Flux melting is the process by which volatile substances, primarily water and carbon dioxide, are introduced into hot rock that would otherwise remain solid, lowering its melting temperature enough to generate magma. It is the dominant mechanism behind volcanism at subduction zones, where one tectonic plate dives beneath another, and it accounts for the fiery chains of volcanoes that ring the Pacific Ocean. The process is deceptively simple in concept but layered in detail, because the source, transport, and chemistry of those volatiles all shape the magma that eventually reaches the surface.
Why Adding Water Makes Rock Melt
Mantle rock at the depths where subduction-zone magma forms is hot, often well above 1,000 °C, but still below the temperature required to melt dry peridotite (the dominant rock type in the upper mantle). Dry peridotite needs roughly 1,300 °C or more to begin melting under the pressures found beneath volcanic arcs. Water changes the equation dramatically by disrupting the mineral crystal lattice, breaking bonds between silicon and oxygen atoms, and making it far easier for the rock to transition from solid to liquid. Experiments using a fertile peridotite composition show that when the rock is saturated with water, the solidus (the temperature at which melting first begins) drops to somewhere between 950 °C and 1,000 °C at pressures around 3 GPa, which corresponds to roughly 90–100 km depth.1Journal of Geophysical Research: Solid Earth. The Water‐Saturated Solidus and Second Critical Endpoint of Peridotite: Implications for Magma Genesis Within the Mantle Wedge Other experimental work places vapor-saturated melting starting as low as 800 °C at higher pressures, around 3.2 GPa.2Earth and Planetary Science Letters. The influence of H2O on mantle wedge melting
That temperature drop of several hundred degrees is enormous. It means the mantle wedge, the triangle of hot rock sitting above a subducting plate, does not need any extra heat source to produce magma. The heat is already there. What it needs is water. Once enough water arrives, partial melting begins, and a small fraction of the rock converts to liquid. This liquid is buoyant relative to the surrounding solid, so it percolates upward, eventually pooling in magma chambers beneath volcanoes.
Where the Water Comes From
The water that triggers flux melting does not come from the surface directly. It arrives locked inside the minerals of a subducting oceanic plate. When ocean crust forms at mid-ocean ridges and then sits on the seafloor for millions of years, seawater infiltrates cracks, reacts with the basaltic rock, and produces a suite of hydrous minerals: amphibole, chlorite, epidote, lawsonite, and various micas. These minerals contain water bound into their crystal structures, essentially trapping it in a solid form that can survive the journey deep into the Earth.
As the slab descends, increasing pressure and temperature gradually destabilize those hydrous minerals, causing them to break down and release their water. This does not happen all at once. Experiments on oceanic basalt show that amphibole dehydration alone is spread over a depth interval of at least 20 km, and other hydrous phases like mica, epidote, chloritoid, and lawsonite release their water across an even wider range of depths.3PubMed. Water sources for subduction zone volcanism: new experimental constraints Most of the slab crust loses its water efficiently beneath the volcanic front, because the slab’s temperature path crosses mineral-breakdown reactions at depths shallower than about 150 km.4PubMed Central. Water transport to the core–mantle boundary In cooler slabs, minerals like lawsonite can hold on to water somewhat deeper, but even then, nearly complete dehydration is expected by around 300 km depth.
The result is a conveyor belt of water delivery. As the plate sinks, successive mineral breakdown reactions release pulses of fluid at different depths, feeding the mantle wedge above with a more or less continuous supply of volatiles. The depth at which the most water is released determines where the overlying mantle begins to melt, and by extension, where the volcanic arc sits on the surface.
How Fluids Move Through the Mantle Wedge
Once released from the slab, the water-rich fluid has to travel upward through dozens of kilometers of solid mantle rock before it can trigger melting in the zone where magma forms. How this happens is one of the less settled questions in the field, but modeling work and experiments have clarified the broad picture.
Two-phase flow models, which track both the fluid and the solid mantle matrix it moves through, show that subduction dynamics strongly shape the fluid’s path. Two main pathways emerge: a “trenchward” route that sends fluids toward the shallow part of the system, and an “arcward” route that carries fluids further inland. Which pathway dominates depends on how deep the water is released. A shallow release around 40 km depth tends to push water along the trenchward path, while a deeper release around 120 km sends it along the arcward path, with a horizontal migration distance of roughly 300 km from the trench.5Journal of Geophysical Research: Solid Earth. Water Migration in the Subduction Mantle Wedge: A Two‐Phase Flow Approach
How the fluid moves matters just as much as where it goes. High-pressure experiments exploring slab-fluid interaction with mantle rock have shown two end-member scenarios. In one, fluid percolates through the mantle as a diffuse, porous flow, reacting extensively with the surrounding rock. This creates new hydrous minerals that soak up many of the dissolved elements the fluid was carrying, essentially filtering out much of the slab’s chemical signature. In the other scenario, fluid travels through focused channels, reacting far less with the mantle along the way. That focused flow preserves the trace-element fingerprint acquired when the fluid was extracted from the slab.6Geology. Focused fluid transfer through the mantle above subduction zones The chemistry of the resulting magma depends heavily on which transport mode dominated.
