August 2017 Issue Table of Contents
The behavior of boron during the early evolution of the Solar System provides the foundation for how boron reservoirs become established in terrestrial planets. The abundance of boron in the Sun is depleted relative to adjacent light elements, a result of thermal nuclear reactions that destroy boron atoms. Extant boron was primarily generated by spallation reactions. In the initial materials condensing from the solar nebula, boron was predominantly incorporated into plagioclase. Boron abundances in the terrestrial planets exhibit variability, as illustrated by B/Be. During planetary formation and differentiation, boron is redistributed by fluids at low temperature and during crystallization of magma oceans at high temperature.
Keywords: solar nebula, condensation, accretion, magma ocean, moderately volatile elements
The majority of articles in this special issue relate to the behavior of boron in terrestrial environments. The Earth makes up less than 0.02% of the mass of the Solar System and, therefore, it seems sensible to explore the behavior of boron outside of our terrestrial home. Since the pioneering work that laid the foundation of cosmochemistry (e.g. Goldschmidt 1926), significant inroads have been made into understanding the synthesis of boron in the cosmos, its behavior in extraterrestrial environments since the formation of the Solar System, and its link to the exploration for life outside of Earth.
Boron provides a unique perspective on Solar System processes. This is because of the unusual nuclear synthesis reactions involved in its production; its valence and size characteristics, which are distinctly different from those of many other moderately volatile elements (e.g. K+, Na+); its mobility during fluid/rock interactions; its significant partitioning into a vapor phase; and its isotope systematics, which are sensitive to a variety of processes and conditions (e.g. pH, temperature). The access to numerous planetary environments through meteorites and returned samples and the development of new analytical techniques (e.g. secondary ion mass spectrometry, laser ablation inductively coupled plasma mass spectrometry) have given us the ability to look at boron behavior backward in time and upward beyond Earth. Here, we examine the behavior of boron and its stable isotopes during the first 180 million years of our Solar System, from the formation of the first particles to the primordial differentiation of large and small rocky planetary bodies.
Behavior of Boron During Solar Nebular Processes
Bulk Boron Content of the Solar System
Abundances of elements in the Solar System are obtained by spectral analysis of the photosphere of the Sun—the Sun contains >99% of the mass of the Solar System—and from the compositions of CI chondrites, which contain solar abundances of the condensable elements. Abundances are typically reported in wt% element or in atoms per 106 silicon atoms. Zhai and Shaw (2009) assumed that the boron content of carbonaceous chondrite matrices is representative of the Solar System abundance and reported the average for seven samples: 0.69 ± 0.09 ppm and 16.9 ± 2.2 atoms per 106 Si. More recently, Lodders et al. (2009) reported an average CI content of 0.775 ± 0.078 ppm and a bulk Solar System content of 18.8 atoms per 106 Si, in agreement with Zhai and Shaw (1994). The abundance of boron (and Be and Li) in the Solar System is substantially depleted relative to adjacent light elements (e.g. H, He, C, O) and is the result of thermal nuclear reactions that destroy boron isotopes during hydrogen burning. Boron was primarily generated by spallation reactions between high-energy particles in galactic cosmic rays and C, N, and O nuclei in the interstellar medium (Reeves et al. 1970), rather than thermonuclear reactions. Subtleties of boron formation processes and boron behavior in the solar nebula are revealed by the examination of primitive Solar System materials.
