Obsidian can be melted, though the process is more nuanced than melting an ice cube or even a typical rock. Because obsidian is already a glass rather than a crystalline mineral, it doesn’t flip from solid to liquid at one clean temperature. Instead, it gradually softens as you heat it, passing through a transition zone around 680 °C before eventually flowing like thick honey at higher temperatures. What makes remelting obsidian tricky isn’t reaching the right temperature; it’s what the trapped water and gases inside the glass do on the way up.
Why Obsidian Doesn’t Melt Like a Normal Rock
Most rocks are made of interlocking crystals with orderly atomic structures. Heat a crystal to its melting point and the structure collapses all at once, producing a liquid. Obsidian skips that step entirely. It formed when silica-rich lava cooled so fast that atoms never had time to arrange themselves into crystals. The result is a solid that is structurally a frozen liquid, with atoms locked in a disordered arrangement. Researchers studying rhyolite-obsidian lava flows have confirmed that obsidian occupies the rapidly cooled outer margins of a flow, while the slower-cooling interior crystallizes into rhyolite, a true crystalline rock.1ScienceDirect. Cooling and crystallization of rhyolite–obsidian lava: Insights from micron-scale projections on plagioclase microlites
Because obsidian is already amorphous, reheating it doesn’t produce a dramatic melting event. Instead, it undergoes what materials scientists call a glass transition: a gradual shift from a brittle solid to a rubbery, then viscous, state. The glass transition temperature (Tg) for obsidian sits at roughly 680 °C, about 100 degrees lower than that of moldavite, a tektite glass with a different chemical makeup.2ScienceDirect. Glass chemistry of tektites At and above Tg, obsidian stops behaving like a solid and starts to deform under its own weight. Push the temperature higher still, toward 1,000–1,200 °C, and the glass becomes fluid enough to pour.
The Role of Trapped Water
Here is where melting obsidian gets complicated in practice. Obsidian contains dissolved water molecules locked inside its glassy structure. These aren’t bubbles you can see; the water is chemically bound at the molecular level. Measurements using infrared spectroscopy and density analysis across 34 different obsidian sources worldwide show that this structural water can range from as little as 0.07% to as much as 1.66% by weight.3Journal of Archaeological Science: Reports. Variability in obsidian structural water content and its importance in the hydration dating of cultural artifacts That might sound trivial, but in a glass being heated to near-melting temperatures, even a small percentage of dissolved water has dramatic effects.
As the temperature climbs past the glass transition point, the trapped water starts converting to steam. Because the surrounding glass is viscous but not yet fully liquid, the steam can’t escape easily. It inflates tiny bubbles inside the softened glass, causing the material to expand, sometimes enormously. This is essentially the industrial process for making perlite, a lightweight construction material. When obsidian-like volcanic glass is heated rapidly to around 850–900 °C, the internal water vaporizes and puffs the glass up to many times its original volume. The expansion depends on the ratio of dissolved to crystallized water in the rock, the size of the pieces being heated, and the firing conditions, including temperature ramp rate and hold time.4IOP Conference Series Materials Science and Engineering. Obsidian expansion kinetics
So if your goal is to gently melt obsidian into a smooth, bubble-free liquid, you need to account for degassing. Heating too fast produces a foamy, pumice-like mass. Heating slowly and holding the temperature for extended periods gives the water time to diffuse out before the glass becomes fully mobile. Industrial and experimental settings that aim for a clean melt tend to ramp the temperature gradually and may hold the glass at intermediate temperatures to allow volatiles to escape before pushing into the fully liquid range.
What Temperatures Actually Produce a Flowing Liquid
The Tg of roughly 680 °C is just the beginning of the story. At that temperature, obsidian becomes deformable but not pourable. Think of it like warming a stick of butter: at some point it bends instead of snapping, but it isn’t yet a puddle. For obsidian to flow freely enough that you could pour it, you generally need temperatures above 1,000 °C and often closer to 1,200 °C or higher, depending on the specific composition. Obsidian’s high silica content (typically 70–75% SiO₂) makes it far more viscous than low-silica glasses. Silica creates long, tangled molecular chains that resist flow, which is why obsidian’s parent magma, rhyolitic lava, is famously sluggish and pasty compared to the runny basaltic lava you see in Hawaiian eruptions.
