A preprint reports a topological surface-acoustic-wave chip that registered concentration-linked changes in salt and glucose solutions while using 0.04 μL for a single measurement. Its resonance frequency—the frequency at which the device’s acoustic response is strongest—shifted as the liquid concentration changed.
At a 0.01% concentration variation, the reported frequency shift was 2 kHz for NaCl and 1 kHz for glucose. The work is a laboratory demonstration and contains no human or animal data.
A chip built to trap sound
The researchers constructed valley-Hall topological phononic crystals, coupled a straight waveguide to a resonant cavity, and examined the liquid-loaded device through numerical simulations and transmission measurements.
The modeled device used a 128° Y-cut LiNbO₃ substrate, holes 4 μm deep and a 15 μm-high liquid tank. The fabricated chip had a 1.8 mm × 1.2 mm footprint and interdigital transducers at both waveguide ends.
Here, the paper’s use of “topological” refers to a calculated contrast between two unit-cell types: Type-A and Type-C showed opposite energy-flow directions. The authors identify this pattern as a signature of valley modes and a topological phase transition.
Two computational approaches were reported to produce closely agreeing maps of the device’s wave behavior. The calculated interface-state fields concentrated acoustic energy near the solid–liquid interfaces, the substrate surface and the holes instead of spreading through the liquid bulk.
The signal followed concentration
Under liquid loading, the measured transmission frequency was 29.51 MHz, compared with 30.22 MHz in simulation—a 2.4% difference. The paper attributes the gap to liquid loading and ambient temperature, but reports no uncertainty interval.
For NaCl solutions from 0% to 2%, higher concentration corresponded to a higher resonance frequency. Experimentally, the frequency moved from 29.12 to 29.50 MHz, giving an average sensitivity of 190 kHz per percentage point; the simulation predicted 30.53 to 30.95 MHz and 210 kHz per percentage point, with a 4.84% simulation–experiment deviation.
For glucose solutions from 1% to 10%, the reported average sensitivity was 101.8 kHz per percentage point in experiment and 109.4 kHz per percentage point in simulation, a 7.5% difference. Higher concentration again corresponded to a higher resonance frequency.
A prototype, not a clinical test
The report does not state how many devices or measurements were used, how many repeats were made, or provide inferential tests, confidence intervals or standard deviations. That makes the repeatability of the reported shifts difficult to judge.
The tests used deionized water and aqueous NaCl and glucose solutions. There was no matched conventional or non-topological surface-acoustic-wave comparator, so the study does not establish that the topological design performs better than other sensor designs or that it has clinical diagnostic accuracy.
Localization and reduced radiation loss were inferred from calculated fields, and the study did not test topological protection with a matched disorder challenge or a long-term experiment.
The authors present the device as a potential basis for stable micro-volume surface-acoustic-wave biosensing chips, but biomedical application remains prospective. Performance in complex biological matrices, the effects of controlled temperature and liquid-layer thickness, and repeatability across independently fabricated devices still need to be evaluated.
Paper data and sources
Original title: Experimental Realization of Topological Surface Acoustic Wave Resonances under Surface Liquid Loading for Micro-volume sample Sensings
Authors: Bowei Wu, Tingfeng Ma, Hanbang Deng, Shuanghuizhi Li
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-20
DOI: Not available
Original paper · Full text