Nanoscale chemical heterogeneities control magma viscosity and failure

Nanoscale chemical heterogeneities control magma viscosity and failure | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Nanoscale chemical heterogeneities control magma viscosity and failure Pedro Valdivia Munoz, Alessio Zandonà, Jessica Löschmann, Dmitry Bondar, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3891365/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 11 Jun, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Abstract 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 SiO 2 -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 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction 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 occurs 20 . 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 models 36 , 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. Results and discussion 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 (Fe 3+ /∑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 (Fe 3+ /∑Fe tot = 0.27) and AND100ox (Fe 3+ /∑Fe tot = 0.71) represent isochemical analogues of AND100 with lower and higher Fe 3+ /∑Fe tot ratios, respectively (Table 1). AND65 (Fe 3+ /∑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 , TiO 2 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. 1 a) 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. 1 b), 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. 1 b) 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. 2 b) 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 log 10 η ∞ 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 SiO 2 60.38 (0.36) 60.47 (0.40) 60.56 (0.35) 62.52 (0.19) 65.91 (0.34) TiO 2 0.79 (0.04) 0.80 (0.07) 0.81 (0.06) 0.56 (0.04) 0.04 (0.03) Al 2 O 3 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) Na 2 O 3.50 (0.15) 3.46 (0.10) 3.40 (0.10) 3.49 (0.11) 3.82 (0.14) K 2 O 1.58 (0.07) 1.65 (0.07) 1.66 (0.06) 1.70 (0.05) 1.75 (0.06) P 2 O5 0.18 (0.05) 0.18 (0.05) 0.18 (0.05) 0.16 (0.05) 0.04 (0.03) Fe 3+ /∑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) a Ratios derived using Mossa software 45 . b NBO/T calculated after Prata et al. (2019). c Values derived using Eq. 4 and Brillouin spectroscopy data. d Derived via DSC measurements (see Methods or SI). e Fitted fragility indices ( m ) using the Mauro–Yue–Elli- son–Gupta–Allan (MYEGA, Eq. 5) parametrization using η ∞ = -2.93. Our results (Fig. 2 a) 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 , TiO 2 and MnO) from AND100 would result in a viscosity increase of 0.7 ± 0.05 log 10 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 log 10 Pa s (⁓ 20 times), or a rise of ⁓1.5 ± 0.1 log 10 Pa s (⁓ 30 times) if starting from AND100red, within the same temperature range. Moreover, the only increase in iron oxidation, from Fe 3+ /∑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. 2 b; 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_MP 660 to a more distinct set of well-visible diffraction rings in AND100_MP 723 and especially AND100_MP 808 . Notably, AND100_MP 808 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 SiO 2 , 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 Al 2 O 3 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 TiO 2 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 Al 2 O 3 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 SiO 2 . This behaviour is consistent with well-established observations in glass-ceramic materials, exhibiting Al-enriched shells around TiO 2 - and/or ZrO 2 -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 TiO 2 -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. 2 b). For AND100_MP 660 , our initial viscosity measurement ( η = 10 12 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 Fe 3+ /∑Fe tot = 0.7). This viscosity is roughly twice greater than the viscosity parametrization of AND100 (Fe 3+ /∑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_MP 723 , the final viscosity value ( η = 10 11.9 Pa s; Fe 3+ /∑Fe tot = 0.79) is almost twice the pure melt parametrization of AND65 (Fe 3+ /∑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_MP 808 , the final viscosity value ( η = 10 10.17 Pa s; Fe 3+ /∑Fe tot = 0.83) slightly surpassed the melt parametrization of AND0 at 808 ℃ ( η = 10 10 ; 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 Fe 3+ /∑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_MP 723 and AND100_MP 880 , which are shown to have well-formed titanomagnetite nanolites, the remaining iron contained in the SiO 2 -enriched matrix may be predominantly in the Fe 3+ 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 log 10 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 SiO 2 -enriched regions (up to ⁓70 SiO 2 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., SiO 2 -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. 3. Conclusions 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 SiO 2 -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. Materials and methods 4.1 Synthesis of starting glasses AND100, AND65 and AND0 were synthesized by mixing powder reagents (SiO 2 , TiO 2 , Al 2 O 3 , Fe 2 O 3 , MnO, MgO, CaCO 3 , Na 2 CO 3 , K 2 CO 3 and P 2 O 5 ) 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 CO 2 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 Au 80 Pd 20 open capsule at 1275 ℃ and 1 atm for 24 hours, using a gas mixing furnace with a gas mixture of 95% CO 2 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 Pt 80 Rh 20 cylindric crucible (62 mm in height, and 32 mm inner diameter) and stirred at γ̇ = 10 s −1 using a Pt 80 Rh 20 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 Pt 80 Rh 20 crucible under a constant N 2 (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 ) ≈ 10 12 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) 57 Co 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 Fe 3+ /Fe 2+ 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, $$\begin{array}{c}m=43\cdot \frac{K}{G}-31\#\left(7\right)\end{array}$$ Declarations 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. References 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). Additional Declarations There is NO Competing Interest. Supplementary Files ExtraImages.docx Supplementary Images Valdiviaetal.2024Supp.Inf.PV.docx Supplementary Information insituTEM.mov In-situ high-temperature TEM insituTEMofAND1001.mp4 Supplementary Video Cite Share Download PDF Status: Published Journal Publication published 11 Jun, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3891365","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":271189031,"identity":"3449818a-85cf-493c-8f9a-1410d6672238","order_by":0,"name":"Pedro Valdivia Munoz","email":"data:image/png;base64,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","orcid":"","institution":"University of Bayreuth","correspondingAuthor":true,"prefix":"","firstName":"Pedro","middleName":"Valdivia","lastName":"Munoz","suffix":""},{"id":271189032,"identity":"1666d223-716f-4db6-98d3-189f38f999bb","order_by":1,"name":"Alessio Zandonà","email":"","orcid":"https://orcid.org/0000-0003-0091-9546","institution":"Clausthal University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Alessio","middleName":"","lastName":"Zandonà","suffix":""},{"id":271189033,"identity":"7ebdde3a-7c49-4415-833a-af11f4dba521","order_by":2,"name":"Jessica Löschmann","email":"","orcid":"","institution":"Clausthal University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Jessica","middleName":"","lastName":"Löschmann","suffix":""},{"id":271189034,"identity":"e49c2c5e-bdc0-4f61-89b1-c12e7167f73b","order_by":3,"name":"Dmitry Bondar","email":"","orcid":"","institution":"University of Bayreuth","correspondingAuthor":false,"prefix":"","firstName":"Dmitry","middleName":"","lastName":"Bondar","suffix":""},{"id":271189035,"identity":"e15fd2b7-c07c-4fc3-ba7e-665fdf34a4b2","order_by":4,"name":"Cécile Genevois","email":"","orcid":"","institution":"CNRS - CEMHTI laboratory","correspondingAuthor":false,"prefix":"","firstName":"Cécile","middleName":"","lastName":"Genevois","suffix":""},{"id":271189036,"identity":"943fa22a-0e72-45af-8039-4e931c39fdbd","order_by":5,"name":"Aurélien Canizarès","email":"","orcid":"","institution":"CEMHTI-CNRS","correspondingAuthor":false,"prefix":"","firstName":"Aurélien","middleName":"","lastName":"Canizarès","suffix":""},{"id":271189037,"identity":"6b7990b8-605f-4fae-91dd-d15a951a71ca","order_by":6,"name":"Mathieu Allix","email":"","orcid":"https://orcid.org/0000-0001-9317-1316","institution":"Université d'Orléan","correspondingAuthor":false,"prefix":"","firstName":"Mathieu","middleName":"","lastName":"Allix","suffix":""},{"id":271189038,"identity":"5b18af05-4800-480b-ab06-521f512593b4","order_by":7,"name":"Nobuyoshi Miyajima","email":"","orcid":"https://orcid.org/0000-0002-6226-5675","institution":"Bayerisches Geoinstitut, Universitaet Bayreuth,","correspondingAuthor":false,"prefix":"","firstName":"Nobuyoshi","middleName":"","lastName":"Miyajima","suffix":""},{"id":271189039,"identity":"b8ba0b9f-90e5-44d7-b54c-782e0a4e6fd4","order_by":8,"name":"Alexander Kurnosov","email":"","orcid":"","institution":"Bayerisches Geoinstitut, Universität Bayreuth","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Kurnosov","suffix":""},{"id":271189040,"identity":"179d673f-12c4-41e5-a6cf-53f313dcf3c9","order_by":9,"name":"Tiziana Boffa-Ballaran","email":"","orcid":"","institution":"Bayerisches Geoinstitut","correspondingAuthor":false,"prefix":"","firstName":"Tiziana","middleName":"","lastName":"Boffa-Ballaran","suffix":""},{"id":271189041,"identity":"6725691d-3bd2-411d-a71c-83dff8c1f257","order_by":10,"name":"Fabrizio Di Fiore","email":"","orcid":"https://orcid.org/0000-0002-6749-280X","institution":"Istituto Nazionale di Geofisica e Vulcanologia","correspondingAuthor":false,"prefix":"","firstName":"Fabrizio","middleName":"Di","lastName":"Fiore","suffix":""},{"id":271189042,"identity":"5cbe27da-8579-433d-8b50-2b6e652d1044","order_by":11,"name":"Alessandro Vona","email":"","orcid":"https://orcid.org/0000-0002-5483-5623","institution":"Università degli Studi Roma Tre","correspondingAuthor":false,"prefix":"","firstName":"Alessandro","middleName":"","lastName":"Vona","suffix":""},{"id":271189043,"identity":"2a37eef0-cd4f-4a2c-870e-026673e4c033","order_by":12,"name":"Claudia