Banded iron formations (BIFs; also called banded ironstone formations) are distinctive units of sedimentary rock consisting of alternating layers of iron oxides and iron-poor chert. They can be up to several hundred meters in thickness and extend laterally for several hundred kilometers. Almost all of these formations are of Precambrian age and are theorized to record the oxygenation of the Earth's oceans. Some of the Earth's oldest rock formations, which formed about 3,700 million years ago (Ma), are associated with banded iron formations.
Banded iron formations are theorized to have formed in sea water as the result of oxygen production by photosynthetic cyanobacteria. The oxygen combined with dissolved iron in Earth's oceans to form insoluble iron oxides, which precipitated out, forming a thin layer on the ocean floor. Each band is similar to a varve, resulting from cyclic variations in oxygen production.
Banded iron formations were first discovered in northern Michigan in 1844. Banded iron formations account for more than 60% of global iron reserves and provide most of the iron ore presently mined.[1] Most formations can be found in Australia, Brazil, Canada, India, Russia, South Africa, Ukraine, and the United States.
Description
A typical banded iron formation consists of repeated, thin layers (a few millimeters to a few centimeters in thickness) of silver to black iron oxides, either magnetite (Fe3O4) or hematite (Fe2O3), alternating with bands of iron-poor chert, often red in color, of similar thickness.[2][3][4][5] A single banded iron formation can be up to several hundred meters in thickness and extend laterally for several hundred kilometers.[6]
Banded iron formation is more precisely defined as chemically precipitated sedimentary rock containing greater than 15% iron. However, most BIFs have a higher content of iron, typically around 30% by mass, so that roughly half the rock is iron oxides and the other half is silica.[6][7] The iron in BIFs is divided roughly equally between the more oxidized ferric form, Fe(III), and the more reduced ferrous form, Fe(II), so that the ratio Fe(III)/Fe(II+III) typically varies from 0.3 to 0.6. This indicates a predominance of magnetite, in which the ratio is 0.67, over hematite, for which the ratio is 1.[5] In addition to the iron oxides (hematite and magnetite), the iron sediment may contain the iron-rich carbonates siderite and ankerite, or the iron-rich silicates minnesotaite and greenalite. Most BIFs are chemically simple, containing little but iron oxides, silica, and minor carbonate,[6] though some contain significant calcium and magnesium, up to 9% and 6.7% as oxides respectively.[8][9]
When used in the singular, the term banded iron formation refers to the sedimentary lithology just described.[2] The plural form, banded iron formations, is used informally to refer to stratigraphic units that consist primarily of banded iron formation.[10]
A well-preserved banded iron formation typically consists of macrobands several meters thick that are separated by thin shale beds. The macrobands in turn are composed of characteristic alternating layers of chert and iron oxides, called mesobands, that are several millimeters to a few centimeters thick. Many of the chert mesobands contain microbands of iron oxides that are less than a millimeter thick, while the iron mesobands are relatively featureless. BIFs tend to be extremely hard, tough, and dense, making them highly resistant to erosion, and they show fine details of stratification over great distances, suggesting they were deposited in a very low-energy environment; that is, in relatively deep water, undisturbed by wave motion or currents.[3] BIFs only rarely interfinger with other rock types, tending to form sharply bounded discrete units that never grade laterally into other rock types.[6]
Banded iron formations of the Great Lakes region and the Frere Formation of western Australia are somewhat different in character and are sometimes described as granular iron formations or GIFs.[8][6] Their iron sediments are granular to oolitic in character, forming discrete grains about a millimeter in diameter, and they lack microbanding in their chert mesobands. They also show more irregular mesobanding, with indications of ripples and other sedimentary structures, and their mesobands cannot be traced out any great distance. Though they form well-defined, discrete units, these are commonly interbedded with coarse to medium-grained epiclastic sediments (sediments formed by weathering of rock). These features suggest a higher energy depositional environment, in shallower water disturbed by wave motions. However, they otherwise resemble other banded iron formations.[8]
The great majority of banded iron formations are Archean or Paleoproterozoic in age. However, a small number of BIFs are Neoproterozoic in age, and are frequently,[9][11][12] if not universally,[13] associated with glacial deposits, often containing glacial dropstones.[9] They also tend to show a higher level of oxidation, with hematite prevailing over magnetite,[11] and they typically contain a small amount of phosphate, about 1% by mass.[11] Mesobanding is often poor to nonexistent[14] and soft-sediment deformation structures are common. This suggests very rapid deposition.[15] However, like the granular iron formations of the Great Lakes, the Neoproterozoic occurrences are widely described as banded iron formations.[9][11][15][5][16][17]
Banded iron formations are distinct from most Phanerozoic ironstones. Ironstones are relatively rare and are thought to have been deposited in marine anoxic events, in which the depositional basin became depleted in free oxygen. They are composed of iron silicates and oxides without appreciable chert but with significant phosphorus content, which is lacking in BIFs.[12]
