In a Yonsei University lab outside Seoul, a polymer ribbon the size of a paperclip is wrapped around a rat’s sciatic nerve. The leg has just been wounded in a controlled, post-surgical sort of way. A handheld ultrasound wand presses against the skin above the wound. The two faces of the ribbon, made of materials with opposite electrical affinities, shiver against each other a few hundred thousand times a second. A tiny alternating field flickers along the nerve. The rat does not limp.

A few weeks later, the ribbon is gone. It has dissolved into the surrounding tissue, and the rat goes about its business. The paper landed in Nature Biomedical Engineering on January 9. The team, from Yonsei University, Sungkyunkwan University and Samsung Medical Center, pitched the device as a drug-free alternative to opioids for postoperative pain. They had also tested it in pigs.

The physics powering all of this is the same thing that makes your sweater shock you when you pull it over your head.

A failed gadget tech, looking for a real job

Triboelectric nanogenerators were not supposed to end up in nerves. They were discovered, more or less by accident, in 2011, when graduate students in Zhong Lin Wang’s lab at Georgia Tech were studying piezoelectric materials and noticed that two pieces of polymer, rubbed together, produced an unreasonable amount of voltage. A fingernail-sized device could put out a few milliwatts. Enough, Wang would later say, to run a pacemaker.

For roughly a decade that line was the whole pitch. Shoes that charge phones. Floor tiles that power LEDs as commuters walk on them. Smart fabric that scavenges energy from a billowing shirt. None of it became a product. The conversion efficiency was decent in the lab and terrible in a sweaty insole. The output impedance was so high you needed bespoke power-management circuitry for trivial amounts of harvested energy. Chemical & Engineering News asked, in 2021, whether the technology would ever find its niche.

The patent record tells the same story from a different angle. There are 201 US utility grants since 2010 with “triboelectric” in the title. Of those, exactly one before 2024 mentioned a nerve, a neurostimulator, an electroceutical or an implant: a 2010 cardiac sensor filed by Medtronic that essentially used the effect as a strain gauge. Then, starting in January 2024, five more issued in rapid succession. Every single one describes an ultrasound-driven generator sealed inside a medical implant. Every single one comes from South Korea.

What changed: the receiver, not the transmitter

The trick is that nobody had thought of triboelectric generators as the receiving end of an acoustic link.

Focused ultrasound is a clinical workhorse. It is what destroys uterine fibroids without an incision and what HistoSonics is using to slosh liver tumors apart with cavitation bubbles. The wand sits in the cabinet of every radiology department. What the Korean consortium realized is that two dissimilar polymers, sealed in a hermetic shell with a thin gas layer between them, are a near-perfect receiver for it. The ultrasound flexes the shell, the polymers slap together hundreds of thousands of times a second, electrons hop from one to the other, and the device delivers a usable voltage straight into a stimulation circuit. No battery. No inductive coil whose efficiency falls off a cliff a few centimeters into tissue. No photovoltaics, which are useless inside a human.

A 2024 paper out of the same Yonsei group, in Advanced Materials Technologies, made the engineering case explicitly. They built the implant shell out of polyether ether ketone (PEEK), the medical-grade polymer used in spinal-fusion cages, because its acoustic impedance is a near-match for soft tissue. Transmission efficiency, they measured, was 99.94%. The implant produced 11.5 V in water and 8.75 V through biological tissue, comfortably within the power budget of a peripheral nerve stimulator. It held that output for more than five hours of continuous ultrasound.

The 2023 prerequisite paper, in Advanced Materials, came from a wider consortium: Sungkyunkwan, POSTECH, Yonsei, Gachon and a Suwon-based startup called Energymining. They sealed the same device inside a titanium can, which absorbs most ultrasound, and showed it still drove a stimulator strong enough to treat overactive bladder in a rat model. The point was to prove the receiver works even when the package is hostile.

The patent layer

Energy Mining Co., Ltd., based in Suwon, holds three of the five recent US grants. The first, US 11,878,190, issued in January 2024, covers a triboelectric generator made from a biodegradable nano-composite whose dissolution can be triggered, at a chosen time, by a separate focused-ultrasound pulse, plus the neurostimulation therapy device that uses it. The second, US 12,186,565, issued in January 2025, claims the titanium-packaged ultrasonic version. The third, US 12,609,556, issued in April 2026, covers a multi-layered architecture for the same family of devices. A fourth grant from April 2026, US 12,609,638, is held by Sungkyunkwan University and adds a ferroelectric composite layer to boost output. The fifth, US 12,355,372, issued in July 2025, belongs to the Daegu Gyeongbuk Institute of Science and Technology and covers a rotational variant for nerve stimulation.

The point is not that any one of these will dominate on its own. The point is that the same handful of Korean institutions and one Korean startup now own the foundational US patent positions for a specific class of medical implant that no Western company has filed against. As of last week, the OpenAlex corpus contains 1,655 papers from 2025 alone with “triboelectric nanogenerator” in the indexed text, up from four papers in 2012. The literature has been compounding for a decade. The patents have just started to catch up, and they are catching up in one country.

Why an R&D director should care

The implant market this lands in is not small. Spinal-cord stimulators and sacral neuromodulators together are a multi-billion-dollar category dominated by Medtronic, Abbott, Boston Scientific and Nevro, all of whom sell devices that get explanted and replaced every few years when their primary cells run down. A device with no battery, that can be made to dissolve when treatment is over, and that a patient charges by holding a sonogram wand against their skin would not be a product update. It would be a new category.

It is also a different cost structure. The bill of materials for a biodegradable polymer ribbon is measured in dollars. A primary-cell-powered stimulator is measured in hundreds. The Korean groups are not coy about where they are pointing this: a postoperative-pain device that displaces opioids in surgical recovery, a bladder-control device that does not need a generator pocket cut into the abdomen, and eventually anything else where a few hours a day of nerve stimulation would beat a permanent implant.

There is plenty here for a skeptic to flag. The opioid framing is doing work in the grant-writing. “No adverse effects during full resorption” is not yet a regulatory pathway. The animal models are still small. None of the five US patents has a corresponding human trial registered on ClinicalTrials.gov. But the engineering question—can you build a battery-free, ultrasound-driven implant that delivers useful voltage and disappears when you are done with it—has now been answered five different ways, in print, by overlapping groups, with patent claims to match.

The same physics that has failed to find a market for fifteen years was, it turns out, just waiting for focused ultrasound to grow up.

Method note

Patent figures come from a curated copy of US utility grant XML from the USPTO, covering grants published through May 19, 2026. The “triboelectric” filter is a title-string match across all utility grants. The medical subset filters those titles for “nerve”, “neuro”, “electroceutical”, “implant”, “cardiac” or “pacemaker”. Assignees were verified one by one against the USPTO record. The literature counts come from OpenAlex, full-text search across roughly 357 million scholarly works. The Yonsei rat-and-pig nerve-block result is Wireless and bioresorbable triboelectric nerve block system for postoperative pain control, Kim et al., Nature Biomedical Engineering, January 9, 2026. The PEEK-shelled implant paper is Jeon et al., Advanced Materials Technologies, 2024 (DOI 10.1002/admt.202400317). The capacitance-matched bladder stimulator paper is Kim et al., Advanced Materials, 2023 (DOI 10.1002/adma.202307194). The 2014 breathing-driven pacemaker result is Zheng et al., Advanced Materials, 2014 (DOI 10.1002/adma.201402064). Every specific number in this post can be reproduced from the citations above.