REE in Space
Since they are some of the last elements to crystallize from a melt, and because they are sorted in such unique ways during geological processing of minerals, the distribution of rare earth elements in various bodies in the solar system provides important insight in to both what processes took place in the early solar system, and in what order those processes occured.
Rare Earth Elements: A Key to Understanding the Structure and History of Planets
Rare Earth Elements: A key to understanding geological sorting processes in the solar system.
By Robert E. Beauford 27 April, 2011
The subject of rare earth elements has burst into public consciousness recently with explosive increase in the demand for rare earth elements in high tech communications and green technology applications. Because of limited supplies, a lack of trained specialists outside of China, and the critical importance of these materials to modern society and technology, they have also become a very high research and funding priority of the US and other governments. (Haxel et al., 2006) Demand for knowledge of REEs in geoscience graduates is at an unparalleled high point, and is rising steadily. A basic knowledge of these materials, the geological processes that governs their distribution in the landscape, their role in modern technology, and their applications to the interpretation of geological processes, both on earth and space, is now a critical part of any planetary science or geoscience professional’s toolbox. (Long et al., 2010)
Figure 1: The rare earth elements are a series of 17 elements including scandium, yttrium, and the lanthanide series from lanthanum to lutetium. Periodic Table from MNDMF, Ontario Geological Survey, Kenora District, Recommendations for Exploration, untitled document, graphic attributed as unpublished report, Sinton, 2005
To the space and planetary scientist, these elements provide a powerful geochemical tool for understanding the processes of planetary differentiation and of igneous petrogenetic processes in deep or previously existing rock units we will never see. In a sense, the distribution of incompatible elements within a body provides a time capsule of geological processes in the early solar system and a real, though somewhat clouded, window into the actual movement of rocks and minerals as they were sorted in the proto planetary and early planetary environment. This paper will explore and clarify what these elements are, and in what context they serve some of these functions.
The Distribution of Rare Earth Elements in the Solar System
The rare earth elements are a series of 17 elements, specifically Scandium, Yttrium and the elements from Lanthanum to Lutetium, comprising the lanthanide series on the periodic table. Promethium, due to its unstable state, exists in such small quantities in the solar system at any given moment that it is usually neglected from consideration. Scandium is considered to be part of the group by the IUPAC, but it and Yttrium are rarely listed among catalogued rare earths in a geochemical interpretive capacity. (Connelly et al., 2005)
One of the first things pointed out in most papers and texts discussing the rare earth’s as a group, is the ‘fact’ that rare earths are not particularly rare. From a geoscience perspective, when discussing earth’s crustal rocks, this is basically true. From a space and planetary science perspective, however, the statement is neither accurate nor appropriate. Rare earth elements, in fact, are among the rarest elements in the solar system. They are also some of the rarest elements in the earth, when it is considered as whole. In fact, 12 of the 20 rarest stable elements in the solar system and in the earth are listed among the REE group. (Grevesse et al., 1988)
The term ‘rare earth,’ however, does not originate from this scarcity. Due to their responsiveness to geological sorting processes, these elements are significantly enriched (though seldom concentrated) in earth’s crustal rock. Instead, the term rare, when used in regards these elements, derives from how uncommon the host rocks in which these elements are concentrated occur. Rare earth concentrating minerals are primarily found in low temperature, late-crystallizing rocks formed by hydrothermal and igneous processes at the extreme margins of the planetary sorting processes. (Long et al., 2010) These environments include carbonatites and peralkaline igneous rocks. (Walters et al., 2010)
Outside of these rare concentrating environments, rare earth elements are primarily found as widely distributed, randomly trapped elements dispersed in silicate minerals in which they are very incompatible, and are present only as disruption in a crystal lattice. To a significant extent, the elements are merely trapped by the growing minerals rather than meaningfully incorporated. With only a few notable exceptions among the silicates, these elements are uniformly incompatible. Only in uncommon circumstances, and in a few low temperature minerals among the carbonatites and phosphates, will these elements actually form stable minerals. (Walters et al., 2010) Because these properties make their concentrations sensitive to very small changes in chemical compatibility, it is exactly this tendency of the REEs to be excluded and to be sparsely and randomly distributed in silicates that makes them so useful as a planetary scientist or geochemists tool.