Behavior of Boron During Condensation
It is thought that the inner Solar System was once completely vaporized and that the solid materials we see today condensed from that vapor. At high temperatures, the dominant boron-bearing gas species in this vapor are BO, HBO, and HBO2 (Lauretta and Lodders 1997). Given the bulk composition of the Solar System and the thermodynamic properties of the elements, the equilibrium condensation sequence can be calculated, starting with the highly refractory platinum-group elements and continuing with the refractory lithophile elements, the major elements, the moderately volatile elements, and the volatile elements (Fig. 1). Different mineral phases become stable with decreasing temperature, which affects the condensation temperatures of the elements in addition to their volatility. In a solar gas at a total pressure of 10−3 atm, the first major phase to condense is corundum (Al2O3) at 1,730 K. It starts to react with Ca in the gas to form hibonite (CaAl12O19) at 1,700 K, followed by grossite (CaAl4O7) at 1,660 K, gehlenite (Ca2Al2SiO7), spinel (MgAl2O4), and clinopyroxene [Ca(Mg,Ti,Al)(Si,Al)2O6] (Grossman 2010). Calculations by Lauretta and Lodders (1997) showed that boron should initially condense into feldspar as danburite (CaB2Si2O8) and then as reedmergnerite (NaBSi3O8). In both components, boron is substituting for Al in tetrahedral coordination. Condensation temperatures of elements are typically reported as the temperature at which the element is halfway to complete condensation. The 50% condensation temperature for boron was estimated to be ~910 K by Zhai (1995), essentially identical to the value of 908 K given by Lodders et al. (2009). Chaussidon and Jambon (1994) proposed that the condensation temperature of boron could be ~700 K if it condensed mainly as borides and borates. For comparison, refractory Ca–Al-rich inclusions (CAIs), which are some of the oldest materials in the Solar System (Fig. 1), condensed at temperatures between 1,300 K and 1,700 K (Fig. 1). The ancient ages of CAIs are reflected in their preservation of extinct radionuclide systems (10Be, 26Al, 53Mn, and 129I), even if they were reheated and became partially molten. Note that although the inclusion is not a primary gas–solid condensate, it consists of phases predicted to condense at high temperatures from a gas of solar composition.
Behavior of Boron During Ca–Al-rich inclusion (CAI) Remelting
As a moderately volatile element, boron is depleted in CAIs. Based on modal mineralogy and the average boron contents of the phases (assuming the boron content of spinel to be negligible) in Allende Type B1 inclusion 3529-41 (Fig. 2), Chaussidon et al. (2006) calculated a bulk boron content of 0.125 ppm (~0.18 times the content of CI chondrites). Chaussidon et al. (2006) noted that the volatility of boron is similar to that of sodium and suggested that, like sodium, boron could have been introduced to CAIs during low-temperature alteration in the nebula prior to remelting and then redistributed during fractional crystallization. Therefore, the bulk boron content of the CAIs represents an upper limit for primary boron concentrations.
Boron has two stable isotopes, 10B and 11B, with 11B/10B ≈ 4. Throughout this paper, boron isotope ratios will be referred to in terms of their 11B/10B values in ‰, such that,
The standard commonly used is NIST SRM 951 boric acid. Refractory inclusions have negative 11B/10B values (negative 11B/10B reflects enrichment in the light isotope (10B) relative to the standard), a feature that is not consistent with Rayleigh isotopic fractionation that would accompany evaporation from a melt. Instead, refractory inclusions exhibit 10B excesses correlated with Be/B ratios. The daughter product of 10Be decay is 10B (half-life = 1.5 My), so these excesses in 10B are interpreted as evidence that CAIs incorporated live 10Be that decayed in situ to 10B (McKeegan et al. 2000; MacPherson et al. 2003). The short-lived 10Be radionuclide is produced by spallation reactions, not stellar nucleosynthesis. Also important are the 10Be/9Be ratios that are derived from analyses of CAIs. If these ratios were uniform, that would indicate that the 10Be excesses were inherited from the dense interstellar cloud that became the solar nebula. They are not uniform, however, and are not correlated with 26Al/27Al ratios (MacPherson et al. 2003). This suggests that 10Be formed by local irradiation processes within the nebula (Srinivasan and Chaussidon 2013).
Boron During Chondrule Formation
Chondrules are a major component of many chondritic meteorites, ranging in abundance from 0% to 80%. The majority of chondrules are submillimeter igneous spheres that mainly consist of ferromagnesian silicates and glass. An image of a glass-rich chondrule from the Murchison meteorite, with a texture that strongly indicates that it was once a molten droplet, is shown in Figure 3. Although many models for chondrule formation have been proposed (e.g. low-velocity impacts between molten planetesimals), those involving formation in the solar nebula are the most accepted (Brearley and Jones 1998) (Fig. 1). The mineral assemblages in chondrules have been re-equilibrated to varying degrees within the solar nebula or on their parent bodies after accretion. Therefore, they provide another window to nebular and parent body behavior of boron. The behavior of boron and 11B/10B in these products of nebular processes has been explored using secondary ion mass spectrometry (Brearley and Layne 1996; Chaussidon and Robert 1998; Shearer 2002). Chondrules in unequilibrated chondrites (e.g. the Semarkona meteorite) that experienced minimal degrees of parent-body alteration probably provide the best view of nebular processes and boron reservoirs. The B/Be ratio of moderately volatile/refractory element in the least metamorphosed chondrules are generally less than solar (Brearley and Layne 1996). The strong inverse correlation between beryllium and boron and the limited correlation between chondrule surface area and boron contents suggests that volatile loss during chondrule formation was not a major control (Brearley and Layne 1996). In this case, the sub-solar B/Be in the precursor material may reflect fractionation of boron from beryllium during condensation from the solar nebula at T > 900 °C (Lauretta and Lodders 1997). Meibom et al. (2001) observed that in a metal-rich chondrite, the B/Be in chondrules varied over two orders of magnitude and was negatively correlated with the Ca/Si ratio (Meibom et al. 2001). These observations indicate that the B/Be in chondrules represents both the precursor material of the chondrules and their temperature of formation.
