MIT physicists build terahertz microscope to reveal hidden quantum motions in superconductors

Physicists at MIT have built a new type of microscope that uses terahertz light to observe quantum motions inside superconducting materials that were previously impossible to detect. The instrument, described in a paper published in Nature, fills a specific observational gap that has frustrated condensed matter physicists for decades: the inability to watch what happens at the atomic scale inside a superconductor while it is actively in its superconducting state.

Superconductors are materials that conduct electricity with zero resistance below a certain critical temperature. Understanding exactly how electrons pair up and flow without losing energy is one of the central problems in modern physics. The short answer is that we know it happens, and we have a theoretical framework called BCS theory that explains it for conventional superconductors, but high-temperature superconductors, which work at much warmer temperatures, remain poorly understood at a fundamental level.

Why terahertz light was the right tool for this problem

Terahertz radiation sits between microwave and infrared light on the electromagnetic spectrum, with frequencies between 0.1 and 10 terahertz. Its particular value for studying superconductors comes from energy matching. The quantum energy gaps and collective oscillation modes inside superconductors happen to correspond to the energy range that terahertz photons carry. Visible light is too energetic and disrupts the delicate quantum states being studied. Terahertz light interacts with those states without destroying them.

What the MIT team built is a near-field terahertz microscope. Standard terahertz imaging systems are limited by diffraction to a spatial resolution of roughly 300 micrometres, which is too coarse to see what is happening at the scale of individual atomic domains inside a material. Near-field microscopy gets around the diffraction limit by bringing an extremely sharp metal tip within a few nanometres of the sample surface and using the tip to confine the terahertz field to a spot size determined by the tip's geometry rather than the light's wavelength. The MIT instrument achieves spatial resolution of approximately 20 nanometres while maintaining terahertz frequency sensitivity.

What the microscope actually observed

The team used the microscope to study a cuprate superconductor, a class of copper-oxide materials that become superconducting at temperatures up to around 135 kelvin, far warmer than conventional superconductors but still extremely cold by everyday standards. Cuprates have been known since 1986 but their mechanism of superconductivity has never been fully explained.

The microscope revealed spatial variations in the terahertz response across the surface of the material at a resolution that shows individual nanometre-scale domains with different quantum properties. These variations had been theoretically predicted but never directly imaged. The data showed that the material is not uniformly superconducting at the nanoscale. Instead it contains patches where electron pairing behaviour differs measurably from neighbouring regions, a heterogeneity that standard bulk measurement techniques average out and therefore cannot detect.

Why nanoscale heterogeneity in superconductors matters

One of the persistent puzzles in cuprate superconductivity research is why these materials have a so-called pseudogap state above their superconducting transition temperature, where some electron pairing seems to occur but full superconductivity does not. Several competing theoretical models exist to explain the pseudogap, and distinguishing between them requires exactly the kind of spatial, nanometre-resolution data that the MIT microscope can now provide.

Professor Nuh Gedik, the senior author on the MIT paper, noted in an accompanying press release that the ability to map terahertz responses at nanometre resolution opens direct access to the spatial structure of quantum order parameters that theorists have been arguing about for nearly four decades. That is a specific claim with specific implications: if the microscope can definitively map how electron pairing varies across nanoscale domains in the pseudogap phase, it gives theorists data that can actually falsify competing models rather than simply being consistent with all of them.

Potential applications beyond fundamental research

Room-temperature superconductivity remains one of the most sought-after goals in materials science. A material that conducts electricity without resistance at ambient conditions would eliminate most of the energy losses in electrical transmission infrastructure, which globally account for approximately 8 to 10 percent of all electricity generated. The terahertz microscope does not by itself bring that goal closer, but it gives researchers a way to test whether candidate materials actually behave as expected at the quantum level, which is a bottleneck that has slowed progress.

In quantum computing, superconducting qubits are currently the dominant hardware platform used by companies including IBM, Google, and various national laboratories. The coherence times of those qubits, which determine how long a quantum computation can run before errors accumulate, depend on material properties at exactly the nanometre scale that the MIT microscope can now probe. Using the instrument to identify and eliminate sources of decoherence in qubit materials is a near-term application that the team has identified as a next experimental priority.

The current instrument operates at cryogenic temperatures, which is necessary both for studying superconductors and for maintaining the sensitivity of the terahertz detector. The MIT team is now working on a version that can operate at higher sample temperatures, which would allow the microscope to study a wider range of quantum materials including topological insulators and two-dimensional materials like twisted bilayer graphene, where nanoscale terahertz imaging has never previously been possible.