No classification scheme for banded iron formations has gained complete acceptance.[6] In 1954, Harold Lloyd James advocated a classification based on four lithological facies (oxide, carbonate, silicate, and sulfide) assumed to represent different depths of deposition,[2] but this speculative model did not hold up.[6] In 1980, Gordon A. Gross advocated a twofold division of BIFs into an Algoma type and a Lake Superior type, based on the character of the depositional basin. Algoma BIFs are found in relatively small basins in association with greywackes and other volcanic rocks and are assumed to be associated with volcanic centers. Lake Superior BIFs are found in larger basins in association with black shales, quartzites, and dolomites, with relatively minor tuffs or other volcanic rocks, and are assumed to have formed on a continental shelf.[18] This classification has been more widely accepted, but the failure to appreciate that it is strictly based on the characteristics of the depositional basin and not the lithology of the BIF itself has led to confusion, and some geologists have advocated for its abandonment.[3][6] However, the classification into Algoma versus Lake Superior types continues to be used.[19][20]
Occurrence
Banded iron formations are almost exclusively Precambrian in age, with most deposits dating to the late Archean (2800–2500 Ma) with a secondary peak of deposition in the Orosirian period of the Paleoproterozoic (1850 Ma). Minor amounts were deposited in the early Archean and in the Neoproterozoic (750 Ma).[6][5] The youngest known banded iron formation is an Early Cambrian formation in western China.[17] Because the processes by which BIFs are formed appear to be restricted to early geologic time, and may reflect unique conditions of the Precambrian world, they have been intensively studied by geologists.[6][5]
Banded iron formations are found worldwide, in every continental shield of every continent. The oldest BIFs are associated with greenstone belts and include the BIFs of the Isua Greenstone Belt, the oldest known, which have an estimated age of 3700 to 3800 Ma.[6][21] The Temagami[22] banded iron deposits formed over a 50-million-year period, from 2736 to 2687 Ma, and reached a thickness of 60 meters (200 feet).[23] Other examples of early Archean BIFs are found in the Abitibi greenstone belts, the greenstone belts of the Yilgarn and Pilbara cratons, the Baltic Shield, and the cratons of the Amazon, north China, and south and west Africa.[6]
The most extensive banded iron formations belong to what A.F. Trendall calls the Great Gondwana BIFs. These are late Archean in age and are not associated with greenstone belts. They are relatively undeformed and form extensive topographic plateaus,[3] such as the Hamersley Range.[24][25][26] The banded iron formations here were deposited from 2470 to 2450 Ma and are the thickest and most extensive in the world,[5][27] with a maximum thickness in excess of 900 meters (3,000 feet).[8] Similar BIFs are found in the Carajás Formation of the Amazon craton, the Cauê Itabirite of the São Francisco craton, the Kuruman Iron Formation and Penge Iron Formation of South Africa, and the Mulaingiri Formation of India.[6]
Paleoproterozoic banded iron formations are found in the Iron Range and other parts of the Canadian Shield.[6] The Iron Range is a group of four major deposits: the Mesabi Range, the Vermilion Range, the Gunflint Range, and the Cuyuna Range. All are part of the Animikie Group and were deposited between 2500 and 1800 Ma.[28] These BIFs are predominantly granular iron formations.[6]
Neoproterozoic banded iron formations include the Urucum in Brazil, Rapitan in the Yukon, and the Damara Belt in southern Africa.[6] They are relatively limited in size, with horizontal extents not more than a few tens of kilometers and thicknesses not more than about 10 meters (33 feet).[11] These are widely thought to have been deposited under unusual anoxic oceanic conditions associated with the "Snowball Earth."[3]
Origins
Banded iron formation provided some of the first evidence for the timing of the Great Oxidation Event, 2,400 Ma.[30][31] With his 1968 paper on the early atmosphere and oceans of the Earth,[32] Preston Cloud established the general framework that has been widely, if not universally,[33][34] accepted for understanding the deposition of BIFs.[6][5]
Cloud postulated that banded iron formations were a consequence of anoxic, iron-rich waters from the deep ocean welling up into a photic zone inhabited by cyanobacteria that had evolved the capacity to carry out oxygen-producing photosynthesis, but which had not yet evolved enzymes (such as superoxide dismutase) for living in an oxygenated environment. Such organisms would have been protected from their own oxygen waste through its rapid removal via the reservoir of reduced ferrous iron, Fe(II), in the early ocean. The oxygen released by photosynthesis oxidized the Fe(II) to ferric iron, Fe(III), which precipitated out of the sea water as insoluble iron oxides that settled to the ocean floor.[32][30]
Cloud suggested that banding resulted from fluctuations in the population of cyanobacteria due to free radical damage by oxygen. This also explained the relatively limited extent of early Archean deposits. The great peak in BIF deposition at the end of the Archean was thought to be the result of the evolution of mechanisms for living with oxygen. This ended self-poisoning and produced a population explosion in the cyanobacteria that rapidly depleted the remaining supply of reduced iron and ended most BIF deposition. Oxygen then began to accumulate in the atmosphere.[32][30]