Water content actually lowers viscosity. The more dissolved water the obsidian contains, the more easily it flows at a given temperature. This is one reason volcanic eruptions of rhyolitic magma can be so explosive: the magma is viscous enough to trap water under enormous pressure, and when that pressure drops, the water flashes to steam all at once. In a laboratory or furnace setting, though, you’re working at atmospheric pressure, so the water tends to bubble out rather than stay dissolved and reduce viscosity. The practical result is that remelting dry obsidian (obsidian that has already lost most of its water) requires higher temperatures or longer hold times to achieve the same fluidity as the original magma had underground.
Can You Make Obsidian Again by Cooling the Melt
In principle, yes. If you melt obsidian and cool it fast enough, you get glass again. Whether that glass qualifies as “obsidian” depends on how strict your definition is. Geologists define obsidian as a naturally occurring volcanic glass with a specific chemical composition. A synthetic glass made by melting obsidian in a furnace would have the same chemistry and the same amorphous structure, but most geologists would call it a synthetic volcanic glass rather than obsidian, the way a lab-grown diamond is chemically identical to a mined one but carries a different label.
The key variable is cooling rate. Cool the melt slowly and crystals will nucleate and grow, turning the glass into a crystalline rock. This is exactly what happens in the interior of thick lava flows, where the insulated center cools over weeks or months and crystallizes into rhyolite, while only the rapidly quenched exterior remains glassy obsidian.1ScienceDirect. Cooling and crystallization of rhyolite–obsidian lava: Insights from micron-scale projections on plagioclase microlites In a furnace, if you let the crucible sit and cool overnight, you might end up with a partially crystalline mass rather than the clean, conchoidal-fracturing glass you started with. Quenching the melt quickly, by pouring it onto a metal plate or plunging the crucible into water (carefully, to avoid thermal shock shattering the container), favors glass formation.
Another factor is the lost water. The original obsidian’s dissolved water lowered its viscosity and may have helped suppress crystallization during the initial volcanic quench. Once that water has boiled off in your furnace, the remelted glass is drier and behaves slightly differently on cooling. It’s still possible to get a glassy product, but the window of cooling rates that produce glass versus crystals shifts somewhat.
Obsidian Versus Manufactured Glass
People often wonder how obsidian compares to the glass in a window or a bottle. Both are amorphous silicates, but the similarities mostly end there. Manufactured soda-lime glass, the everyday variety, typically contains about 70% silica, 15% soda (Na₂O), and 10% lime (CaO), plus minor ingredients. Obsidian has a similar silica content but a very different balance of the other oxides. The higher sodium oxide and lower calcium and magnesium oxide content in obsidian compared to some other natural glasses like tektites account for its lower glass transition temperature and higher rate of thermal expansion.2ScienceDirect. Glass chemistry of tektites
The thermal expansion coefficient for obsidian has been measured at around 6.3 × 10⁻⁶ per degree Kelvin, compared to 3.7 × 10⁻⁶ for moldavite tektite glass.2ScienceDirect. Glass chemistry of tektites In practical terms, obsidian expands more when heated than some other natural glasses, which means it’s more sensitive to rapid temperature changes. Heating one spot of an obsidian slab while the rest stays cool creates internal stresses that can crack or shatter it. Anyone attempting to melt obsidian needs to heat the entire piece uniformly and ramp up gradually, or work with small pieces and powder that can tolerate uneven heating without fracturing.
Window glass, by contrast, is engineered for thermal stability, and borosilicate glass (the kind in laboratory beakers and some cookware) is specifically designed to resist thermal shock. Obsidian was never optimized for anything; it’s whatever chemistry the volcano happened to produce, which makes its thermal behavior less predictable and more variable from source to source.
Why People Try to Melt Obsidian
The question isn’t purely academic. There are several real-world reasons someone might want to heat or melt obsidian, and the outcome depends heavily on the goal.
- Knifemaking and tool production: Some modern knappers heat-treat obsidian at sub-melting temperatures (a few hundred degrees Celsius) to relieve internal stresses and make the glass easier to flake. This is well below the glass transition point and doesn’t involve melting. At higher temperatures, archaeologists have studied how obsidian artifacts change shape and surface texture, which helps them recognize heat-altered pieces in the archaeological record.
- Perlite production: The construction and horticulture industries use expanded perlite as insulation and a lightweight soil amendment. The feedstock is a hydrated volcanic glass closely related to obsidian, and the entire process depends on flash-heating the glass to exploit its trapped water content.4IOP Conference Series Materials Science and Engineering. Obsidian expansion kinetics
- Art and glassblowing: A handful of artists and experimenters have attempted to remelt obsidian for casting or glassblowing. The main challenges are the high viscosity (compared to the manufactured glass that glassblowers normally use), the degassing problem, and the difficulty of achieving a homogeneous melt without specialized equipment.