The fluids themselves are not pure water. Evidence from natural samples and experiments indicates that saline fluids resembling seawater, carrying dissolved salts and carbon, migrate from the dehydrating and decarbonating slab into the mantle wedge.7PubMed Central. Mantle wedge infiltrated with saline fluids from dehydration and decarbonation of subducting slab These are complex chemical cocktails, and the dissolved constituents they carry help determine the eventual composition of arc magmas.
Carbon Dioxide as a Second Flux
Water gets most of the attention, but carbon dioxide also lowers the melting temperature of mantle rock, and it does so at greater depths where water-based fluxing has largely already played out. Experiments on carbonated peridotite show that CO₂ can initiate melting at depths up to about 330 km beneath mid-ocean ridges, producing very small fractions of carbonatite liquid, on the order of 0.03 to 0.3 percent melt.8PubMed. Melting in the Earth’s deep upper mantle caused by carbon dioxide That sounds trivially small, and it is relative to the volume of the mantle, but even tiny melt fractions matter. They are extremely mobile and enriched in incompatible elements, and they can metasomatize (chemically alter) large volumes of mantle rock as they migrate upward.
Carbonate-rich melts in the mantle tend to be short-lived, because they react with the surrounding peridotite and are eventually consumed through decarbonation reactions that release CO₂ as a free gas phase.9Geoscience Frontiers. CO2, carbonate-rich melts, and brines in the mantle – Section: Fluid origin and metasomatism: from carbonate-rich melts to CO2 ± brine fluids The CO₂ released can then act as a flux itself, lowering the solidus of neighboring rock and potentially triggering further melting. This creates a kind of chain reaction in which carbon migrates through the deep mantle in pulses, leaving a trail of altered rock chemistry behind.
In subduction zones, the story gets more interesting because the slab delivers both water and carbon. The two volatiles interact in the mantle wedge, and which one dominates the fluxing depends on depth and temperature. Water is the primary flux at the relatively shallow depths (80–150 km) where most arc magma forms. Carbon dioxide takes over at deeper levels, influencing the production of unusual magma types like carbonatites and certain alkalic basalts that erupt far from plate boundaries.
What Kind of Magma Does Flux Melting Produce?
Flux melting does not produce just one flavor of magma. The composition depends on how depleted or enriched the mantle source was before the fluids arrived, how much fluid was added, and what elements that fluid carried. A global analysis of primitive arc melts, those compositions that appear to have last been in chemical equilibrium with mantle olivine, identified six principal types across 30 volcanic arcs worldwide. These include calc-alkaline basalts and andesites, tholeiitic basalts, highly depleted tholeiitic andesites, shoshonites, and low-silica basalts.10Geochemistry, Geophysics, Geosystems. The global systematics of primitive arc melts
These different magma types reflect different combinations of mantle conditions. Tholeiitic and calc-alkaline basalts equilibrated at higher temperatures and pressures, between about 1,220 °C and 1,350 °C at depths corresponding to 1.0 to 2.5 GPa. Calc-alkaline andesites, which are richer in silica, last equilibrated at lower temperatures and shallower depths. The variation comes down to how the mantle wedge was preconditioned (how much it had already been depleted by previous melting events), the amount and type of slab-derived fluid added, and whether the rising melt reacted with the cooler lithospheric mantle on its way up.10Geochemistry, Geophysics, Geosystems. The global systematics of primitive arc melts
One practical consequence: the magmas produced by flux melting tend to be significantly wetter than those from other settings like mid-ocean ridges. Mafic arc magmas contain roughly 4 weight percent water on average, a figure that has implications for everything from how viscous the magma is to how explosively it erupts.11Elsevier / Earth and Planetary Science Letters. Why do mafic arc magmas contain ∼4 wt% water on average? That dissolved water weakens the crystal structure of minerals like olivine, making the mantle flow more easily, and it depolymerizes the melt itself, keeping it relatively fluid despite its increasingly silica-rich composition.
The Chemical Signature of Flux-Derived Magma
Geologists can often tell that a volcanic rock was born from flux melting by looking at its trace-element chemistry. Subduction-zone magmas carry a distinctive pattern: they are enriched in fluid-mobile elements like barium, strontium, lead, and lithium, and they are depleted in elements like niobium and tantalum that tend to stay behind in the residual mantle. This pattern does not show up in magmas produced by simple decompression melting at mid-ocean ridges, which is what makes it a useful fingerprint.