Romano","email":"","orcid":"","institution":"RomaTre University","correspondingAuthor":false,"prefix":"","firstName":"Claudia","middleName":"","lastName":"Romano","suffix":""},{"id":271189044,"identity":"8c38201b-eec3-420d-84e9-f6f3dbb6065c","order_by":13,"name":"Joachim Deubener","email":"","orcid":"https://orcid.org/0000-0002-3474-7490","institution":"Clausthal University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Joachim","middleName":"","lastName":"Deubener","suffix":""},{"id":271189045,"identity":"db31e614-48cd-42d3-ada8-b8613be2875d","order_by":14,"name":"Emily Bamber","email":"","orcid":"","institution":"University of Turin","correspondingAuthor":false,"prefix":"","firstName":"Emily","middleName":"","lastName":"Bamber","suffix":""},{"id":271189046,"identity":"137820a2-a095-43ba-8992-144167836834","order_by":15,"name":"Danilo Di Genova","email":"","orcid":"","institution":"National Research Council (CNR)","correspondingAuthor":false,"prefix":"","firstName":"Danilo","middleName":"Di","lastName":"Genova","suffix":""}],"badges":[],"createdAt":"2024-01-23 14:40:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3891365/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3891365/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43247-025-02424-9","type":"published","date":"2025-06-12T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":50750539,"identity":"cc3a3dd2-f115-4f15-85c0-a7dfa10aa8ba","added_by":"auto","created_at":"2024-02-06 17:30:48","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":462827,"visible":true,"origin":"","legend":"\u003cp\u003ea)\u003cem\u003e \u003c/em\u003e\u0026nbsp;Results of an \u003cem\u003ein-situ\u003c/em\u003e heating Raman spectroscopy experiment performed on the sample AND100 at 723 ℃, revealing the formation of titanomagnetite crystals during an isothermal hold slightly above \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg \u003c/em\u003e\u003c/sub\u003e(654 ℃; Table 1). Shaded areas highlight the position of the most intense Raman features of titanomagnetite. b) Images acquired during an \u003cem\u003ein-situ\u003c/em\u003e heating TEM experiment (videos of the experiment are available in the Supplementary Information section), detailing the nanostructural development of the sample AND100 from a homogeneous glass to a heterogeneous amorphous material (350 °C) and then a nanocrystallized glass (550 °C and 750 °C). FFTs (insets) and post-experiment EDS (Supp. Inf. Fig. S5) confirm the formed crystals to be magnetite-like. Temperatures (Supp. Inf. Fig. S6) are affected by electron irradiation and should be therefore considered only in qualitative terms.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/e317ffa3474048657721bb75.jpeg"},{"id":50750538,"identity":"e2e8f329-c6f4-4a29-8693-6c792bff5dbc","added_by":"auto","created_at":"2024-02-06 17:30:48","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":289220,"visible":true,"origin":"","legend":"\u003cp\u003eSummary of viscosity results. a) Coloured symbols represent homogeneous (HOM) nanolite-free viscosity data. Solid lines illustrate our MYEGA parametrization results for AND100, AND65 and AND0 (parameters are provided in Table 1). b) The impact of heterogeneities and iron oxidation state on the viscosity of AND100 compositions. White symbols represent nanolite-bearing viscosity data. The heterogeneous (HET) region, shaded in grey, corresponds to the expected viscosity increase attributed to nanolite formation. Red arrows represent the increase in measured viscosity during micropenetration experiments.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/8a1b29b800c76383d5e253c9.jpeg"},{"id":50751540,"identity":"ae190f45-a49b-4c70-a888-f43cdac22c97","added_by":"auto","created_at":"2024-02-06 17:38:48","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1090314,"visible":true,"origin":"","legend":"\u003cp\u003eSTEM-HAADF and EDS elemental map of sample AND100_MP\u003csub\u003e808\u003c/sub\u003e showing Fe, Ti, Si, Al, and Fe+Ti+Al distribution. STEM-HAADF and EDS elemental maps of sample AND100_MP\u003csub\u003e660 \u003c/sub\u003eand AND100_MP\u003csub\u003e723 \u003c/sub\u003ecan be found in the Supplementary Information (Fig. S11 and Fig. S12, respectively).\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/d01d3420318a10b6e072280a.png"},{"id":50751539,"identity":"901d9e7f-212d-46e2-be88-318fb3e10226","added_by":"auto","created_at":"2024-02-06 17:38:48","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":180492,"visible":true,"origin":"","legend":"\u003cp\u003eAverage SiO\u003csub\u003e2\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e, Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and FeO\u003csub\u003etot\u003c/sub\u003e contents of: a) the residual amorphous matrix, b) nanolites and the compositionally differentiated halo surrounding them, c) nanolites alone. Values are normalized to the bulk composition of the sample (see also Table S3). The data was obtained from the analysis of STEM-EDS mappings (Fig. 3 and Table S3) performed on samples treated at 660 °C, 723 °C and 808 °C during viscosity measurements. Error bars correspond to ±1σ considering 20-50 ROIs per sample.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/21d9af20a3ebbcd458fd634c.jpeg"},{"id":50750540,"identity":"5d46ed9a-0060-40cd-925b-32f601dbf430","added_by":"auto","created_at":"2024-02-06 17:30:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":903540,"visible":true,"origin":"","legend":"\u003cp\u003eViscosity for AND100 post-micropenetration (MP) samples as a function of temperature and FeO\u003csub\u003etot \u003c/sub\u003e+ TiO\u003csub\u003e2\u003c/sub\u003e content of their EDS matrix composition (Fig. 4; Supp. Inf. Table S4). Viscosity values are normalized to our parametrization for AND100 (Table 1; Figure 2). MP sub-index represent the experimental micropenetration temperature in ℃.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/82b1678f6452f4c6615072e9.png"},{"id":84560882,"identity":"dccec48b-bcbf-48d2-94b5-72dd6bea0365","added_by":"auto","created_at":"2025-06-13 13:06:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4058727,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/da07f7b0-a47b-4bad-9f59-9beb182a3533.pdf"},{"id":50750537,"identity":"de3aa97b-1bae-4619-a6c5-c26971fe04a7","added_by":"auto","created_at":"2024-02-06 17:30:48","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1421637,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Images\u003c/p\u003e","description":"","filename":"ExtraImages.docx","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/fa62b3dab15805dc4dce6e65.docx"},{"id":50750542,"identity":"df4b4ee6-2162-422c-8a2d-b52c86555954","added_by":"auto","created_at":"2024-02-06 17:30:48","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":4080189,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003c/p\u003e","description":"","filename":"Valdiviaetal.2024Supp.Inf.PV.docx","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/d33eb546a32999e0f77a9809.docx"},{"id":50750544,"identity":"f2ea4593-8fba-4edd-956d-3c4df50aea07","added_by":"auto","created_at":"2024-02-06 17:30:49","extension":"mov","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":47841729,"visible":true,"origin":"","legend":"\u003cp\u003eIn-situ high-temperature TEM\u003c/p\u003e","description":"","filename":"insituTEM.mov","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/ed4633f4280e122cda012a3b.mov"},{"id":50750546,"identity":"ff6fd9d7-1736-49b8-a4b5-97b95c7feaa4","added_by":"auto","created_at":"2024-02-06 17:30:50","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":47841535,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Video\u003c/p\u003e","description":"","filename":"insituTEMofAND1001.mp4","url":"https://assets-eu.researchsquare.com/files/rs-3891365/v1/76444b652204b3296f4da7b1.mp4"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Nanoscale chemical heterogeneities control magma viscosity and failure","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eExplosive eruptions are recognized as one of the most hazardous natural phenomena on Earth \u003csup\u003e1\u003c/sup\u003e, 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 \u003csup\u003e2\u0026ndash;5\u003c/sup\u003e. This phenomenon, known as magma fragmentation, can be triggered by bubble overpressure linked to their limited expansion\u003csup\u003e6,7\u003c/sup\u003e or induced by sufficiently high strain rates relative to the structural relaxation time, primarily dictated by the chemical composition of the melt \u003csup\u003e2,8,9\u003c/sup\u003e. Ocasionally, sub-Plinian and Plinian volcanic activities are fueled by andesite and basalt magmas \u003csup\u003e10\u0026ndash;13\u003c/sup\u003e. Nevertheless, these kind of magmas face challenges in attaining the requisite for magma fragmentation due to their relatively low viscosity \u003csup\u003e14\u0026ndash;19\u003c/sup\u003e, unless a substantial volume of microlite crystallization occurs\u003csup\u003e20\u003c/sup\u003e. 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\u003csup\u003e21\u0026ndash;23\u003c/sup\u003e. Interestingly, these nanolites can be present in natural volcanic products erupted during explosive events \u003csup\u003e24\u0026ndash;28\u003c/sup\u003e. 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 \u003csup\u003e21,22,27,29\u0026ndash;35\u003c/sup\u003e. In this study, we present the first \u003cem\u003ein-situ\u003c/em\u003e 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 models\u003csup\u003e36\u003c/sup\u003e, to overestimate the nanolite-free melt viscosity of compositions prone to nanocrystallization\u003csup\u003e22,23,25,37\u003c/sup\u003e. 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.