Figure 2: In plots of the crustal abundances of minerals in the solar system and in the earth, the relatively unaltered distribution of the REEs in relation to each other, as nearly all other minerals shift in abundance and concentration around them, illustrates the incompatibility of the REEs and their resistance to capture and concentration in crystallizing minerals. Crustal abundances from: http://pubs.usgs.gov/fs/2002/fs087-02/
In earth’s upper crust, the more common REE’s are as or more common than nickel, copper, lead, or tin, and even the rarest of the REEs are more common than silver, gold, platinum, or mercury. (Long et al., 2010) They are primarily dispersed in the crustal rock, however, at low ppm levels of distribution, rather than concentrated like these comparative metals. Only in the most extraordinary geological circumstances are they concentrated to recoverable proportions in rock (Walters et al., 2010) and it is not likely that these circumstances have been achieved on any other rocky body in the solar system.
The active, long term, and ongoing recycling of earth’s crustal rock allows a degree of sorting of incompatible elements by repeated exclusion that is not duplicable in the short lived cycles of geological activity that seem to predominate in the other inner solar system planets, and certainly have not occurred in the frozen rocky moons or even the largest commentary or asteroidal bodies. As for the geologically active low temperature moons of the outer solar system, however, all bets are off. Substantial concentration of what are, within the silicate realms, incompatible elements, may be possible by both inclusion and exclusion in the poorly constrained hydrous and icy magmas of truly low temperature outer body lithologic cycles.
The notion that REEs move as a group in the mineral environment almost cannot be overstated. So extreme is this behavior, in comparison to most other minerals, that the largest difficulty in mining and recovery falls, not in locating or removing them from the ground, despite the rarity of their concentrated deposits, but rather in the process of sorting the elements from each other. (Long et al., 2010) Seven REEs were discovered and isolated from rocks at one mine in Ytterby, Sweden, where the elements shared and substituted in the crystal lattice of local minerals. (Kean, 2010) The history of REE discovery, between the late 1700s and mid-1900s, was by and large the long and tedious process of separating the elements. REE rich minerals are never found with a single REE. Rather, a half dozen or more elements will be found in a given mineral in mutual substitution.
What Sorts REEs? Fractional Crystallization and Partial Melting
The sorting of rare earth elements, at planetary and large regional scales, happens largely through fractional crystallization and partial melting during the heating and cooling of rocks.
Figure 3: Rare earth elements are incompatible in most common silicates. As a result, they will remain in the melt during fractional crystallization. Illustration modified from Wikimedia commons: http://en.wikipedia.org/wiki/File:Fractional_crystallization.svg
Fractional crystallization occurs as melted rocks cool in magma chambers or conduits below ground. Cooling magma will form specific minerals at different temperatures as the melt cools. The forming minerals will remove specific elements from the melt, leaving a remaining magma that differs in composition from the original. An example of this would be the formation of the high temperature mineral, olivine, which ranges in composition from Mg2SiO4 to Fe2SiO4 and crystallizes at temperatures from approximately 1900C (Mg2SiO4) to ~1200C (Fe2Si O4). (Foulger and Jurdy, 2007) As olivine crystallizes from a melt (magma), iron, magnesium, silicon, and oxygen, the components of olivine, are all going to be removed from the melt and sequestered in the mineral. The melt will thus be enriched in every other element, including REEs, compared to its state before olivine formed, and depleted in the olivine forming elements. Since silicon and oxygen are extremely abundant in the rocky bodies of the inner solar system, the oxygen and silicon depletion will probably have little effect upon subsequent mineral formation from the melt. If an adequate amount of iron and magnesium are removed, however, the resulting melt remainder may produce a very different group of minerals as it continues to cool.
The above example is oversimplified, since several minerals are generally crystallizing at once from a melt, but the example serves to illustrate the process. Two important secondary processes are occurring during fractional crystallization that will become very important to the understanding of the role of REE systematics in the interpretation of geological histories. First, iron and magnesium may be replaced, or substituted, by other elements in the crystal lattice that are almost, but not quite as good a fit. Manganese and nickel are common replacements in this particular example, and would be termed ‘compatible’ elements in olivine, since they fit readily into the crystal lattice. (Righter and Drake, 2004) The second major additional process that may be occurring during fractional crystallization is the trapping of incompatible elements. This can happen regardless of the ability of the element to fit into the forming crystal lattice. The forming crystals simply grow around and entrap a certain percentage of incompatible elements, which are awkwardly or incompletely bound to the surrounding atoms.
