Silicides, in Fulgurites and Ferrosilicon

Unusual Compounds Formed under Reducing Conditions in Nature and Industry


This Rock of the Month feature is more detailed than the usual issue, as it illustrates two separate samples, and attempts to portray a small family of unusual minerals in three distinct settings.

The following images display a large fulgurite [a glassy mass formed by the melting of soil and sand in a lightning strike in southeast Ontario]; two photomicrographs of the silicide assemblage generated in the strike; and three images of silicides occurring in a sample of ferrosilicon [an artificial additive in the steel industry]. Finally, the chemistry of the silicides is considered briefly, and reference made to published descriptions of silicides in fulgurites, ferrosilicon and meteorites. Coincidentally, masses of ferrosilicon, found in unexpected locations, have on more than one occasion been mistaken for meteorites.

The fulgurite in which rare alloy beads were found was previously described here as Rock of the Month 37, July 2004.

fulgurite in section [58 kb]
Fig. 1. Here is a close-up image of an "arm" of the glass-filled fulgurite. Just below the illustrated face, a sawn section revealed silvery spheroids of curious alloys, ranging from tens of microns to 7 mm in diameter. Photo by Karyn Gorra.

"Rock of the Month # 44, posted February 2005" --- This sample (2220) was obtained by Mark Stanley of Sumar Minerals (Norwood, Ontario: Mark kindly made available small but critical fragments from this mass for ongoing research. The sample is from a large fulgurite developed in the Foymount area of southeastern Ontario.

Rare instances excepted, the silicides (compounds of silicon with other elements in the absence of oxygen) do not appear to occur naturally on Earth except as metastable and thus geologically short-lived products of extreme processes, such as lightning strikes. The enormous effective temperatures attainable through lightning strikes on scales of millimetres to tens of metres are able, where reducing conditions exist, to drive off the oxygen from quartz (silica) to leave metallic silicon, which may in turn react with iron to generate a range of iron silicides.


Essene and Fisher (1986) provided a detailed discussion of the formation of glassy fulgurite on a morainal ridge in southeast Michigan, U.S.A. Metallic micron- to cm-scale metallic globules rich in native Si are present, having unmixed from a silica-rich liquid. The Winans Lake fulgurite, 20 km north of Ann Arbor, was sampled in 1984. Masses of fulgurite were found at intervals over a 30-m length, along a low ridge of soil and Pleistocene glacial till. Notably, large metallic spheroids occur in dark glass near a charred tree root. The metallic assemblage includes native Si, Fe3Si7, FeSi, and FeTiSi2. Associated minerals include TiP and sub-micron blebs of native Au in skeletal crystals of Si. The fulgurite glass contains graphite, quartz and rare ZrO2 (baddeleyite). The fulgurite glass itself is markedly heterogeneous with 82-99 wt.% SiO2. Synthetic fulgurites can be generated by triggered lightning-strike experiments, leading to the formation of mm-scale silicate spheroids with metastable assemblages generated in events on time scales of µs to s (Rietmeijer et al., 1999).

silicides, 80X [54 kb] silicides, 160X [31 kb]

Fig. 2. Left: A rounded ovoid mass of silicides, 1.2x0.9 mm, displaying curving intergrowths of two phases. Large polished section, 80X magnification, long-axis field-of-view 1.4 mm, PPL-RL.

Fig. 3. Right: A close-up of silicide intergrowths in a large (6.8x5.5 mm) silvery spheroid of these highly-reduced phases, within the vesicular glass of the fulgurite. Large polished section, 160X magnification, long-axis field-of-view 0.7 mm, offset XP-RL.


