22 March 2002More precious than diamonds, lumps of iron fall from the sky
by Art WinfreeThis column started in November with astronomy-related inquiries partly because this writer lacks background, experience, or advantage in this area. Naiveté can be an asset, since the interloper from over the hill is apt to see things from a peculiar viewpoint that sometimes supplements an otherwise-familiar perspective. But more importantly for present purposes, the advantage to a would-be columnist is that writing about fresh encounters with the unfamiliar cannot degenerate, as financial advice columns and scholarly literature often do, into a morass of mutually contradictory citations about nuances. Such writing therefore has a chance of being readable and memorable, and the questions raised might prove approachable without specialized background and sophisticated equipment.This writer sports innumerable areas of naiveté suitable to such a task. Lest just one of them bore you, we move now from the cosmos to things that it drops on our heads, about which I profess truly stunning ignorance, except for the little awareness acquired in the course of this engagement. For example, I read that meteorite iron was the only iron of decent quality available to the many civilizations predating the Iron Age. While celestial iron (the only kind in those days) played important symbolic roles in religion and art, its refashioning into knives and swords may have catalyzed the improvements of furnace technology that enabled transition from the Bronze Age to the Iron Age about 3000 years ago.
While visiting for the Tucson Gem and Mineral Show early February 2001, my brother Bob presented me a flattish 60 gram lump of rusted iron with the challenge to determine whether it is in fact a meteorite as purported by the Russians from whom he traded it. They said it is a fragment of the 23,000 kg meteorite
A 1957 postage stamp of the USSR, featuring a painting by artist PI Medvedev, who was beginning to sketch the Sikhote-Alin mountains when the unbelievable transpired.
that blasted holes in the Siberian tundra, just west of the Sikhote-Alin mountains and northeast of Vladivostok, on the morning of 12 Febr 1947, 10:38 AM local time. (Hmmm... why morning? Always? Do meteorites ever strike between noon and midnight, i.e., on the backside of the Earth as it speeds along its orbit?) Knowing nothing whatever of any of this, it seemed to me to have potential for a "GamesWorth" of investigation, according to my daily habit of one hour on something unfamiliar that that comes unbidden to my curious attention. I will tell this one just simply as a chronicle of my own experience, implicitly recommending you try something of the sort yourself to see how different is the outcome with your own meteorite. The next column will tell some of my own experiences with other meteorites recently acquired for diversification.
Given the rules of the game --- its just me against the lump, no books allowed until I have "engaged the enemy" --- the first hard question is of course, "How to even get started??" Well, simple things first, then maybe some better ideas will come. Any adequate plan you can implement today beats a later perfect plan. Test the basics: is this even iron? I think so: magnets love it. But it has no effect on a compass nor does it attract fine iron dust: it is not magnetized. After wire-brushing off a lot of what looks like rust, I determined the remaining lump's weight, then its volume by submersing it a cup of in water that I had so topped up that one more drop would overflow, and then topping it up again with drops from the captured overflow. The excess volume confirmed density near pure alpha-iron's 7.86 grams per cubic cm. So if it is some commercial alloy, the other metals are pretty similar or pretty dilute.
What else than iron might be in it? With a wire cutter I pinched off a silvery bit. It weighs 12.8 mg on my wife's gemological balance. Dissolved that bit in HCl. The yellow solution tests positive for iron, of course, with phenanthroline (turns blood red) and for nickel with dimethyl glyoxime (makes orange fluffy precipitate). By diluting pale green nickel chloride samples until they react similarly, I guessed [Ni] in range 5-15% of alloy weight. Conclusion: mostly iron, hammered and/or heated to lose of any magnetism, and at least 5 % nickel. Not unlike meteorites, I imagine, but maybe not unlike furnace slag, either. Nothing conclusive here, though different outcomes from these simple tests might have precluded interpretation as meteorite iron.
