Microscopic astronomy: sophisticated laboratory techniques help divine the complex history of solar-system objects.
The split between mineralogists and astronomers continued to widen well into the 20th century, and it seemed that these scientists were forever destined to go their separate ways. They were unfamiliar with each others' work and only rarely attended the same scientific meetings.
However, the advent of the space age brought a reversal of this divergence. With lunar rocks and peculiar new meteorites awaiting study, the hybrid field of cosmochemistry began to flourish, bridging the gap between the disparate disciplines. It's still true that many of the chemists, mineralogists, petrologists, and physicists who dominate cosmochemistry have never looked through a telescope. Instead, their analytical tools are optical microscopes, electron microscopes, electron and ion microprobes, and mass spectrometers - devices completely alien to mountaintop domes. These are the indispensable workhorses of modern laboratories, and with them a generation of researchers has learned much about our solar system.
Our continued existence on Earth depends completely on the dependable light and warmth provided by the Sun, so it is hardly surprising that scientists have long worried about its constancy. The 11-year periodicity known as the sunspot cycle was recognized in 1843, but references to naked-eye sunspots were recorded before 800 B.C. in the Chinese Book of Changes.
Has the Sun's cyclic behavior been constant for the last 3,000 years? How about since the Sun settled on the main sequence? The answer to both questions is no. Between 1645 and 1715 almost no sunspots were seen. This period of low sunspot activity coincided with reports of a very weak corona during total solar eclipses and few reports of auroras from northern Europe. The inconstancy. of the Sun was established.
Tree rings provide additional evidence of solar variability. Favorable climate induces a tree to add thick layers around its trunk; drought causes thin layers to be added. A cross section of a trunk can thus be used to deduce climatic conditions in past centuries.
A tree trunk also bears a record of activity taking place high above its branches. The Earth is constantly bombarded by galactic cosmic rays - mainly fast-moving atomic nuclei ejected perhaps millions of years ago during supernova explosions. Upon hitting molecules in the atmosphere, cosmic rays produce showers of secondary particles: protons, neutrons, and mesons. Some of the neutrons are captured in the atmosphere by atoms of nitrogen-14 (having seven protons and seven neutrons), which then eject a proton and are transformed into carbon-14 (six protons and eight neutrons). Carbon-14 (14C) is a radioactive isotope with a half-life of 5,730 years, and it combines with molecular oxygen to form carbon dioxide (C[O.sub.2]). As trees take up C[O.sub.2] to aid photosynthesis, atoms of 14C eventually make their way into tree rings.
During periods of intense solar activity the Earth's magnetic field is carried farther into space, where it shields the atmosphere more effectively from cosmic rays. At these times the ratio of 14C to the stable isotope 12C measured in tree rings is low. When the Sun is quiet,more cosmic rays bombard the atmosphere and the ratio goes up. Thus, scientists find in the rings a record of solar activity stretching back thousands of years. Surprisingly, the Sun seems to be quiet most of the time, with only occasional spurts of high activity.
Solar activity has even been recorded in Precambrian siltstones that formed as sediments on the bottom of a glacial lake in South Australia some 680 million years ago. During those ancient summers, abundant glacial runoff made the lake turbulent and only coarse sedimentary particles were able to settle onto the lakebed. When the lake froze over in winter the water beneath calmed, allowing small particles to drift slowly to the bottom. Year after year, coarse layers alternated with fine ones, sediments that later were compressed into solid rock. Examination of the rock reveals sequences of coarse layers that were repeated every 11 years, evidence for an ancient sunspot cycle. Apparently, solar variability had affected the Earth's climate, indirectly causing periodic seasonal variations in the abundance of glacial meltwater at these sites.
Studies of returned lunar samples also tell us something about the Sun's past. Because it has virtually no atmosphere or magnetic field, the Moon is constantly bombarded by the solar wind. High-speed protons (hydrogen nuclei) and ions of heavier elements like helium, neon, argon, and krypton slam into the lunar surface and become implanted in exposed mineral grains. When propelled by solar flares to much greater energies, these ions can penetrate the lunar regolith ("soil") to depths of several centimeters and even trigger isotopic transmutations. Meteoritic impacts complicate the analysis somewhat by kicking some grains onto the surface and burying others. However, it appears that over the last three billion years the solar wind's isotopic composition has remained constant to within a factor of two or three.