- Scientific research: Petrologists remelt obsidian and other volcanic glasses in laboratories to study magma behavior, measure viscosity at different temperatures, and understand how dissolved water affects eruption dynamics.
For perlite production, the foaming caused by trapped water is the entire point. For nearly every other application, it’s a nuisance that has to be managed.
Common Misconceptions About Melting Obsidian
One persistent myth is that obsidian is somehow unbreakable or extraordinarily heat-resistant, perhaps because of its association with sharp, durable tools and its dramatic black appearance. In reality, obsidian is a fairly fragile glass. It shatters easily on impact (that’s why it makes such sharp edges when fractured), and it’s no more resistant to heat than many other silicate glasses. Its Tg of around 680 °C is lower than that of borosilicate laboratory glass and far lower than the melting points of crystalline minerals like quartz (which needs roughly 1,700 °C to melt at atmospheric pressure).
Another misconception is that you could melt obsidian in a campfire or a standard pottery kiln and pour it into a mold like liquid metal. A well-built campfire can reach 600–800 °C in its hottest zones, which is enough to soften obsidian slightly but nowhere near enough to produce a pourable liquid. A high-temperature pottery kiln operating at 1,200–1,300 °C could potentially get obsidian to flow, but you’d still need to deal with the foaming from trapped water, the difficulty of containing molten glass in a standard ceramic crucible (which the glass may bond to), and the high viscosity of the melt. It’s doable, but messy and unpredictable without experience.
A third misconception, common in gaming and fantasy communities, is that obsidian can be cast like metal into swords, arrowheads, or other shapes by simply pouring the melt into a mold. While you can technically cast obsidian glass, the resulting piece would be riddled with internal stresses unless annealed (slowly cooled through the glass transition zone in a controlled way). Without annealing, a cast obsidian object would spontaneously crack or shatter within hours or days as those stresses equilibrate. Real obsidian tools throughout history were always knapped (chipped to shape), never cast, because the fracture properties of naturally quenched glass are what make it useful in the first place.
How Source Chemistry Affects What Happens in the Furnace
Not all obsidian is the same. The structural water content alone varies by more than a factor of twenty across different geological sources, from 0.07% to 1.66%.3Journal of Archaeological Science: Reports. Variability in obsidian structural water content and its importance in the hydration dating of cultural artifacts An obsidian sample from one volcano might foam dramatically at 900 °C while a sample from another source barely expands at all, simply because one contains twenty times more trapped water. The alkali and alkaline earth oxide ratios also differ from flow to flow, shifting the glass transition temperature, viscosity curve, and thermal expansion behavior.
This variability means there’s no single recipe for melting obsidian. The temperature at which it starts to soften, the rate at which volatiles escape, and the viscosity of the resulting liquid all depend on where the obsidian came from. Researchers working with obsidian in laboratory furnaces routinely characterize their specific sample’s composition and water content before designing a heating protocol. For a hobbyist or artist, the practical takeaway is that you can’t assume one piece of obsidian will behave like another, especially if they’re from different sources. Starting with small test pieces and ramping temperature slowly is the safest approach.
The Devitrification Problem
Even without human intervention, obsidian doesn’t stay glassy forever. Over geological time, the disordered atomic structure slowly rearranges toward a more stable crystalline state, a process called devitrification. Water molecules from the environment gradually diffuse into the glass surface, and tiny crystal clusters called spherulites nucleate and grow outward. This is why very old obsidian (millions of years) is rare; most of it has partially or fully crystallized into perlite or other altered forms.
When you heat obsidian in a furnace, you’re accelerating this competition between glass and crystal. Hold the temperature in the right range for long enough and the glass will crystallize rather than remain amorphous when cooled. The window between “hot enough to flow” and “hot enough to nucleate crystals” depends on composition and especially on water content. A drier glass (one whose water has been driven off during heating) tends to be more resistant to crystallization, because water acts as a flux that lowers the energy barrier for crystal nucleation. Paradoxically, this means that slowly preheating obsidian to drive off water before ramping to full melting temperature can help you produce a cleaner glass on cooling, since the drier melt is less prone to crystallizing during the cool-down.
For anyone attempting to remelt and cast obsidian, the practical implication is that timing matters almost as much as temperature. Holding the melt at high temperature for too long invites crystallization. Quenching too fast invites thermal shock cracking. The sweet spot is a moderately fast cool that passes through the crystallization-prone temperature range quickly but then slows through the glass transition zone to relieve internal stresses. Glassmakers call this annealing, and it’s as important for obsidian glass as it is for any manufactured variety.