The enrichment in fluid-mobile elements happens because the slab-derived fluids selectively dissolve and carry those elements out of the subducting plate. Research on serpentinites, rocks formed when mantle rock is hydrated, shows that the mineral antigorite acts as a major carrier of these elements within subduction zones. Two generations of antigorite have been identified in ancient examples: an older, coarser generation formed at lower temperatures (165–250 °C) that concentrated lithium, rubidium, barium, and cesium, and a younger generation formed at higher temperatures (425–475 °C) that accumulated boron, arsenic, molybdenum, lead, strontium, and light rare earth elements.12PubMed Central. Origin of arc magmatic signature: A temperature-dependent process for trace element (re)-mobilization in subduction zones As these minerals break down at increasing depth, they release their chemical cargo into the fluids that go on to flux the mantle wedge.
The story gets more nuanced when you consider how the fluids themselves evolve. At high enough pressures and temperatures, the distinction between a water-rich fluid and a silicate melt vanishes, and the system passes through a supercritical state. When that supercritical fluid separates back into distinct aqueous and melt components, the trace-element signatures partition unevenly. Sediment-derived trace elements preferentially remain in the melt-like component, while the aqueous fluid inherits mainly the fluid-mobile elements.13PubMed Central. Separation of supercritical slab-fluids to form aqueous fluid and melt components in subduction zone magmatism This means different parts of the mantle wedge can be fluxed by chemically distinct fluids, even when those fluids originated from the same slab, producing volcanic rocks with subtly different geochemical signatures along a single arc.
Flux Melting Versus Decompression Melting
Flux melting is one of two major ways the Earth produces magma. The other is decompression melting, where hot mantle rock rises toward the surface, the pressure drops, and the rock crosses its solidus without any added volatiles. Mid-ocean ridges are the classic example: mantle material wells up to fill the gap as plates spread apart, and the drop in pressure alone is enough to generate basaltic melt.
The difference is not always clean-cut. In some tectonic settings, both mechanisms operate simultaneously or in sequence. A study of the Fukue Volcano Group in southwestern Japan found evidence for two distinct magma sources: one generated by fluid-fluxed melting of the mantle driven by volatiles derived from the deep mantle transition zone, and another produced by decompression melting of a hydrous mantle plume rising from that same zone.14Elsevier. Fluid-fluxed melting of mantle versus decompression melting of hydrous mantle plume as the cause of intraplate magmatism over a stagnant slab: Implications from Fukue Volcano Group, SW Japan The low mantle potential temperature (around 1,300 °C) and the hydrous nature of one magma type strongly suggested fluid-fluxed origins, while the other bore the hallmarks of decompression. Both were found in the same volcanic field.
A useful rule of thumb: if the volcano sits above a subduction zone, flux melting is almost certainly the primary driver. If it sits on a mid-ocean ridge, decompression melting dominates. At hotspots and other intraplate settings, the situation varies, and both processes may contribute.
Why Flux-Melted Magma Erupts Explosively
The water that triggers flux melting in the first place travels upward with the resulting magma, dissolved in the silicate liquid under high pressure. As the magma rises through the crust and pressure drops, that water can no longer stay dissolved. It exsolves, forming bubbles of steam that expand rapidly. This is the fundamental driver of explosive volcanic eruptions.
Because flux-melted magmas start out with so much more dissolved water than ridge basalts, they are inherently more explosive. The roughly 4 weight percent water in a typical arc magma translates to enormous volumes of gas once it reaches shallow depths.11Elsevier / Earth and Planetary Science Letters. Why do mafic arc magmas contain ∼4 wt% water on average? This is why the most violent eruptions on Earth, from Mount Pinatubo in 1991 to Krakatoa in 1883, occur at subduction-zone volcanoes. The flux melting process essentially loads the gun by building volatile-rich magma at depth, and decompression during ascent pulls the trigger.
The high water content also affects the magma’s evolution as it stalls in crustal magma chambers. Water promotes the crystallization of minerals that are poor in silica, which concentrates silica in the remaining liquid. This is one reason arc volcanoes produce the full spectrum of compositions from basalt through andesite to rhyolite, while mid-ocean ridges mostly produce basalt. The water-driven fractional crystallization pathway steers the chemistry toward more silica-rich, more viscous, and ultimately more explosive end members.
Tracking the Fluid Pathway With Earthquakes
Flux melting happens tens to hundreds of kilometers below the surface, well out of reach for direct observation. But geophysicists can trace the movement of fluids indirectly, because fluids weaken rock and trigger earthquakes. A study combining seismic observations and geodynamic modeling from western Greece found that earthquake patterns effectively track the flow of slab-derived fluids from their source at depths greater than 80 km to shallow depths below 40 km. Between source and sink, the fluids flow updip along the sealed plate interface, facilitating earthquakes within the slab itself. In some locations, the seal breaks and fluids escape through vents into the mantle wedge, reducing both the fluid supply and the earthquake activity updip.15Science Advances / PubMed Central. Earthquakes track subduction fluids from slab source to mantle wedge sink
This seismic imaging approach is powerful because it gives a real-time snapshot of where fluids are moving in an active subduction zone. The patterns observed in Greece match what has been found at other subduction zones worldwide, reinforcing the general model of slab dehydration feeding the mantle wedge. When the seal along the plate interface is intact, fluids stay channeled within the slab and drive a particular type of seismicity. When the seal fails, fluids flood into the wedge, and the conditions for flux melting are met.