Almost everything we know about the Moon has been learned since the first lunar samples were returned in July 1969. Of course, by then the Moon's surface had been photographed in great detail from the ground and by spacecraft. Astronomers were acquainted with details about the Moon's mass, bulk density, and the relative ages of major surface components. And thanks to the alpha-backscatter experiment on Surveyor 5, the dark surface of Mare Tranquillitatis was suspected of being basalt. But before Apollo the bulk composition of the lunar interior was unknown. The Moon's age was also unknown, and conflicting theories of the Moon's origin abounded.
Answers to many questions had to await laboratory analysis of the lunar samples. These include 382 kilograms of rock and soil collected by the Apollo astronauts at six sites from the center of the Moon's near side and 0.321 kg of material collected by three Soviet Luna robot landers from the near side's eastern edge. The samples also include (at this writing) 12 small lunar meteorites discovered in Antarctica and Australia, totaling 2.05 kg, that probably derived from at least six separate craters - some perhaps located on the Moon's far side.
The lunar surface is visually and compositionally divided into two major regions: dark maria and bright highlands. The maria cover 17 percent of the surface and occur within large impact basins. They are basaltic lava flows, analogous to those that cover the Earth's ocean floor Age-dating of returned mare basalts indicates that they formed between 3.2 and 4.2 billion years ago, though some unsampled lunar basalts may be as young as 2.5 billion year. The basalts formed by partial melting of iron- and magnesium-rich rocks in the lunar interior and were carried to the surface as rising magma. Maria cover relatively little of the far side (4 percent, versus 30 percent on the near side) because the far-side crust is much thicker (see page 32), which hindered the ascent of the basaltic magma.
The light-colored highlands consist of anorthosite, a low-density rock containing high concentrations of silicon, calcium, and aluminum. Lunar scientists believe that the outer regions of the early Moon were once completely molten, forming a vast magma ocean. Anorthosite crystallized from the magma and, because it was less dense than the liquid, floated upward to form the highland crust. If this scenario is true, the anorthosite should be quite ancient. Using the relative abundance of isotopes of rubidium and strontium, and of samarium and neodymium, geochemists have found that rare rock fragments among the highland samples are 4.3 to 4.5 billion years old - they are, indeed, pieces of the original lunar crust.
Analysis of lunar samples also indicates that the Moon is depleted in metallic elements such as iron, nickel, iridium, and platinum relative to the Earth. Therefore, if the Moon has an iron core it must be a rather small one, probably only 350 to 500 kilometers in radius. Yet paleomagnetic studies show that mineral grains within highland rocks crystallized in the presence of a magnetic field; this indicates that the Moon likely had a molten, convecting iron core 3 1/2 to 4 billion years ago.
But how did the Moon form in the first place? A crucial piece of information comes from analysis of the isotopic composition of its oxygen. This common element, with eight protons in its nucleus, has three stable isotopes: 16O, the most abundant, contains eight neutrons; 17O, the least abundant, contains nine neutrons; and 18O contains 10 neutrons. The ratios of these isotopes vary widely among bodies in the solar system (as seen at right), implying that objects forming at different heliocentric distances had their own distinct compositions. Interestingly, the oxygen isotopic composition of lunar samples matches that of the Earth, implying that the Moon formed in the Earth's vicinity.
Other constraints on lunar origin include the similarity of the Moon's bulk density to the gravity-corrected (uncompressed) value for the Earth's mantle, and the Moon's dearth of "volatile" elements (those that melt at low temperatures such as thallium, mercury, and bismuth), which indicates that either the Moon or the matter forming it was once very hot. The overall lunar composition is similar, but not identical, to that of the Earth's mantle; for example, their proportions of oxidized iron (FeO) are 13 and 8 percent, respectively.