\u003c/p\u003e"},{"header":"2. Results and discussion","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003e2.1 Preliminary in-situ observation of nanolite formation during heating\u003c/h2\u003e\n\u003cp\u003eWe produced four anhydrous andesitic melts and one transition-metal-free analog (Table\u0026nbsp;1). The sample AND100 (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.64) was designed to mirror the andesitic chemical composition of the magma erupted at Sakurajima volcano (Okumura et al. 2022). Samples AND100red (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.27) and AND100ox (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.71) represent isochemical analogues of AND100 with lower and higher Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e ratios, respectively (Table\u0026nbsp;1). AND65 (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.70) and AND0 were produced based on the composition of AND100, from which 35% and 100% of the total transition metal content (FeO\u003csub\u003etot\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e 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 \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eInitially, we conducted a comprehensive exploration of the thermal response of AND100 composition from a fundamental standpoint, performing \u003cem\u003ein-situ\u003c/em\u003e high-temperature measurements. High-temperature Raman spectra (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ea) reveal that AND100 is unstable against thermal treatments, leading to the formation of titanomagnetite nano-sized crystals already ⁓70 ℃ above the glass transition temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e), which was determined by differential scanning calorimetry (Table\u0026nbsp;1; Supp Inf. Section S2.1). This is evidenced by the appearance of vibrational features assignable to Fe-Ti-oxides such as titanomagnetite \u003csup\u003e37\u0026ndash;40\u003c/sup\u003e, 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 \u003cem\u003ein-situ\u003c/em\u003e 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 Zandon\u0026agrave; 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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb), 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 (\u0026gt;\u0026thinsp;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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb) 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 \u003cem\u003ein-situ\u003c/em\u003e 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 \u003csup\u003e23\u003c/sup\u003e. They therefore provide an accurate overview of processes that should be expected to occur in deeply supercooled melts and reheated glasses.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n\u003ch2\u003e2.2 The viscosity of homogeneous and crystal-free andesitic melts\u003c/h2\u003e\n\u003cp\u003eThe \u003cem\u003ein-situ\u003c/em\u003e high-temperature measurements revealed the high reactivity of andesitic glasses and melts at a temperature above \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e, emphasizing the need to apply a meticulous experimental methodology in deriving accurate pure melt viscosity \u003csup\u003e22,23,42\u003c/sup\u003e. 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 M\u0026ouml;ssbauer spectroscopy before and after experiments (Supp. Inf. Sections S2.1, S2.2, S3 and S5).\u003c/p\u003e\n\u003cp\u003eThe 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 M\u0026ouml;ssbauer and Raman results (Supp. Inf. Sections S2.1, S2.2, S3 and Table \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003e), as well as micropenetration (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb) and post-micropenetration TEM results (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e; 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.\u0026nbsp;5), with log\u003csub\u003e10\u003c/sub\u003e\u003cem\u003e\u0026eta;\u003c/em\u003e\u003csub\u003e\u003cem\u003e\u0026infin;\u003c/em\u003e\u003c/sub\u003e fixed at -2.93 \u003csup\u003e43,44\u003c/sup\u003e to fit the melt fragility index (\u003cem\u003em\u003c/em\u003e). Together with our C-DSC-derived \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e (Table\u0026nbsp;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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e, and the fitting parameters of the corresponding MYEGA parametrizations are summarized in Table\u0026nbsp;1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Electron microprobe chemical compositions, Brillouin spectroscopy results and MYEGA (Eq. 5) fit parameters.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Taba\" border=\"1\"\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eAND100\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eAND100red\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eAND100ox\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eAND65\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eAND0\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSiO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e60.38 (0.36)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e60.47 (0.40)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e60.56 (0.35)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e62.52 (0.19)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e65.91 (0.34)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eTiO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.79 (0.04)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.80 (0.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.81 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.56 (0.04)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.04 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAl\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e16.69 (0.18)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e16.83 (0.22)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e16.79 (0.19)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e17.28 (0.22)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e18.01 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFeO\u003csub\u003etot\u003c/sub\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.77 (0.12)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.87 (0.12)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.76 (0.15)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4.40 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.04 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMnO\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.17 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.18 (0.04)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.18 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.11 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.01 (0.01)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMgO\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.00 (0.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.94 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2.95 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.02 (0.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.21 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eCaO\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.62 (0.09)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.51 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.49 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.76 (0.13)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.26 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eNa\u003csub\u003e2\u003c/sub\u003eO\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.50 (0.15)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.46 (0.10)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.40 (0.10)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.49 (0.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.82 (0.14)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eK\u003csub\u003e2\u003c/sub\u003eO\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.58 (0.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.65 (0.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.66 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.70 (0.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.75 (0.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eP\u003csub\u003e2\u003c/sub\u003eO5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.18 (0.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.18 (0.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.18 (0.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.16 (0.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.04 (0.03)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.64\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.71\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.70\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003eNBO/T\u003c/em\u003e\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.34\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.34\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.34\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.16\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003eK/G\u003c/em\u003e\u003csup\u003ec\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.57 (0.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.57 (0.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.58 (0.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.53 (0.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.48 (0.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e\u003csup\u003e\u003cem\u003ed\u003c/em\u003e\u003c/sup\u003e (℃)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e654 (1)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e645 (1)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e662 (1)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e696 (1)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e737 (1)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003em\u003c/em\u003e\u003csup\u003e\u003cem\u003ee\u003c/em\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e30.5 (0.5)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e30.5 (0.5)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e30.5 (0.5)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e31 (0.4)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e31.8 (0.2)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"6\" align=\"left\"\u003e\n\u003cp\u003e\u003csup\u003ea\u003c/sup\u003eRatios derived using Mossa software \u003csup\u003e45\u003c/sup\u003e.\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"6\" align=\"left\"\u003e\n\u003cp\u003e\u003csup\u003eb\u003c/sup\u003e\u003cem\u003eNBO/T\u003c/em\u003e calculated after Prata et al. (2019).