Under the microscope it almost looks like a miniature scaffold: very thin beams, large voids and a three-dimensional structure that, when compressed, holds surprising loads. To build it, the researchers started from a material that we normally associate with something completely different, the DNAusing it to shape a lattice then coated with silica, the main component of glass. The result is a nanomaterial that, in tests published on Cell Reports Physical Scienceshowed an extraordinary weight-to-resistance ratio.
The comparison that inevitably made the rounds around the world is that with steel. According to researchers at the University of Connecticut, Columbia University and Brookhaven National Laboratory, the resulting nanostructures have a approximately four times higher strength and approximately five times lower density than steel. If you read somewhere five times stronger and four times lighter, the numbers have simply been mixed up.
And yes, even Iron Man is involved in this story. Oleg Gang, one of the authors of the study, explained that he had thought of the superhero’s armor: light enough to allow him to fly and resistant enough not to transform him into a very expensive can on the first try. The armor, however, can wait. The material still exists on scales of a few micrometers.
Because glass stops behaving like the glass we know
The interesting part begins with the apparently least suitable material. Macroscopic glass breaks easily because cracks, scratches and other defects concentrate stresses until they fracture. Drastically reducing the size changes a lot: silica elements just a few nanometers thick are much less likely to contain critical defects and can approach the theoretical strength of the material. The problem, if anything, is building something complex with such small elements.
So the researchers realized via DNA origami structures that self-assemble according to a programmed geometry. A kind of molecular scaffolding, with a precision that normal three-dimensional printing techniques struggle to reach below a few tens of nanometers. A very thin covering of silica was grown on this skeleton. In the 2023 study the structural elements obtained had thicknesses in the order of 4-20 nanometers.
Geometry does a lot of the work. The lattice contains a lot of empty space and therefore little matter, while the very thin silica beams exploit their high resistance at the nanometric scale. It’s one of those cases where removing material, if you remove it in the right places, can make the structure much more interesting.
Furthermore, before the mechanical tests, the structures of the first study were subjected to heat treatment. DNA therefore played above all the role of programmable mold to obtain the silica architecture, rather than being the component responsible for the final properties of the tested lattice.
From the first experiment, DNA stopped being just a template
The story, however, did not stop with the work published in 2023. In 2024 another group built octagonal and elongated DNA lattices coated with silica layers thin to about 1.65 nanometers. In that case the DNA remained inside the structures. Molecular simulations have indicated something curious: The DNA core can help delay the collapse of the tiny beams, limiting their instability and allowing the structure to absorb more energy before failing. The study was published on Matter.
DNA, therefore, can do more than the surveyor who designs the scaffolding and then leaves. In certain configurations it can remain inside and contribute to the mechanical behavior of the material.
In the same year the Gang group expanded the game even further. In a study on Science AdvancesDNA-programmed structures have been used to obtain three-dimensional architectures containing metals, metal oxides and semiconductorsincluding copper, zinc, aluminum, tungsten, indium, tin and platinum. DNA thus begins to resemble a construction platform, capable of establishing where very different materials should be located with a precision that is difficult to achieve through conventional manufacturing.
The problem now is to make it big
There is a fairly obvious industrial distance between a grating a few micrometers wide and a car door. Costs, quantities of DNA required, assembly times and integration with normal manufacturing processes remain substantial obstacles. Gang himself, already presenting the first work, warned that a lot of research would still be needed before transforming this method into a usable technology.
A few centimeters in the right direction, however, have been made. In 2025 the group showed up Nature Communications how to grow three-dimensional lattices programmed from DNA only in desired areas of silicon wafers and surfaces coated with different metal oxides. The structures were organized into patterns tens of micrometers wide over total areas of approximately 10 square millimeters.
The researchers even managed to arrange them in predetermined shapes and embed gold nanoparticles and proteins at defined points in the structure. Here the focus shifts from simply demonstrating an exceptionally strong material to the possibility of integrating these architectures into optical, electronic and sensor devices.
So no superhero armor hanging in the laboratory closets, at least for now. However, the less cinematic part is also the most scientifically interesting: DNA is becoming a tool for building materialsallowing you to design geometries at scales where cutters, printers and traditional processes are starting to have hands that are decidedly too big.
