August 2017 Issue Table of Contents
Boron is rare in the cosmos because its nucleus is “fragile.” So, how does one get from the interstellar medium, where boron was first produced, to Earth’s upper continental crust where boron is concentrated in deposits containing remarkably diverse suites of boron minerals? Processes that led to the formation of continental crust also concentrated boron, which is preferentially incorporated into melts and aqueous fluids. Deposits with high boron-mineral diversity include granitic pegmatites, peralkaline intrusions, boron-enriched skarns, and evaporite deposits. Despite the loss of boron minerals from the geologic record due to their ready solubility in water and breakdown at low temperatures, the increase in boron-mineral diversity with time is real, and is accelerated during supercontinent assembly.
Keywords: boron concentration, mineral diversity, continental crust, supercontinent assembly, boron cycling
Introduction: In the Beginning
Boron, like lithium and beryllium, is rare in the cosmos because its nucleus is “fragile.” So, how does one get from the interstellar medium where boron was first produced to Earth’s upper continental crust where boron has been sufficiently concentrated to produce nearly 300 minerals that contain boron as an essential constituent? Boron is the quintessential crustal element—the continental crust is presently the largest reservoir of boron, exceeding both the oceans and oceanic crust. Processes that concentrated boron also led to growth of the continental crust. Thus, boron is a tracer for the growth of continental crust, particularly the upper component. The present paper links the growth of Earth’s continental crust to a progressive crustal enrichment in boron and an increase in the diversity of boron minerals.
Cosmic scarcity of boron resulted in a low-boron concentration in the Solar System, where boron remains scarce. Boron is not produced by nucleosynthesis, except the small fraction coming from the radioactive decay of 10Be (Liu and Chaussidon 2017). Solar System boron concentration can be approximated from type CI (carbonaceous Ivuna type) carbonaceous chondrites, coming out at 0.775 ± 0.078 parts per million (µg/g) B (Shearer and Simon 2017 this issue). Earth’s primitive mantle is depleted in boron relative to CI chondrites, i.e. 0.19 µg/g (Marschall et al. 2017). There is no reason to suspect that the early Earth was significantly more enriched in boron than other terrestrial planets, because it is unlikely that boron contents would have varied significantly between the different chondrite types that constituted the complex mix from which the Earth and the other terrestrial planets formed. Although estimates for the bulk silicate Moon give 0.0743 µg/g B (Hauri et al. 2015), which is even more depleted than Earth’s primitive mantle, the preliminary determination of 10–100 µg/g B in calcium sulfate veins in Mars’ Gale Crater by the NASA Curiosity rover (Gasda et al. 2016) is consistent with Mars having a bulk boron content comparable to Earth’s bulk boron content.
The Earliest Boron Concentrations
Starting with 0.19 µg/g B in the primitive mantle (Marschall et al. 2017), how can boron be concentrated to reach the enrichments in Earth’s crust today? The current estimated average boron concentration for the upper continental crust is 17 µg/g B (Rudnick and Gao 2014). This figure includes values for several enriched crustal reservoirs, such as boron in terrigenous detritus in pelagic sediments (30–150 µg/g B) (Leeman and Sisson 1996) and granitic pegmatites (e.g. 213–287 µg/g B) (Stilling et al. 2006; Simmons et al. 2016). White and Klein (2014) calculated the bulk boron content of oceanic crust to be 0.8 µg/g from the composition of lavas erupted at the surface plus that of cumulate minerals in the lower crust and upper mantle. These authors gave 1.8 µg/g B as the global average of mid-ocean ridge basalts (MORB), both of which are but modest enrichments from the mantle. It appears unlikely that magmatic processes would, by themselves, lead to the present average of 17 µg/g B in the upper continental crust.