\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"6\" align=\"left\"\u003e\n\u003cp\u003e\u003csup\u003ec\u003c/sup\u003eValues derived using Eq.\u0026nbsp;4 and Brillouin spectroscopy data.\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"6\" align=\"left\"\u003e\n\u003cp\u003e\u003csup\u003ed\u003c/sup\u003eDerived via DSC measurements (see Methods or SI).\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"6\" align=\"left\"\u003e\n\u003cp\u003e\u003csup\u003ee\u003c/sup\u003eFitted fragility indices (\u003cem\u003em\u003c/em\u003e) using the Mauro\u0026ndash;Yue\u0026ndash;Elli- son\u0026ndash;Gupta\u0026ndash;Allan (MYEGA, Eq.\u0026nbsp;5) parametrization using \u003cem\u003e\u0026eta;\u003c/em\u003e\u003csub\u003e\u003cem\u003e\u0026infin;\u003c/em\u003e\u003c/sub\u003e = -2.93.\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eOur results (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea) 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\u003csub\u003etot\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e and MnO) from AND100 would result in a viscosity increase of 0.7\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05 log\u003csub\u003e10\u003c/sub\u003e Pa s (⁓ 5 times) at the eruptive temperature for Sakurajima volcano (900\u0026ndash;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\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1 log\u003csub\u003e10\u003c/sub\u003e Pa s (⁓ 20 times), or a rise of ⁓1.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1 log\u003csub\u003e10\u003c/sub\u003e Pa s (⁓ 30 times) if starting from AND100red, within the same temperature range. Moreover, the only increase in iron oxidation, from Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 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 \u003csup\u003e19,47,48\u003c/sup\u003e. 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\u0026deg;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.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n\u003ch2\u003e2.3 The viscosity of nanolite-bearing andesitic melts\u003c/h2\u003e\n\u003cp\u003eWhile conducting micropenetration measurements on AND100 samples, we observed a time-dependent increase in viscosity at constant temperatures (red arrows in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb; 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.\u003c/p\u003e\n\u003cp\u003eElectron 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 \u003cem\u003ed\u003c/em\u003e-spacings can be associated with the structure of titanomagnetite \u003csup\u003e49,50\u003c/sup\u003e. Specifically, the diffraction patterns evolved from diffused circular halos in the case of AND100_MP\u003csub\u003e660\u003c/sub\u003e to a more distinct set of well-visible diffraction rings in AND100_MP\u003csub\u003e723\u003c/sub\u003e and especially AND100_MP\u003csub\u003e808\u003c/sub\u003e. Notably, AND100_MP\u003csub\u003e808\u003c/sub\u003e 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 \u003cem\u003ein-situ\u003c/em\u003e experiments (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e; Supp. Inf. Fig. S4), suggesting that nucleation and growth of nanolite become more pronounced as temperatures increase further above \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e. 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\u0026thinsp;\u0026plusmn;\u0026thinsp;0.5 to 2.6\u0026thinsp;\u0026plusmn;\u0026thinsp;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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003eWe observe that the Fe-rich regions can be identified with the nanocrystals appearing bright in the STEM-HAADF images (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e, 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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e (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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). 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\u0026nbsp;1 and Supp. Inf. Table S4), evidencing that only a negligible migration of alkalis occurred during EDS measurements. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e 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 SiO\u003csub\u003e2\u003c/sub\u003e, with average values of approximately 70 wt% (Supp. Inf. Table S4), which is ca. 10 wt% higher than the original AND100 composition (Table\u0026nbsp;1). Conversely, a gradual depletion in Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e and FeO\u003csub\u003etot\u003c/sub\u003e 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\u003csub\u003etot\u003c/sub\u003e and TiO\u003csub\u003e2\u003c/sub\u003e in nanolite regions, aligning with the nucleation and growth of titanomagnetite nanolites observed in our high-temperature \u003cem\u003ein-situ\u003c/em\u003e experiments (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Notably, we observed that nanolites exhibited an initial enrichment in Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e 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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e and observed after high-temperature \u003cem\u003ein-situ\u003c/em\u003e TEM experiments (Supp. Inf. Fig. S5), whereas the surrounding matrix is enriched in SiO\u003csub\u003e2\u003c/sub\u003e. This behaviour is consistent with well-established observations in glass-ceramic materials, exhibiting Al-enriched shells around TiO\u003csub\u003e2\u003c/sub\u003e- and/or ZrO\u003csub\u003e2\u003c/sub\u003e-bearing nanocrystals acting (upon further heating) as seeds for the controlled crystallization of the surrounding aluminosilicate matrix \u003csup\u003e51\u003c/sup\u003e. Recent tracer diffusivity data obtained from a TiO\u003csub\u003e2\u003c/sub\u003e-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 \u003csup\u003e52\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTo 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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb). For AND100_MP\u003csub\u003e660\u003c/sub\u003e, our initial viscosity measurement (\u003cem\u003e\u0026eta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10\u003csup\u003e12\u003c/sup\u003e 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 \u003cem\u003e\u0026eta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10\u003csup\u003e12.7\u003c/sup\u003e Pa s (the overall iron oxidation state slightly increased during the measurement to Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.7). This viscosity is roughly twice greater than the viscosity parametrization of AND100 (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.64) at 660 ℃. However, our STEM and EDS results (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e; Table S4) show that, on average, less than 3% of the FeO\u003csub\u003etot\u003c/sub\u003e precipitated in the form of nanocrystals in this sample. As such, neither the presence of solid crystalline particles (\u0026lt;\u0026thinsp;1%; Supp. Inf. Table S4), as suggested by classic particle suspension models \u003csup\u003e20\u003c/sup\u003e, nor the overall compositional variations (e.g., Fe-Ti removal; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e) of the surrounding amorphous phase are sufficient to explain such an increase in viscosity. We argue that the heterogeneous distribution of chemical species (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e) is responsible for the observed surplus in measured viscosity as illustrated in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e. Similar conclusions can be drawn from the viscosity measurements performed at higher temperatures. For AND100_MP\u003csub\u003e723\u003c/sub\u003e, the final viscosity value (\u003cem\u003e\u0026eta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10\u003csup\u003e11.9\u003c/sup\u003e Pa s; Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.79) is almost twice the pure melt parametrization of AND65 (Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.70), although the residual matrix of this sample still contains approximately 40% of the initial FeO\u003csub\u003etot\u003c/sub\u003e content (Supp. Inf. Table S4). For AND100_MP\u003csub\u003e808\u003c/sub\u003e, the final viscosity value (\u003cem\u003e\u0026eta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10\u003csup\u003e10.17\u003c/sup\u003e Pa s; Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e = 0.83) slightly surpassed the melt parametrization of AND0 at 808 ℃ (\u003cem\u003e\u0026eta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;10\u003csup\u003e10\u003c/sup\u003e; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e), although the residual amorphous phase was far from being completely free of transition metal oxides: approximately 25% of the initial FeO\u003csub\u003etot\u003c/sub\u003e content (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e; 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 (~\u0026thinsp;1 vol%) is too negligible to play any relevant rheological role \u003csup\u003e20\u003c/sup\u003e, 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 \u003csup\u003e21,35\u003c/sup\u003e. Moreover, our post-micropenetration M\u0026ouml;ssbauer spectroscopy results (Supp. Inf. Section S3; Table \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003e) indicate that the Fe\u003csup\u003e3+\u003c/sup\u003e/\u0026sum;Fe\u003csub\u003etot\u003c/sub\u003e 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_MP\u003csub\u003e723\u003c/sub\u003e and AND100_MP\u003csub\u003e880\u003c/sub\u003e, which are shown to have well-formed titanomagnetite nanolites, the remaining iron contained in the SiO\u003csub\u003e2\u003c/sub\u003e-enriched matrix may be predominantly in the Fe\u003csup\u003e3+\u003c/sup\u003e state. In this scenario, the remaining iron may acts preferentially as a network former \u003csup\u003e53\u003c/sup\u003e, further increasing the nanoscale local viscosity of the remaining matrix.\u003c/p\u003e\n\u003cp\u003eAs such, a notable surplus in viscosity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e; between 0.2 and 0.8 log\u003csub\u003e10\u003c/sub\u003e units) emerges across all three cases with respect to values expected for homogeneous melts. Considering the very low crystal content of the melts (\u0026lt;\u0026thinsp;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.