With both the wrong size and valence state to replace iron or magnesium in olivine crystals, most REEs will be pushed aside and remain in the melt, but a certain percentage will be trapped within the disrupted crystal lattice of the forming mineral, or in interstitial pockets of mixed minerals microscopically trapped between crystal grain boundaries and trapped within the rock as a whole. Because this is the only way that most silicates succeed in capturing the rare earths, the bulk of the earth’s crustal silicate rocks hold only small quantities of these elements.
The typical example used to illustrate this process is the making of rock candy. Water is boiled, and sugar (~C12H22O11) is dissolved in the water until it can absorb no more sugar. Food coloring is added to the water, and a string is suspended in it. As the water cools, sugar crystals, essentially a mineral, will form by fractional crystallization within this artificially created magma. (Pasachoff, 1996) A tiny amount of food coloring will be trapped within the crystal structure of the forming sugar crystals, imparting a color to the resulting candy. The food coloring is analogous to the distribution of REEs trapped within silicates such as olivine when they form. If the water (magma) is placed in a freezer, it will continue to crystallize out sugar as the temperature drops, and the remaining magma will become more and more depleted in the carbon used to form sugar. When most of the carbon is used up, and the temperature drops adequately, it will begin to crystallize water ice, and perhaps later, a minute amount of salt.
This is a directly analogous process to fractional crystallization. Temperatures drop, minerals crystallize, the composition of the remaining melt changes as elements are absorbed in minerals, and different minerals continue to form from the remaining melt at lower temperatures as the process continues. If this analogy is extended to the behavior of REEs in silicate rocks and melts, the REEs will consistently be excluded from the forming crystals and will be enriched in the remaining fluid.
The inverse, and corollary, of this function is the process of partial melting, in which rare earth elements are enriched by inclusion in melts as they are released from the heating of solid rocks. Rocks melt in the reverse order from which they crystallize. The last minerals to form as magma cools will be the first minerals to melt as it heats. Incompatible elements (such as REEs) in a rock environment will form the weakest links in a crystal structure, and thus the links that will fail first as temperatures rise. As a result, the REEs will join the lowest melting temperature rocks, and be squeezed out of the heated rock as magma of a composition that is distinctly different than the rock environment in which it formed.
Figure 4: Rare earth elements will be enriched in fluids removed during typical partial melting events, and the residue from which the melt was removed will be selectively enriched. Illustration modified from: http://www.nvcc.edu/home/cbentley/105/billy_goat_trip.htm
A good way to visualize partial melting is to imagine a cold plate of macaroni and cheese in a refrigerator. The mac and cheese dinner will start as a flexible, but solid mix of noodles and stiff, cooled cheese sauce. If the dish is heated, the cheese sauce will begin to soften and liquefy, but the noodles will remain much the same. After the dish has been heated, if the plate is tilted and the delicious snack is pressed firmly with a spatula, the rich, buttery cheese sauce, very different in composition than the still firm noodles, will run out of the mix, forming a pool of cheesy magma. If the plate is allowed to cool, the mineral that was once macaroni and cheese will have been differentiated, by the process of partial melting, to form two new minerals, noodlite and cheesite. If this analogy is extended to the partial melting of silicate rocks, the rare earth elements will always preferentially migrate to the melt, and will always be enriched in the magma that is removed compared to the original mix or remaining un-melted minerals.
Three broad statements can be derived from the behavior of the REEs in rocky planetary environments based on their behavior in response to fractional crystallization or partial melting.
Rare Earth Elements as Geochemical Tools
The utility of REEs in understanding the history of geological sorting processes and rock formation arises from several aspects of the chemical behavior of this elemental group. First, all of the REEs are very incompatible, meaning they are uncooperative components during mineral formation. They remain in magmatic melts when minerals crystallize with falling temperatures and are quickly freed from crystal lattices when temperatures rise. This has the effect or keeping the elements relatively evenly distributed within the silicates, relative to each other, even through generations of rock formation and destruction. More compatible materials are trapped out in minerals as they form, and stay there, essentially removing the elements from the overall elemental mix, and sequestering them. Secondly, the REEs move through the rock cycle as a group because of their very similar chemical properties, including an almost uniformly +3 valence, or charge, and their similar very large ionic radius. (webelements.com) Third, their relatively low tendency to deviate from chondritic ratios, and the predictable circumstances in which they will individually do so, makes them interpretable in relationship to each other, in order to pull out fine distinctions in the geological history of a rock group.