Earth’s boron story appears to be inextricably linked to water: the formation of Earth’s oceans would be a prerequisite for concentrating boron. There are two possible sources for the Earth’s oceans: first, volatiles released by degassing of “wet” planetary embryos accreted during the second half of Earth’s formation; or second, volatile-rich comets originating in the outer Solar System. Whatever the source of water, there is evidence that the ocean could have been present during the Hadean. Harrison et al. (2017) cited O and Li isotopic evidence in Hadean zircons from Jack Hills (Western Australia) for a “clement” Earth from at least 4.3 Ga. At this time, chemical weathering could have played a major role in the disaggregation and breakdown of the then-exposed rocks, Earth’s earliest crust. Harrison et al. (2017) concluded that the Jack Hills zircons crystallized from relatively cool, wet, felsic melts partially sourced from sedimentary protoliths at a plate boundary. In contrast, Kemp et al. (2010) argued that the Hf–Pb systematics of the Jack Hills zircons are consistent with protracted reworking of a mafic protocrust formed from solidification of a magma ocean, but with no juvenile additions after extraction from the primordial mantle at 4.4–4.5 Ga. The protocrust envisaged by Kemp et al. (2010) would probably not contain more than 2–3 µg/g B. However, exposure to the atmosphere and hydrosphere of the protocrust, whether mafic or felsic, could begin the process of concentrating boron. Chemical weathering released boron present in the protocrust, freeing it to be incorporated in the Earth’s ocean(s).
There are no constraints on boron contents of Earth’s ocean and crust until 3.8 Ga. At this time, rocks start to contain sufficient boron to be expressed mineralogically. Sedimentary precursors to metachert, mica schist, tourmalinite, and amphibolite in the 3.7–3.8 Ga Isua supracrustal belt (Greenland) contain enough boron to form tourmaline during amphibolite-facies metamorphism at 3.55 Ga (Fig. 1) (Grew et al. 2015). No bulk boron data have been reported for the Isua rocks, but comparable amphibolite-facies metapelites in other areas reveal 7–80 µg/g B (Leeman and Sisson 1996). Isua tourmalinites that contain 30%–50% modal tourmaline (Grew et al. 2015) would have a bulk boron content of ~10,000–16,000 µg/g.
The 3.7–3.8 Ga Isua supracrustal belt is the oldest geologic entity containing minerals for which boron is an essential constituent. No convincing evidence has been reported for boron concentrations or boron minerals prior to 3.8 Ga. Nonetheless, the presence of significant boron in the Isua supracrustal belt presumes some prehistory of concentrating boron. Furukawa and Kakegawa (2017 this issue) infer that volcanoes were a source of boron in the proto-arc, which presumes boron had already been incorporated in the source rocks of erupting magmas, i.e. boron could have originated from preexisting crust or seawater. It is possible that there were concentrations of boron in proto-arcs before Isua, which are plausible at Hadean plate boundaries (Harrison et al. 2017). But no such boron concentrations have been preserved.
Crustal Growth and Increasing Boron Mineral Diversity
Continental Crust Growth
Virtually all investigators agree that the volume of continental crust has increased with time, but opinions differ on the calculated rate of growth, the extent of crustal destruction, and the proportion of recycling. Most of the proposed growth curves lie in the area bounded by the curve for ‘crustal growth with some recycling’ and the curve for ‘present-day exposure’ (Hawkesworth et al. 2013) (Fig. 1A).
