Carbon nanotubes are hollow cylinders of carbon atoms with walls only one atom thick in their most extreme form, yet they possess remarkable mechanical strength, thermal conductivity and electrical properties. When millions of them grow simultaneously on a substrate, they rise in parallel like blades of grass, forming what researchers call an array or, more evocatively, a nanotube forest. The way this forest stands, how tall it grows, how tightly packed the individual tubes are, and how well they align with one another, determines how the material performs in real devices. Thermal interface materials rely on dense, well-aligned arrays to channel heat away from hotspots in electronics. Energy storage electrodes benefit from high surface area and controlled porosity. Flexible electronics demand forests that can bend and recover without collapsing. In each case, the macroscopic usefulness of the material traces directly back to those three structural descriptors: height, density and alignment.

The trouble is that these descriptors cannot be tuned independently. Growing nanotube forests typically relies on chemical vapour deposition, a process in which a carbon feedstock gas decomposes on catalytic nanoparticles, allowing tubes to grow upward from the surface. Water-assisted chemical vapour deposition, the variant used in this study, introduces controlled amounts of water vapour to keep the catalyst active and extend growth. But the process is a web of interdependent variables. Raising the growth temperature might accelerate catalysis and increase height, while simultaneously altering how quickly the catalyst particles sinter or deactivate, which in turn changes tube density. Adjusting gas flows, treatment times, or the timing of water injection can shift several structural features at once. A modification that nudges height toward its target may push density or alignment further away, forcing researchers into laborious trial-and-error cycles. The scale of the search space is staggering: with eight processing variables and even ten candidate settings per variable, the number of possible combinations reaches one hundred million.

To tame this complexity, the research team, drawn from Huazhong University of Science and Technology and Beihang University, developed an AI-assisted inverse-design framework. Inverse design flips the usual logic of materials synthesis. Instead of running an experiment and measuring what comes out, the researcher specifies the outcome first, the desired height, density and alignment, and asks an algorithm to find the processing conditions most likely to deliver it. The team began by assembling an experimental database of two hundred distinct recipes for water-assisted chemical vapour deposition, each recording the full set of processing conditions alongside the structural features of the resulting array. A neural network then absorbed this dataset, learning the intricate mapping from processing space to structural space, including the nonlinear couplings that make manual optimization so difficult.

On top of this learned model, the researchers deployed an optimization algorithm that searches processing space in reverse, proposing recipes predicted to hit a prescribed structural target. Crucially, the framework was designed not to find one answer but many. For each of three distinct structural targets, repeated computational searches generated two hundred candidate solutions. Rather than accepting the single best-scoring recipe, the team deliberately selected three candidates whose processing conditions differed substantially from one another. This is where the mountain analogy becomes more than a metaphor: the optimization landscape contains many peaks that satisfy the same structural criteria, and a navigator equipped with a learned map can explore routes far from the well-worn trail surrounding a familiar recipe.

The experimental validation was rigorous. Each of the three selected recipes for each target was tested in three independent growth batches, producing twenty-seven validation runs in total. Of these, twenty-six achieved structural descriptor matching scores above ninety-five percent, meaning the measured height, density and alignment of the grown arrays came within a few percent of the prescribed targets. The consistency across independent batches matters as much as the scores themselves, because it demonstrates that the AI-identified recipes are not statistical flukes but reproducible synthesis protocols. For all three targets, the framework had successfully converted a specification into working laboratory instructions, and it had done so through multiple, demonstrably different routes.

Yet the story acquires its most intriguing twist at smaller length scales. When the researchers examined the nanotubes themselves under electron microscopy, the forests that looked identical at the array level turned out to be built from different trees. For one target, the three successful routes produced arrays with closely matched overall features, but the mean outer diameters of the individual nanotubes measured approximately 3.8, 4.8 and 7.2 nanometres respectively. The wall structures of the tubes, whether single-walled or multi-walled and how many concentric layers they contained, also differed between routes. Similar forests, in other words, need not contain identical trees. This observation carries practical weight: applications sensitive to nanotube diameter or wall number, such as electronic transport or optical absorption, may respond differently to arrays that satisfy the same coarse structural specification.