Most planetary researchers now feel that the giant-impact model, sometimes dubbed the "Big Whack," satisfies all these constraints best. In this scenario, depicted at lower right, a Mars-size body struck the Earth with a glancing blow 4.4 to 4.5 billion years ago. (It's assumed that prior to the collision both bodies had differentiated, that is, melted to form a metal core and silicate mantle.) Upon impact the metal core of the projectile merged with that of the Earth. Meanwhile, vaporized silicate debris from the impactor's mantle was flung into Earth orbit, where it quickly cooled and accreted to form the Moon.
The similarity between the Moon Earth's oxygen-isotope ratios implies that the impactor must have originally accreted in our general neighborhood, about 1 astronomical unit (a.u.) from the Sun. The similar density of the Moon and Earth's mantle indicates that that lunar rocks consist mainly of silicates, with very little "core" material (dense metals) present -just as the giant-impact hypothesis predicts. The model also explains the lack of lunar volatiles, since these elements would have boiled off from the hot material flung into space.
However, the Big Whack does not provide all the answers, and more data will be needed before we can truly claim to understand the Moon's origin. If such an impact did occur, did it leave the Moon with enough iron to form a core? Why are certain siderophile elements like molybdenum, which have a natural affinity for iron, even rarer in lunar material than in the Earth's mantle? And do our lunar samples retain any geochemical signatures of the impacting object or its target? Microscopic astronomy may one day provide these answers.
Before Mariner 4's successful flyby of Mars in 1965, the red planet was a puzzling place. Telescopic observations had revealed a world with seasonally variable dark and light areas and white polar caps. The Mariners and Vikings showed Mars to be a world rich in geologic diversity, replete with impact craters, polar ice, dust deposits, volcanoes, landslides, sand dunes, and canyon systems. Networks of dry river valleys indicate that water once flowed on the ruddy surface; this in turn implies that billions of years ago the Martian atmosphere was thicker and surface temperatures appreciably higher than they are now. In 1976 the Viking landers measured the composition of the atmosphere and of the Martian soil. Their instruments sought, but did not find, evidence for Martian microorganisms or organic compounds.
Despite the deluge of data, much about Mars remained unknown. The landers obtained only limited compositional information; more sophisticated analyses would require having pieces of Mars returned to Earth - presumably by expensive, roundtrip space missions.
As it happens, cratering events on Mars long ago have provided us with at least 10 Martian samples for free. These are the SNC (pronounced "snick") meteorites; the acronym derives from three distinct igneous meteorite groups: shergottites (basalts consisting of the minerals pyroxene and plagioclase), nakhlites (mainly pyroxene) and Chassigny (a unique meteorite dominated by the iron- and magnesium-rich silicate mineral olivine).
We cannot be certain that these meteorites are from Mars, but the evidence is strong. First, the SNCs have relatively young crystallization ages, about 1.3 billion years (though some samples may have crystallized as recently as 180 million years ago). The paucity of impact craters on Mars's largest volcanoes suggests these towering peaks were probably still active a billion years ago; hence volcanic rocks could have been produced on Mars at the time that the SNC samples crystallized. In contrast, asteroids are so small that they must have radiated away all of their heat shortly after forming; igneous rocks younger than about 4.5 billion years are not to be expected on asteroids unless they are impact melts.
Second, the relative isotopic abundances of nitrogen and argon trapped within glass in some shergottites are the same as those in the Martian atmosphere as measured by Viking. Third, highly oxidized (ferric) iron, the primary component of rust, occurs within some minerals in SNC meteorites; this is the same component responsible for the red color of Martian soil.
If the SNC meteorites truly come from Mars, what have we learned about the planet? Isotopic and elemental analyses of these meteorites, coupled with compositional models of the magmas that created them, indicate that Mars differentiated about 4.5 billion years ago to form a sulfur-rich core and a mantle both denser and richer in iron than Earth's. Paleomagnetic studies suggest that Mars had an active magnetic field during the period when the SNC samples were crystallizing. Martian meteorites have oxygen-isotope ratios similar to those of the Earth and Moon, suggesting that oxygen in the inner solar system was relatively homogeneous when the planets formed.