The 11B/10B values of chondrules from several chondrites range from −50‰ to +44‰ (i.e. 11B/10B between 3.86 and 4.23) and are correlated with B/Si ratios (e.g. Chaussidon and Robert 1998). This heterogeneity of 11B/10B observed in individual chondrules has not been reported in bulk meteorite analyses (Zhai et al. 1996). Further, this heterogeneity contrasts with measurements by Hoppe et al. (2001) that exhibited a much more limited variation in 11B/10B. Liu and Chaussidon (2017) concluded that to resolve these contrasting results required further analytical effort.
Although the results of Zhai et al. (1996), Chaussidon and Robert (1998), and Hoppe et al. (2001) contrast with regard to the variability of 11B/10B, they are all consistent with a 11B/10B ratio for chondrules of ~4. However, this value is not consistent with boron being solely produced by galactic cosmic ray spallation, which yields a 11B/10B value of 2.5. Chaussidon and Robert (1998) and Liu and Chaussidon (2017) concluded that boron in the Solar System was a product of mixing of the boron produced by galactic cosmic ray spallation (11B/10B = -400‰) and the boron produced by a low-energy spallation process that favors the production of 11B over 10B (11B/10B = +110‰).
Boron During Parent-Body Metamorphism
Both chondrules and chondrites have experienced parent-body processes involving interaction with fluid phases and reheating (e.g. Brearley and Jones 1998). Brearley and Layne (1998) illustrated that chondrules modified by low-temperature alteration have B/Be ratios that are greater than solar. This fractionation above the solar ratio reflects the relative mobility of boron compared to beryllium. The mobility of boron on the parent body is further suggested by the enrichment of boron in carbonaceous chondrite matrix (0.69 ppm) relative to chondrules (0.23 ppm) and the positive correlation of boron with another fluid-mobile element, sulfur (Zhai and Shaw 1994). Krot et al. (1995) proposed that an iron–alkali–halogen vapor/fluid phase was responsible for asteroidal parent-body alteration and that this phase would have been responsible for the transport of boron. The characteristics of this fluid phase (e.g. pH, temperature) would systematically fractionate boron isotopes, but such fractionation associated with aqueous alteration of chondrites has yet to be examined in detail.
Behavior of Boron During the Accretion of the Terrestrial Planets
According to the solar nebula theory, planetary development initiated with the condensation and formation of primitive aggregates of undifferentiated solar nebular material followed by accretion and differentiation of this primitive material (Fig. 1). As discussed above, the memory of this primitive material is preserved in undifferentiated meteorites, such as chondrites. Contemporary models advocate that the planets are the end products of a hierarchical accretionary process that first assembles a large number of kilometer-sized planetesimals from an initial protoplanetary disk of gas and dust (e.g. Elkins-Tanton 2012; Chambers 2014). Growth during this stage was controlled by surface, electromagnetic, and electrostatic forces. These small bodies then coalesced into protoplanets—planets whose evolution was controlled primarily by gravitational interactions. The formation of Mars-size embryos occurred in as little as 105 years. The final stages of planetary accretion, which may have persisted for 108 years, are characterized by large, discrete impact events, such as the impact that formed the Moon, and subsequent events recorded on the lunar surface (e.g. Elkins-Tanton 2012). The relationship between the estimated composition of the planetary body and the primordial Solar System composition provides clues to planetary formation processes. Boron for bulk planetary bodies provides some insights into how the volatile reservoirs of the solar nebula were incorporated into the mantles of planetary bodies. As the terrestrial planets have undergone substantial differentiation and evolution since their assembly, determining the composition of their precursors and understanding the processes that were instrumental in their formation is an important, but formidable, task.