The broader implication is a shift in what inverse synthesis aims to accomplish. Traditional optimization seeks a single best recipe, but this work demonstrates that the goal can be reframed as identifying a family of experimentally accessible alternatives, each satisfying the same structural target while differing in the practical demands they place on a laboratory. One route might use a temperature that a particular furnace reaches easily; another might favour gas flows that are cheaper or safer. The AI framework functions as a navigation aid built from accumulated experimental knowledge, and its learned relationship between processing and structure can be reused to screen additional targets without repeating the full experimental investment. The initial database of two hundred recipes remains a significant cost, and extending the approach to other catalysts, feedstocks or growth methods would require new experimental data and model retraining, but within its trained domain the framework offers a reusable map of synthesis space.

For a field that has long treated nanotube growth as something of an artisanal craft, the study offers a glimpse of a more flexible future, one in which engineers specify the structure they need and choose among several validated paths to reach it, constrained only by the equipment and materials at hand. The work, led by first author Lei Zhu, a doctoral student at Huazhong University of Science and Technology, with Professor Ming Xu as corresponding author, suggests that the one hundred million possible combinations of processing variables are not an obstacle but an opportunity: hidden within that vast space lie many recipes for the same material, and artificial intelligence is now capable of finding them, testing them against reality, and revealing that even at the nanoscale, there is more than one way to grow a forest.

Subject of Research: AI-assisted inverse design of carbon nanotube array synthesis by chemical vapour deposition

Article Title: AI finds different recipes for carbon nanotube forests with matching structures

Article References: AI finds different recipes for carbon nanotube forests with matching structures. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: carbon nanotubes, inverse design, artificial intelligence, chemical vapour deposition, nanomaterials, machine learning, materials synthesis, nanotube arrays, neural networks, National Science Review, nanotechnology, structural optimization

Intermetallic compounds are ordered alloys in which nickel and gallium atoms occupy fixed positions in a crystalline lattice, a configuration that endows the material with catalytic properties distinct from either element alone. Nickel-gallium phases have attracted sustained interest since researchers demonstrated nearly a decade ago that they can catalyze the synthesis of methanol from carbon oxides, and subsequent work has explored their use in deoxygenation reactions for upgrading biomass-derived compounds. The obstacle has always been manufacturing. Conventional routes require extended heating at elevated temperatures to drive gallium atoms into the nickel lattice and establish the ordered intermetallic structure, an energy-intensive process that also coarsens the microstructure and can limit the active surface area available for catalysis.

The new approach replaces heat with mechanical energy delivered at the microscale. When ultrasound propagates through a liquid, it alternately compresses and stretches the medium, and if the acoustic pressure amplitude is sufficient, microscopic cavities nucleate, grow, and collapse violently in a phenomenon known as acoustic cavitation. The collapse of a cavitation bubble concentrates acoustic energy into a vanishingly small volume, generating transient local temperatures of thousands of degrees and enormous pressures, along with intense microjets and shock waves that impinge on nearby solid surfaces. Materials chemists have exploited these effects for decades to accelerate reactions and clean or activate surfaces, but the new study adds a crucial control knob: how readily cavitation bubbles attach to and attack the metal surface in the first place.

That knob is surface roughness, and the team tunes it with an electrochemical etch that could hardly be simpler. By polarizing the nickel in an electrolyte, hydrogen gas is generated at the surface, and the streams of evolving bubbles scour the metal, gradually texturing it into a landscape of pits and protrusions. The researchers allowed this hydrogen bubble etching to proceed for intervals ranging from zero to five minutes, producing a graded series of nickel surfaces whose roughness increased with etching time. Roughness matters for two reasons simultaneously. First, it governs how well liquid gallium wets the nickel, because a rougher surface offers more crevices for the liquid metal to infiltrate and more contact area across which gallium atoms can diffuse toward the nickel lattice. Second, it dictates where cavitation bubbles preferentially nucleate, since gas trapped in surface cavities acts as a seed for the acoustic bubbles that deliver the mechanical hammering.