The synthetic silicides featured here are from a sample recovered from an urban setting, in the northeastern part of Toronto, the largest city in Canada. The main fragment of sample 2222, shown below, is roughly 11.0x9.3x9.0 cm in size and weighs 5.9 kg, for a bulk specific gravity near 6.4. This is lowered somewhat by the presence of minor slag inclusions. Named silicides have higher calculated specific gravities, with several examples in the range 7.08 to 7.37.

ferrosilicon block [35 kb] ferrosilicon 2222a1.jpg [91 kb] ferrosilicon 2222b1.jpg [54 kb]

Fig. 4. Left: Note the pronounced flow lines and the pitted appearance due to the presence of tiny bubbles: features found also in common silicate slags, but very rarely observed to this degree in meteorites. The Canadian two-dollar coin has a diameter of 27 mm.

Fig. 5. Centre: Very fine-grained eutectoid intergrowths of iron silicides hosting a high concentration of equant crystals of a titanium phase: this is deduced to be the carbide, TiC **. 160X magnification, long-axis field-of-view 0.7 mm, PPL-RL. Polished mount 2222-A.

Fig. 6. Right: Iron silicide intergrowths and more of the TiC crystallites. 160X magnification. Polished mount 2222-B.

** The electron microprobe I used at the time could not detect the characteristic x-rays of light elements such as C and O (Z<9). Thinking overmuch on normal terrestrial rocks, I imagined the crystals would be the dioxide, TiO2. However, TiO2 in nature occurs in 3, and as many as 5 polymorphs. The most common is rutile (tetragonal), followed by anatase (tetragonal) and brookite (orthorhombic). The crystal habits tend to be prismatic to acicular. High-pressure polymorphs have since been identified, namely akaogiite (monoclinic) and "TiO2 II" (orthorhombic, not yet accepted as a distinct mineral species).

On reflection, the tiny, evidently cubic crystals in the silicide-dominated, highly reduced assemblage of phases in the ferrosilicon are very likely to be not an oxide but a carbide, TiC. TiC has a cubic close-packed structure, and is extremely hard and refractory. TiC occurs in nature as the extremely rare mineral khamrabaevite.

The iron silicides

Preliminary electron microprobe analysis suggests that the metal phase in the Foymount fulgurite is largely a mixture of Fe3Si7 and FeSi. Such phases were described from Michigan by Essene and Fisher (1986) but were not fully described, thus achieving official mineral status, for a further two decades. They were eventually characterized. They may be metastable, in terms of millennia to millions of years, but then there are quite a few water-soluble minerals, and common water-ice is itself classified a mineral, so silicides too should qualify!

fulgurite and ideal silicides [8 kb]

Fig. 7. This chart displays the ideal compositions of six hypothetical iron silicides, and actual electron microprobe data for two silicides analysed in a large bead of melted alloys in the large fulgurite. Mineral names are provided where known. Another example is perryite (Fe5Si2), ideal formula 16.75 wt.% / 28.57 at.% Si, plotting near the top left of the chart, near gupeiite / suessite.

Stefano et al. (2020) describe another large fulgurite from Michigan, formed in 2014. The fulgurite glass contained small (up to 0.2 mm) globules of iron silicides. By this time, the phases with compositions shown in the above chart had been named. Compositionally (we have no crystallographic measurements for the Foymount sample) the silicides may be identified with luobusaite (Fe3Si7) and naquite (aka fersilicite, FeSi). It is worth noting that these two Fe silicides were first fully described as fine intergrowths in the Luobusa ophiolite from Tibet, part of a belt of deep ophiolitic assemblages enriched in minerals both highly reduced and high-pressure in origins. This contrasts with the high-energy, high-temperature of fulgurite formation on and immediately beneath the Earth's surface.

Other silicides include linzhiite (FeSi2), xifengite (Fe5Si3) and gupeiite (Fe3Si). Other silicides have been described in nature and generated in materials science, as in the semiconductor industry, but the iron silicides are the most diverse and widespread, presumably because of the great abundance of these two elements.


Perhaps coincidentally, Lloyd Montgomery Pidgeon, who in 1941 pioneered the Pidgeon process for the production of Mg metal by combining dolomite with ferrosilicon, was born in Markham, Ontario in 1906 (Hoy-Petersen et al., 1990; Anon, 1999; Avedesian and Baker, 1999). Markham lies just northeast of Toronto, and thus is not far from the site of the second sample featured here, which was initially mistaken for a meteorite. Berzelius was the first to isolate Si in 1808: within two years it was realized that Si could be used to harden steel. Ferrosilicon was first produced on an industrial scale in 1889 (Committee on Raw Materials, 1987), the year in which the first modern study of FeSi alloys was published.