Experimental Work and the Wet Solidus
Much of what we know about flux melting comes from laboratory experiments in which tiny samples of mantle-like rock are subjected to extreme pressures and temperatures in piston-cylinder or multi-anvil presses, with carefully controlled amounts of water or CO₂ added. These experiments are the only way to directly measure the solidus temperature under different volatile conditions, because no one can stick a thermometer into the mantle wedge.
The experimental picture has been refined considerably in recent years. Investigations of the water-saturated peridotite solidus at 3 to 6 GPa have bracketed the onset of melting more tightly than older studies, placing the wet solidus between 950 °C and 1,000 °C at 3 GPa.1Journal of Geophysical Research: Solid Earth. The Water‐Saturated Solidus and Second Critical Endpoint of Peridotite: Implications for Magma Genesis Within the Mantle Wedge Meanwhile, work on the water-undersaturated solidus, the condition where some water is present but the rock is not fully saturated, shows that this boundary is hotter than previously thought.16PubMed. Experimental constraints on the damp peridotite solidus and oceanic mantle potential temperature That finding matters because mantle rock in most of the wedge is not drenched in water; it receives varying doses from the slab below. Getting the “damp” solidus right is crucial for predicting where melting begins in real subduction zones and how much melt is produced.
These experiments also reveal subtle transitions. At pressures above about 3–4 GPa, the boundary between hydrous fluid and silicate melt disappears entirely, and the system enters a supercritical regime where a single phase carries both dissolved silicates and water.1Journal of Geophysical Research: Solid Earth. The Water‐Saturated Solidus and Second Critical Endpoint of Peridotite: Implications for Magma Genesis Within the Mantle Wedge Below this critical endpoint, it makes sense to talk about “fluid” triggering “melting.” Above it, the fluid and the melt are the same thing, and the whole concept of a solidus becomes blurred.
How Volatile Recycling Has Changed Over Earth’s History
Flux melting depends on the delivery of volatiles into the mantle, which means its importance has shifted as the Earth’s tectonic regime evolved. Global mass-balance modeling of volatile recycling through subduction suggests that the efficiency of water recycling back into the deep mantle was lower in the past, even though carbon recycling may have been more consistent.17Annual Review of Earth and Planetary Sciences. Subduction-Driven Volatile Recycling: A Global Mass Balance The decoupling between water and carbon recycling is attributed to the evolving thermal state of subduction zones: hotter slabs in the early Earth would have dehydrated at shallower depths and released water before it could penetrate deep, while carbon-bearing minerals, which are stable to higher temperatures, could survive longer.
This has implications for how volcanism has shaped the atmosphere and oceans over billions of years. If early subduction zones were less efficient at recycling water into the deep mantle, more of the water released from the slab would have triggered flux melting at shallow depths, potentially producing a different style of arc volcanism than what we see today. Meanwhile, the carbon that did make it deeper could have contributed to mantle metasomatism and to the production of unusual deep-sourced melts. The interplay between flux melting and volatile cycling is one of the processes that has regulated Earth’s surface conditions over geologic time, gradually adjusting the distribution of water and carbon between the planet’s interior and its ocean-atmosphere system.
When Flux-Derived Magmas Transform the Crust
Flux melting does not end once magma leaves the mantle. As volatile-rich arc magmas intrude into the lower crust, they can trigger a secondary round of melting in the surrounding crustal rock. The heat and fluids delivered by the incoming magma lower the solidus of the country rock, causing it to partially melt and mix with the intruding magma. Geologists call this process MASH, for melting, assimilation, storage, and homogenization. In the classic model, mantle-derived basaltic magma is the heat engine that drives crustal melting.
But fieldwork on Kangaroo Island in Australia revealed a purely crustal version of the process: a felsic MASH zone where granite magmas intruded into the crust and triggered in-place crustal melting without any direct involvement of mantle magma.18Lithos. A felsic MASH zone of crustal magmas — Feedback between granite magma intrusion and in situ crustal anatexis The feedback loop between intrusion and anatexis (partial melting of pre-existing rock) produced a complex zone of mixed magmas entirely within the continental crust. This is a reminder that flux-driven processes set in motion at the mantle level can cascade upward, altering crustal composition and building the continental crust over time in ways that extend well beyond the initial slab-derived water pulse.