Once a textured nickel substrate was immersed in liquid gallium and subjected to low-power ultrasound at 40 degrees Celsius, intermetallic layers formed with remarkable speed. Cross-sectional microscopy revealed a clear and initially intuitive trend: the thickness of the nickel-gallium layer grew as the etching time increased, peaking at samples etched for two minutes, where the intermetallic reached 8.2 micrometers. Beyond that optimum, however, the trend reversed. On the most aggressively textured surfaces, etched for five minutes, the cavitation became so intense that the freshly formed intermetallic layer could not withstand the mechanical assault and began to fragment and peel away from the substrate. The authors attribute this to cavitation-induced exfoliation, a self-limiting destruction in which the very bubbles that accelerate alloying also tear the product loose once the surface crosses a roughness threshold.

This non-monotonic relationship between roughness, cavitation efficiency, and layer integrity is the conceptual heart of the paper. It converts surface preparation into a genuine synthesis variable rather than a mere pretreatment. Too smooth, and gallium wets poorly while cavitation nucleates sparsely, leaving the alloying sluggish. Too rough, and the reaction runs hot and fast but destroys its own product. Somewhere in between lies a window in which wettability and cavitation intensity are jointly optimized, and the two-minute etch sits squarely inside it. The finding echoes earlier observations that cavitation can be positioned selectively on appropriately patterned surfaces, but it extends that principle from controlling where bubbles form to controlling how much intermetallic compound ultimately accumulates.

The practical payoff was evaluated in an electrochemical reaction of enormous contemporary relevance: the hydrogen evolution reaction, the cathodic half of water splitting that underpins green hydrogen production. Testing in one molar sodium hydroxide, a standard alkaline electrolyte, the team measured Tafel slopes, which describe how rapidly the current density increases as overpotential is applied and serve as a proxy for reaction kinetics. The optimized nickel-gallium surface etched for two minutes delivered a Tafel slope of 70.6 millivolts per decade, substantially lower than the 116.7 millivolts per decade recorded for pristine nickel under identical conditions. A lower Tafel slope means less additional voltage is needed for each tenfold increase in current, translating directly into improved energy efficiency. The treated electrode also maintained its performance over extended operation, indicating that the intermetallic layer is not merely an active but fragile coating.

The result is notable partly because liquid gallium and its alloys have become one of the most fashionable platforms in materials science, prized for a low melting point near room temperature, metallic conductivity, and a fluidity that allows them to flow, deform, and react in ways solid metals cannot. Recent studies from other groups have harnessed ultrasound to drive liquid metals into contact with reactive solids, dramatically accelerating the reaction of aluminum with water and even fracturing biometals through liquid metal layers inserted into multilayer coatings. The present work extends that emerging toolkit in a different direction, using ultrasound not to smash a material apart but to build an ordered alloy on demand, and using a cheap electrochemical pretreatment to dial in the outcome.

For catalyst designers, the implications are twofold. Synthetically, the protocol offers intermetallic compounds under mild conditions, avoiding furnace treatments that consume energy and can degrade fine structures, and offering a turnaround time of a quarter hour that makes systematic composition and thickness screening far more practical. Mechanistically, the study demonstrates that wettability and acoustic cavitation can be co-engineered through a single geometric parameter, suggesting that similar roughness-based control could transfer to other liquid-metal-solid combinations beyond nickel and gallium. The work was supported by the Natural Science Foundation of Shandong Province and several provincial research and development programs, reflecting the region’s investment in advanced light alloy technology and aluminum recycling, fields where liquid metal processing knowledge is directly transferable.