Ferrosilicon is used as a graphitization accelerator in cast iron manufacture, and as an alloying element, deoxidizer and heat source in steel making. It also reduces Cr oxides in slag, returning the Cr metal to the steel. The bulk compositions of ferrosilicon vary with the application, and as a function of time, as various industrial processes were discarded and others adopted (Committee on Raw Materials, 1987). Thus the precise identification of the likely origin of a given piece of ferrosilicon would entail significant knowledge of the industry.

Beyond the realm of metallurgy, the high density of ferrosilicon renders it useful in dense media separation, tested for materials as diverse as diamond -bearing rocks and manganese ores. Given the odd places where ferrosilicon "meteorwrongs" appear, one wonders whether other uses have been assigned to it, e.g., ballast in boats.

Ferrosilicon compositions vary according to the purpose. One bulk analysis (Morsi et al., 2002) of ferrosilicon is as follows; Si (74.8%), Fe (23.36%), C (0.09%), Ca (0.31%), S (0.003%), P (0.031%) and Al (1.41%, total 100.004%). The silica raw material for ferrosilicon production is typically quartz sand or quartzite: industry standards are set for common impurities, including iron and aluminium. The main raw materials of ferrosilicon are quartzite and reductants - Fe is added as steel scrap or high-grade Fe ore or pellets. The raw materials and the processing impart a variety of impurities. Thus SiC inclusions typically contribute 0.15-0.20% C in the product (Committee on Raw Materials, 1987). Sample 2222 has a bulk composition near 75 wt.% Si, 18 wt.% Fe and 7 wt.% Ti, plus traces of P, S and other elements.

Silicides in Meteorites

Silicides occur in particularly reduced meteorite groups. Thus it is no surprise that they occur in some ureilites, those unusual carbonaceous achondrites. Suessite, nickeliferous Fe3Si, occurs in the North Haig ureilite (Rubin, 1997). The Dar Al Gani 319 polymict ureilite is a complex brecciated achondrite with 7 recognized classes of clasts (Ikeda et al., 1990). It contains a range of alloys, phosphides and silicides, the latter including suessite and perryite (Fe5Si2)

Silicides occur also in enstatite chondrites. EH chondrites are characterized by Si enrichments in kamacite, the silicide perryite, and alkali metal content in sphalerite and chalcopyrite (Lin and El Goresy, 2002). As an example, the Qingzhen unequilibrated enstatite chondrite contains perryite with 12.1-13.4 wt.% Si.

Note that ferrosilicon, being heavy, of metallic lustre, and in some cases displaying signs of a molten past, occurs as an unusual class of "meteorwrong". The display at the Natural History Museum (London, England) includes some fine meteorwrongs, including brassy ferrosilicon and a large "iron bear" (a mass of metal from an old furnace, enclosing a cm-scale ovoid blob of dark slag).


Based on the available evidence, it is likely that large, kg-size masses of silicide are almost certainly industrial ferrosilicon, whereas small, mm- to cm-scale silicide spherules are quite likely of fulgurite origin .

Slags in general are poorly documented, being waste products rather than the desired end-products of industrial processes. Ferrosilicon is for the most part an intermediate commodity in industry, rather than an end-product. One may wonder whether peculiar variants, such as sample 2222, may not be generated from time to time in failed smelting ventures.

An interesting report on silicides from the southern Urals of Russia (Basu et al., 2000) describes spherical to irregular masses of unusual mineral assemblages, 0.1 to 10 mm in diameter, found in layers of eroded Pleistocene sediments. The minerals included Fe silicides with carbides of Ti, Fe and Si. Based on the data presented herein, and the findings of Essene and Fisher (1986), we may wonder whether these minerals, and similar alluvial occurrences elsewhere, represent eroded remnants of fulgurites, as opposed to meteoritic materials.