Some details of Cloud's original model were abandoned. For example, improved dating of Precambrian strata has shown that the late Archean peak of BIF deposition was spread out over tens of millions of years, rather than taking place in a very short interval of time following the evolution of oxygen-coping mechanisms. However, his general concepts continue to shape thinking about the origins of banded iron formations.[3] In particular, the concept of the upwelling of deep ocean water, rich in reduced iron, into an oxygenated surface layer poor in iron remains a key element of most theories of deposition.[6][35]
The few formations deposited after 1,800 Ma[36] may point to intermittent low levels of free atmospheric oxygen,[37] while the small peak at 750 million years ago may be associated with the hypothetical Snowball Earth.[38]
Formation processes
The microbands within chert layers are most likely varves produced by annual variations in oxygen production. Diurnal microbanding would require a very high rate of deposition of 2 meters per year or 5 km/Ma. Estimates of deposition rate based on various models of deposition and sensitive high-resolution ion microprobe (SHRIMP) estimates of the age of associated tuff beds suggest a deposition rate in typical BIFs of 19 to 270 m/Ma, which are consistent either with annual varves or rhythmites produced by tidal cycles.[6]
Preston Cloud proposed that mesobanding was a result of self-poisoning by early cyanobacteria as the supply of reduced iron was periodically depleted.[30] Mesobanding has also been interpreted as a secondary structure, not present in the sediments as originally laid down, but produced during compaction of the sediments.[6] Another theory is that mesobands are primary structures resulting from pulses of activity along mid-ocean ridges that change the availability of reduced iron on time scales of decades.[39] In the case of granular iron formations, the mesobands are attributed to winnowing of sediments in shallow water, in which wave action tended to segregate particles of different size and composition.[6]
For banded iron formations to be deposited, several preconditions must be met.[14]
- The deposition basin must contain waters that are ferruginous (rich in iron).
- This implies they are also anoxic, since ferrous iron oxidizes to ferric iron within hours or days in the presence of dissolved oxygen. This would prevent transport of large quantities of iron from its sources to the deposition basin.
- The waters must not be euxinic (rich in hydrogen sulfide), since this would cause the ferrous iron to precipitate out as pyrite.
- There must be an oxidation mechanism active within the depositional basin that steadily converts the reservoir of ferrous iron to ferric iron.
Source of reduced iron
There must be an ample source of reduced iron that can circulate freely into the deposition basin.[6] Plausible sources of iron include hydrothermal vents along mid-ocean ridges, windblown dust, rivers, glacial ice, and seepage from continental margins.[14]
The importance of various sources of reduced iron has likely changed dramatically across geologic time. This is reflected in the division of BIFs into Algoma and Lake Superior-type deposits.[40][41][42] Algoma-type BIFs formed primarily in the Archean. These older BIFs tend to show a positive europium anomaly consistent with a hydrothermal source of iron.[5] By contrast, Lake Superior-type banded iron formations primarily formed during the Paleoproterozoic era, and lack the europium anomalies of the older Algoma-type BIFs, suggesting a much greater input of iron weathered from continents.[9][43][5]
Absence of oxygen or hydrogen sulfide
The absence of hydrogen sulfide in anoxic ocean water can be explained either by reduced sulfur flux into the deep ocean or a lack of dissimilatory sulfate reduction (DSR), the process by which microorganisms use sulfate in place of oxygen for respiration. The product of DSR is hydrogen sulfide, which readily precipitates iron out of solution as pyrite.[31]
The requirement of an anoxic, but not euxinic, deep ocean for deposition of banded iron formation suggests two models to explain the end of BIF deposition 1.8 billion years ago. The "Holland ocean" model proposes that the deep ocean became sufficiently oxygenated at that time to end transport of reduced iron. Heinrich Holland argues that the absence of manganese deposits during the pause between Paleoproterozoic and Neoproterozoic BIFs is evidence that the deep ocean had become at least slightly oxygenated. The "Canfield ocean" model proposes that, to the contrary, the deep ocean became euxinic and transport of reduced iron was blocked by precipitation as pyrite.[31]
Banded iron formations in northern Minnesota are overlain by a thick layer of ejecta from the Sudbury Basin impact. An asteroid (estimated at 10 km (6.2 mi) across) impacted into waters about 1,000 m (3,300 ft) deep 1.849 billion years ago, coincident with the pause in BIF deposition. Computer models suggest that the impact would have generated a tsunami at least 1,000 m (3,300 ft) high at the point of impact, and 100 m (330 ft) high about 3,000 km (1,900 mi) away. It has been suggested that the immense waves and large underwater landslides triggered by the impact caused the mixing of a previously stratified ocean, oxygenated the deep ocean, and ended BIF deposition shortly after the impact.[36]
Oxidation
Although Cloud argued that microbial activity was a key process in the deposition of banded iron formation, the role of oxygenic versus anoxygenic photosynthesis continues to be debated, and nonbiogenic processes have also been proposed.
Oxygenic photosynthesis
Cloud's original hypothesis was that ferrous iron was oxidized in a straightforward manner by molecular oxygen present in the water:[30][14]