The exceptions to uniformity in rare earth element chemistry govern their usefulness as much as what they have in common. Two of the REEs, Cerium and Europium, have non +3 valence charges under certain common circumstances. Also, the differences in ionic radius across the group cause distinct, though subtle, differences in behavior. (White, 2009) This distinction is broadly described through the subdivision of the REEs into LREEs and HREEs, though the slight decrease in ionic radius is continuous across the group, as will be illustrated later. The LREEs, or Light Rare Earth Elements, are the larger and less compatible lower end of the spectrum from Lanthanum to Gadolinium. The HREEs are the more compatible, smaller ionic radius upper end of the spectrum from terbium to lutetium. Understanding the differences in chemical behavior that are caused by these variations comprises much of the interpretive tool kit of rare earth element trace element geochemistry.
In a general sense, it can be said that, due to their incompatibility and subtle variability in response to chemical environments, the REEs are 1) widely available In rock groups that are the target of analysis, 2) highly and predictably responsive and sensitive to sorting processes, and 3) carry in their distribution and relative abundance, a history of the sorting processes that they have undergone.
Because they are not absorbed in significant quantities into forming silicates, they remain present as a dispersed group, available in a wide range of rock types. More compatible trace elements are absorbed into host minerals and essentially removed from the mix at some stage, meaning they can’t be used in the same way. Since they are selectively enriched in any fluid that is left over after rock formation or that leaves a rock group when it is heated, their overall abundance can be used to evaluate the degree to which comparable rock groups have been processed, giving them responsiveness to change. Fluctuations from minor differences in chemical behavior between the individual elements can be used to interpret specific variations in the progenitor processes that result in a specific rock group that is variably enriched in certain REEs over other REEs. This provides the sensitivity to make them a useful metric of change. (White, 2009) Further, because their overall original abundance is known, the REEs can be used to propose not just a comparative history of rock groups, but the total extent to which an enriched rock group represents a portion of a larger pre-sorted whole.
Incompatibility and Partition Coefficients
Much mention has been made in this article of compatibility and incompatibility, and a further clarification of this concept is in order. Compatibility and incompatibility are expressed, in igneous petrology and geochemistry, through a number called a partition coefficient. The partition coefficient of an element in a specific mineral is an expression of its ability to stably form a part of a crystal lattice continuous with that mineral. Put differently, the partition coefficient of an element in a specific mineral represents the amount of an element that enters or remains in a melt during the heating or cooling, versus the portion that is captured within the mineral’s crystal lattice. (White, 2009)
(place holder for figure 5)
Figure 5: Rare earth elements have low partition coefficients in most common silicates because of their large ionic radius and +3 available charge. Note the size of the REEs compared to Fe and Mg. Eu and Ce are both capable of taking on different charges (+2 and +4 respectively) and substituting for ~similarly sized atoms in some silicate minerals. Illustration from http://lablemminglounge.blogspot.com/2010/11/rare-earth-revelry-week-one.html (<<<check out this web site. Its great!)
Figure 6(right): The source of common Eu spikes is obvious in the compatibility plot for plagioclase. The HREE are more compatible than the LREE in a wide range of silicates. Garnet uptakes significant quantities of HREEs. Illustration modified from Rollinson, 1993.
Partition coefficients are not numbers that are unique to specific elements, like their atomic mass, radius, or charge, but rather a number that expresses an element’s likelihood of behaving a particular way in a specific geochemical setting. These numbers are primarily a function of the ionic radius (actual physical size) of an atom of the element compared to the spot in a mineral’s crystal lattice into which it may or may not fit, combined with the atom’s valence electron state, or charge (usually +3) in relation to the charges of the atoms adjacent to the site in which the atom of the elements would have to fit under specific circumstances (ie oxygen fugacity). Some atoms just don’t fit into certain minerals. These are called incompatible. The degree to which they do not fit, and thus the improbability of their being found in the mineral, is expressed numerically as a number: the partition coefficient. (White, 2009)
An element with a low partition coefficient is more likely to remain in the melt, and a number with a higher partition coefficient is more likely to be captured in a crystal lattice and be present in a mineral. In other words, low partition coefficients for a particular mineral express incompatible elements and high partition coefficients express compatible elements. Anything with a partition coefficient greater than 1 is considered compatible. A partition coefficient less than 1 is considered incompatible, though this distinction is somewhat arbitrary.