Estimates of the moderately volatile element and bulk boron content of many of the terrestrial planets have been calculated based on the following: analyses of samples (collected samples, returned samples, meteorites); geophysical and geochemical measurements by spacecraft (in situ and from orbit); cosmochemical models for planetary assembly. Although boron abundances have not been determined directly by many spacecraft observations, the combined observations have been used to calculate indices of a planet’s volatile element characteristics [e.g. K(moderately volatile element)/Th(refractory element)]. Using this index, McCubbin et al. (2012) proposed that the moderately volatile elements were depleted in the order K/ThMercury = K/ThMars > K/ThVenus = K/ThEarth >> K/Th4Vesta ≥ K/ThMoon. How does this sequence in K/Th ratios compare to estimates of B/Be ratios and boron for planets that have been sampled? Can the K/Th relationship be extended to planets that have not been sampled, for which only remotely sensed data are available?
The terrestrial planets for which the bulk boron concentrations, B/Be, and 11B/10B are best known are the Earth and the Moon. Based on mantle-derived magmas and crust compositions, the bulk silicate Earth has approximately 0.25 ppm B (Chaussidon and Jambon 1994) with a B/Be of approximately 4.5 (Shearer 2002 and references within; Chaussidon and Jambon 1994). The 11B/10B of oceanic basalts range from −7.40‰ to +0.6‰ (Chaussidon and Jambon 1994; Zhai et al. 1996). In individual glass suites, 11B/10B is correlated with boron concentration, water content, and dD and is interpreted as the result of assimilation and fractionation processes (Chaussidon and Jambon 1994). This would imply that the bulk silicate Earth has a 11B/10B at the lower end of this range (−7.40‰). Marschall et al. (2017) estimated a value of −7.1‰ for the bulk silicate Earth. Estimates for the boron content of the bulk silicate Moon vary dramatically from 0.013 ppm to 0.54 ppm. Perhaps a better estimate is that based on the differences between the most common lunar basalts (very low- to low-Ti mare basalts) and unaltered terrestrial oceanic basalts. Whereas the terrestrial basalts (corrected for fractional crystallization and alteration) range from 0.5 ppm to 1.3 ppm, the low-Ti primordial lunar volcanic glasses have boron concentrations in the range of 0.11 ppm to 0.40 ppm (Shearer 2002). These volcanic glasses also have B/Be values that are approximately 2. These comparisons indicate that, along with a lower moderate volatile element content implied by the K/Th ratio, the Moon has a lower bulk boron content and lower B/Be than the Earth. The reported 11B/10B of the Moon ranges from −4.13‰ to −4.87‰.
Boron, B/Be and 11B/10B data for other differentiated planetary bodies (e.g. 4 Vesta, Mars) from which we have samples are far more limited. Data from Zhai et al. (1996) (lunar and meteorite samples) and Shearer (2002) (only meteorites) reveal that eucrites exhibit boron content variations from 0.11 ppm to 0.92 ppm, B/Be ratios between 0.87 ppm and 2.4 ppm, and have 11B/10B values from −6.90‰ to −1.60‰. Although the database for martian basalts is limited, a comparison can still be made with regard to Earth. Examination of clinopyroxenes shows that the B/Be ratios in martian magmas are significantly higher than in pyroxenes from terrestrial ocean island basalts (Lentz et al. 2001).
Based upon boron abundances and B/Be for Earth, Moon, 4 Vesta, and Mars, it appears that the B/Be values correlate with the estimated K/Th. Extending this correlation to Venus and Mercury allows approximations of the relative bulk boron and B/Be of these planetary bodies. These comparisons suggest that B/BeCI chondrite >> B/BeMercury = B/BeMars > B/BeVenus = B/BeEarth >> B/BeMoon > B/Be4Vesta, and the order of the extent of boron depletion relative to CI chondrite is 4 Vesta > Moon > Earth = Venus > Mars = Mercury >> CI chondrite.
These estimates of bulk planetary boron contents, B/Be ratios, and measurements of d11B provide valuable information concerning accretion of the rocky planets. All of the rocky planets are depleted in boron and have lower B/Be and B/Si relative to CI chondrites. The extent of the boron depletion factor is similar to those for elements with similar condensation temperatures, such as K, Rb, Cs, and Ge. Therefore, boron was a moderately volatile element during planetary accretion, and the degree of volatility observed reflects either the accreting material or the temperature of accretion. The degree of boron loss was not a function of heliocentric distance from the Sun. The d11B values for the Earth, Moon, and 4 Vesta are similar to each other and are distinctly different from the values predicted by the production of boron strictly from galactic cosmic rays.
