STM-D-0824Paper2007Published and peer-reviewed

Production of Ball-Lightning-Like Luminous Balls by Electrical Discharges in Silicon

Gerson Silva Paiva · Antonio Carlos Pavão · Elder Alpes de Vasconcelos · Odim Mendes · Eronides Felisberto da Silva

Summary and citation · read the original at the source

In one page

Ball lightning has always had a credibility problem: thousands of witnesses, almost no instruments. Gerson Paiva, Antonio Pavão and their colleagues at the Federal University of Pernambuco took the problem to a workbench. They struck an ordinary electric arc on a wafer of pure silicon, and glowing balls flew off it — orange-white, roughly golf-ball sized, spinning, bouncing, trailing smoke, and lasting for seconds rather than the fraction of a second a hot spark manages. The point of the experiment is its plainness. It uses no exotic driver, nothing that could not happen when lightning strikes silicon-bearing soil, and so it demonstrates the mechanism John Abrahamson and James Dinniss proposed in Nature in 2000: a strike reduces silica in the ground to silicon vapour, the vapour condenses into a fluff of nanometre-scale particles, and the ball glows for as long as that silicon burns back to oxide in the air. The natural phenomenon acquires a laboratory twin anyone can make.

Why it matters hereChapter 9 treats the long-lived luminous ball as a real object with a mechanism, and this is the paper that moved it from testimony to bench work — a repeatable recipe with a named chemistry behind it. It is also chapter 1 in miniature: an anomaly stops being an anomaly the day someone can make one on demand and hand the recipe to the next laboratory.

What it claims

01Electric arc discharges performed in pure silicon generate luminous balls with lifetimes on the order of seconds. That duration is the whole point — hot fragments thrown from an arc cool in about a second, so a glow that persists for several seconds is being fed by something other than the heat it left the electrode with.Published abstract, first sentence; companion 2010 paper, section 1
Published and peer-reviewed
02The laboratory balls reproduce several properties usually reported for natural ball lightning. In the authors’ follow-up description of the same apparatus, the objects are 1 to 4 centimetres across, live up to 8 seconds, spin, leave spiral smoke trails, bounce off surfaces, burn polystyrene on contact, ignite ethanol-soaked cotton, and decay leaving no trace behind.Published abstract, first sentence; companion 2010 paper, section 1 and Figures 1 and 3, on this site at /library/stm-e674ef992b
Published and peer-reviewed
03The experiment is deliberately austere: it does not rely on energy sources and excitation mechanisms that are improbable in the natural phenomenon. Where earlier laboratory analogues needed microwave cavities or sustained high-power drivers that lightning does not supply, this one needs a low-voltage arc struck on silicon — the follow-up paper gives 20 to 25 volts at 100 to 140 amperes, with the electrode lifted one to two millimetres — which is the kind of event a ground strike can plausibly produce.Published abstract, second sentence; companion 2010 paper, section 1, apparatus description
Published and peer-reviewed
04The result clearly demonstrates the role of vaporization and oxidation of silicon, as proposed by the Abrahamson-Dinniss theory for ball-lightning formation. In that picture a lightning strike reduces silicon dioxide in the soil to silicon vapour, the vapour condenses into a filamentary network of nanoparticles, and the object glows for as long as that finely divided silicon oxidises back to silica in air — chemistry supplying the persistence that a plasma alone cannot.Published abstract, second sentence
Published and peer-reviewed
05The energy carried by these balls is measurable, and was measured. Weighing the silica ash each ball leaves, reading its temperature from its own emission spectrum, and computing the heat of the silicon oxidation reaction with quantum chemistry, the same group returned 31.9 joules per ball and a mean energy density of 3.9 megajoules per cubic metre — inside the range other researchers infer from the damage natural ball lightning does.Companion 2010 paper, section 3, equations 1 and 2 and Table 1, on this site at /library/stm-e674ef992b
Published and peer-reviewed
06What to watch: the bench ball is a twin, not yet proof of identity, and two named measurements would tighten it. One is the electrostatic side — Stephan’s charge bound says a natural ball moving horizontally near grounded metal carries little net charge, so measuring the charge on a silicon ball in flight tests whether the two objects are alike in that respect as well. The other is the natural event itself: an optical spectrum of a real ball lightning showing silicon, iron and calcium lines from soil would close the loop the laboratory has opened.Published abstract, read against the Abrahamson-Dinniss mechanism it tests; Stephan’s bound on this site at /library/stm-c3a97e1fe2
What to watch

The way in

https://doi.org/10.1103/PhysRevLett.98.048501Physical Review Letters 98, 048501, published 26 January 2007, under the APS default licence. The record is closed: Unpaywall reports no open copy on 2026-09-08, the APS harvest interface returns not authorized for this DOI, and the green location Semantic Scholar lists at the Brazilian INPE repository serves no HTTPS endpoint. None of the paper is reproduced here. This page is written from the published abstract — read on 2026-09-08 both from the National Library of Medicine record, PubMed 17358820, and from OpenAlex, which agree word for word — together with the bibliographic record, and from the authors’ own follow-up paper on the same apparatus, Energy density calculations for ball-lightning-like luminous silicon balls, Physics-Uspekhi 53 (2010) 209, which is on this site at /library/stm-e674ef992b and was read for that sheet. Locators say which of the two each claim comes from. The authors wrote from the Departamento de Química Fundamental and the Departamento de Física, Universidade Federal de Pernambuco, Recife, with a co-author at INPE. When the Letter itself can be read, this sheet should be rewritten from it.

How to cite it

Gerson Silva Paiva, Antonio Carlos Pavão, Elder Alpes de Vasconcelos, Odim Mendes, Eronides Felisberto da Silva (2007) Production of Ball-Lightning-Like Luminous Balls by Electrical Discharges in Silicon. doi:10.1103/PhysRevLett.98.048501

Where it sits in the curriculum

Plasmoids, charge clusters and the orbs The evidence ladder