Challenges remain before the method matures beyond the laboratory. The intermetallic layers were grown on planar nickel substrates, and scaling to the high-surface-area porous electrodes favored in industrial electrolyzers will require confirming that cavitation and wetting behave comparably inside complex three-dimensional architectures. The balance between layer growth and cavitation-induced exfoliation is also inherently delicate, and real-world electrodes must tolerate far harsher current densities and gas evolution than laboratory tests impose. Even so, the demonstration that a two-minute bubble etch can more than halve the kinetic penalty of a hydrogen-evolving electrode, achieved with a household temperature, a beaker of gallium, and a modest ultrasonic source, is the kind of result that invites rapid replication. If roughness-engineered cavitation proves general, the slow, hot furnaces that have guarded the intermetallic catalysts may soon find themselves increasingly idle.

Subject of Research: Ultrasound-assisted synthesis of nickel-gallium intermetallic surfaces for electrocatalysis

Article Title: Ultrasound-driven rapid synthesis of Ni–Ga intermetallic surfaces modulated by electrolytic surface texturing under mild conditions

Article References: Bi, X., Guo, X., Xu, X., Song, W., Yao, S., Yang, X., Jia, S., Shi, D., Huo, Y., Zhao, J., & Dai, H. (2026). Ultrasound-driven rapid synthesis of Ni–Ga intermetallic surfaces modulated by electrolytic surface texturing under mild conditions. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13838-x

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13838-x

Keywords: nickel-gallium intermetallics, ultrasound, acoustic cavitation, liquid gallium, surface texturing, hydrogen bubble etching, electrocatalysis, hydrogen evolution reaction, Tafel slope, surface wettability, materials synthesis, alkaline water splitting

The challenge the team set out to solve is fundamental rather than merely technical. Two-dimensional intermetallic compounds, in which two or more metal species arrange themselves into ordered crystallographic sublattices, and high-entropy metallic alloys, in which four or more elements share a single disordered lattice, are both thermodynamically precarious when squeezed into a few atomic layers. Their intrinsic instability means that conventional vapor-phase or solution-based growth tends to produce three-dimensional particles rather than flat crystals, and kinetic factors during growth are equally difficult to control. Without a way to stabilize flat nuclei and steer their lateral expansion, the rich emergent physics predicted for these systems remained largely out of experimental reach.

The researchers call their solution an adsorption–conversion strategy, and its mechanics are worth unpacking. In a typical growth run, a metal halide precursor such as copper chloride is deposited onto a metallic matrix, for example copper germanide, in the presence of a potassium chloride salt flux. As the precursor decomposes on the hot surface, the liberated metal atoms do not simply sit there: they embed themselves into the metallic matrix below, converting the local composition into a new intermetallic phase. Meanwhile, the residual chlorine atoms adsorb onto the exposed crystal faces. Theoretical calculations show that this double action is what tips the energetic balance in favor of two dimensions.

The numbers from the team’s density functional theory calculations illustrate the effect vividly. For the copper–germanium system, embedding additional copper into a copper germanide matrix to form the richer Cu3Ge phase lowers the surface energy of the (001) facet from 0.069 electronvolts to 0.044 electronvolts, a critical first step that triggers the conversion of bulky three-dimensional Cu2Ge particles into flat Cu3Ge nanostructures. Chlorine adsorption then drives the surface energy down further, to 0.01 electronvolts. Because different crystal facets respond differently to chlorine, the energy gap between competing faces widens, and in the language of classical crystal morphology dating back to Wulff’s 1901 analysis, the crystal is pushed to grow laterally as thin plates rather than isotropically as cubes. The potassium chloride flux plays a supporting role, enhancing the wettability and fluidity of the molten system so that the flat nuclei can spread into large, uniform nanoflakes.

Control experiments confirm that both ingredients are essential. When only potassium chloride is present, the growth yields nothing but three-dimensional particles. Replace it with copper chloride, and two-dimensional nucleation begins. The team extended the same logic to iron, demonstrating through ab initio molecular dynamics simulations of the iron chloride and iron germanide interface that the identical mechanism of metal incorporation and chlorine-mediated surface stabilization operates there too, underscoring the generality of the approach rather than a lucky quirk of one chemistry.

