STM-D-1015Paper2003Published and peer-reviewed

Neutron emission from a fast plasma focus of 400 Joules

Patricio Silva · José Moreno · Leopoldo Soto · Lipo Birstein · Roberto E. Mayer · Walter Kies

Abstract and summary · read the original at the source · none found

In one page

A plasma focus is a beautifully plain fusion machine: two coaxial electrodes, a capacitor bank and a puff of deuterium. Fire the bank and a sheet of current peels off the insulator, races down the barrel, folds over the end of the anode and collapses onto the axis into a pinch hot and dense enough that deuterium nuclei fuse. Patricio Silva, José Moreno, Leopoldo Soto, Lipo Birstein, Roberto E. Mayer and Walter Kies, at the Chilean Nuclear Energy Commission, built one that stores only a few hundred joules — and made it fast, with an 880 nanofarad bank, 38 nanohenries of circuit inductance and a current that reaches its peak in about 300 nanoseconds. Charged to 30 kilovolts and filled with deuterium, it was read out with a silver activation counter, and the yield peaked at about 1.06 million neutrons a shot at 9 millibar. The lesson matters for anybody building on a bench: fusion-relevant plasma does not need a large machine, it needs a fast one.

Why it matters hereChapter 9 is about what a self-organising knot of plasma can be made to do, and this is the smallest end of the plasma-focus family — a device you can put on a table that still produces real fusion neutrons. Chapter 12 needs exactly that: a compact, high-density source, because a compact source is the enabler every downstream architecture on this site is waiting on.

What it claims

01The device is a fast, low-energy plasma focus: an 880 nanofarad capacitor bank charged to between 20 and 35 kilovolts, storing 176 to 539 joules, with a circuit inductance of 38 nanohenries and a current rise time of about 300 nanoseconds.Abstract, sentence 2
Published and peer-reviewed
02For discharges at a 30 kilovolt charging voltage the maximum neutron yield measured was 1.06 million plus or minus 0.13 million neutrons per shot, at a deuterium filling pressure of 9 millibar.Abstract, final sentence
Published and peer-reviewed
03The neutrons were counted with a silver activation counter and the total yield was mapped against deuterium filling pressure, so the optimum operating pressure of this machine is a measured curve rather than a parameter carried over from larger devices.Abstract, sentences 3 and 4
Published and peer-reviewed
04In a plasma focus a high pulsed voltage is applied to a low-pressure gas between coaxial cylindrical electrodes; the discharge starts over the surface of the insulator that partially covers the anode, the current sheath is then magnetically accelerated along the electrodes, and after it runs over their ends the plasma is compressed into a small cylindrical column. Best efficiency requires the pinch compression to coincide with peak current.Introduction, opening paragraph, as carried in the CoLab record for this DOI
Published and peer-reviewed
05Run on deuterium, a plasma focus produces deuterium-deuterium fusion reactions that emit fast neutron pulses near 2.5 megaelectronvolts, in bursts lasting tens to hundreds of nanoseconds, alongside beams of ions and electrons and ultrashort X-ray pulses. Because the source only exists while the machine is firing, it does not carry the storage and handling problems of an isotopic neutron source such as californium-252 or americium-beryllium.Introduction, opening paragraph, as carried in the CoLab record for this DOI
Published and peer-reviewed
06What to watch: the yield-versus-pressure curve of a hundreds-of-joules device is the datum the scaling argument turns on — whether neutron yield keeps tracking the same power of the discharge current all the way down into the tens of joules, where this same Santiago laboratory took the experiment next with PF-50J.Abstract, read against the group’s later tens-of-joules results
What to watch

Read it · abstract

Abstract

The neutron emission from a small and fast plasma focus operating in deuterium is presented. The system operates at low energy in the hundred of joules range (880 nF capacitor bank, 38 nH, 20–35 kV, 176–539 J, ∼300 ns current rise time). The neutrons were measured by means of a silver activation counter, and the total neutron yield versus deuterium gas filling pressure was obtained. For discharges operating at 30 kV charging voltage, the maximum neutron yield was (1.06±0.13)×10⁶ neutrons per shot at 9 mbar.

Patricio Silva, José Moreno, Leopoldo Soto, Lipo Birstein, Roberto E. Mayer and Walter Kies, all of the Comisión Chilena de Energía Nuclear, Casilla 188-D, Santiago, Chile, Neutron emission from a fast plasma focus of 400 Joules, Applied Physics Letters 83, issue 16, pages 3269 to 3271, 20 October 2003. The published article is at doi.org/10.1063/1.1621460.

(Abstract only — see the rights note above. On this site, the same Santiago group taking the same experiment down to tens of joules is at /library/stm-914ae7ba20 and /library/stm-0291726f47, their survey of what small pinch devices can do is at /library/stm-bf5c2a25ae, the runaway-electron model of hard X-rays from a small focus is at /library/stm-89b2dbcf3b, the Lee model code that the whole field fits its current traces with is at /library/stm-a9cc428c6d, and the PF-24 argon-doping neutron study is at /library/stm-ef8e9744ca.)

The way in

https://doi.org/10.1063/1.1621460SOURCE NOT REACHED — this page is written from the authors’ own abstract and the bibliographic record, and says so. Published as Applied Physics Letters volume 83, issue 16, pages 3269 to 3271, dated 20 October 2003, by the plasma physics group of the Comisión Chilena de Energía Nuclear in Santiago. Unpaywall reports a green open-access deposit of the submitted manuscript in the Chilean and Ibero-American repository network, but on 8 September 2026 none of the three mirrors answered: CONICET Digital returned a service-unavailable page, the LA Referencia handle 10533/174369 resolved to a 404 at the ANID repository, and the AECID Americanae record did not respond. The AIP article page and its PDF answer 403. The abstract below is the version deposited by AIP with Crossref, with the lost superscript in the neutron figure restored. Claims 1 to 3 and claim 6 are taken from that abstract; claims 4 and 5 are taken from the opening paragraph of the article’s own introduction as carried in the CoLab bibliographic record for this DOI. No Creative Commons statement appears in any record, so AIP’s copyright stands and nothing beyond the abstract is reproduced here.

How to cite it

Patricio Silva, José Moreno, Leopoldo Soto, Lipo Birstein, Roberto E. Mayer, Walter Kies (2003) Neutron emission from a fast plasma focus of 400 Joules. doi:10.1063/1.1621460

Where it sits in the curriculum

Plasmoids, charge clusters and the orbs Lattice confinement fusion