The study, led by Christian Ronny Singer together with Harald Behrens, Ingo Horn, Martin Oeser, Stefan Weyer, and François Holtz, tackles a problem that has frustrated geochemists for decades. When a pegmatite-forming melt becomes saturated with water, a separate supercritical fluid exsolves from the magma. This fluid can, in principle, ferry incompatible elements such as lithium and boron through the crystallizing rock, potentially enriching pockets of the pegmatite to ore-grade concentrations. Fluid inclusions trapped in natural pegmatite minerals have been reported with lithium concentrations of up to 12,000 to 16,000 ppm and boron concentrations reaching more than 10 weight percent as B2O3, hinting at just how metal-rich these fluids can become. But measuring what actually happens at the interface between melt and fluid, on the timescales and at the temperatures where it happens, has been nearly impossible.

The Hannover team’s solution is elegantly simple in concept and fiendishly demanding in execution. They loaded a gold capsule with two synthetic glass cylinders: a blue-tinted source melt, rich in lithium and boron and modeled on a late-stage pegmatite composition containing 1.7 percent Li2O and 2.5 percent B2O3, and a chemically similar sink melt that was initially almost free of both elements. Between the two cylinders they packed a coarse-grained porous filling of crushed quartz or zircon, saturated with an aqueous chloride solution. Because the two melts never touched, any lithium or boron that appeared in the sink melt must have traveled through the fluid phase alone, exactly mimicking the way elements would move through an exsolved fluid in a cooling pegmatite dyke. The capsules were then held at 850 degrees Celsius and 100 megapascals in a rapid-heat, rapid-quench cold-seal pressure vessel for durations ranging from one hour to four days.

The analytical challenge was formidable. Lithium and boron concentrations spanning more than three orders of magnitude, from a few parts per million to over 8,000 ppm, were mapped using femtosecond laser ablation coupled to sector field ICP mass spectrometry, while isotope ratios were measured with a multicollector ICP-MS setup that switched detector configurations depending on concentration. The team even manufactured synthetic fluid inclusions inside pre-treated quartz cylinders placed within the packing material, freezing and ablating them to glimpse the composition of the fluid itself during the run. Fluid inclusion analyses from the first experiment yielded average concentrations of about 419 ppm lithium and 532 ppm boron, values that turned out to be surprisingly low compared with what equilibrium partitioning models predicted.

That surprise became one of the study’s central findings: the transport of lithium and boron through the fluid was remarkably inefficient. In the sink melts, lithium concentrations generally stayed below 130 ppm and boron rarely exceeded 600 ppm even near the fluid-melt interface, despite the source melt containing thousands of times more of both elements. The amounts transferred increased steadily with experimental duration, pointing to a sluggish, kinetically limited mobilization of the elements from the source. The culprit, the researchers conclude, is the small reactive surface area between melt and fluid. In the capsule, the melt crept only slowly into the pore spaces of the mineral packing, so the fluid contacted far less melt than it would in nature, where fluid bubbles nucleate throughout the magma and coalesce into an interconnected network. In a sense, the experiments demonstrate the importance of the reactive interface by showing what happens when it is starved.

Several other parameters left their fingerprints on the results. Experiments run under a thermal gradient, with the source end at roughly 830 degrees Celsius and the sink end near 770 degrees, transferred more lithium and boron than isothermal runs, a pattern the authors attribute to possible thermodiffusion, the Soret effect, or localized convection at the fluid-melt interface. The choice of packing material mattered too: quartz dissolved aggressively into the silica-undersaturated melts, polymerizing them and apparently promoting lithium release, whereas zircon proved nearly inert. An NaCl-bearing fluid appeared to enhance boron transport relative to a CsCl fluid, while the cesium-bearing fluid left both melts strongly enriched in cesium, consistent with published fluid-melt partition coefficients for that element.

