Home » Posts tagged 'electron microscopy'

Tag Archives: electron microscopy


High resolution imaging of crystal zoning reveals events occurring in the months and days before an eruptions

Development of new techniques enables quantification of major and trace element zoning in lava phenocysts to nanometre scale

Whilst the long-term evolution of a magmatic system may occur over many thousands of years, changes immediately preceding volcanic eruptions may occur on the timescale of months, days and minutes. This kind of resolution is not obtainable using classical radiometric dating, being is constrained by the half-lives of elements. Within the last decade, petrologists have increasingly turned to the technique of diffusion chronometry; here, the blurring in chemical zoning within lava phenocrysts is used to estimate the duration between the last perturbation of the magmatic system and the eruption. The better the resolution of the electron microscope image, the smaller the timescale that can be calculated from a crystal.

Saunders and co-workers interrogated plagioclase and orthopyroxene crystals from the 1980 eruption of Mt St Helens using the FEG-EPMA at the University of Bristol, a high-resolution electron microprobe purchased as part of the CRITMAG grant. The FEG-EPMA has a beam size of ~ 30 nm, two orders of magnitude smaller than is common in conventional microprobe analysis. This allows collection of both detailed back scattered electron (BSE) images and major element composition. The latter has a quantitative resolution of 750 nm (0.00075 mm), making FEG-EPMA ideally suited for application to diffusion chronometry. The authors also look at two other methods of microanalysis, NanoSIMS and TOF-SIMS, which cannot image samples but have the advantage of being able to measure a large range of major and trace elements at ≥50 nm resolution.

Using one, or a combination of the three techniques described in the paper, the authors demonstrate is it possible to obtain chemical profiles of zoned minerals with nanoscale precision. Such detail facilitates characterisation of events that occurred in the months and days before historic eruptions.

Saunders K, Buse B, Kilburn MR, Kearns S, & Blundy J (2014) ‘Nanoscale characterisation of crystal zoning’, Chemical Geology, 364, 20-32.




Compositional distribution of erupted lavas is controlled by phase relations of primitive basalts

High pressure experiments demonstrate that variations in water content and depth of differentiation can produce a wide variety of erupted lavas from a single primitive source

Lava suites erupted from individual volcanic centres commonly exhibit a compositional ‘gap’ between basaltic and rhyolitic compositions, where the volume of intermediate eruptives is less than mafic and acidic equivalents.  A study by Melekhova and co-workers explores the distribution of lava compositions erupted from crustal volcanoes, focusing on a case study from the volcanic island of St Vincent in the Lesser Antilles. The crystallisation of cooling basaltic magmas was simulated using high pressure experiments, with synthetic run products analysed using a variety of microanalytical techniques.  The authors discovered that variation in melt fraction (the amount of molten rock remaining in the model system) and melt composition with temperature is controlled by the composition of minerals crystallising from the parent magma. For example, a rapid decrease in melt fraction, and increase in melt SiO2, occurs when the minerals and melt have similar (eutectic-like) compositions, which is the case when little water is present.

Summit of Soufriere St Vincent, Lesser Antilles,

View of lava dome in the summit caldera of Soufriere St Vincent, Lesser Antilles. Credit: Richard Arculus

The experimentally determined phase relations were incorporated into a numerical model, which allowed enabled the team to explore the evolution of a magmatic system over time by simulating the incremental emplacement of small batches of magma beneath a volcano. When model results were compared with natural rocks from St Vincent, the best fit is produced by theoretical runs with water contents mirroring data from recently analysed melt inclusions, and heat content correlating well with the age of the island (~0.4 – 2.0 Ma). Furthermore, calculations show that the observed bimodality in erupted compositions is a natural consequence of the ‘damp’ nature of sub-arc melts.

Although Melekhova et al.’s approach focused on an oceanic island arc volcano, it offers insights into other types of volcanic system; because magmas produced from a given basalt exhibit tractable changes in composition with time, they can be compared to lavas from any igneous terrains where there are good temporal constraints on changing magma (or melt inclusion) chemistry.

Melekhova, E, Annen, CJ & Blundy, JD (2013) ‘Compositional gaps in igneous rock suites controlled by magma system heat and water content’ Nature Geoscience, vol 6, pp. 385-390.