General patterns of REE behavior and concentration
Figure 7: Enrichment of REEs in increasingly processed rock groups.
Constraining parent rock lithologies and petrogenetic environments of melt sources using REEs requires the comparison of often subtle changes within the distribution of the elements across the series, but not all of the interpretations are difficult. First, as previously mentioned, any degree of melting and sorting will tend to result in general REE enrichment in the lowest melting temperature products and in the most processed igneous rock units. As a result, a general pattern can be observed throughout the solar system. Volatile free CI abundances are enriched compared to overall solar system abundances. Rocky planets are enriched compared to CI chondrites. Crusts and mantles of planets are enriched compared to their overall composition. Rocky planetary crusts are enriched compared to their mantles. Highly sorted felsic rocks are more enriched compared to their mafic and ultra-mafic crustal progenitors, and so forth.
A second major trend in REE behavior as it relates to the constraining of source and derivative rock volumes, is the general enrichment or depletion of the HREEs relative to the LREEs. The ‘HREE’ refers to the ‘heavy’ rare earth elements, the group from terbium to lutetium. Heavy refer to their higher atomic mass (Z). (Walters et al., 2010) The LREE refer to the light rare earth elements, lanthanum to gadolinium. These constitute the lower atomic mass set among the overall series. As the atomic number of the REEs increases, there is a steady, though slight decline in ionic radius. Though the change is not dramatic, it has an appreciable effect upon the compatibility of the REE specie from the lower to upper atomic numbers. In other words, because the HREE are smaller, they fit better into the crystal lattices of forming silicates than the LREEs. This means that the HREE are depleted in subsequently forming rocks relative to the LREEs, thus skewing the distribution of REEs heavily towards the light end of the spectrum. The extreme version of this in silicates regards the formation of garnets, in which the HREE are fairly highly compatible by any standard. (see figure 6)
A third major trend in the utility of REEs for the interpretation of igneous petrogenetic relationships is the development of a Europium anomaly in response to the crystallization of plagioclase. While this function is very specific, it is visible in such a large percentage of REE plots that it is worth bringing up. Europium, as previously mentioned, is capable of taking on a +2 valence electron charge in addition to the +3 charge common to the other REEs. As a result, it is able to substitute for calcium in the plagioclase crystal lattice. Plagioclase feldspar is one of the most common minerals in the rocky parts of the solar system. Its crystallization and melting temperatures are moderate, so it is a frequent contributor to both melts and crystallizing volumes of rock during igneous processes. Since plagioclase sequesters Europium in its crystal lattice, rocks crystallized from a volume of melt after Plagioclase has been crystallized out in sufficient quantities, will show a downward ‘spike’ in REE plots where the Europium was removed. (Fowler and Doig, 1983) Similarly, a melt that has entrained previously crystallized plagioclase may show a positive spike in Europium compared to background levels of other REEs. See figure 6 and 11. To summarize, if a crystallization or melting event involved plagioclase, there will be a positive (melting) or negative (crystallization) europium anomaly.
These three examples are only a few of the many possible ways that tiny differences in Ionic radius (size) or charge can be seen to make big differences in compatibility in minerals and in subsequent elemental distribution ratios. These types of distinctions make REEs as a very sensitive tool for seeing and understanding differentiation at all levels from the planetary to the very local
The Zigzag, Normalization, and Normalizing to What?
Rare earth element abundances are almost always normalized before being compared for petrogenetic interpretation. This means that before two sets of rare earth element abundances are compared to each other, they are corrected to a common baseline. When comparing data, normalization is the process, in the broadest sense, of putting data in common terms. When applied to the interpretation of differences in REE distribution in rocks, it is a way of answering the question ‘Compared to what?’ The result of normalizing is that we compare how much two groups have varied from a common starting point rather than the final magnitude of the two groups.
