a sheet of gold one atom thick

The Mechanism

*Two-dimensional materials* — crystalline sheets of solid matter exactly one atom thick — became a major branch of condensed-matter physics in October 2004, when *Andre Geim* and *Konstantin Novoselov* at the University of Manchester isolated single-atom-thick sheets of carbon (which they named *graphene*) by mechanical exfoliation of graphite with adhesive tape. The 2004 paper (*Science* 306: 666) opened a twenty-year programme of synthesis and characterisation of 2D materials: graphene itself, hexagonal boron nitride, the transition-metal dichalcogenides (MoS₂, WSe₂, etc.), the silicene/germanene/stanene family of Group-IV monolayers, phosphorene, the MXenes (2D early-transition-metal carbides and nitrides), borophene, and more than thirty other free-standing single-layer crystalline solids by 2023. The Geim-Novoselov work won the 2010 Nobel Prize in Physics. Until April 2024 *no free-standing two-dimensional sheet of a noble metal had ever been synthesised*. The reason is structural: noble metals (gold, silver, copper, platinum) bond into close-packed three-dimensional crystals with high coordination numbers (each atom in face-centred-cubic gold is bonded to twelve nearest neighbours), and a single-atom-thick free-standing sheet has only six in-plane neighbours per atom — so any attempt to fabricate a free metal monolayer drives the atoms to *reduce their surface energy* by curling up into nanoparticles or coalescing into thicker islands. Many groups, between 2010 and 2023, demonstrated metal *monolayers on a substrate* — metal islands a single atom thick supported on graphene or on a copper foil — but no group, in twenty years, had produced a standalone free 2D sheet of a noble metal that could be peeled off the substrate and handled. The breakthrough came from the *Inorganic Chemistry* group of *Lars Hultman* and *Shun Kashiwaya* at the *Thin Film Physics Division* of *Linköping University* in Sweden. The group worked on *MAX phases* — a family of layered hexagonal ternary compounds with the general formula M_{n+1}AX_n, where *M* is an early transition metal (titanium, vanadium, chromium, etc.), *A* is a Group-13 or Group-14 element (aluminium, silicon, germanium, gallium), and *X* is carbon or nitrogen, with *n* = 1, 2, or 3. The MAX phases have a characteristic crystal structure in which layers of *M_{n+1}X_n* (e.g., Ti₃C₂) are intercalated with single layers of pure *A*-element atoms (e.g., a single layer of pure aluminium between every Ti₃C₂ block). Removing the A-element layer by selective chemical etching of the parent MAX phase leaves the *M_{n+1}X_n* layers as a 2D material — this is the standard route to the family of MXenes (Ti₃C₂, V₂C, Mo₂C, etc.) pioneered by *Yury Gogotsi* and *Michel Barsoum* at Drexel University in 2011. The Kashiwaya group thought: *if the MAX-phase A-element layer is itself a 2D layer of pure metal atoms locked between MXene blocks, and if we can swap out the conventional A-element (aluminium, silicon) for a noble metal (gold, silver), and then selectively etch the surrounding MXene blocks away, we should get a free-standing 2D sheet of the noble metal.* Theoretical work in the early 2020s — and a 2022 first-principles density-functional-theory study by Hultman and Kashiwaya — predicted that the gold-substituted MAX phase *Ti₃AuC₂* should be stable, that gold could be intercalated as a single layer between Ti₃C₂ MXene blocks, and that selective etching of the Ti₃C₂ blocks should leave a 2D gold sheet. The synthesis succeeded across 2022-2023. The first step was preparing *Ti₃SiC₂*, a well-known silicon-containing MAX phase, by powder metallurgy at 1,500 °C. The second step was the *substitution reaction*: heating Ti₃SiC₂ with a thin sputtered gold film at 670 °C in a vacuum furnace for a controlled period, during which the gold atoms diffuse into the silicon layers of the MAX phase and the silicon atoms diffuse out — leaving behind *Ti₃AuC₂*, a MAX phase identical in structure to the parent except that the silicon A-element layers have been swapped for gold A-element layers. The third step — *selective etching of the surrounding Ti₃C₂ MXene blocks while preserving the gold layers* — was the experimental obstacle that had previously defeated everyone who had tried to make 2D gold. Standard MXene etching uses *hydrofluoric acid* or *LiF/HCl* solutions, both of which dissolve gold along with everything else. The Linköping team discovered, in the autumn of 2023, that *Murakami's reagent* — an aqueous mixture of *potassium ferricyanide* (K₃[Fe(CN)₆]) and *potassium hydroxide* (KOH) developed in the late nineteenth century (the work is attributed to the Japanese metallurgist *T. Murakami* and is still used in 2026) for the *acid-etching of Japanese sword blades* to bring out the *hamon* (the differential-quench-line martensite pattern characteristic of high-quality samurai blades, first developed during the late Heian and Kamakura periods, 12th–14th centuries) — selectively dissolves the Ti₃C₂ MXene blocks of Ti₃AuC₂ while leaving the gold layers intact. The etching procedure was: a 100-nanometre-thick film of Ti₃AuC₂ on a sapphire substrate, immersed in Murakami's reagent at room temperature for twelve hours, then rinsed in deionised water, then mechanically peeled off the substrate with a soft-tip carbon-tape lift. The resulting free-standing sheets — when imaged by aberration-corrected scanning transmission electron microscopy at the National Institute for Materials Science in Tsukuba — showed *single-atom-thick close-packed-hexagonal sheets of gold* with lateral dimensions of approximately 100 by 100 nanometres, atomically resolved gold atoms in a 2.85 Å hexagonal lattice, and zero detectable contamination from the etching solution. The 2D gold was named *goldene*. The paper appeared in *Nature Synthesis* on *16 April 2024*: Kashiwaya, S., Shi, Y., Lu, J., Sangiovanni, D.G., Greczynski, G., Magnuson, M., Andersson, M., Rosen, J. & Hultman, L., "Synthesis of goldene comprising single-atom layer gold," *Nature Synthesis* 3: 744-751 (2024). Subsequent characterisation work in 2024 and 2025 established the electronic structure of goldene: the 2D constraint opens a band gap of approximately 1 eV in what is otherwise a noble metal, the work function shifts from gold's bulk 5.1 eV to approximately 4.4 eV in the 2D sheet, and the d-band electrons are markedly more reactive than in bulk gold — properties that make goldene a potential platform for plasmonic catalysis, transparent electrodes, and 2D quantum-confined optoelectronics. As of June 2026 the original *Nature Synthesis* paper has been cited 670 times and the synthesis has been independently reproduced by groups in Germany, China, and the United States; the MAX-phase-substitution-and-Murakami-etch route is being extended to *silverene*, *platinene*, *palladene*, and *coppergene*, a small new family of 2D noble-metal sheets. The route turns on a chemical separation discovered by a Japanese metallurgist in the late nineteenth century for the cosmetic finishing of samurai blades. There is no historical record of any swordsmith of the *Kamakura* or *Muromachi* periods ever predicting that the same reagent, applied a hundred and forty years later to a titanium-and-gold ceramic, would lift the first free-standing two-dimensional sheet of gold off a sapphire substrate in Linköping, Sweden, on the afternoon of *15 December 2023*.