Behavior of Boron During Primordial Differentiation. A Lunar Perspective
Primordial differentiation of planetary bodies perhaps started on small bodies as early as 2 My after CAI formation and may have extended to as late as 4.38 Ga for the Moon (Fig. 1) (Borg et al. 2011). During primordial planetary differentiation (e.g. magma ocean), boron is moderately to highly incompatible in partitioning between major silicate phases and basaltic melt. Among relevant phases during differentiation, the relative Dboron between a mineral phase and basaltic melt is Dfeldspar/melt > Dclinopyroxene/melt > Dspinel/melt > Dolivine/melt > Dorthopyroxene/melt > Dmetal/melt (Shearer 2002). During planetary differentiation boron will preferentially partition into the melt phase (Shearer 2002), and this, even taking into account variations in the composition of a primordial crust, makes boron the quintessential crustal element: CrustB >>> MantleB >> CoreB.
Since the return of samples from the Moon during NASA’s Apollo Program, a fundamental concept for large-scale primordial differentiation has been developed and applied to many of the terrestrial planets, moons, and differentiated asteroids: the concept of ‘magma oceans’. This planetary-scale process involved total to partial melting soon after accretion and occurred over a range of pressure–temperature regimes, enabling the formation of cores, mantles, and crusts (Elkins-Tanton 2012). Perhaps the best-recognized remnants of planetary magma oceans are those preserved on the Moon. The Moon is thought to have formed through accretion of materials generated by a collision between proto-Earth and a Mars-sized body (Theia), after which the Moon differentiated through internal melting followed by solidification of the lunar magma ocean (LMO) many hundreds of kilometers in depth (Warren and Taylor 2014 and references therein). In the LMO theory of lunar differentiation, the initial crystallization of the molten Moon resulted in a sequence of cumulates that included the following: (1) mafic cumulates, mainly containing olivine and pyroxene (with a crystallization sequence of olivine → orthopyroxene ± olivine → clinopyroxene + orthopyroxene ± olivine); (2) flotation cumulates that formed a plagioclase-rich, ferroan anorthositic crust; (3) ilmenite-rich cumulates; and finally (4) very late-stage lithologies characterized by high K, rare-earth element (REE), and P abundances, and therefore given the acronym “KREEP”. The interactions between these primary crystallization products of the LMO are believed to be responsible for the compositional diversity of lunar rocks (e.g. Warren and Taylor 2014). For example, lunar basalts are products of melting of the mafic cumulates that make up the Moon’s mantle. The compositions of these various LMO components provide insights into the behavior of boron during the primordial differentiation of the Moon. Lunar basalts are partial melts of these primordial LMO cumulates and, therefore, provide insights into the LMO process. Light lithophile element data for lunar volcanic glasses, lunar basalts, and crust have been documented by Shearer et al. (1994) and Zhai et al. (1996). A transmitted-light image of lunar high-Ti volcanic glasses is shown in Figure 4. Presumably, this glass represents a partial melt of a lunar mantle source with a significant ilmenite-rich LMO cumulate component.
Within the context of this simple LMO crystallization model, boron contents should increase from a bulk LMO composition of <0.25 ppm at the first mafic cumulate phase and rise to 25 ppm during the last KREEP stage of LMO crystallization (Warren and Taylor 2014). Boron (and Be) should be enriched to greater degrees than Li in the KREEP component of the LMO. This behavior is strictly the result of the contrasting behavior of boron (and Be) compared to Li in olivine. Boron is highly incompatible in basaltic melts, whereas Li is moderately incompatible (Shearer et al. 1994 and references therein). This resulted in the modification of Li/B signatures of the mantle cumulate sources. Based on the pyroclastic glass data from Shearer et al. (1994) and the assumption that the Li/B ratio does not fractionate during source melting, the early cumulate horizon of the LMO (olivine + pyroxene cumulates) had a higher Li/B (>25) than the late-stage products of LMO crystallization such as the ilmenite cumulates (Li/B = 5–20) and KREEP (Li/B = 1.6). Therefore, this ratio decreases with LMO crystallization and may reflect the preference of pyroxene for Li relative to B. The boron concentration and Li/B ratio are useful fingerprints for determining the cumulate horizon from which the basaltic magma originated and can be used to model the extent of LMO cumulate mixing.