Chaussidon and Albarède (1992) proposed a simple model (Fig. 2A) that relates the growth of the crust to seawater d11B, which is the boron isotope composition expressed as per mil deviation from the standard NIST SRM 951 boric acid (Palmer 2017 this issue). The model assumes that the mass of the continental crust is the only reservoir that changes in size over time and that amount of boron recycled back into the mantle is negligible compared to the amount extracted. The d11B of each reservoir varies with isotopic fractionation between two reservoirs, given as D11B = d11B1 − d11B2 in Figure 2A. Assuming values for d11B, boron concentration [B], and the mass (M) given in Figure 2A, Chaussidon and Alberède’s (1992) equations give d11B = –22.6‰ for seawater at 4.3 Ga, which is a plausible age for the ocean (see above), and give d11B = –9.4‰ for the continental crust today. Compare these values to the measured value of –9.1‰ by Marschall et al. (2017). By analyzing Isua tourmalines, Grew et al. (2015) estimated an Archean seawater composition of d11B = +14 ± 15‰ at 3.7–3.8 Ga, whereas the above model gives a seawater composition of d11B = –27‰, which assumes that the mass of continental crust at 3.7–3.8 Ga was about 13% of the present-day mass. That the results are even this close could be fortuitous because the model is sensitive to the many assumptions made, particularly the constancy of seawater boron concentration with time and the starting isotope composition, i.e. –7.1‰ for depleted mantle and the bulk silicate Earth (Marschall et al. 2017). However oversimplified, the model does suggest a mechanism whereby increases in seawater d11B can be related to the growth of the continental crust, which appears to retain much of the boron extracted from the mantle. Because the mass of present-day continental crust is about 30 times that of the ocean, increases in seawater boron concentration would have an order-of-magnitude less impact on its d11B than growth of continental crust. This would justify the assumption of constancy of seawater boron content as a first-order approximation.The boron cycle can be recast into a geologic framework for the period following the onset of modern plate tectonics (Fig. 2B). Extrusion of MORB could be accompanied by the release of boron into seawater, MORB being subsequently altered by reaction with seawater to form altered oceanic crust; extruded mantle is serpentinized. Both the altered MORB and mantle rocks can contain over 100 µg/g B (e.g. Wunder et al. 2005; Boschi et al. 2008; Palmer 2017 this issue). Boron is then cycled through the subduction zone and arc volcanoes, followed much later by incorporation in illite or dissolved in rain water after weathering of crustal rocks and is eventually transported by rivers back to the ocean to end up in pelagic sediments or oceanic crust to enter another cycle. Whether boron is recycled in the subducted slab back into the mantle remains an open question (queried in Fig. 2); modelling suggests that only small amounts of boron remain in dry subducted rocks, but some could survive in phengite (e.g. Konrad-Schmolke and Halama 2014; Palmer 2017 this issue). Evidence that crustal boron could be recycled back into the mantle is from the occurrence of qingsongite (natural cubic BN) in crustal rocks subducted to ~400 km depth (Dobrzhinetskaya et al. 2014).
With the growth of the continental crust, the possibility for boron to be concentrated to levels well above the upper crustal average of 17 µg/g B (Rudnick and Gao 2014) also increases. Such an increase would result in more opportunities for mineralogical diversity. Figure 1 quantitatively relates crustal growth to increasing diversity of boron minerals, which can be expressed in one of two ways. First, by showing the cumulative increase in the total number of boron mineral species inferred to have formed by a given time in Earth’s history, as based on the reported first occurrences in the geological record and for which the requisite age data are available. This applies to 274 boron minerals (Fig. 1B). Second, by showing boron minerals that existed at a given time in Earth’s history, specifically, the number of species inferred to have been present during a given 50 My interval for which the requisite age data are available. This applies to 285 minerals (Fig. 1C). Comparing the proportion of exposed continental crust at a given geologic period (Fig. 1A) with cumulative diversity (Fig. 1B) might suggest that species diversity is simply a matter of exposed area—low species diversity in older rocks is due to their limited exposure. However, for the period between 1,825 Ma and 550 Ma the increase in the number of species inferred to have been present during a given 50 My interval is modest compared to the marked increase in present-day exposure of continental crust. Instead, existing boron mineral diversity somewhat better matches the curve for increase in crust with recycling.
There is another factor involved in the diversity of boron minerals. Eighty-eight (or about 30%) of boron minerals are soluble in water or are broken down at relatively low temperatures (Grew et al. 2016, 2017). An example of this is sassolite (natural boric acid) (Fig. 3). The greatest diversity of ephemeral minerals is found in evaporite deposits (Helvacı and Palmer 2017 this issue), including those from marine settings (e.g. the Angara basin near Irkutsk, in Siberia, Russia, and the Moncton Basin of New Brunswick, Canada) and nonmarine settings (e.g. Death Valley in California, USA). Such ephemeral minerals are largely restricted to the Phanerozoic, during which they contribute significantly to the steep increase in both cumulative (Fig. 1B) and existing diversity (Fig. 1C). Their scarcity in the Precambrian could be more a matter of preservation, e.g. by armoring (Fig. 3A), than of areal exposure.
