Elucidated by High-Temperature In Situ Experiments. Cryst. Growth Des. 23, 4545–4555 (2023).
Zandonà, A. et al. Non-stoichiometric crystal nucleation in a spodumene glass containing TiO2 as seed former: Effects on the viscosity of the residual melt. J. Non. Cryst. Solids 619, 122563 (2023).
Langhammer, D., Di Genova, D. & Steinle-Neumann, G. Modeling the Viscosity of Anhydrous and Hydrous Volcanic Melts. Geochemistry, Geophys. Geosystems 22, 1–26 (2021).
Mauro, J. C., Yue, Y., Ellison, A. J., Gupta, P. K. & Allan, D. C. Viscosity of glass-forming liquids. Proc. Natl. Acad. Sci. U. S. A. 106, 19780–4 (2009).
Prescher, C., McCammon, C. & Dubrovinsky, L. MossA : a program for analyzing energy-domain Mössbauer spectra from conventional and synchrotron sources. J. Appl. Crystallogr. 45, 329–331 (2012).
Araya, N. et al. Shallow magma pre-charge during repeated Plinian eruptions at Sakurajima volcano. Sci. Rep. 9, 1–10 (2019).
Scherrer, M. C., Hess, K., Fehr, K. T. & Dingwell, D. B. Oxidation-state dependence of rheology in peralkaline glasses of phonolitic composition. Geophys. Res. Abstr. 14, 5805 (2012).
Dingwell, D. B. Redox viscometry of some Fe-bearing silicate melts. Am. Mineral. 76, 1560–1562 (1991).
Wechsler, B. A., Lindsley, D. H. & Prewitt, C. T. Crystal structure and cation distribution in titanomagnetites (Fe3-xTixO4). Am. Mineral. 69, 754–770 (1984).
Zinin, P. et al. Raman spectroscopy of titanomagnetites: Calibration of the intensity of Raman peaks as a sensitive indicator for their Ti content. Am. Mineral. 96, 1537–1546 (2011).
Kleebusch, E., Patzig, C., Höche, T. & Rüssel, C. The evidence of phase separation droplets in the crystallization process of a Li2O-Al2O3-SiO2 glass with TiO2 as nucleating agent – An X-ray diffraction and (S)TEM-study supported by EDX-analysis. Ceram. Int. 44, 2919–2926 (2018).
Fielitz, P., Helsch, G., Borchardt, G. & Deubener, J. Al-26 and O-18 tracer diffusion in a titania-coated sodium aluminosilicate glass. J. Non. Cryst. Solids 614, 122400 (2023).
Tuheen, M. I., Wilkins, M. C., McCloy, J. & Du, J. The structures of iron silicate glasses with varying iron redox ratios from molecular dynamics simulations and EXAFS analysis. J. Non. Cryst. Solids 624, 122713 (2024).
Kennedy, B. M., Jellinek, A. M., Russell, J. K., Nichols, A. R. L. & Vigouroux, N. Time-and temperature-dependent conduit wall porosity: A key control on degassing and explosivity at Tarawera volcano, New Zealand. Earth Planet. Sci. Lett. 299, 126–137 (2010).
Thivet, S. et al. Metastable liquid immiscibility in the 2018–2021 Fani Maoré lavas as a mechanism for volcanic nanolite formation. Commun. Earth Environ. 4, (2023).
Hornby, A. J. et al. Nanoscale silicate melt textures determine volcanic ash surface chemistry. 1–10 (2024) doi:10.1038/s41467-024-44712-6.
Dubosq, R. et al. Bubbles and atom clusters in rock melts: A chicken and egg problem. J. Volcanol. Geotherm. Res. 428, (2022).
Hughes, E. C. et al. Low analytical totals in EPMA of hydrous silicate glass due to sub-surface charging: Obtaining accurate volatiles by difference. Chem. Geol. 505, 48–56 (2019).
Behrens, H. et al. Structural relaxation mechanisms in hydrous sodium borosilicate glasses. J. Non. Cryst. Solids 497, 30–39 (2018).
Di Genova, D., Romano, C., Alletti, M., Misiti, V. & Scarlato, P. The effect of CO2 and H2O on Etna and Fondo Riccio (Phlegrean Fields) liquid viscosity, glass transition temperature and heat capacity. Chem. Geol. 377, 72–86 (2014).
Douglas, R. W., Armstrong, W. L., Edward, J. . & Hall, D. A penetration viscometer. Glas. Technol. 6, 52–55 (1965).
Meerlender, G. Viskositäts-Temperaturverhalten des Standardglases I der DGG. Glas. Ber. 47, 1–3 (1974).
Di Fiore, F. et al. Experimental insights on the shear-induced crystallization of a phonotephrite magma. Chem. Geol. 637, 121682 (2023).
Di Fiore, F., Vona, A., Costa, A., Mollo, S. & Romano, C. Quantifying the influence of cooling and shear rate on the disequilibrium rheology of a trachybasaltic melt from Mt. Etna. Earth Planet. Sci. Lett. 594, 117725 (2022).
Al-Mukadam, R., Di Genova, D., Bornhöft, H. & Deubener, J. High rate calorimetry derived viscosity of oxide melts prone to crystallization. J. Non. Cryst. Solids 536, 119992 (2020).
Yue, Y. The iso-structural viscosity, configurational entropy and fragility of oxide liquids. J. Non. Cryst. Solids 355, 737–744 (2009).
Zheng, Q., Mauro, J. C. & Yue, Y. Reconciling calorimetric and kinetic fragilities of glass-forming liquids. J. Non. Cryst. Solids 456, 95–100 (2017).
Whitfield, C. H., Brody, E. M. & Bassett, W. Elastic moduli of NaCl by Brillouin scattering at high pressure in a diamond anvil cell. Rev. Sci. Instrum. 47, 942–947 (1976).
Cassetta, M. et al. Estimating the viscosity of volcanic melts from the vibrational properties of their parental glasses. Sci. Rep. 11, 1–14 (2021).
Sinogeikin, S. et al. Brillouin spectrometer interfaced with synchrotron radiation for simultaneous x-ray density and acoustic velocity measurements. Rev. Sci. Instrum. 77, (2006).
Berg, S. et al. Ilastik: Interactive Machine Learning for (Bio)Image Analysis. Nat. Methods 16, 1226–1232 (2019).
Angell, C. A. Formation of Glasses from Liquids and Biopolymers. Science (80-. ). 267, 1924–1935 (1995).