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e; 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 SiO\u003csub\u003e2\u003c/sub\u003e-enriched regions (up to ⁓70 SiO\u003csub\u003e2\u003c/sub\u003e wt%; Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e and Supp Inf. Table S4), thereby significantly increasing the overall viscosity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). 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.\u003c/p\u003e\n\u003cp\u003eOur 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 \u003csup\u003e54\u003c/sup\u003e. Additionally, the presence of nanoscale liquid immiscibility (phase separation) under eruptive conditions has already been observed in lavas from the 2018\u0026ndash;2021 Fani Maor\u0026eacute; eruption \u003csup\u003e55\u003c/sup\u003e. 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 \u003csup\u003e23,33,37\u003c/sup\u003e. 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 \u003csup\u003e28,33\u003c/sup\u003e, indicating that Fe-rich nanolites could serve as proto-fragmentation surface for future ash particles. Indeed, recent investigations \u003csup\u003e56\u003c/sup\u003e 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., SiO\u003csub\u003e2\u003c/sub\u003e-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 \u003csup\u003e57\u003c/sup\u003e, 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 \u003csup\u003e22,33\u003c/sup\u003e 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.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"3. Conclusions","content":"\u003cp\u003eWe present the first \u003cem\u003ein-situ\u003c/em\u003e 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 SiO\u003csub\u003e2\u003c/sub\u003e-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.\u003c/p\u003e"},{"header":"4. Materials and methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Synthesis of starting glasses\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAND100, AND65 and AND0 were synthesized by mixing powder reagents (SiO\u003csub\u003e2\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e, Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, Fe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, MnO, MgO, CaCO\u003csub\u003e3\u003c/sub\u003e, Na\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e, K\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e and P\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e5\u003c/sub\u003e) 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 ~\u0026thinsp;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\u0026deg;C to eliminate CO\u003csub\u003e2\u003c/sub\u003e from the carbonate compounds. Following decarbonization, the material was transferred to a Pt crucible and melted for 24 hours at 1400\u0026deg;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\u0026deg;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 Au\u003csub\u003e80\u003c/sub\u003ePd\u003csub\u003e20\u003c/sub\u003e open capsule at 1275 ℃ and 1 atm for 24 hours, using a gas mixing furnace with a gas mixture of 95% CO\u003csub\u003e2\u003c/sub\u003e and 5% CO. The resulting melt was rapidly quenched in water by melting the Pt wire.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Electron microprobe analyses (EMPA)\u003c/h2\u003e \u003cp\u003eThe major elemental composition (Si, Ti, Al, Fe\u003csub\u003etot\u003c/sub\u003e., 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\u0026nbsp;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 \u0026micro;m 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 \u003csup\u003e58\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Micropenetration viscometry\u003c/h2\u003e \u003cp\u003eWe conducted micropenetration (MP) viscometry measurements on plane-parallel and polished glass chips of 2\u0026ndash;3 mm in thickness. These measurements were carried out utilizing a vertical dilatometer (B\u0026auml;hr VIS 404) at the Institute of Non-Metallic Materials, TU Clausthal (Germany). We measured the indentation rate of a sapphire sphere (r\u0026thinsp;=\u0026thinsp;0.75 mm) during isothermal dwells at temperatures controlled using an S-type thermocouple (Pt-PtRh) placed at ~\u0026thinsp;1.5 mm from the sample surface. The temperature error is estimated to be \u0026plusmn;\u0026thinsp;5\u0026deg;C considering the accuracy of the S-type thermocouple and its distance from the sample \u003csup\u003e59\u003c/sup\u003e. We followed standard procedures \u003csup\u003e22,23,37,60\u003c/sup\u003e 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.\u0026nbsp;1 \u003csup\u003e61\u003c/sup\u003e:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}\\eta =\\frac{9F}{32 \\sqrt{2r} \\sqrt{{L}^{3}}}t\\#\\left(1\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eη\u003c/em\u003e is the Newtonian viscosity (Pa s), \u003cem\u003eF\u003c/em\u003e is the applied force (N), \u003cem\u003et\u003c/em\u003e is the time (s), \u003cem\u003er\u003c/em\u003e is the radius of the sphere (m) and \u003cem\u003eL\u003c/em\u003e is the indentation depth (m). The dilatometer was previously calibrated using a standard glass DGG-1, reproducing the certified viscosity data \u003csup\u003e62\u003c/sup\u003e with a deviation of \u0026plusmn;\u0026thinsp;0.1 in log units. Data points are reported in the text according to the scheme \u003cem\u003eSampleName\u003c/em\u003e_MP\u003csub\u003e\u003cem\u003eTemperature\u003c/em\u003e\u003c/sub\u003e, with temperature representing the final experimental temperature expressed in \u0026deg;C, and the duration in minutes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Concentric cylinder (CC) viscometry\u003c/h2\u003e \u003cp\u003eHigh-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 \u003csup\u003e63\u003c/sup\u003e. Temperature monitoring was carried out using a factory-calibrated S-type thermocouple, with a precision of \u0026plusmn;\u0026thinsp;2\u0026deg;C. The concentric cylinder was previously calibrated using a standard glass NIST 717a, reproducing the certified viscosity data with a deviation of \u0026plusmn;\u0026thinsp;0.03 in log units \u003csup\u003e64\u003c/sup\u003e. To ensure thorough thermo-chemical homogenization, the glass materials were loaded into Pt\u003csub\u003e80\u003c/sub\u003eRh\u003csub\u003e20\u003c/sub\u003e cylindric crucible (62 mm in height, and 32 mm inner diameter) and stirred at γ̇ = 10 s\u003csup\u003e\u0026minus;1\u003c/sup\u003e using a Pt\u003csub\u003e80\u003c/sub\u003eRh\u003csub\u003e20\u003c/sub\u003e spindle (3.2 and 42 mm in diameter and length, respectively) at 1435\u0026deg;C at air oxygen fugacity and ambient pressure for 5 hours. Subsequently, the temperature was lowered by steps of 25\u0026ndash;50\u0026deg;C down to 1150 and 1130\u0026deg;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 (~\u0026thinsp;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.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Differential scanning calorimetry\u003c/h2\u003e \u003cp\u003eConventional differential scanning calorimetry (C-DSC) measurements were performed at the Institute of Non-Metallic Materials, TU Clausthal (Germany). Around 15 mg (\u0026plusmn;\u0026thinsp;5) of glass was placed in a Pt\u003csub\u003e80\u003c/sub\u003eRh\u003csub\u003e20\u003c/sub\u003e crucible under a constant N\u003csub\u003e2\u003c/sub\u003e (5.0) flow rate of 20 ml min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. 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 (\u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e) of 10 and 20 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Additionally, we used\u0026thinsp;~\u0026thinsp;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\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003eThe 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\u0026deg;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 \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e = 20 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e until it reached a temperature slightly above \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003epeak\u003c/em\u003e\u003c/sub\u003e, namely \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003emax\u003c/em\u003e\u003c/sub\u003e. Subsequently, we cooled the melt to 100\u0026deg;C at rates of \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003ec\u003c/em\u003e\u003c/sub\u003e = 10 or 20 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. 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., \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e = \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003ec\u003c/em\u003e\u003c/sub\u003e). From the measured heat flow during the matching upscan, we extracted the characteristic temperatures \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003epeak\u003c/em\u003e\u003c/sub\u003e. For further details see the methodology presented in Valdivia et al. (2023).\u003c/p\u003e \u003cp\u003eFor F-DSC experiments, we followed the methodology described above employing a \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e = \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003ec\u003c/em\u003e\u003c/sub\u003e = 1,000 ℃ s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (60,000 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Subsequently, we conducted a series of measurements on the same sample, using the same chip, at 10,000 ℃ s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (600,000 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) to investigate the impact of nanocrystallization on the characteristic temperatures \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003epeak\u003c/em\u003e\u003c/sub\u003e due to consecutive thermal treatments.\u003c/p\u003e \u003cp\u003eFollowing the theoretical background discussed elsewhere \u003csup\u003e37,65\u0026ndash;67\u003c/sup\u003e, viscosity values were derived from C- and F-DSC data using the relationship between the matching heating rate (\u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e) of the measurement and the shift factors \u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e \u003cem\u003eand K\u003c/em\u003e\u003csub\u003e\u003cem\u003epeak\u003c/em\u003e\u003c/sub\u003e \u003csup\u003e37,38\u003c/sup\u003e expressed in Eq.