Nonetheless, if discussion were restricted to the refractory minerals, there would still be an increase in diversity of boron minerals, even during the Phanerozoic (Figs. 1B, 1C). This increase is punctuated by steps in cumulative diversity and both steps and spikes in existing diversity. The three oldest steps and spikes in cumulative and existing diversity correspond not only to the collisional phases of the supercontinent cycles of Kenorland, Nuna, and Rodinia (Cawood and Hawkesworth 2013), but also, albeit more approximately, to the humps in the curve for preservation of juvenile crust at ~2,500 Ma, ~1,700 Ma, and ~1,100 Ma (Fig. 1A). For example, pegmatites of the lithium–cesium–tantalum (LCT) family, such as the pegmatites in southeastern Manitoba (at Tanco) and neighboring Ontario (the Fort Hope pegmatite field) that contributed significantly to the step at 2,640–2,650 Ma (Kenorland), are typical for settings of crustal thickening associated with horizontal tectonic processes of subduction and continental collision (Cerný et al. 2012). Localities contributing to steps at 1,825–1,950 Ma (Nuna) and at 1,010 Ma (Rodinia) include localities with diverse suites of boron minerals: the Tayozhnoye deposit in the Aldan shield (central Siberia), the Långban-type deposits in the Bergslagen ore region (Sweden), and the Kalahari manganese field (South Africa) (Fig. 1B). The latter two deposits have been linked to plate tectonic processes. However, not all localities of high-boron mineral diversity bear a simple relationship to supercontinent assembly. The skarn at Pitkäranta (Karelia, Russia) is interpreted to be associated with an early stage of continental rifting (Amelin et al. 1997), whereas the peralkaline intrusion at Mont Saint-Hilaire (Quebec, Canada) is associated with hotspot activity (Chen and Simonetti 2015).
Two steep increases in diversity between 2,700 Ma and 1,850 Ma are followed by a near leveling off between 40 and 50 boron mineral species from 1,850 Ma to 550 Ma (Fig. 1C) This implies that many of the minerals first reported in the geologic record prior to 1,850 Ma have been forming ever since, in part because the geologic environments for these minerals, such as those of the LCT pegmatites, have formed throughout geologic time, and that, in part, these minerals can occur in more than one environment. In contrast, spikes at 1,825 Ma, 1,525 Ma, 1,025 Ma, and 925 Ma can largely be attributed to localities of high diversity that feature boron minerals occurring at no more than 4–5 localities worldwide (in many cases, just one, e.g. Långban). During the Phanerozoic, diversity in refractory boron minerals again increases markedly. There are several major spikes due to the development of boron-rich skarns and peralkaline intrusions, including Mont Saint-Hilaire. These latter types of deposits show greater boron mineral diversity than their Precambrian analogues (Grew et al. 2016). With an estimated ~200 boron minerals remaining to be discovered in Earth’s crust (Grew et al. 2017), further increases in the diversity shown in Figure 1B and 1C can be expected.
Italo Campostrino, Francesco Demartin, and Rainer Thomas are thanked for permission to reproduce their photographs, Marco Ciriotti for permission to use a photograph in his book (Campostrini et al. 2011), and Martin Palmer for assistance with application of the Chaussidon and Albarède (1992) model and fruitful discussions on secular variation of boron isotope compositions. Ralf Halama and Marc Chaussidon are thanked for their constructive reviews.