Only a few boron isotopic compositions have been obtained for samples of lunar basalt and primordial crust (Zhai et al. 1996). The limited range of d11B varies from −4.13‰ to −4.87‰ for all lithologies derived from mafic cumulates, flotation cumulates, and KREEP. These values overlap the range determined for terrestrial oceanic basalts (Chaussidon and Jambon 1994). The limited variation in 11B/10B in basalts derived from LMO cumulates contrasts with isotopic fractionations observed in other moderately volatile elements. For example, some KREEP basalts exhibit 37Cl enrichments of up to +50‰, which has been attributed to degassing of 35Cl during the last stages of LMO crystallization (Boyce et al. 2015).
Interpreting the First 180 Million Years
The story of boron during the first 180 million years of Solar System history can still be read through the subsequent evolutionary conditions experienced by a planet—planet size, oxygen fugacity, availability and state of H-species, stagnant-lid versus plate-tectonic thermal regimes, stability and composition of atmosphere. These evolutionary conditions influenced the nature of the boron reservoirs (from mineral- to planetary scales) and the planetary boron cycle itself. The initial processes of nebular, accretionary, and primordial differentiation shaped the starting points for boron contents and d11B in the terrestrial planets (Fig. 1). Spallation by galactic cosmic rays is an important boron-producing process, but on its own will not yield appropriate 11B/10B for the Solar System. Low-energy spallation is a possible mechanism for producing more 11B-rich components. These processes still need to be further quantified with respect to the production of 11B. In addition, refractory inclusions show evidence of in situ decay of 10Be to 10B, and there is some variation of the initial 10Be/9Be ratio in CAIs. The boron concentration and d11B in chondrites and chondrules may suggest a high degree of heterogeneity of boron isotopes in the chondrite-forming portion of the Solar System. The mobility of boron in parent-body fluids at relatively low temperatures may have led to a heterogeneous distribution of boron and fractionation of boron isotopes. Still, there remain questions concerning the condensation temperature of boron and the real variability of d11B. The d11B remains a potential future tool for better understanding the conditions of chondrite parent-body aqueous alteration. Boron isotope systematics are sensitive to pH and temperature changes and can fractionate during phase separation. During primordial differentiation in a magma ocean, boron will generally behave incompatibly and increase in concentration during crystallization. It provides an index for magma ocean cumulate horizons and potential cumulate mixing. Late-stage magma ocean volatile loss has been called upon to explain fractionation of isotopes of other moderately volatile elements on the Moon (e.g. d37Cl). However, boron shows limited degrees of isotopic fractionation during this magma ocean process.
The coauthors appreciate the guidance and reviews of Ed Grew, Don Burnett, and Katharina Lodders that substantially improved concepts presented in this manuscript with regards to both clarity and correctness. This work was supported by the NASA Emerging Worlds Program through grants NNX16AI26G (SBS) and NNX13AH85G (CKS). I (CKS) would like to express appreciation to CJS for his input, creativity, and support during the preparation of this manuscript.
Borg LE, Connelly JN, Boyet M, Carlson RW (2011) Chronological evidence that the Moon is either young or did not have a global magma ocean. Nature 477: 70-72
Boyce JW and 6 coauthors (2015) The chlorine isotope fingerprint of the lunar magma ocean. Science Advances 1, doi: 10.1126/sciadv.1500380
Brearley AJ, Jones RH (1998) Chondritic meteorites. Reviews in Mineralogy 36: 3.1-3.398
Brearley AJ, Layne GD (1996) Light lithophile element abundances in microchondrules in CH chondrites: Insights into volatile behavior during chondrules formation. Lunar and Planetary Science XXVI: 167-168
Chambers JE (2014) Planet Formation. In: Davis, AM (ed) Planets, Asteroids, Comets and the Solar System. Treatise on Geochemistry, Volume 2 (second edition), Elsevier, pp 55-72