\u0026nbsp;2:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}{log}_{10}\\eta \\left({T}_{onset,peak}\\right)={K}_{onset,peak}-{log}_{10}\\left({q}_{h}\\right)\\#\\left(2\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e = 11.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.15 and \u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003epeak\u003c/em\u003e\u003c/sub\u003e = 9.84\u0026thinsp;\u0026plusmn;\u0026thinsp;0.20 \u003csup\u003e37,38\u003c/sup\u003e. It is important to mention that when \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh\u003c/em\u003e\u003c/sub\u003e is 10 ℃ min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, \u003cem\u003eη\u003c/em\u003e(\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e)\u0026thinsp;\u0026asymp;\u0026thinsp;10\u003csup\u003e12\u003c/sup\u003e Pa s, and therefore, \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset \u0026asymp;\u003c/em\u003e\u003c/sub\u003e \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg.\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.6 Brillouin Spectroscopy\u003c/h2\u003e \u003cp\u003eBrillouin spectroscopy (BLS) measurements were performed at the Bayerisches Geoinstitut (University of Bayreuth, Germany). Plane-parallel glass plates with a thickness of ~\u0026thinsp;50 \u0026micro;m 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\u0026ndash;Perot interferometer \u003csup\u003e68\u003c/sup\u003e coupled with a single-pixel photon counter detector. Measurements were conducted using a symmetric forward scattering configuration \u003csup\u003e68,69\u003c/sup\u003e with a scattering angle of θ\u0026thinsp;=\u0026thinsp;79.8\u0026deg;. The accuracy of the scattering angle was established through calibration with a reference silica glass. Conversion of frequency shifts (\u003cem\u003eΔω\u003c/em\u003e) to longitudinal (\u003cem\u003ev\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e) and shear (\u003cem\u003ev\u003c/em\u003e\u003csub\u003e\u003cem\u003es\u003c/em\u003e\u003c/sub\u003e) sound velocities was carried out using Eq.\u0026nbsp;3:\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}v=\\frac{\\varDelta \\omega \\lambda }{2sin\\left(\\theta /2\\right)}\\#\\left(3\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eλ\u003c/em\u003e is the laser wavelength and \u003cem\u003eθ\u003c/em\u003e is the angle between the incident and scattered beams \u003csup\u003e68,70\u003c/sup\u003e. We collected 8 spectra for each sample at different rotation angles (from \u0026minus;\u0026thinsp;180\u0026deg; to +\u0026thinsp;180\u0026deg;) to factor for uncertainties. Finally, we calculated the \u003cem\u003eK\u003c/em\u003e/\u003cem\u003eG\u003c/em\u003e factor using the ratio between \u003cem\u003ev\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003ev\u003c/em\u003e\u003csub\u003e\u003cem\u003es\u003c/em\u003e\u003c/sub\u003e (Eq.\u0026nbsp;4):\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}\\frac{K}{G}={\\left(\\frac{{v}_{p}}{{v}_{s}}\\right)}^{2}-\\frac{4}{3}\\#\\left(4\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.7 (High-temperature) Raman spectroscopy\u003c/h2\u003e \u003cp\u003eGlasses 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\u0026ndash;1300 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. 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.\u003c/p\u003e \u003cp\u003eAdditionally, we performed \u003cem\u003ein-situ\u003c/em\u003e 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\u0026ndash;2000 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Acquisition parameters included an integration time of 60 to 120 seconds, and a laser power of 20 mW.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.8 M\u0026ouml;ssbauer spectroscopy\u003c/h2\u003e \u003cp\u003eM\u0026ouml;ssbauer measurements were performed at the Bayerisches Geoinstitut (University of Bayreuth, Germany). Glass samples of ~\u0026thinsp;4 mm diameter and ~\u0026thinsp;600 \u0026micro;m thickness were measured before and after experiments at room temperature (293 K) using a constant acceleration M\u0026ouml;ssbauer spectrometer equipped with a high specific activity (370 MBq) \u003csup\u003e57\u003c/sup\u003eCo point source within a 12 \u0026micro;m thick Rh matrix. Calibration of the velocity scale was relative to a 25 \u0026micro;m thick α-Fe foil, and data were gathered within the range of \u0026plusmn;\u0026thinsp;5 mm s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, with acquisition durations of 2 to 3 days each. The recorded spectra were fitted with the full transmission integral using the MossA software \u003csup\u003e45\u003c/sup\u003e. Finally, the resulting Fe\u003csup\u003e3+\u003c/sup\u003e/Fe\u003csup\u003e2+\u003c/sup\u003e ratios were calculated using the relative area associated to each iron species.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.9 (High-temperature) Transmission electron microscopy (TEM) analyses\u003c/h2\u003e \u003cp\u003eWe performed TEM analyses on AND100 samples post-micropenetration experiments and \u003cem\u003ein-situ\u003c/em\u003e 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 \u003csup\u003e71\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003cem\u003eIn-situ\u003c/em\u003e heating experiments were performed at the CNRS CEMHTI in Orl\u0026eacute;ans (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 \u003csup\u003e41\u003c/sup\u003e 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\u0026deg;C for 1 h to remove possible volatile contaminants. The subsequent \u003cem\u003ein-situ\u003c/em\u003e experiment involved manual heating at 1 ℃ s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to various temperatures between 200\u0026deg;C and 750\u0026deg;C, where isothermal dwells of 30\u0026ndash;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).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.10 Viscosity parametrization\u003c/h2\u003e \u003cp\u003eThe 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 \u003cem\u003eη\u003c/em\u003e(\u003cem\u003eT\u003c/em\u003e), via the Mauro\u0026ndash;Yue\u0026ndash;Ellison\u0026ndash;Gupta\u0026ndash;Allan (MYEGA) equation (Eq.\u0026nbsp;5) \u003csup\u003e44\u003c/sup\u003e:\u003cdiv id=\"Eque\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Eque\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}{log}_{10}\\eta \\left(T\\right)={log}_{10}{\\eta }_{\\infty }+\\left(12-{log}_{10}{\\eta }_{\\infty }\\right)\\frac{{T}_{g}}{T}exp\\left[\\left(\\frac{m}{12-{log}_{10}{\\eta }_{\\infty }}-1\\right)\\left(\\frac{{T}_{g}}{T}-1\\right)\\right]\\#\\left(5\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({log}_{10}{\\eta }_{\\infty }=-2.93\\pm 0.3\\)\u003c/span\u003e\u003c/span\u003e is the logarithmic viscosity at infinite temperature \u003csup\u003e43,44\u003c/sup\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{g}\\)\u003c/span\u003e\u003c/span\u003e is the glass transition temperature determined by C-DSC (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eonset\u003c/em\u003e\u003c/sub\u003e at \u003cem\u003eq\u003c/em\u003e\u003csub\u003e\u003cem\u003eh,c\u003c/em\u003e\u003c/sub\u003e = 10 ℃ min\u003csup\u003e\u0026minus;1\u003c/sup\u003e) and \u003cem\u003em\u003c/em\u003e is the melt fragility defined in Eq.\u0026nbsp;6 \u003csup\u003e72\u003c/sup\u003e as the slope of viscosity curve evaluated at \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e:\u003cdiv id=\"Equf\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equf\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}m={\\left.\\frac{\\partial {log}_{10}\\eta }{\\partial {T}_{g}/T}\\right|}_{T={T}_{g}}\\#\\left(6\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe melt fragility parameter, \u003cem\u003em\u003c/em\u003e, can be determined by fitting Eq.\u0026nbsp;5 to our viscosity datasets and the measured \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{g}\\)\u003c/span\u003e\u003c/span\u003e via C-DSC. Additionally, \u003cem\u003em\u003c/em\u003e also can be inferred from BLS measurements using the empirical relationship introduced by Cassetta et al. (2021) (Eq.\u0026nbsp;7) and recently employed by Di Genova et al. (2023) for the viscosity of peridotitic melts,\u003cdiv id=\"Equg\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equg\" name=\"EquationSource\"\u003e\n$$\\begin{array}{c}m=43\\cdot \\frac{K}{G}-31\\#\\left(7\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003ePV and DDG acknowledge the funding by Deutsche Forschungsgemeinschaft (DFG) project DI 2751/2\u0026ndash;1. DDG acknowledges the funding from the European Research Council (ERC) under the European Union\u0026rsquo;s Horizon 2020 research and innovation programme (NANOVOLC, ERC Consolidator Grant \u0026ndash; 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 M\u0026ouml;ssbauer facilities at the Bayerisches Geoinstitut.\u003c/p\u003e\n\u003cp\u003eAll authors declare that they have no conflicts of interest.\u003c/p\u003e\n\u003ch2\u003eAuthor contributions\u003c/h2\u003e\n\u003cp\u003eP.V. drafted the original manuscript, synthesised the starting materials, processed the experimental data, performed the analyses, derived the viscosity models, performed M\u0026ouml;ssbauer 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.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eLoughlin, S. C., Sparks, S., Brown, S. K., Jenkins, S. F. \u0026amp; Vye-Brown, C. \u003cem\u003eGlobal volcanic hazards and risk\u003c/em\u003e. \u003cem\u003eGlobal Volcanic Hazards and Risk\u003c/em\u003e (Cambridge University Press, 2015). doi:10.1017/CBO9781316276273.\u003c/li\u003e\n\u003cli\u003ePapale, P. Strain-induced magma fragmentation in explosive eruptions. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e397\u003c/strong\u003e, 425\u0026ndash;428 (1999).\u003c/li\u003e\n\u003cli\u003eGonnermann, H. M. Magma Fragmentation. \u003cem\u003eAnnu. Rev. Earth Planet. Sci.\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 431\u0026ndash;458 (2015).\u003c/li\u003e\n\u003cli\u003eCashman, K. V. \u0026amp; Scheu, B. \u003cem\u003eMagmatic Fragmentation\u003c/em\u003e. \u003cem\u003eThe Encyclopedia of Volcanoes\u003c/em\u003e (Elsevier Inc., 2015). doi:10.1016/b978-0-12-385938-9.00025-0.