In addition to those mentioned above, it should be said that the ferrosilicon mass described here was recovered by John Rucklidge and Richard Herd; that the bulk analysis of a chip of the material was provided by Mike Gorton; and that this note is a work in progress. The text was partially upgraded in December 2023: the literature on silicides has expanded greatly since the original article was written in 2005, but this description remains a factual record.


ANON (1999) Lloyd Montgomery Pidgeon. Northern Miner 85 no.43, 5.

AVEDESIAN,MM and BAKER,H (editors) (1999) Magnesium and Magnesium Alloys. ASM Specialty Handbook, ASM International, 314pp.

BASU,S, MURTY,SVS, SHUKLA,PN and SHUKLA,AD (2000) Origin of silicides (unknown meteorites?). Meteoritics & Planetary Science 35, A22-23.

COMMITTEE ON RAW MATERIALS (1987) Ferro-silicon and the Steel Industry. International Iron and Steel Institute, Committee on Raw Materials, Brussels, 89pp.

ESSENE,EJ and FISHER,DC (1986) Lightning strike fusion: extreme reduction and metal-silicate liquid immiscibility. Science 234, 189-193.

HAUGEROD,T and SKANE,O (1990) Magnesium. In `Ullmann's Encyclopedia of Industrial Chemistry' (Elvers,B, Hawkins,S and Schulz,G editors), 5th edition, volume A15, 559-580.

HOY-PETERSEN,N, AUNE,T, VRALSTAD,T, ANDREASSEN,K, OYMO,D, HAUGEROD,T and SKANE,O (1990) Magnesium. In "Ullmann's Encyclopedia of Industrial Chemistry" (Elvers,B, Hawkins,S and Schulz,G editors), 5th edition, volume A15, 559-580.

IKEDA,Y, PRINZ,M and NEHRU,CE (2000) Lithic and mineral clasts in the Dar Al Gani (DAG) 319 polymict ureilite. Antarctic Meteorite Research 13, National Institute of Polar Research, Tokyo, 349pp., 177-221.

LIN,Y and EL GORESY,A (2002) A comparative study of opaque phases in Qingzhen (EH3) and MacAlpine Hills 88136 (EL3): representatives of EH and EL parent bodies. Meteoritics & Planetary Science 37, 577-599.

MORSI,IM, EL BARAWY,KA, MORSI,MB and ABDEL-GAWAD,SR (2002) Silicothermic reduction of dolomite ore under inert atmosphere. Can.Metall.Q. 41, 15-28.

RIETMEIJER,FJM, KARNER,JM, NUTH,JA and WASILEWSKI,PJ (1999) Nanoscale phase equilibrium in a triggered lightning-strike experiment. Eur.J.Mineral. 11, 181-186.

RUBIN,AE (1997) Mineralogy of meteorite groups. Meteoritics & Planetary Science 32, 231-247.

STEFANO,CJ, HACKNEY,SA and KAMPF,AR (2020) The occurrence of iron silicides in a fulgurite: implications for fulgurite genesis. Can.Mineral. 58, 115-123.

Written by Graham Wilson, posted on 17 February 2005, format upgraded 19 May 2013, content added 14,26 March 2020,
format upgraded and select detail added on 24,29-31 December 2023

Visit the Turnstone Meteorite Index

or the broader "Rock of the Month" Archives!

or the detailed "Rock of the Month Index"

Classification of Sample Descriptions available on this Site

Class/Group/Family Topics
The "Rock of the Month"
Fulgurites --- #37 --- Fulgurite from Ontario
Fulgurites --- #44 --- Silicides in fulgurite and ferrosilicon (this page)
Fulgurites --- #123 --- Fulgurite, Grant county, Nebraska, U.S.A.
Fulgurites --- #124 --- Fulgurite, Florence, Arizona, U.S.A.
Fulgurites --- #124 --- Fulgurite, Oswego, New York, U.S.A.