In the Classroom
Periodic Tables of Elemental Abundance Steven I. Dutch* Department of Natural and Applied Sciences, University of Wisconsin–Green Bay, Green Bay, WI 54311-7001
In 1985 I was asked by a colleague to make a presentation to the local section of the American Chemical Society on some geological topic that might also be of interest to chemists. While researching the possibilities I read the articles by Lewis (1) and Wetherill (2) and was impressed with the way simple concepts in chemistry and physics could be applied to understand the composition of the solar system. I prepared a talk called “Recipe for a Small Planet”, combining ideas from those papers with other concepts of planetary geology, which was very successful. My colleague suggested I submit the talk as a possible speaking tour talk, which I did. I have since done four speaking tours for the American Chemical Society on a variety of topics, including “Recipe for a Small Planet”. One of the problems I faced in preparing my talk was illustrating elemental abundance patterns over many orders of magnitude in a way that portrayed both abundance and chemical affinities. The natural choice for abundance was a logarithmic scale. I decided to represent each element with a circle whose radius is proportional to the logarithm of its atomic abundance. The area of each circle is proportional to the square of its radius, so the visual impact of abundance actually increases faster than the logarithm, but of course not as fast as a linear scale. The figures represented in this article are recently updated and computer-generated versions of my original figures. These figures have been well received, and so I have decided to make them generally available for anyone who might find them useful. The figures are also available at http://www.uwgb.edu/~dutchs/geochem.htm and may be used freely for educational purposes provided the source is credited. Creating the Figures The most important step in creating the diagrams is finding good tables of elemental abundance. A brief scan of commonly available sources illustrates just how far we are from having extremely precise data on elemental abundances. Tabulations are commonly incomplete, and those tabulations that do attempt to be complete must rely on multiple sources that may vary in accuracy and methodology. Fortunately, the logarithmic scale used in these figures is not overly sensitive to variations less than an order of magnitude. A more serious problem is the tendency of many modern publications to interpolate rare elements on the basis of solar or chondritic composition. While this sort of interpolation is justified, it introduces a certain logical circularity into figures designed to show how solar, chondritic, lunar, and terrestrial elemental abundances are related. I initially used data from Ehmann (3), Goles (4), Green (5), Mason and Moore (6 ), and Turekian (7 ). For the figures presented here I used the more recent tabulations of Newsom (8). The overall appearance of the old and new diagrams is not greatly different. Additional articles relevant to cosmic and planetary abundances in the *Email:
[email protected].
356
Journal of Chemical Education include Douglas (9), Ehmann (10), Fleischer (11), Hendrickson (12), Huntress (13), Mueller (14), and Viola (15). If the data are not in atomic proportions initially, they can easily be converted using a spreadsheet (elemental data on a spreadsheet can be put to myriad educational uses and never go to waste). Solar and Stellar Abundance The sun is a typical star whose composition is known from spectroscopy. The composition of the sun in atomic abundance is shown in Figure 1. Silicon is commonly used as a basis for comparison because it allows convenient comparison of solar and planetary element abundances. From Figure 1 we observe four patterns: 1. An overwhelming abundance of light elements. 2. A strong preference for even-numbered elements. 3. A relative peak in abundance at iron, followed by a steady decrease. 4. Elements 3–5, lithium, beryllium, and boron, are very low in abundance.
These patterns reflect nucleosynthesis (element formation) in the stars (Cox [16 ]). A normal star like the sun converts hydrogen into helium by a series of proton–proton collisions followed by emission of particles. The sun cannot make elements heavier than helium; the other elements formed in earlier generations of stars. The sun is probably a third-generation star. (Incidentally, even though the sun cannot make elements heavier than helium, it can make use of those that are already there. One energy process in the sun called the CNO cycle adds four protons to a carbon nucleus to create a heavier nucleus, which splits to yield a helium nucleus plus carbon. In essence, the carbon is a nuclear catalyst for creating helium.)
Figure 1. Element abundances in the sun. Note the preponderance of light elements, the relative abundance peak at iron, and the preference for even-numbered elements.