\u003c/li\u003e\n\u003cli\u003eCassidy, M., Manga, M., Cashman, K. V. \u0026amp; Bachmann, O. Controls on explosive-effusive volcanic eruption styles. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 2839 (2018).\u003c/li\u003e\n\u003cli\u003eKueppers, U., Scheu, B., Spieler, O. \u0026amp; Dingwell, D. B. Fragmentation efficiency of explosive volcanic eruptions: A study of experimentally generated pyroclasts. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e153\u003c/strong\u003e, 125\u0026ndash;135 (2006).\u003c/li\u003e\n\u003cli\u003eMueller, S., Scheu, B., Spieler, O. \u0026amp; Dingwell, D. B. Permeability control on magma fragmentation. \u003cem\u003eGeology\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 399\u0026ndash;402 (2008).\u003c/li\u003e\n\u003cli\u003eDingwell, D. B. Volcanic Dilemma: Flow or Blow? \u003cem\u003eScience (80-. ).\u003c/em\u003e \u003cstrong\u003e273\u003c/strong\u003e, 1054\u0026ndash;1055 (1996).\u003c/li\u003e\n\u003cli\u003eOkumura, S., Nakamura, M., Nakano, T., Uesugi, K. \u0026amp; Tsuchiyama, A. Shear deformation experiments on vesicular rhyolite: Implications for brittle fracturing, degassing, and compaction of magmas in volcanic conduits. \u003cem\u003eJ. Geophys. Res. Solid Earth\u003c/em\u003e \u003cstrong\u003e115\u003c/strong\u003e, 1\u0026ndash;10 (2010).\u003c/li\u003e\n\u003cli\u003eValdivia, P., Marshall, A. A., Brand, B. D., Manga, M. \u0026amp; Huber, C. Mafic explosive volcanism at Llaima Volcano: 3D x-ray microtomography reconstruction of pyroclasts to constrain shallow conduit processes. \u003cem\u003eBull. Volcanol.\u003c/em\u003e \u003cstrong\u003e84\u003c/strong\u003e, 2 (2022).\u003c/li\u003e\n\u003cli\u003eMarshall, A. A. \u003cem\u003eet al.\u003c/em\u003e The mafic Curacaut\u0026iacute;n ignimbrite of Llaima volcano, Chile. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e421\u003c/strong\u003e, 107418 (2021).\u003c/li\u003e\n\u003cli\u003eBamber, E. C. \u003cem\u003eet al.\u003c/em\u003e Pre- and syn-eruptive conditions of a basaltic Plinian eruption at Masaya Volcano, Nicaragua: The Masaya Triple Layer (2.1 ka). \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e392\u003c/strong\u003e, 106761 (2020).\u003c/li\u003e\n\u003cli\u003eParfitt, E. A. A discussion of the mechanisms of explosive basaltic eruptions. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e134\u003c/strong\u003e, 77\u0026ndash;107 (2004).\u003c/li\u003e\n\u003cli\u003eArzilli, F. \u003cem\u003eet al.\u003c/em\u003e Magma fragmentation in highly explosive basaltic eruptions induced by rapid crystallization. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1023\u0026ndash;1028 (2019).\u003c/li\u003e\n\u003cli\u003eLa Spina, G., Arzilli, F., Burton, M. R., Polacci, M. \u0026amp; Clarke, A. B. Role of volatiles in highly explosive basaltic eruptions. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eVetere, F. \u003cem\u003eet al.\u003c/em\u003e Viscosity of andesite melts and its implication for magma mixing prior to Unzen 1991-1995 eruption. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e175\u003c/strong\u003e, 208\u0026ndash;217 (2008).\u003c/li\u003e\n\u003cli\u003eRobert, G., Whittington, A. G., Stechern, A. \u0026amp; Behrens, H. The effect of water on the viscosity of a synthetic calc-alkaline basaltic andesite. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e346\u003c/strong\u003e, 135\u0026ndash;148 (2013).\u003c/li\u003e\n\u003cli\u003eRichet, P., Lejeune, A. M. M., Holtz, F. \u0026amp; Roux, J. Water and the viscosity of andesite melts. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e128\u003c/strong\u003e, 185\u0026ndash;197 (1996).\u003c/li\u003e\n\u003cli\u003eLiebske, C., Behrens, H., Holtz, F. \u0026amp; Lange, R. A. The influence of pressure and composition on the viscosity of andesitic melts. \u003cem\u003eGeochim. Cosmochim. Acta\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 473\u0026ndash;485 (2003).\u003c/li\u003e\n\u003cli\u003eMader, H. M., Llewellin, E. W. \u0026amp; Mueller, S. P. The rheology of two-phase magmas: A review and analysis. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e257\u003c/strong\u003e, 135\u0026ndash;158 (2013).\u003c/li\u003e\n\u003cli\u003eOkumura, S. \u003cem\u003eet al.\u003c/em\u003e Rheology of nanocrystal-bearing andesite magma and its roles in explosive volcanism. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 241 (2022).\u003c/li\u003e\n\u003cli\u003eValdivia, P. \u003cem\u003eet al.\u003c/em\u003e Are volcanic melts less viscous than we thought? The case of Stromboli basalt. \u003cem\u003eContrib. to Mineral. Petrol.\u003c/em\u003e \u003cstrong\u003e178\u003c/strong\u003e, 45 (2023).\u003c/li\u003e\n\u003cli\u003eScarani, A. \u003cem\u003eet al.\u003c/em\u003e A chemical threshold controls nanocrystallization and degassing behaviour in basalt magmas. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 284 (2022).\u003c/li\u003e\n\u003cli\u003eZellmer, G. F. \u003cem\u003eet al.\u003c/em\u003e Inferring the effects of compositional boundary layers on crystal nucleation, growth textures, and mineral chemistry in natural volcanic tephras through submicron-resolution imaging. \u003cem\u003eFront. Earth Sci.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1\u0026ndash;7 (2016).\u003c/li\u003e\n\u003cli\u003eDi Genova, D. \u003cem\u003eet al.\u003c/em\u003e A compositional tipping point governing the mobilization and eruption style of rhyolitic magma. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e552\u003c/strong\u003e, 235\u0026ndash;238 (2017).\u003c/li\u003e\n\u003cli\u003eEichheimer, P. \u003cem\u003eet al.\u003c/em\u003e Combined numerical and experimental study of microstructure and permeability in porous granular media. \u003cem\u003eSolid Earth Discuss.\u003c/em\u003e 1\u0026ndash;25 (2020) doi:10.5194/se-2019-199.\u003c/li\u003e\n\u003cli\u003eKnafelc, J. \u003cem\u003eet al.\u003c/em\u003e Havre 2012 pink pumice is evidence of a short-lived, deep-sea, magnetite nanolite-driven explosive eruption. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 1\u0026ndash;11 (2022).\u003c/li\u003e\n\u003cli\u003eBarone, G. \u003cem\u003eet al.\u003c/em\u003e Nanoscale surface modification of Mt. Etna volcanic ashes. \u003cem\u003eGeochim. Cosmochim. Acta\u003c/em\u003e \u003cstrong\u003e174\u003c/strong\u003e, 70\u0026ndash;84 (2016).\u003c/li\u003e\n\u003cli\u003eMujin, M. \u0026amp; Nakamura, M. A nanolite record of eruption style transition. \u003cem\u003eGeology\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 611\u0026ndash;614 (2014).\u003c/li\u003e\n\u003cli\u003eMujin, M., Nakamura, M. \u0026amp; Miyake, A. Eruption style and crystal size distributions: Crystallization of groundmass nanolites in the 2011 Shinmoedake eruption. \u003cem\u003eAm. Mineral.\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 2367\u0026ndash;2380 (2017).\u003c/li\u003e\n\u003cli\u003eMujin, M. \u0026amp; Nakamura, M. Late-stage groundmass differentiation as a record of magma stagnation, fragmentation, and rewelding. \u003cem\u003eBull. Volcanol.\u003c/em\u003e \u003cstrong\u003e82\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eDi Genova, D., Caracciolo, A. \u0026amp; Kolzenburg, S. Measuring the degree of \u0026ldquo;nanotilization\u0026rdquo; of volcanic glasses: Understanding syn-eruptive processes recorded in melt inclusions. \u003cem\u003eLithos\u003c/em\u003e \u003cstrong\u003e318\u003c/strong\u003e\u0026ndash;\u003cstrong\u003e319\u003c/strong\u003e, 209\u0026ndash;218 (2018).\u003c/li\u003e\n\u003cli\u003eDi Genova, D. \u003cem\u003eet al.\u003c/em\u003e In situ observation of nanolite growth in volcanic melt: A driving force for explosive eruptions. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eC\u0026aacute;ceres, F. \u003cem\u003eet al.\u003c/em\u003e Can nanolites enhance eruption explosivity? \u003cem\u003eGeology\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 997\u0026ndash;1001 (2020).\u003c/li\u003e\n\u003cli\u003eC\u0026aacute;ceres, F. \u003cem\u003eet al.\u003c/em\u003e Oxide nanolitisation-induced melt iron extraction causes viscosity jumps and enhanced explosivity in silicic magma. (2024) doi:10.1038/s41467-024-44850-x.\u003c/li\u003e\n\u003cli\u003eGiordano, D., Russell, J. K. \u0026amp; Dingwell, D. B. Viscosity of magmatic liquids: A model. \u003cem\u003eEarth Planet. Sci. Lett.\u003c/em\u003e \u003cstrong\u003e271\u003c/strong\u003e, 123\u0026ndash;134 (2008).\u003c/li\u003e\n\u003cli\u003eDi Genova, D., Zandona, A. \u0026amp; Deubener, J. Unravelling the effect of nano-heterogeneity on the viscosity of silicate melts: Implications for glass manufacturing and volcanic eruptions. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e545\u003c/strong\u003e, 120248 (2020).\u003c/li\u003e\n\u003cli\u003eStabile, P. \u003cem\u003eet al.\u003c/em\u003e The effect of iron and alkali on the nanocrystal-free viscosity of volcanic melts: A combined Raman spectroscopy and DSC study. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e559\u003c/strong\u003e, 119991 (2021).\u003c/li\u003e\n\u003cli\u003eShebanova, O. N. \u0026amp; Lazor, P. Raman spectroscopic study of magnetite (FeFe2O4): a new assignment for the vibrational spectrum. \u003cem\u003eJ. Solid State Chem.\u003c/em\u003e \u003cstrong\u003e174\u003c/strong\u003e, 424\u0026ndash;430 (2003).\u003c/li\u003e\n\u003cli\u003eJubb, A. M. \u0026amp; Allen, H. C. Vibrational Spectroscopic Characterization of Hematite, Maghemite, and Magnetite Thin Films Produced by Vapor Deposition. \u003cem\u003eACS Appl. Mater. Interfaces\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 2804\u0026ndash;2812 (2010).\u003c/li\u003e\n\u003cli\u003eZandon\u0026agrave;, A. \u003cem\u003eet al.\u003c/em\u003e Crystallization Mechanism of Gel-Derived SiO2-TiO2 Amorphous Nanobeads Elucidated by High-Temperature In Situ Experiments. \u003cem\u003eCryst. Growth Des.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 4545\u0026ndash;4555 (2023).\u003c/li\u003e\n\u003cli\u003eZandon\u0026agrave;, A. \u003cem\u003eet al.\u003c/em\u003e Non-stoichiometric crystal nucleation in a spodumene glass containing TiO2 as seed former: Effects on the viscosity of the residual melt. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e619\u003c/strong\u003e, 122563 (2023).\u003c/li\u003e\n\u003cli\u003eLanghammer, D., Di Genova, D. \u0026amp; Steinle-Neumann, G. Modeling the Viscosity of Anhydrous and Hydrous Volcanic Melts. \u003cem\u003eGeochemistry, Geophys. Geosystems\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1\u0026ndash;26 (2021).\u003c/li\u003e\n\u003cli\u003eMauro, J. C., Yue, Y., Ellison, A. J., Gupta, P. K. \u0026amp; Allan, D. C. Viscosity of glass-forming liquids. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 19780\u0026ndash;4 (2009).\u003c/li\u003e\n\u003cli\u003ePrescher, C., McCammon, C. \u0026amp; Dubrovinsky, L. MossA : a program for analyzing energy-domain M\u0026ouml;ssbauer spectra from conventional and synchrotron sources. \u003cem\u003eJ. Appl. Crystallogr.\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 329\u0026ndash;331 (2012).\u003c/li\u003e\n\u003cli\u003eAraya, N. \u003cem\u003eet al.\u003c/em\u003e Shallow magma pre-charge during repeated Plinian eruptions at Sakurajima volcano. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1\u0026ndash;10 (2019).\u003c/li\u003e\n\u003cli\u003eScherrer, M. C., Hess, K., Fehr, K. T. \u0026amp; Dingwell, D. B. Oxidation-state dependence of rheology in peralkaline glasses of phonolitic composition. \u003cem\u003eGeophys. Res. Abstr.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 5805 (2012).\u003c/li\u003e\n\u003cli\u003eDingwell, D. B. Redox viscometry of some Fe-bearing silicate melts. \u003cem\u003eAm. Mineral.\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 1560\u0026ndash;1562 (1991).\u003c/li\u003e\n\u003cli\u003eWechsler, B. A., Lindsley, D. H. \u0026amp; Prewitt, C. T. Crystal structure and cation distribution in titanomagnetites (Fe3-xTixO4). \u003cem\u003eAm. Mineral.\u003c/em\u003e \u003cstrong\u003e69\u003c/strong\u003e, 754\u0026ndash;770 (1984).\u003c/li\u003e\n\u003cli\u003eZinin, P. \u003cem\u003eet al.\u003c/em\u003e Raman spectroscopy of titanomagnetites: Calibration of the intensity of Raman peaks as a sensitive indicator for their Ti content. \u003cem\u003eAm. Mineral.\u003c/em\u003e \u003cstrong\u003e96\u003c/strong\u003e, 1537\u0026ndash;1546 (2011).\u003c/li\u003e\n\u003cli\u003eKleebusch, E., Patzig, C., H\u0026ouml;che, T. \u0026amp; R\u0026uuml;ssel, C. The evidence of phase separation droplets in the crystallization process of a Li2O-Al2O3-SiO2 glass with TiO2 as nucleating agent \u0026ndash; An X-ray diffraction and (S)TEM-study supported by EDX-analysis. \u003cem\u003eCeram. Int.\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 2919\u0026ndash;2926 (2018).\u003c/li\u003e\n\u003cli\u003eFielitz, P., Helsch, G., Borchardt, G. \u0026amp; Deubener, J. Al-26 and O-18 tracer diffusion in a titania-coated sodium aluminosilicate glass. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e614\u003c/strong\u003e, 122400 (2023).\u003c/li\u003e\n\u003cli\u003eTuheen, M. I., Wilkins, M. C., McCloy, J. \u0026amp; Du, J. The structures of iron silicate glasses with varying iron redox ratios from molecular dynamics simulations and EXAFS analysis. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e624\u003c/strong\u003e, 122713 (2024).\u003c/li\u003e\n\u003cli\u003eKennedy, B. M., Jellinek, A. M., Russell, J. K., Nichols, A. R. L. \u0026amp; Vigouroux, N. Time-and temperature-dependent conduit wall porosity: A key control on degassing and explosivity at Tarawera volcano, New Zealand. \u003cem\u003eEarth Planet. Sci. Lett.\u003c/em\u003e \u003cstrong\u003e299\u003c/strong\u003e, 126\u0026ndash;137 (2010).\u003c/li\u003e\n\u003cli\u003eThivet, S. \u003cem\u003eet al.\u003c/em\u003e Metastable liquid immiscibility in the 2018\u0026ndash;2021 Fani Maor\u0026eacute; lavas as a mechanism for volcanic nanolite formation. \u003cem\u003eCommun. Earth Environ.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eHornby, A. J. \u003cem\u003eet al.\u003c/em\u003e Nanoscale silicate melt textures determine volcanic ash surface chemistry. 1\u0026ndash;10 (2024) doi:10.1038/s41467-024-44712-6.\u003c/li\u003e\n\u003cli\u003eDubosq, R. \u003cem\u003eet al.\u003c/em\u003e Bubbles and atom clusters in rock melts: A chicken and egg problem. \u003cem\u003eJ. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e428\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eHughes, E. C. \u003cem\u003eet al.\u003c/em\u003e Low analytical totals in EPMA of hydrous silicate glass due to sub-surface charging: Obtaining accurate volatiles by difference. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e505\u003c/strong\u003e, 48\u0026ndash;56 (2019).\u003c/li\u003e\n\u003cli\u003eBehrens, H. \u003cem\u003eet al.\u003c/em\u003e Structural relaxation mechanisms in hydrous sodium borosilicate glasses. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e497\u003c/strong\u003e, 30\u0026ndash;39 (2018).\u003c/li\u003e\n\u003cli\u003eDi Genova, D., Romano, C., Alletti, M., Misiti, V. \u0026amp; Scarlato, P. The effect of CO2 and H2O on Etna and Fondo Riccio (Phlegrean Fields) liquid viscosity, glass transition temperature and heat capacity. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e377\u003c/strong\u003e, 72\u0026ndash;86 (2014).\u003c/li\u003e\n\u003cli\u003eDouglas, R. W., Armstrong, W. L., Edward, J. . \u0026amp; Hall, D. A penetration viscometer. \u003cem\u003eGlas. Technol.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 52\u0026ndash;55 (1965).\u003c/li\u003e\n\u003cli\u003eMeerlender, G. Viskosit\u0026auml;ts-Temperaturverhalten des Standardglases I der DGG. \u003cem\u003eGlas. Ber.\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 1\u0026ndash;3 (1974).\u003c/li\u003e\n\u003cli\u003eDi Fiore, F. \u003cem\u003eet al.\u003c/em\u003e Experimental insights on the shear-induced crystallization of a phonotephrite magma. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e637\u003c/strong\u003e, 121682 (2023).\u003c/li\u003e\n\u003cli\u003eDi Fiore, F., Vona, A., Costa, A., Mollo, S. \u0026amp; Romano, C. Quantifying the influence of cooling and shear rate on the disequilibrium rheology of a trachybasaltic melt from Mt. Etna. \u003cem\u003eEarth Planet. Sci. Lett.\u003c/em\u003e \u003cstrong\u003e594\u003c/strong\u003e, 117725 (2022).\u003c/li\u003e\n\u003cli\u003eAl-Mukadam, R., Di Genova, D., Bornh\u0026ouml;ft, H. \u0026amp; Deubener, J. High rate calorimetry derived viscosity of oxide melts prone to crystallization. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e536\u003c/strong\u003e, 119992 (2020).\u003c/li\u003e\n\u003cli\u003eYue, Y. The iso-structural viscosity, configurational entropy and fragility of oxide liquids. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e355\u003c/strong\u003e, 737\u0026ndash;744 (2009).\u003c/li\u003e\n\u003cli\u003eZheng, Q., Mauro, J. C. \u0026amp; Yue, Y. Reconciling calorimetric and kinetic fragilities of glass-forming liquids. \u003cem\u003eJ. Non. Cryst. Solids\u003c/em\u003e \u003cstrong\u003e456\u003c/strong\u003e, 95\u0026ndash;100 (2017).\u003c/li\u003e\n\u003cli\u003eWhitfield, C. H., Brody, E. M. \u0026amp; Bassett, W. Elastic moduli of NaCl by Brillouin scattering at high pressure in a diamond anvil cell. \u003cem\u003eRev. Sci. Instrum.\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 942\u0026ndash;947 (1976).\u003c/li\u003e\n\u003cli\u003eCassetta, M. \u003cem\u003eet al.\u003c/em\u003e Estimating the viscosity of volcanic melts from the vibrational properties of their parental glasses. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1\u0026ndash;14 (2021).\u003c/li\u003e\n\u003cli\u003eSinogeikin, S. \u003cem\u003eet al.\u003c/em\u003e Brillouin spectrometer interfaced with synchrotron radiation for simultaneous x-ray density and acoustic velocity measurements. \u003cem\u003eRev. Sci. Instrum.\u003c/em\u003e \u003cstrong\u003e77\u003c/strong\u003e, (2006).\u003c/li\u003e\n\u003cli\u003eBerg, S. \u003cem\u003eet al.\u003c/em\u003e Ilastik: Interactive Machine Learning for (Bio)Image Analysis. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 1226\u0026ndash;1232 (2019).\u003c/li\u003e\n\u003cli\u003eAngell, C. A. Formation of Glasses from Liquids and Biopolymers. \u003cem\u003eScience (80-. ).\u003c/em\u003e \u003cstrong\u003e267\u003c/strong\u003e, 1924\u0026ndash;1935 (1995).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"andesite, viscosity, nanolite, differential scanning calorimetry, Brillouin spectroscopy, Raman spectroscopy, TEM","lastPublishedDoi":"10.21203/rs.3.rs-3891365/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3891365/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eExplosive 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 SiO\u003csub\u003e2\u003c/sub\u003e-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.\u003c/p\u003e","manuscriptTitle":"Nanoscale chemical heterogeneities control magma viscosity and failure","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-06 17:30:43","doi":"10.21203/rs.3.rs-3891365/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-earth-and-environment","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsenv","sideBox":"Learn more about [Communications Earth and Environment](https://www.nature.com/commsenv/)","snPcode":"","submissionUrl":"","title":"Communications Earth \u0026 Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"90370d0d-0969-4b54-9fe9-959fe064cfab","owner":[],"postedDate":"February 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":28584479,"name":"Earth and environmental sciences/Solid Earth sciences/Volcanology"},{"id":28584480,"name":"Earth and environmental sciences/Solid Earth sciences/Petrology"}],"tags":[],"updatedAt":"2025-06-13T13:06:05+00:00","versionOfRecord":{"articleIdentity":"rs-3891365","link":"https://doi.org/10.1038/s43247-025-02424-9","journal":{"identity":"communications-earth-and-environment","isVorOnly":false,"title":"Communications Earth \u0026 Environment"},"publishedOn":"2025-06-12 00:00:00","publishedOnDateReadable":"June 12th, 2025"},"versionCreatedAt":"2024-02-06 17:30:43","video":"","vorDoi":"10.1038/s43247-025-02424-9","vorDoiUrl":"https://doi.org/10.1038/s43247-025-02424-9","workflowStages":[]},"version":"v1","identity":"rs-3891365","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3891365","identity":"rs-3891365","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}