Amelin YV, Larin AM, Tucker RD (1997) Chronology of multiphase emplacement of the Salmi rapakivi granite-anorthosite complex, Baltic Shield: implications for magmatic evolution. Contributions to Mineralogy and Petrology 127: 353-368
Boschi C, Dini A, Früh-Green GL, Kelley DS (2008) Isotopic and element exchange during serpentinization and metasomatism at the Atlantis Massif (MAR 30°N): insights from B and Sr isotope data. Geochimica et Cosmochimica Acta 72: 1801-1823
Campostrini I, Demartin F, Gramaccioli CM, Russo M (2011) Vulcano: Tre secoli di mineralogia. Associazione Micro-mineralogica Italiana, Cremona, Italy, 344 pp
Cawood PA, Hawkesworth CJ (2013) Temporal relations between mineral deposits and global tectonic cycles. In: Jenkin GRT (and 6 coeditors) Ore Deposits in an Evolving Earth. Geological Society, London, Special Publications, 393, pp 9-21
Chaussidon M, Albarède F (1992) Secular boron isotope variations in the continental crust: an ion microprobe study. Earth and Planetary Science Letters 108: 229-241
Chen W, Simonetti A (2015) Isotopic (Pb, Sr, Nd, C, O) evidence for plume-related sampling of an ancient, depleted mantle reservoir. Lithos 216-217: 81-92
Dobrzhinetskaya LF and 6 coauthors (2014) Qingsongite, natural cubic boron nitride: the first boron mineral from the Earth’s mantle. American Mineralogist 99: 764-772
Gasda PJ and 9 coauthors (2016) First observations of boron on Mars and implications for Gale Crater geochemistry. American Geophysical Union Fall 2016 Meeting, Abstract P21D-04
Grew ES and 6 coauthors (2015) Boron isotopes in tourmaline from the ca. 3.7–3.8 Ga Isua supracrustal belt, Greenland: sources for boron in Eoarchean continental crust and seawater. Geochimica et Cosmochimica Acta 163: 156-177
Grew ES, Krivovichev SV, Hazen RM, Hystad G (2016) Evolution of structural complexity in boron minerals. Canadian Mineralogist 54: 125-143
Grew ES, Hystad G, Hazen RM, Krivovichev SV, Gorelova LA (2017) How many boron minerals occur in Earth’s upper crust? American Mineralogist 102, in press, doi: 10.2138/am-2017-5897
Harrison TM, Bell EA, Boehnke P (2017) Hadean zircon petrochronology. Reviews in Mineralogy & Geochemistry 83: 329-363
Hauri EH, Saal AE, Rutherford MJ, Van Orman JA (2015) Water in the Moon’s interior: truth and consequences. Earth and Planetary Science Letters 409: 252-264
Hawkesworth C, Cawood P, Dhuime B (2013) Continental growth and crustal record. Tectonophysics 609: 651-660
Kemp AIS and 7 coauthors (2010) Hadean crustal evolution revisited: new constraints from Pb–Hf isotope systematics of the Jack Hills zircons. Earth and Planetary Science Letters 296: 45-56
Konrad-Schmolke M, Halama R (2014) Combined thermodynamic–geochemical modeling in metamorphic geology: boron as tracer of fluid–rock interaction. Lithos 208-209: 393-414
Leeman WP, Sisson VB (1996) Geochemistry of boron and its implications for crustal and mantle processes. Reviews in Mineralogy 33: 645-707
Liu M-C, Chaussidon M (2017) The cosmochemistry of boron isotopes. In: Marschall HR and Foster GL (eds) Boron Isotopes – The Fifth Element. Advances in Isotope Geochemistry, Springer, in press
Marschall HR and 5 coauthors (2017) The boron and lithium isotopic composition of mid-ocean ridge basalts and the mantle. Geochimica et Cosmochimica Acta 207: 102-138
Rudnick RL, Gao S (2014) Composition of the continental crust. In: Rudnick RL (ed) The Crust. Treatise on Geochemistry, Volume 4 (second edition), Elsevier, pp 1-51
Simmons W and 6 coauthors (2016) Bulk composition of Mt. Mica pegmatite, Maine, USA: implications for the origin of an LCT type pegmatite by anatexis. Canadian Mineralogist 54: 1052-1070
Stilling A, Cˇerný P, Vanstone, PJ (2006) The Tanco pegmatite at Bernic Lake, Manitoba. XVI. Zonal and bulk compositions and their petrogenetic significance. Canadian Mineralogist 44: 599-623
Thomas R, Davidson P, Beurlen H (2012) The competing models for the origin and internal evolution of granitic pegmatites in the light of melt and fluid inclusion research. Mineralogy and Petrology 106: 55-73
White WM, Klein EM (2014) Composition of the oceanic crust. In: Rudnick RL (ed) The Crust. Treatise on Geochemistry, Volume 4 (second edition), Elsevier, pp 457-496
Wunder B, Meixner A, Romer RL, Wirth R, Heinrich W (2005) The geochemical cycle of boron: constraints from boron isotope partitioning experiments between mica and fluid. Lithos 84: 206-216