The breadth of the resulting materials library is arguably the paper’s most striking feature. Using the adsorption–conversion route, the researchers synthesized ultrathin single crystals based on iron, cobalt, nickel, copper and palladium, spanning ordered intermetallics such as Fe3Sn2, Pd2Bi3, Ni3Sn2 and Co6Ga, multicomponent compounds such as CoFe2Ge2 and (Fe, Co, Cu)3Sn2, and genuine high-entropy alloys including (Cr, Fe, Co, Cu)3(Ge, Sn, Sb)2. Atomic-resolution scanning transmission electron microscopy confirmed that the flakes are high-quality single crystals with well-ordered atomic arrangements, and energy-dispersive mapping showed homogeneous elemental distributions without discernible aggregation even in the most compositionally complex samples. Most of the twenty-five characterized nanoflake compositions could be thinned below five nanometers, with the thinnest, Pd2Bi3, reaching just 1.9 nanometers.

Synthesis did not proceed blindly; the team distilled clear design principles that should guide future attempts. The valence electron concentration of the constituent elements governs which crystal lattice is stable, with higher values favoring close-packed hexagonal and face-centered structures and lower values permitting body-centered arrangements, while substrate symmetry can occasionally impose hexagonal stacking even on bcc-forming chemistries through quasi-van-der-Waals epitaxy. Oxygen affinity proved to be the other decisive variable. An Ellingham-type analysis of oxidation free energies explains why early transition metals such as vanadium, chromium and manganese are difficult to incorporate, since their high oxygen affinity leads to oxide contamination, whereas late transition metals like copper, palladium and nickel form readily. Even the difficult cases are not hopeless: when a hygroscopic precursor such as vanadium dichloride is rigorously dried in an oxygen-free environment, two-dimensional vanadium antimonide can indeed be grown.

The payoff for all this synthetic effort shows up in the electrical measurements. The ultrathin alloys exhibit conductivities reaching 3 × 10^7 siemens per meter, a figure that places them among the most conductive thin-film materials ever reported, and an ultralow temperature coefficient of resistance of just 2 parts per million per kelvin. That second number is the one likely to excite engineers: a resistance that barely changes between room temperature and cryogenic conditions is exactly what precision electronics, sensors and metrology instruments demand, and high-entropy alloys have long been theorized to achieve such behavior by balancing multiple electron-scattering channels. Magnetotransport measurements on CoFe2Ge2 flakes of roughly 9 and 22 nanometers revealed room-temperature ferromagnetism in both thickness regimes, opening a path toward combining high conductivity, temperature-stable resistance and magnetism in a single ultrathin platform.

What emerges from this work is less a single new material than a versatile materials platform. By establishing a reproducible, mechanistically understood route to two-dimensional single crystals of intermetallics and high-entropy alloys across much of the transition-metal series, the researchers have handed condensed-matter physicists a fresh playground in which ordered and disordered metallic lattices can be stacked, patterned and tuned at the nanometer scale. The same platform may serve technologists pursuing interconnects, magnetic devices and precision resistors. Given that the field of two-dimensional metals only reached the ångström thickness limit in 2025, the rapid arrival of a general synthesis for ultrathin intermetallics and high-entropy alloys suggests that the metallic frontier of two-dimensional materials science is now opening in earnest.

Subject of Research: Synthesis of two-dimensional intermetallic compounds and high-entropy metallic alloys via an adsorption–conversion growth strategy

Article Title: Adsorption–conversion synthesis of ultrathin intermetallics and high-entropy alloys

Article References: Zhang, P., Zhao, F., Wang, X., Wang, H., Si, K., Gong, Y., Li, B., Wei, J., Jia, Y., Gao, B., Liu, Z., Zou, X., & Duan, X. (2026). Adsorption–conversion synthesis of ultrathin intermetallics and high-entropy alloys. Nature Materials. https://doi.org/10.1038/s41563-026-02752-x

Image Credits: AI Generated

DOI: 10.1038/s41563-026-02752-x

Keywords: two-dimensional materials, intermetallic compounds, high-entropy alloys, adsorption–conversion synthesis, chloride-mediated growth, surface energy, single crystals, electrical conductivity, temperature coefficient of resistance, ferromagnetism, nanoflakes, materials synthesis