A new fabrication method guides liquid gallium beneath graphene after lithography, producing durable superconducting devices that maintain performance through standard chip processing.
(Nanowerk Spotlight) Graphene is promoted as a material that could reshape electronics. It is only one atom thick, conducts electricity with very low loss, and has strong, flexible bonds between carbon atoms. These traits have led engineers to imagine transistors that switch faster, sensors that use less power, and circuits that tolerate bending and stretching.
Despite this promise, graphene lacks an important feature found in many advanced devices: it does not become superconducting on its own. Superconductivity is the state where electrical resistance drops to zero at low temperature. It is essential for quantum computers, magnetic sensors, and electrical standards that depend on perfect signal stability.
In parallel, superconducting electronics have moved toward thinner and more scalable materials, including metal films only a few atoms thick and stacked two-dimensional structures that merge different quantum or electronic traits. These systems allow information to move faster and with less energy loss than conventional devices, but they face strict fabrication limits. The materials can be damaged by high temperatures or by chemical treatments used in standard lithography.
That conflict between material performance and fabrication durability has slowed progress toward devices that combine low dimensional conductors like graphene with the frictionless current flow of a superconductor.
Researchers tried several ways to make graphene superconducting. One method brought graphene into contact with a superconducting metal to let paired electrons move across the boundary. Another stacked two sheets of graphene at a precise angle so electrons slowed down and paired. A third method slid metal atoms between the graphene and the surface it rested on, a process known as intercalation.
All of these showed signs of superconductivity in small or controlled settings, but none survived the real steps used in large scale chip fabrication. High heat, chemical etching, or plasma treatments broke down the fragile metal layers or disrupted electronic behavior before measurements could be made.
Without a stable process, superconducting graphene remained a laboratory curiosity instead of a practical building block.
Instead of adding metal before fabrication and risking damage during later steps, the researchers built the entire device structure first and only then introduced the superconducting layer. They did this by using lithographically patterned channels to guide liquid gallium under the graphene after all cleaning and etching were complete. This preserved the metal layer and allowed the combined structure to be tested without further degradation.
Schematic design of liquid metal intercalation of epitaxial graphene Hall bars. a) Graphene Hall bar design: Schematic representation of the graphene Hall bar device on a SiC substrate, including a reservoir area for liquidmetal intercalation and reference Hall bar structures. b) Defect tuning by plasma treatment: controlled plasma treatment applied to selectively introduce lattice defects in the graphene layer. c) placement of liquidmetal droplet: a liquid gallium (Ga) droplet is deposited onto the reservoir area, initiating the intercalation process. d) Liquid metal intercalation: Gallium atoms diffuse through the graphene layers and migrate along the SiC-graphene buffer layer (GBL) interface. e) Formation of the metal-intercalated graphene Hall bar: intercalated device consisting of decoupled quasi-freestanding bilayer graphene (QFBLG) and a confined gallium layer (2DGa) between the QFBLG and the SiC substrate. (Image: Reproduced from DOI:10.1002/adma.202511992, CC BY)
The work begins with graphene that is grown on a silicon carbide wafer in a process known as epitaxial growth. Heating silicon carbide causes the top layers to reorganize into a smooth sheet of graphene attached to a buffer layer. This approach works over entire wafers and avoids handling or transferring the material, which helps maintain the graphene’s quality.
Next, the researchers pattern the graphene into Hall bar devices. A Hall bar is a simple structure shaped like a capital “T” with metal contacts at its ends and sides, allowing researchers to measure both the flow of current and the Hall voltage that forms when a magnetic field is applied. Thin channels are etched from each Hall bar to a nearby area known as the reservoir. Only the reservoir is exposed to a brief plasma step, which introduces defects into the buffer layer beneath the graphene. Those defects act as openings where metal atoms can enter.
Once the device layout is complete, a small drop of gallium is placed on the reservoir and warmed to just above its melting point, which is about 30 degrees Celsius. The liquid gallium seeps into the defects and begins to spread beneath the graphene. Because the reservoir and channels were patterned before intercalation, the metal moves along controlled paths and reaches the device area without exposing the rest of the surface to plasma damage.
Optical microscopy reveals the motion of the gallium as a visible front, which advances along the channels with a steady pace. The rate depends on the topography of the underlying silicon carbide, which is made up of flat terraces separated by narrow atomic steps. Gallium moves more easily along a terrace than across a step because extra energy is needed to overcome the height difference. This energy barrier, known as the Ehrlich Schwoebel barrier, causes the diffusion to be directional. By aligning diffusion channels with the terraces, the researchers increased uniformity and prevented gaps in the metal layer.
Raman spectroscopy confirms what happens once the gallium reaches the device. In Raman spectroscopy, laser light scatters from the sample and shifts in energy depending on how the atoms vibrate. These energy shifts are measured in inverse centimeters, a unit that indicates vibrational energy by counting how many wave cycles fit into one centimeter. Before intercalation, graphene produces a characteristic peak called the G peak at about 1602 inverse centimeters.
After intercalation, this peak moves down to about 1586 inverse centimeters, which shows that strain has been removed and the buffer layer has separated from the graphene. Another peak called the 2D peak also changes shape, which matches the formation of two layers of graphene in some places due to the metal beneath. New low energy peaks appear at around 21 and 52 inverse centimeters, which are known signatures of a thin gallium film.
Further measurements confirm this picture from the electronic side. Angle resolved photoemission spectroscopy shows that electron bands in graphene shift after intercalation, with the Dirac point moving about 0.35 electronvolts below the Fermi level.
The Dirac point is where the energy of electrons changes from hole like to electron like. Its shift indicates that electrons from the gallium have moved into the graphene layer. The number of additional charge carriers is about 8.1 times 10¹² per square centimeter.
X ray photoemission spectroscopy detects signals from both metallic gallium and silicon gallium bonds, which supports a model where the first layer of gallium bonds to the silicon while the layers above act like a metal.
Transport measurements make the role of the metal clear. A reference Hall bar made from untreated epitaxial graphene on the same chip shows the quantum Hall effect. In this effect, the voltage across the device locks to precise values at certain magnetic fields while the resistance along the device drops to zero.
The intercalated device shows something very different. As the temperature is lowered, the device’s resistance falls sharply and reaches zero between about 3.0 and 3.25 kelvin. Zero resistance is the defining feature of superconductivity. A magnetic field of about 100 millitesla destroys this state, which aligns well with the known properties of thin gallium films.
The Hall voltage above the superconducting transition includes a part that does not match a simple Hall effect. It is symmetric with respect to the magnetic field and even changes sign. The researchers attribute this to current jetting, which happens when current prefers certain paths in an uneven conductor and mixes the measured voltages.
After removing this effect, the remaining Hall voltage points to a carrier density above 10¹⁵ electrons per square centimeter, which is far higher than in the graphene alone and supports the idea that most conduction happens in the gallium layer.
This method provides a way to embed a superconducting film beneath graphene using tools that match existing chip workflows. It avoids the damage caused by earlier attempts because the metal is added only after all heating, cleaning, and patterning steps are finished.
The diffusion paths can be designed with lithography and tuned to match the natural structure of the wafer. The result is a device that can be built in arrays, measured with standard equipment, and modified for different metals or thicknesses. The study shows how a thin film of liquid metal can be guided under a two-dimensional material to give it new electronic behavior in a form that is stable, scalable, and ready for further development.
For authors and communications departmentsclick to open
Lay summary
Prefilled posts
ORCID information
Stefan Wundrack (Physikalisch-Technische Bundesanstalt)
, 0000-0003-4811-1239 corresponding author, first author
Nanowerk Newsletter
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
