Neutron Transmutation Doping

Neutron transmutation doping, or NTD, turns a silicon wafer into its own dopant source by firing it with a reactor's neutron flux and letting nuclear physics do what no diffusion furnace or ion implanter can — convert silicon atoms directly into phosphorus, one nucleus at a time, with a uniformity no chemical doping method can match.

Natural silicon is already carrying the dopant it needs — just in the wrong element. About 3.1 percent of ordinary silicon atoms are the isotope silicon-30, and when one of those nuclei captures a thermal neutron it becomes silicon-31, an unstable isotope that beta-decays with a 2.6-hour half-life straight into phosphorus-31 — a completely ordinary, stable, n-type dopant atom sitting exactly where the silicon atom used to be. No implant, no diffusion, no masking: the dopant is manufactured in place, atom by atom, throughout the entire volume of the ingot.

NTD: Doping Silicon With a Reactor a neutron converts silicon-30 into phosphorus-31, in place, atom by atom 1 · Si-30 nucleus Si-30 neutron 3.1% of natural silicon is Si-30 inside a reactor core or neutron source 2 · Si-31, unstable Si-31 2.6-hour half-life beta decay electron emitted 3 · P-31 donor P-31 stable, n-type, exactly where the Si atom sat neutron flux penetrates a whole ingot evenly — no concentration gradient to fight a post-irradiation anneal repairs lattice damage and activates the new phosphorus atoms

The payoff is uniformity no chemical doping step can touch. Melt-grown Czochralski crystals carry resistivity striations from the growth process itself, and even a careful ion implant or diffusion can only control dopant concentration from the surface inward, leaving radial and axial gradients across a thick wafer. A neutron beam, by contrast, passes through the entire thickness of a boule with almost no attenuation, so the phosphorus it creates lands with the same density everywhere the flux reached — resistivity uniformity within a few percent across an entire ingot, rather than the much looser spread typical of as-grown or diffused material.

NTD is almost exclusively a float-zone silicon technique, for a reason. Float-zone crystal growth — pulling the ingot through a localized RF-heated molten zone with no crucible touching the melt at all — produces silicon with none of the oxygen contamination a quartz Czochralski crucible introduces, and it's also grown with phosphorus deliberately left out, specifically so NTD can add it back under full control afterward. That combination of crucible-free purity and intentionally undoped starting material is what makes float-zone the material of choice for the high-voltage power devices — thyristors, power diodes, high-power IGBTs — that actually need NTD's resistivity precision; these parts conduct thousands of amps and block kilovolts, where a resistivity variation that a logic chip would never notice directly changes breakdown voltage and current-sharing across a large-area device.

The irradiation step has to happen at a reactor, which shapes who uses it. Research or production reactors capable of delivering a controlled, uniform thermal-neutron dose to a full ingot are a specialized, limited resource, so NTD wafers typically travel to a dedicated irradiation facility and back before final device processing — an extra logistics step that ordinary logic wafers never take, and one more reason NTD stays confined to the power-semiconductor niche where its resistivity control is worth that detour.

Read NTD against Dash necking and COPs as the same crystal-physics story told a third way: where Dash necking removes a defect and COP formation is one the crystal creates on its own, NTD is the fab reaching back into the ingot after the fact and using nuclear physics, not chemistry, to fix a uniformity problem no later process step could solve.

Take neutron transmutation doping further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app