The headline result, however, belongs to the isotopes. Both lithium and boron each have two stable isotopes, and the way their ratios shift during transport can encode the history of fluid-melt interaction. In the sink melts, the team measured δ7Li values ranging from roughly 4 to 23 per mil, consistently heavier than the source melt and, in some experiments, heavier than even the inherent signature of the sink glass itself. A conservative mixing model showed that these values cannot be explained by simple blending of source and sink compositions. Instead, the fluid must have been preferentially enriched in the heavier isotope, lithium-7, through equilibrium isotope fractionation between melt and fluid. The physical driver is coordination chemistry: in the silicate melt, lithium sits in tetrahedral coordination with oxygen, while in the low-density supercritical fluid it forms mostly neutral lithium-fluorine and lithium-chlorine complexes with lower coordination numbers. Heavy isotopes favor the phase with the stronger, lower-coordination bonding environment, and here that phase is the fluid.

Equally telling is what the team did not find. There was no evidence of kinetic isotope fractionation during transport through the fluid, the kind of effect that would enrich the sink in the lighter isotope-6 because it diffuses faster. The researchers attribute this to the short transport distance of about six centimeters and to the fact that lithium moves as a complex rather than a bare ion, making the relative mass difference between isotopic complexes negligible. The only hint of a kinetic effect appeared in the one-hour experiment, where a δ7Li gradient within the sink melt suggested that the lighter isotope had outrun the heavier one before isotopic equilibration could erase the signal. Boron, by contrast, showed no resolvable isotope fractionation at all, within an analytical uncertainty of about 4.7 per mil, likely because boron occupies trigonal coordination both in the melt and in the fluid as boric acid species, leaving little bonding-energy difference for isotopes to exploit.

Perhaps the most consequential insight is the decoupling of chemical and isotopic equilibration. The chemical exchange of lithium and boron between melt and fluid was demonstrably slow in the experiments, yet the lithium isotope ratios appear to have equilibrated rapidly at the interface. This means that lithium isotopes in pegmatite minerals may faithfully record the influence of an exsolved fluid even when the minerals themselves crystallized under strong disequilibrium, as is widely believed to occur in rapidly cooled, undercooled pegmatite melts. It gives geochemists a new degree of confidence when interpreting δ7Li values in pegmatite quartz, micas, and tourmalines as tracers of fluid activity.

The implications ripple outward to the real world of lithium exploration. Because the fluid preferentially carries the heavy isotope, fluids escaping a crystallizing pegmatite should drive the δ7Li of surrounding wall rocks upward while leaving the pegmatite itself isotopically lighter, a pattern that field studies in Tibet and elsewhere have already invoked to explain the light lithium isotope signatures of spodumene-bearing pegmatites. The experiments also confirm that in chloride-poor systems lithium remains compatible in the melt, so residual melts can keep enriching themselves in lithium even as fluid is lost, potentially pushing the melt toward supersaturation and the crystallization of spodumene or petalite. For a society betting its energy future on lithium, understanding how nature concentrates this element in the final throes of a cooling magma is more than an academic exercise, and this gold-capsule glimpse of fluid-melt handoff brings that understanding a decisive step closer.

Subject of Research: Experimental investigation of lithium and boron transport and isotope fractionation between melt and fluid in pegmatitic systems

Article Title: A novel experimental approach to investigate element transport and isotope fractionation of Li and B in pegmatitic systems during fluid–melt interaction

Article References: Singer, C. R., Behrens, H., Horn, I., Oeser, M., Weyer, S., & Holtz, F. (2026). A novel experimental approach to investigate element transport and isotope fractionation of Li and B in pegmatitic systems during fluid–melt interaction. European Journal of Mineralogy, 38(4), 497-518. https://doi.org/10.5194/ejm-38-497-2026

Image Credits: AI Generated

DOI: 10.5194/ejm-38-497-2026

Keywords: pegmatites, lithium isotopes, boron, fluid-melt interaction, isotope fractionation, experimental petrology, fluid exsolution, rare-metal deposits, silicate melts, laser ablation ICP-MS, geochemistry, green energy