There is NO Competing Interest.

ExtraImages.docx
Supplementary Images
Valdiviaetal.2024Supp.Inf.PV.docx
Supplementary Information
insituTEM.mov
In-situ high-temperature TEM
insituTEMofAND1001.mp4
Supplementary Video

Download PDF

Version 1

posted

You are reading this latest preprint version

Nanoscale chemical heterogeneities control magma viscosity and failure

Status:

Version 1

Abstract

Figures

1. Introduction

2. Results and discussion

2.1 Preliminary in-situ observation of nanolite formation during heating

2.2 The viscosity of homogeneous and crystal-free andesitic melts

2.3 The viscosity of nanolite-bearing andesitic melts

3. Conclusions

4. Materials and methods

4.1 Synthesis of starting glasses

4.2 Electron microprobe analyses (EMPA)

4.3 Micropenetration viscometry

4.4 Concentric cylinder (CC) viscometry

4.5 Differential scanning calorimetry

4.6 Brillouin Spectroscopy

4.7 (High-temperature) Raman spectroscopy

4.8 Mössbauer spectroscopy

4.9 (High-temperature) Transmission electron microscopy (TEM) analyses

4.10 Viscosity parametrization

Declarations

Acknowledgements

Author contributions

References

Additional Declarations

Supplementary Files

Status:

Version 1