Recently, curators at NASA's Johnson Space Center discovered that an unusual meteorite designated ALH 84001, collected in Antarctica a decade ago, is actually from Mars. However, unlike the other SNCs, which are relatively young basalts, ALH 84001 appears to be a piece of ancient Martian crust. The stone is all the more remarkable because it appears to have been immersed in liquid water at least once and perhaps several times (see page 12).
In January 1801 the Sicilian observer Giuseppe Piazzi chanced upon the first asteroid, Ceres, as a faint "star" moving slowly through Taurus. By 1847 seven more objects had been discovered orbiting between Jupiter and Mars, and since then only the year 1945 has gone by without adding to the thousands now known.
Asteroids are too small to be resolved (except by the Hubble Space Telescope), but over the years astronomers have learned much about them using remote-sensing techniques and, more recently, spacecraft flybys of 951 Gaspra and 243 Ida (January issue, page 20). Based on the spectra of light reflected from their surfaces, many asteroids appear compositionally quite similar to meteorites. Dynamical calculations prove that the gravity of Jupiter can redirect asteroids with heliocentric distances of 2.5 a.u. into Earth-crossing orbits. So the logical conclusion is that the chunks of iron and stone that fall to Earth daily have in fact migrated inward from the distant asteroid belt.
Cosmochemists have their own reasons for believing that most meteorites are derived from asteroids. For example, the abundance of noble gases implanted in meteorites that once lay on the surface of their parent bodies is close to what an object should accumulate in the middle of the asteroid belt, about 3 a.u. from the Sun. Also, most meteorites are extremely ancient, 4.55 billion years old, so they formed along with the planets and, by extension, the asteroids.
Most of the meteorites that fall to Earth are chondrites - compositionally primitive rocks rich in both silicates and metallic iron-nickel. Although chondrites have never melted, most show evidence of significant heating: mineral compositions are uniform, textures are recrystallized, igneous glass is absent, and much of their volatile elements has been lost. We don't know what heated the asteroids, but three promising candidates are the decay of short-lived radioactive isotopes (such as aluminum-26, with a half-life of 720,000 years), an intense T Tauri solar wind embedded in a magnetic field, and massive, energetic impacts.
Of the remaining chondrites that were never heated, some show evidence of extensive alteration by water. In these rocks tiny silicate grains and igneous glass have been transformed into claylike particles. One particular group of chondrites contains 17 percent water, all locked up inside the clay. The conclusion is unmistakable: some asteroids had sopping-wet surfaces.
A third geologic process that affects every asteroid is shock metamorphism - a point amply driven home by the Galileo spacecraft's recent photographs of Ida, which show it to be heavily cratered. Impacts create shock waves that deform crystals, break rocks apart, and cause localized melting of metal, sulfides, and silicates.
Studies of small metal grains in chondrites suggest that many asteroids must have been catastrophically disrupted by collisions and then gravitationally reassembled. The size and composition of grains of the iron-nickel mineral taenite are related to the rate at which they cooled from high temperatures. Based on this relationship, many meteorites cooled at rates between 1 [degrees] and 100 [degrees] Celsius per million years.
However, meteorites rich in solar-wind-implanted gases (and thus presumably from asteroids' surfaces) contain taenite grains with widely varying cooling rates: some dropped only a few tenths of a degree per million years, others did so thousands of times faster. The ones that took longest to cool must have resided near the center of an asteroid, the rapid ones near the surface.
The only plausible way to mix these dissimilar materials is to disrupt the asteroid by a large impact in a way that jumbled the fragments but kept most of them from reaching escape velocity. This inference is consistent with computer models of collisions, which indicate that many large asteroids should have suffered catastrophic disruption followed by gravitational reassembly.