Journal of Chemical Education • Vol. 76 No. 3 March 1999 • JChemEd.chem.wisc.edu
In the Classroom
After helium, there are two ways to form heavier nuclei. We could add particles one at a time, except that all nuclei of mass 5 are extremely unstable. Or we could collide helium atoms, except that all nuclei of mass 8 are also very unstable. At first glance it looks as if stars cannot create elements heavier than helium. However, when massive stars run low on hydrogen, they begin to collapse under their own gravity. Eventually, their interiors become so hot and dense that three-way collisions of helium nuclei can occur. The result is a nucleus with six protons and six neutrons: carbon. In still more massive stars, helium nuclei (alpha particles) collide with existing nuclei to make oxygen, neon, and so on. If it were not for the nuclear bottlenecks at mass 5 and 8, it might be much easier to create heavy nuclei and the universe might have run out of fuel for nucleosynthesis before we evolved to see it. The abundance of light nuclei is due to the increasing difficulty of building heavier nuclei in stars. The preference for even numbers is ultimately due to pairing of nucleons of opposing spin, one manifestation of which is the construction of nuclei by successive collisions of alpha particles. Oddnumbered elements form by collisions of single particles with nuclei. The jump from helium to carbon explains the rarity of lithium, beryllium, and boron. Lithium, beryllium, and boron form by spallation (knocking pieces off heavier nuclei) and tend also to be destroyed by nuclear reactions in stars. In the later stages of evolution of stars much more massive than the sun, additional cycles of nuclear fusion form elements heavier than carbon. The end result of fusion is iron, the most tightly bound nucleus. Iron nuclei cannot yield energy by either fusion or fission, and nuclei beyond iron form via two processes. One, the s-process (for slow) involves stray collisions between nuclei and other atomic particles. Obviously, the more particles added, the rarer the element will be. The other process is the r-process (for rapid—and how!). In a massive star the core eventually becomes so massive its gravity overcomes the electrical repulsion between nuclei and the core collapses to become a neutron star, effectively a giant atomic nucleus. The core of the star collapses in milliseconds and the surrounding star matter falls in. Several times the mass of the sun slams
Figure 2. Element abundances in chondritic meteorites, thought to be the most primitive remaining samples of inner-solar-system material. Note that light elements are depleted compared to the sun, because they mostly formed gases in the warm inner solar system, but the abundance pattern is otherwise a close match to the solar abundance pattern. Chondrite meteorite material is thought to reflect the bulk composition of the inner planets, including Earth.
106
into a neutron star at a tenth of the speed of light. The results are, to put it mildly, impressive. Nuclear reactions run rampant, building nuclei as heavy as plutonium and probably far beyond (we still find tiny traces of the plutonium; see Hoffman and others [17]). The outer layers of the star are blasted off and we look straight into the thermonuclear core of the star. The star brightens by hundreds of millions of times to become a type II supernova. The relative abundance peak of iron is due to its being the most tightly bound nucleus and the end of stellar energy production. The tailing off of heavy elements beyond iron is due to the steadily increasing difficulty of constructing heavy nuclei in both the s- and r-processes. The Earth After any large construction job, there are always scraps of building material lying about. If the sun and solar system formed from the same material, we would expect the raw material of the planets to match the composition of the sun, minus those elements that would remain as gases (Cameron [18]; Lewis [1]; Taylor [19]). We find a good match to the composition of the inner planets in a class of stony meteorites called chondrites, which are thought to be the most primitive remaining inner solar system material (Fig. 2). Chondrites are considered the raw material of the inner solar system and probably reflect the bulk composition of Earth (Grosman [20]; Kerridge and Matthews [21]). About halfway between Mars and Jupiter, it is cold enough for ice to persist even in a vacuum, so the moons of the outer planets are made of both rock and ice. On the outermost planets, methane, ammonia, and finally nitrogen also condensed as solids. Comets are probably leftover raw material from the cold outermost solar system, which is one reason they are so intensively studied. Comets are to the outermost solar system what chondrites are to the inner. The continental crust of the earth differs radically from the overall composition of the earth (Fig. 3). The elements across the middle of the periodic table, the transition elements, are very depleted relative to the earth as a whole; and
Figure 3. Element abundances in the continental crust of the earth differ radically from the chondritic abundance pattern. Large cations, especially those with large ionic charges, are dramatically enriched while transition metals are strongly depleted. This pattern reflects accumulation of large cations in the crust owing to repeated partial melting of rocks in the earth’s interior.