Element maps: A leap forward for microtextural analysis

Quantitative mapping of elements in a lava sample provides insights into the kinetics of explosively-erupting magmatic systems

As magmas ascend from depth towards the surface they undergo decompression and cooling, the former being responsible for the release of dissolved volatiles in the form of volcanic gases; all three factors induce crystals to form. The speed at which magma travels towards the surface affects the rate of crystallisation, and as such, the study of the textures of explosively erupted lavas can reveal quantitative information about magmatic ascent rates and crystallisation history.

Muir and co-workers from the University of Bristol present a novel method of analysing lava microtextures. They harness a powerful new technique (EDS element mapping) whereby a rock sample is bombarded with electrons. Sample interaction with electrons produces a variety of emissions, some of which are X-rays with wavelengths characteristic to the elemental composition of the target. The combination of different elements translates into an energy spectrum which is analysed to determine the abundance of specific elements. This is repeated many thousands of times over a small area to build up a ‘map’ for each element, where the intensity of colour is proportional to the elemental concentration. Individual mineral phases can be identified and isolated; calculations are then performed to obtain the relative proportions, sizes and distributions of glass, minerals and bubbles.

SH154_AlK copy

Aluminium element map of a lava sample erupted from the Mt St Helens Dome on 29th Mar 1984. The image, taken with an electron microscope, reveals zoning in plagioclase crystals.

The technique of EDS element mapping is applied to lavas from Mt St Helens erupted bewteen 1980 – 1986. The observed trends in microtextures are similar to those previously published, with groundmass crystallinity displaying a sharp increase after the catastrophic eruption in the summer of 1980, before increasing more gradually during the next dome-building phase of activity.

EDS element mapping presents significant advantages over the previous method of manually extracting data from greyscale backscattered electron images through faster data processing, reduction in operator time and accurate identification of all textural components. The authors also highlight the potential for coupling of this technique with developing technologies, such as field emission gun (FEG) sources, which would radically reduce acquisition time and enable better spatial resolution at small crystal sizes.

Muir, DD, Blundy, JD & Rust, AC 2012, ‘Multiphase petrography of volcanic rocks using element maps: a method applied to Mount St. Helens, 1980–2005′ Bulletin of Volcanology.



Lava microtextures reflect sub-volcanic geometry

Analysis of microtextures in lava samples can be used to calculate key eruption characteristics, such as magma chamber depth and discharge rate

Microscopic imaging of erupted lavas reveals such rocks comprise a mix of glass (rapidly quenched molten rock), bubbles (originally filled with gases), and crystals. The arrangement of these components is described as the rock’s ‘texture’ and is highly variable between different eruptions and volcanoes. Lavas from explosive eruptions, where lava is rapidly erupted to the surface, are dominated by bubbles; in contrast, slowly extruded samples have time to crystallise tiny minerals (microphenocrysts) and degas dissolved volatiles.

MSH lava texture

Back scattered electron (BSE) image of an erupted lava from Mt St Helens showing distribution of vesicles (voids which were originally filled with gas – now black), glass (light grey) and crystals (geometric shapes in glass). Image width is 50 µm. Photo credit: Kathy Cashman

Melnik and co-workers used crystal size distributions (CSD), a way of analysing crystal size and ’roundness’, as a method of quantifying volcanic texture. The study determined the relationship between the volume fraction of crystals in a sample, and parameters such as the cross-sectional area of the volcanic conduit and crystal nucleation rate. CSD can therefore be used as a measure of how magma travels through the sub-volcanic system at different depths and to track the ascent of molten rock to the surface.

The model is applied to a lava sample from the 1980 eruption of Mt St Helens volcano, USA. High resolution imaging allows the study of small (<10 µm) crystals and extends the depth range of the calculations to ~8 km, which agree with other estimates from seismic imaging and petrologic studies. Not only does the new model of Melnik et al. pave the way for more advanced understanding of how lava ascend through volcanic conduits, it also allows estimation of discharge rate and conduit geometry for prehistoric or unmonitored eruptions.

Melnik, OE, Blundy, JD, Rust, AC & Muir, DD (2011) ‘Subvolcanic plumbing systems imaged through crystal size distributions’ Geology, vol 39(4), pp. 403 – 406.