Science1 publisher3 min readPublished
University of Twente puts a SQUID on a silicon pyramid to bring it right up to the sample
University of Twente researchers built a SQUID sensor on a silicon pyramid tip that still works above 1 tesla, about 20,000 times Earth's field. Etching the tip from silicon could let labs make these sensors in batches instead of by hand.
The Scientist · Science desk
What happened
- University of Twente researchers described a SQUID sensor built on the tip of a silicon pyramid in Physical Review Applied, with PhD candidate Thijs Roskamp as first author.
- The finished pyramid stands upright on the end of a thin, flexible arm that scans a surface the way a record needle follows a groove.
- The sensor keeps working above 1 tesla, roughly 20,000 times Earth's magnetic field, although strong fields normally destroy superconductivity.
- Only two of the four wires framing the tip are needed to measure; the other two let the team steer the sensor and tune its sensitivity during a scan.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- capability If the apex sits as close to the sample as described, scanning SQUID maps could keep nanoscale detail that a several-micrometer gap erases for flat chip sensors.
- capability Quantum-materials experiments that need strong fields could keep a SQUID running past the point where the most sensitive sensors stop working.
- cost If wafer fabrication yields well, the cost of a SQUID tip drops from one specialist's hand labour per sensor to a share of one processed wafer.
The University of Twente has worked on scanning SQUID microscopy for years [16]. A SQUID, a superconducting ring that registers even the smallest magnetic changes, is sensitive enough to map where currents and fields run in quantum materials [1]. What limits it is distance. Magnetic fields weaken quickly away from their source, and a ring lying flat in a chip is held several micrometers off the sample by the chip around it. The Twente account says that gap is where the detail is lost [2].
"That is why we put the sensor on a pyramid," Hans Hilgenkamp said. "On top of that pyramid the sensor can be brought right up to the material we want to look at, with nothing else in the way." [3]
The pyramid is hollowed out of silicon. A liquid etch cuts pits shaped like upside-down pyramids. A very thin glass-like film laid over each pit ends up slightly thicker at the sharp edges, so when the film is removed, material remains only there, as a frame of fine wires running along the pyramid's edges [4]. Erwin Berenschot and Niels Tas developed this technique, corner lithography, in Twente's Mesoscale Chemical Systems group [5]. After the surrounding silicon is removed, the wires get a layer of niobium, and a beam of charged particles cuts two constrictions into the ring at the apex [7]. "We come from superconductivity, our colleagues from 3D nanofabrication," Roskamp said [8].
Of the two free wires on each tip, Roskamp said: "We can add functions to those without having to make the sensor all over again." [12]
The pyramid has to beat two different baselines. Against a flat chip SQUID, its advantage is height above the sample [2]. Other groups already make SQUID-on-tip sensors "like these" by pulling a glass tube into a fine needle and evaporating a superconductor onto it. The article describes that as handwork that yields one sensor at a time, with limited reproducibility and little room for anything extra on the tip [13]. Against those tips, Twente's case is about manufacture. The sensor is made in the MESA+ NanoLab, and phys.org frames the route as a move from handwork to a whole wafer [14]. I'd expect a steady supply of matched tips to matter more to a working microscope lab than any single record-setting tip.
The evidence supports a working geometry with nothing between the ring and the sample, made by a fabrication route aimed at wafers [3][7][14]. The account does not report a tip-to-sample distance, a spatial resolution, a flux sensitivity, a scan of an actual quantum material, or how far above 1 tesla the sensor keeps working [10]. Its headline says the sensor "could reveal" nanoscale magnetic patterns in quantum materials [15].
What to watch
- A published scan of a quantum material with the pyramid tip, stating tip height and spatial resolution next to a flat chip SQUID.
- Yield and device-to-device spread across a full wafer of pyramid SQUIDs from the MESA+ NanoLab.
- How far above 1 tesla the sensor keeps working, and how much sensitivity it retains there.