Normalization serves two purposes. First, it removes variations in the starting points of the data sets that can hide meaningful changes behind ‘noise,’ and secondly, it chooses and establishes a specific common starting point against which changes in two or more data sets can be compared. Establishing a common starting point, against which to consider change, is vital for building meaningful relationships between connected events. If a person reports that car A traveled 8 kilometers, and that car B traveled 10 kilometers. This tells us absolutely nothing about the relative positions of these two cars or about the nature of the change that has occurred. If, however, a report states that car A and B started at the same point, and that A traveled 8 km to the north and B traveled 10 km to the north, we can compare the nature of these changes in a meaningful way.
Figure 8: Solar system elemental abundances in atoms per 1012 atoms of hydrogen. This illustration shows the zigzag pattern of elemental distribution in the solar system.
Rare earth elements are frequently compared by plots of abundances (see figure 8). These plots, if they are not normalized, are difficult to interpret visually. All of the rare earth elements were formed by processes of stellar nucleosynthesis in previous generations of stars, before the solar system cooled from a cloud of hot, nebular gas. The cosmochemical processes that created the mix of elements currently found in the solar system resulted in a strong preference for even numbered elements over odd numbered elements. (Ryan and Norton, 2010) The result is a zigzag pattern in the elemental abundances. This pattern overwhelms visual distinctions in an elemental plot. In addition, since these elements are enriched or depleted variously over time, establishing a particular starting time or common rock group against which to compare eliminates all changes prior to this selected start moment. Each of these changes eliminates noise (or meaningless variation) from the results, thus revealing significant and meaningful changes.
Figure 9: Illustration of the smoothing effect of the normalization process on an elemental abundance plot. In addition to making meaningful comparisons possible in order to interpret REE abundances, normalization visually pulls meaningful distinctions out of the background noise. (Dividing a set of numbers by itself is not, of course, technically normalization, but it nicely illustrates the process.) Chondrite abundances are from Korotev after Anders and Grevesse, 1989.
Normalization of REE elemental abundances is a simple process. It is accomplished by simply dividing the elemental abundances in the measured data set by the abundances of the standard to which we are normalizing. Anything can be used as a set against which to normalize, but terrestrial MORBs (mid-oceanic ridge basalts) and chondritic abundances are the most common. Of these two, chondritic elemental abundances are the more commonly used, since they are the closest analog we possess to the actual original abundances in the solar system before any sorting process took place. As such, they provide an excellent starting place from which to compare very different rock groups such as those from different planets or from very different geological settings. Normalized data should never be reported without clearly indicating to which exact set of numbers the data has been normalized.
Finding meaning in REE abundance data
Looking at real numbers in a couple of different ways will illustrate several of the processes discussed in this article. Figure 10 is a chart of rare earth element abundances compiled in parts per million, from real data, on 3 different rock groups. The first rock group listed is terrestrial (Earth) d-MORB or depleted mid-oceanic ridge basalt. This rock type is a rough approximation of the earth’s mantle. (Peltier, 1989) The second type is an average representation of the bulk continental crust of the earth, based on averaged representative rock groups. The third type is Lunar KREEP basalt. KREEP is a rare earth element enriched basalt type found on the moon by the Apollo missions. The data listed for purposes of normalization are elemental REE abundances from the CI chondrite type, which have been corrected to represent a volatile depleted source.
Figure 10: Real data. In this form, it is difficult to find meaning in REE data. The first three columns were each divided by the chondritic values listed in the fourth column to achieve the normalized figures on the right hand side of the chart.
One thing should be immediately apparent from this chart: no relationships between the numbers are immediately and intuitively apparent. A more detailed study will show several significant facts. First, no enrichment or depletion from element to element can be seen within the data sets or between data sets. The numbers, beyond their gross magnitudes, are a nearly meaningless jumble. Second, the zigzag pattern of solar abundances can be seen within several of the pre-normalized data sets in the alternating small and large numbers as we proceed down several of the left-hand columns. And third, normalization keeps the overall difference in magnitude between the data sets, but changes the distribution of numbers within the sets rather dramatically.
Figure 11: This chart illustrates the same data as in figure 10, but normalized and plotted by abundance. Each form of presentation offers different advantages, but plotting by abundance lends itself to petrogenetic interpretation.