Chaussidon M, Jambon A (1994) Boron content and isotopic composition of oceanic basalts: geochemical and cosmochemical implications. Earth and Planetary Science Letters 121: 277-291
Chaussidon M, Robert F (1998) 7Li/6Li and 11B/10B variations in chondrules from the Semarkona unequilibrated chondrite. Earth and Planetary Science Letters 164: 577-589
Chaussidon M, Robert F, McKeegan KD (2006) Li and B isotopic variations in an Allende CAI: evidence for the in situ decay of short-lived 10Be and for the possible presence of the short-lived nuclide 7Be in the early solar system. Geochimica et Cosmochimica Acta 70: 224-245
Elkins-Tanton LT (2012) Magma oceans in the inner solar system. Annual Review of Earth and Planetary Sciences 40: 113-139
Goldschmidt VM (1926) Probleme und Methoden der Geochemie. Gerlands Beiträge zur Geophysik 15: 38-50
Grossman L (2010) Vapor-condensed phase processes in the early solar system. Meteoritics and Planetary Science 45: 7-20
Hoppe P, Lodders K, Strebel R, Amari S, Lewis RS (2001) Boron in presolar silicon carbide grains from supernovae. The Astrophysical Journal 551: 478-485
Krot AN, Scott ERD, Zolensky ME (1995) Mineralogical and chemical modification of components in CV3 chondrites. Nebular or asteroidal processes? Meteoritics and Planetary Science 30: 748-775
Lauretta DS, Lodders K (1997) The cosmochemical behavior of beryllium and boron. Earth and Planetary Science Letters 146: 315-327
Lentz RCF, McSween HY Jr, Ryan J, Riciputi LR (2001) Water in martian magmas: clues from light lithophile elements in shergottite and nakhlite pyroxenes. Geochimica et Cosmochimica Acta 65: 4551-4565
Liu M-C, Chaussidon M (2017) The cosmochemistry of boron isotopes. Marschall HR and Foster GL (eds) Boron Isotopes – The Fifth Element. Advances in Isotope Geochemistry, Springer, in press
Lodders K, Palme H, Gail HP (2009) Abundance of the elements in the solar system. In: Trumper JE (ed.) Solar System. Landolt-Börnstein, New Series, Group VI, Astronomy and Astrophysics 4B. Springer-Verlag, doi: 10.1007/978-3-540-88055-4_34
Marschall HR and 6 coauthors (2017) The boron and lithium isotopic composition of mid-ocean ridge basalts and the mantle. Geochimica et Cosmochimica Acta 207: 102-138
MacPherson GJ, Huss GR, Davis AM (2003) Extinct 10Be in Type A calcium-aluminum-rich inclusions from CV chondrites. Geochimica et Cosmochimica Acta 67: 3165-3179
McCubbin FM, Riner MA, Vander Kaaden KE, Burkemper LK (2012) Is Mercury a volatile-rich planet? Geophysical Research Letters 39, doi:10.1029/2012GL051711
McKeegan KD, Chaussidon M, Robert F (2000) Incorporation of short-lived 10Be in calcium-aluminum-rich inclusions from the Allende meteorite. Science 289: 1334-1337
Meibom A, Petaev MI, Krot AN, Keil K (2001) Beryllium and boron measurements in QUE 94411 chondrules: possible constraints on formation temperature. Meteoritics and Planetary Science Supplement, 36, A130
Reeves H, Fowler WA, Hoyle F (1970) Galactic cosmic-ray origin of Li, Be, and B in stars. Nature 226: 727-729
Shearer CK (2002) The behavior of beryllium during solar system and planetary evolution: evidence from planetary materials. Reviews in Mineralogy and Geochemistry 50: 77-120
Shearer CK, Layne GD, Papike JJ (1994) The systematics of light lithophile elements (Li, Be, and B) in lunar picritic glasses: implications for basaltic magmatism on the Moon and the origin of the Moon. Geochimica et Cosmochimica Acta 58: 5349-5362
Srinivasan G, Chaussidon M (2013) Constraints on 10 Be and 41 Ca distribution in the early solar system from 26Al and 10Be studies of Efremovka CAIs. Earth and Planetary Science Letters 374: 11-23
Warren PH, Taylor GJ (2014) The Moon. In: Davis, AM (ed) Planets, Asteroids, Comets and the Solar System. Treatise on Geochemistry, Volume 2 (second edition), Elsevier, pp 213-250
Zhai M (1995) Boron cosmochemistry. Part II: Boron nucleosynthesis and condensation temperature. Meteoritics and Planetary Science 30: 733-737
Zhai M, Shaw DM (1994) Boron cosmochemistry. Part I: Boron in meteorites. Meteoritics and Planetary Science 29: 607-615
Zhai M, Nakamura E, Shaw DM, Nakano T (1996) Boron isotope ratios in meteorites and lunar rocks. Geochimica et Cosmochimica Acta 60: 4877-4881