The + anode is deliberately tiny — a thin wire or sharp edge. The intense electric field at that small radius (~30 kV across a few cm gap) exceeds air's breakdown threshold, ripping electrons off nearby N₂ and O₂ molecules. This creates a cloud of positive ions near the anode.
The electric field between the electrodes accelerates the N₂⁺ ions toward the − cathode. Because the cathode is a large flat plate (or foil skirt), its surface field is weak — no secondary corona forms there. The ions gain kinetic energy as they cross the gap.
As the ions race toward the cathode, they collide repeatedly with neutral air molecules (the mean free path in air is only ~70 nm). Each collision transfers momentum. Billions of these micro-collisions per second create a macroscopic ionic wind flowing from anode → cathode.
Air is pushed toward the cathode. By Newton's third law, the whole device experiences a net thrust in the opposite direction — toward the small anode. This is real, measurable force. Typical lifters produce ~1–5 mN/W, enough to levitate lightweight balsa/foil structures.
Dr. Charles Buhler — NASA's lead electrostatics expert at Kennedy Space Center — and his co-founder Andrew Aurigema claim to have tested ~2,000 device variations over a decade. Their setup uses stacked conductive plates with dielectrics, energized at 30–40 kV. They report millinewton-scale forces that persist in vacuum chambers (~10⁻⁶ torr), which they say eliminates ionic wind as an explanation. Buhler frames the theoretical basis as third-order QED perturbation theory — momentum transfer from interactions with quantum vacuum fluctuations. They've evolved from simple asymmetric capacitors to thin-film and liquid-coated surface configurations. Most provocatively, they claim thrust persists after power is switched off.
At 10⁻⁶ torr, there are still ~10¹⁰ molecules per cm³. For millinewton-level forces from a 30+ kV device, you don't need much residual gas to generate ionic wind. Outgassing from dielectrics, styrofoam (used in non-vacuum tests), and epoxy under high voltage can continuously replenish the local gas pressure near the electrodes. This is exactly the failure mode that killed the EmDrive — Tajmar et al. showed that what NASA Eagleworks measured as "thrust" vanished in properly controlled vacuum tests. Buhler's team acknowledges styrofoam explodes in vacuum and use different configs there, but the vacuum quality and gas monitoring are not independently verified.
High-voltage devices create strong electric fields that extend well beyond the device itself. These stray fields can exert forces on the vacuum chamber walls, the torsion wire, the support structure, or any grounded surface nearby. This produces real, measurable forces on the test apparatus that look like thrust but are actually the device pushing against its own test fixture. Even Faraday shielding doesn't fully eliminate this at 30+ kV — fringe fields leak through gaps and connections. This is one of the most common confounders in electrostatic thrust measurements and requires extremely careful null-testing that hasn't been independently verified here.
High-voltage dielectrics dissipate energy as heat (dielectric loss). Thermal gradients create convection currents (even in low-pressure gas) and cause mechanical warping of the device and its mount. A few millinewtons is a tiny force — the weight of a fraction of a paperclip. Radiometric effects (thermal transpiration in partial vacuum) can produce forces of this magnitude. The claim that thrust persists after power-off is especially suspicious: it's consistent with a thermal artifact cooling slowly, or with stored charge in dielectrics producing decaying electrostatic coupling.
The EmDrive followed a nearly identical trajectory: credentialed researchers, small anomalous forces, vacuum tests, theoretical frameworks invoking novel physics. When Martin Tajmar's group at TU Dresden conducted rigorous tests with proper controls (including a "null" device that shouldn't produce thrust but did), the signal vanished — it was interaction between the device's power cables and Earth's magnetic field. Buhler's work has been presented only at APEC (a conference for alternative propulsion enthusiasts) and in popular media — no peer-reviewed paper with full methodology has been published, and no independent replication has been reported.