Graphing the three columns on the right of the chart in figure 10 in a simple line-plot of abundances makes a totally new set of facts instantly and intuitively clear. As a result, meaningful interpretations of petrogenetic context for these rocks can be made. Both the average continental Earth crust and the Lunar KREEP basalt become obviously enriched compared to the d-Morb or chondritic norms, but the KREEP immediately jumps out as far more REE enriched than the terrestrial continental crust. Second, both the continental crust and the KREEP show an appreciable depletion of HREEs compared to LREEs. As we learned previously, this can be attributed to the crystallization of several different minerals during the history of the melts from which these rocks cooled and formed, including olivine, pyroxene, and garnet.
Third, there is an immediately obvious negative Europium spike in the KREEP basalts that is not present in the average continental crust of the Earth. This might be interpreted to mean a couple of things. For instance, it appears that all fractional crystallization that occurred prior to the sorting of terrestrial Earth crust from marine basalts seems to have occurred at well above the relatively low ~650C that is necessary for the crystallization of plagioclase. (Deer et al., 2001) This observation is borne out by brief consideration of the mafic and ultra-mafic materials of the oceanic crust of Earth compared to the generally grano-dioritic composition continental crust. We can also see that the lunar KREEP seems to be a derivative rock group that might have remained as a magma after comparatively low temperature fractional crystallization of a parent melt in a magma chamber removed substantial quantities of plagioclase.
It should be pointed out, here, that the point of playing out this scenario and drawing these conclusions is to illustrate how the REEs are used as interpretive tools. To establish, with any certainty, whether these are the right conclusions to be drawn from these particular sets of rocks would require a substantive search of literature and the placing of these data within a broader context. It should be apparent, however, that the REEs are a powerful tool, giving an instant and visually intuitive insight into the processual history of a rock group even with brief and basic plotting.
Figure 12: This chart illustrates the loss of information from the d-MORB line in the figure 11. In addition to correcting this problem by graphing the information on a separate and appropriate scale, the problem could be solved by using a logarithmic scale on the Y-axis or possibly by normalizing all of the data to a different standard.
One more thing about this rare earth element graph should be pointed out in an instructive capacity. If the relative magnitudes of the numbers involved are sufficiently different, data can be lost using this approach to visualization and interpretation. In figure 11, the d-MORB line appears to approximate the chondritic norm. This is only the result, however, of the much greater enrichment of the other two data sets by comparison. Plotting the d-MORB abundances alone (figure 12) gives a much different result, and invites additional insights and interpretations. This loss of information could also be corrected by using a logarithmic scale to express the ratios on the Y axis of figure 11.
Summary and Conclusion
The rare earth elements provide, to both the planetary scientist and the terrestrial geologist, a significant trace element geochemical tool set for understanding planetary differentiation and igneous petrogenetic processes. The collective low partition coefficients of these elements in most silicate environments, along with their variable compatibility from element to element, gives them a unique sensitivity to aspects of mineralogical change that are not otherwise easily deduced from a retrospective viewpoint. Their overall resistance to capture in high temperature silicate crystallization and their tendency to move as a unit due to similar large ionic radius and similar valence electron configurations also ensures their broad availability to the researcher in a variety of settings. Though there are several other incompatible elements in the periodic table that offer some of these characteristics, the REEs provide a distinct continuum in ionic radius that provides a type of environmental sensitivity not seen elsewhere in the elemental toolbox.
Anders E. and Grevesse N., 1989, Abundances of the elements: Meteoritic and solar, Geochimica et Cosmochimica Acta 53, 197-214. summarized and adjusted for volatiles by Randy Korotev at http://meteorites.wustl.edu/goodstuff/ree-chon.htm.
Beauford, R. E., 2011. Mining the Moon for Rare Earth Elements - Is It Really Possible, retrieved 26 April, 2011. http://www.rareearthelements.us/ree_in_space
Connelly, N. G., Damhus T., Hartshorn R. M., and Hutton A. T., (2005). Nomenclature of Inorganic Chemistry: IUPAC Recommendations 2005. Cambridge: RSC Publ. http://www.iupac.org/publications/books/rbook/Red_Book_2005.pdf. Retrieved 26 April, 2011.
Deer W. A., Howie R. A. and Zussman J., 2001. Rock Forming Minerals. Volume 4A. Framework Silicates: Feldspars 2nd ed. London, Bath: Geological Society of London.