JChemEd.chem.wisc.edu • Vol. 76 No. 3 March 1999 • Journal of Chemical Education
357
Sg
In the Classroom
certain elements very much enriched, notably, the alkali metals, lead, uranium, and some rare earths. The transition elements are largely in the earth’s mantle and core. The enriched elements tend to be large cations, especially those with large electrical charges. This pattern reflects repeated partial melting of the mantle and accumulation of the residue in the crust. The mantle is made up mostly of silicates of iron and magnesium, ions with medium-sized radii and ionic charges of +2. Ions with large radii cannot fit mechanically very well into the mantle minerals and tend to escape at every opportunity, typically by concentration in partial melts. Ions with large charges also require extensive substitution to be incorporated, and they also accumulate in melts. Plate tectonics and repeated partial melting have created the granitic continental crust of the earth. Cox (22) is a good recent reference on terrestrial geochemistry. The Moon In composition, the moon looks quite different from chondritic material and the continental crust (Fig. 4). Hydrogen is enormously depleted, reflecting the almost total absence of water on the moon. Among the metals, the right side of the periodic table is strongly depleted. These elements tend to be more volatile, and the moon is strongly depleted in volatile materials. Generally, the lower the boiling point of an element in a vacuum, the lower its abundance on the moon. This pattern suggests the moon formed in a hotter region of the solar system than did the earth. In the last decade, a catastrophic origin for the moon has become widely accepted: the early earth suffered a grazing impact with a Mars-sized protoplanet and some of the ejecta accreted in orbit around the earth to form the moon (Hartmann [23]; Taylor [24 ]). Computer simulations of solar system accretion suggest that the early solar system did not accrete nine large planets but hundreds of Moon-to-Mars-sized protoplanets, which then accreted in a series of large collisions (Wetherill [2]). A collision origin for the moon avoids the fatal physical flaws that plagued all previous theories of lunar origin, and also could account for the chemical evidence for formation in a hotter environment than the earth. The heat of the impact would have allowed an additional opportunity for volatile elements to escape, as well. Large catastrophic impacts also imply that the planets may be composed of mixtures of materials formed in a variety of temperature zones in the early solar system. It is hardly surprising that the details of planetary chemistry and evolution are far more complex than the simple picture presented here. Rather, the astonishing thing is that simple physical and chemical reasoning can go so far in explaining the observed composition and geology of the inner solar system. Literature Cited 1. Lewis, J. S. Sci. Am. 1974, 230, 50. 2. Wetherill, G. W. Sci. Am. 1981, 244, 162–170.
358
Figure 4. Elemental abundances on the moon. Elements to the right of nickel are strongly depleted relative to the solar and chondritic abundances, whereas some heavier elements like molybdenum are enriched. Volatile elements are systematically depleted on the moon; the lower the vaporization temperature of an element, the more strongly depleted it is. Depletion of volatile elements is one line of evidence that suggests the moon formed as a separate body in a hotter part of the inner solar system than the earth.
3. Ehmann, W. D. In Encyclopedia of the Chemical Elements; Hampel, C. A., Ed.; Reinhold: New York, 1969; pp 567–576. 4. Goles, G. G. In Handbook of Geochemistry, Vol. 1; Wedepohl, K. H., Ed.; Springer: Berlin, 1969; pp116–133. 5. Green, J. Elements: In Encyclopedia of Geochemistry and Environmental Sciences; Fairbridge, R. W., Ed.; Bowden, Hutchinson and Ross: Stroudsburg, PA, 1972; pp 268–300. 6. Mason, B.; Moore, C. B. Principles of Geochemistry, 4th ed.; Wiley: New York, 1982. 7. Turekian, K. K. Chemistry of the Earth; Holt, Rinehart and Winston: New York, 1972. 8. Newsom, H. E. In AGU Reference Shelf 1, Global Earth Physics: A Handbook of Physical Constants; Ahrens, T. J., Ed.; American Geophysical Union: Washington, DC, 1995; pp 159–189. 9. Douglas, J. E. J. Chem. Educ. 1992, 69, 907. 10. Ehmann, William D. J. Chem. Educ. 1961, 38, 53. 11. Fleischer, M. J. Chem. Educ. 1954, 31, 446. 12. Hendrickson, E. J. Chem. Educ. 1988, 65, 986. 13. Huntress, W. T. Jr. J. Chem. Educ. 1976, 53, 204. 14. Mueller, Robert E. J. Chem. Educ. 1965, 42, 294. 15. Viola, V. E. J. Chem. Educ. 1994, 71, 840. 16. Cox, P. A. The Elements: Their Origin, Abundance and Distribution; Oxford University Press: New York, 1989. 17. Hoffman, D. C.; Lawrence, F. O.; Mewherter, J. L. Nature 1971, 234, 132. 18. Cameron, A. G. W. Sci. Am. 1975, 233, 32. 19. Taylor, S. R. Solar System Evolution: A New Perspective: An Inquiry into the Chemical Composition, Origin and Evolution of the Solar System; Cambridge University Press: New York, 1992. 20. Grosman, L. Sci. Am. 1975, 232, 30. 21. Kerridge, J. F.; Matthews, M. S. Meteorites and the Early Solar System; University of Arizona Press: Tucson, 1988. 22. Cox, P. A. The Elements on Earth: Inorganic Chemistry in the Environment; Oxford University Press: New York, 1995. 23. Hartmann, W. K. Nat. Hist. 1989, November, 68. 24. Taylor, J. G. Sci. Am. 1994, 271, 40.
Journal of Chemical Education • Vol. 76 No. 3 March 1999 • JChemEd.chem.wisc.edu