COMETS AND BEYOND
Comets are the quintessential telescopic objects. The icy heart of a comet, the nucleus, is generally 5 to 20 km in diameter and shrouded by a huge, diffuse coma of expelled gas and dust. The first detailed image of a comet nucleus was not obtained until 1986, when the Giotto spacecraft flew within 600 km of Halley's. Comets often exhibit two rather distinct tails. The plasma tail consists mainly of ionized carbon monoxide (C[O.sup.+]), molecular nitrogen ([N.sub.2]), cyanogen ([C.sub.2][N.sub.2]), carbon dioxide, and water; it always points directly away from the Sun, pushed outward by the solar wind. The dust tail consists of fine grains emitted from the comet nucleus.
The idea that comet nuclei are "dirty snowballs" explains the spectral signatures of water ice and other volatile compounds detected in plasma tails. However, it is the dust that allows cosmochemists to discern additional details about the nature of the nucleus. Mass spectrometers flown on the Giotto and Vega spacecraft during their Halley rendezvous found a blizzard of microscopic dust grains rich in carbon, hydrogen, oxygen, and nitrogen. Magnesium-rich silicates such as olivine and pyroxene were abundant too, sulfur-bearing particles less so.
The Halley flybys may be fading into history, but cosmochemists can continue their cometary studies thanks to in-terplanetary dust particles. These tiny flecks of matter are collected high in the stratosphere by specially equipped U-2 aircraft, and they probably represent samples of both comets and asteroids. The "cometary" particles are highly porous aggregates of olivine, pyroxene, iron sulfide, glass, and carbon. Compositionally, they resemble Halley dust particles and chondritic meteorites, with elemental abundances similar to those of the Sun. These chemical resemblances underscore the primitive nature of such materials, and researchers are tantalized by the clues they might provide about the origin of the solar system.
Any deep-sky observer will tell you that the Horsehead Nebula in Orion, the Trifid Nebula in Sagittarius, and the Rosette Nebula in Monoceros surely rank among the most beautiful objects in the heavens. Each of these showpieces consists of thick concentrations of dust silhouetted against a bright background of starlit nebulosity. The dust consists of grains expelled from dying stars (mainly red giants) or material ejected from novae and supernovae from which dust can condense. Each source imprints the grains with distinct isotopic ratios of many different elements.
During star formation dust grains adjacent to the nascent star are destroyed and their isotopic compositions homogenized. This process occurred in our own solar system as well. On the other hand, some interstellar dust grains, particularly those at large heliocentric distances, could have survived the rigors of star formation and preserved their unique isotopic signatures. Recently, such grains have been extracted from primitive meteorites. They differ greatly in size, from about 10 angstroms ([10.sup.-9] meter) to 10 microns ([10.sup.-5] meter), and variously consist of carbon (as diamond and graphite), corundum ([Al.sub.2][O.sub.3]), silicon carbide (SiC), and titanium carbide (TiC). Titanium carbide occurs as tiny inclusions within silicon carbide, making these assemblages "interstellar grains within interstellar grains."
These five grain types involve hardy materials able to survive the burn-down-the-haystack-to-find-the-needle approach taken by cosmochemists in their destructive dissection of the meteorite hosts. Gentler procedures of meteorite dissection are being explored; these hold out the promise of discovery of new kinds of interstellar grains that will help make our sampling of the cosmos more complete.
Anders, E., and E. Zinner, "Interstellar grains in primitive meteorites: Diamond, silicon carbide, and graphite" (Meteoritics, September 1993, pages 490-514).
Friedman, H., Sun and Earth (Scientific American Books, 1986).
Heiken, G., D. Vaniman, and B. M. French, editors, Lunar Sourcebook: A User's Guide to the Moon (Cambridge University Press, 1991).
Norton, O. R., Rocks from Space (Mountain Press Publishing, 1994).
Warren, P. H., "Lunar and Martian meteorite delivery services" (Icarus, October 1994, pages 338-363).
Alan E. Rubin studies meteorites at the University of California, Los Angeles. He wrote "Whence Came the Moon?" in the November 1984 issue of this magazine.