Foulger G. R. and Jurdy D. M. (eds), 2007, Plates, Plumes and Planetary Processes, GSA Special Paper 430
Fowler A. D., and Doig R., 1983, The significance of europium anomalies in the REE spectra of granites and pegmatites, Mont Laurier, Quebec Geochimica et Cosmochimica Acta, Volume 47, Issue 6, June 1983, Pages 1131-1137.
Grevesse, N., Anders, E., Waddington J., (ed.) in Cosmic Abundances of Matter, Amer. Inst. Phys., New York, p. 1. (1988). referenced from Wikipedia data page retrieved 26 April, 2011; http://en.wikipedia.org/wiki/Abundances_of_the_elements_(data_page)
Haxel G, Hedrick J, Orris J. 2006. Rare earth elements critical resources for high technology. Reston (VA): United States Geological Survey. USGS Fact Sheet: 087‐02. Available at http://pubs.usgs.gov/fs/2002/fs087-02/fs087-02.pdf
Kean, S., 2010, Ytterby: The Tiny Swedish Island That Gave the Periodic Table Four Different Elements, Elements: Each one has a story, retrieved 26 April, 2011. http://www.slate.com/id/2258112/entry/2260782
Long, K.R., Van Gosen, B.S., Foley, N.K., and Cordier, Daniel. 2010. The principal rare earth elements deposits of the United States—A summary of domestic deposits and a global perspective: U.S. Geological Survey Scientific Investigations Report - 2010. Available at http://pubs.usgs.gov/sir/2010/5220/
Neal C. R., and Kramer G., 2003, The Composition of KREEP: A detailed study of KREEP Basalt 15386, LPS2003.
Pasachoff N., 1996, Marie Curie: and the Science of Radioactivity, Oxford University Press.
Peltier W. R., 1989, Mantle Convection, Plate Tectonics, and Global Dynamics, Gordon and Breach Science Publishers. Switzerland.
Rollinson H. R., 1993, Using Geochemical Data: Evaluation, Presentation, Interpretation, Longman, UK.
Ryan S. G., and Norton A. J., 2010, Stellar Evolution and Nucleosynthesis. Cambridge: Cambridge University Press.
Righer K., and Drake M. J., 2004, in Holland H. D. and Turekian K. K. (eds.), The Mantle and Core, Vol, 2, Treatise on Geochemistry, Elsevier, Amsterdam.
White W. M., 2009 Geochemistry, Chapter 7: Trace Elements, online textbook, http://www.imwa.info/white-geochemistry.html
Walters, A., Lusty, P., Chetwyn, C. and Hill, A., 2010: Rare Earth elements: Commodity profile. British Geological Survey. Available online at www.mineralsuk.com
Wedepohl H. K., 1995, The composition of the continental crust, Geochimica et Cosmochimica Acta, 59 (7), pp. 1217-1232. summarized on a periodic table at http://meteorites.wustl.edu/goodstuff/terrestrial_abundances.htm by Randy Korotev.
Workman and Hart, 2004, Major and Trace Element Composition of the Depleted MORB Mantle (DMM), http://www.whoi.edu/science/GG/people/shart/DMM.pdf;
The ratios of rare earth elements in chondrite meteorites provides an 'absolute' baseline against which terrestrial or other variations in REEs may be compared. The chondrite abundances represent approximately the original REE content of the solar system, and thus a pre-sorting figure for the earth, mars, and so on.
Variations among rare earth abundances have been used to build an understanding of how the highly differentiated howardite, eucrite and diogenite meteorites relate to one another and to their probable parent asteroids, Vesta and the Vestoids.
The rare earth elements, and in fact, every element in our solar system above silicon, in atomic weight, requires a larger and/or hotter sun than ours to form. Every atome from which we, the planets, and the sun are formed, originated in larger, more massive stars that predate our own solar system, These stars have long since run the cource of their lives, and our sun has been around for about 4.56 to 4.57 billion years. The ratios of the various elements in our solar system are determined by the combination of formations processes, meaning previous generations of stars, that contributed to the dust and gas within which our solar sytem condensed. Dust and gas in other parts of the galaxy have been contributed to by different combinations of large and small stars. A recent discovery showed a portion of space in which exceedingly high levels of REEs were present.
Rare Earth Element Space and Planetary Science Related Links