Well, what do its insides look like? Silvery, I already know from pinching off a corner. More ambitious now, I hacksawed off a bigger corner (see the mm scale) and polished the scar into a curved mirror. This revealed nothing in particular. How to develop some revelations? Might I see stress lines or inclusions or "cosmic ray damage" or something if I etch the polished surface? Exposed the whole piece to dilute HNO3 for a few minutes: this removes more rust and hopefully etches the polished part to develop any crystal grain boundaries, or who knows what. What turned up was unexpected: little pits and fine scratches, which do not show well here, but they seemed remarkably parallel. Now we have stumbled onto something potentially interesting. Every solid metal I ever examined consists of tiny grains (mm or less), each of which is a single crystal, with a definite orientation: i.e., commercial metals are always polycrystalline. Is it conceivable that this whole exposed area (several mm square) is just one grain of this (presumptively) polycrystalline metal lump?? That would be the biggest grain I ever saw, if so. Treating other pieces of iron similarly in my amateurish way, I see nothing of the sort: crystal grains are all microscopic, presumably due to rapid cooling from the melt. Either I am reading too much into these long parallel scratches, or this thing was cooled incredibly slowly.
Getting curious now and still more ambitious, I picked a side that is already tolerably flat and held it against a sander belt until too hot for my fingers, then immersed in cold water, and repeated until an area about 3x2 cm was nicely flat and shiny. Then a series of SiC sandpapers glued to a lathe-rotated plate brought up a polish. I took it as far as 1500 grit, spoiled only by isolated swirly scratches that I guess were made by grains of rust or inclusions that keep breaking off from edges and pits. As a "before" controlimage, I examined this surface in a surplus dissecting microscope that I had taken apart to clean the lenses and re-align. Then photographed by simply sticking my Nikon KoolPix digital camera against the eyepiece. And then painted 10% HN03 (this time in EtOH to minimize exposure to water) onto that mirror for 5 minutes until the former mirror surface looks dull. Again, why does it get dull? Maybe it is not etching at the same rate everywhere and the variations occur on an optically fine scale?
Amazing result! Abrasive scratches vanished, I guess as part of a thin layer that dissolved away, and the underlying plane reveals myriad fine very parallel lines crossing the whole face. There are also a few fainter sets of such parallel lines, each set keeping to one orientation all over the face. The whole 3x2 cm lump seems uniformly oriented! Can this entire face be exposing a single crystal of iron? That is something I never would have expected: my prize Discovery of the Year, if it is true. A crystal that big would take (I would imagine) years, at least, to grow from a cooling melt, and no one in the iron industry has cause to exercise that sort of patience.
So this no longer seems a likely product of Russian furnaces. But if this little lump has been sitting outdoors in the rain, then it must not have been made many decades ago. One does not find elemental iron in nature on our wet planet, made oxygen-rich by plant life. Oooo... "been made" or "been exposed to this environment"? Could it have arrived from a very different environment just a few decades ago?
But how to check further? Rotating the the glass plate that holds the lump on the microscope stage, I see that tiny pits in the etched face reflect light from a distant point source only at certain angles. Drawing an arrow on the plate, I mark its positions against the stage rim as I rotate . After a few full rotations I turn on the room lights to see where the marks are. They cluster 90 degrees apart, as though the lump has cubic symmetry and my sanded face is fortuitously near a crystal plane! When I sanded it, naturally I chose the biggest, flattest area, but never suspected that might have been anything but a random surface determined by rusting. Is it conceivable that this lump broke off a bigger crystal nearly along a cleavage plane? Seems a wild guess, but it would account for there being a good choice for sanding, and that choice exposing a 4-fold symmetry like iron's cubic face-centered lattice.
Next I re-sanded and polished the whole piece, this time to expose also some sides perpendicular to the 3x2 cm face: if these also have scratch lines (they might not), they might run in all directions, or if parallel, at some peculiar tilt to the big flat surface,, and that will be the end of fanciful speculations about a giant crystal of iron fortuitously oriented to the flat side of the lump. But guess what? There are etched lines and they are perfectly parallel and they do run nearly perpendicular to the originally polished face into the exposed "depth"!
Because the whole lump has 90-degree symmetries aligned to the widest flattest originally-rusty surface, I continue to guess it is a single crystal grain. And I guess it really is from space, since I really doubt that any such big crystal could grow on Earth, and if it did, I can't imagine it lasting very long outdoors without rusting away.
In this respect it reminds me of diamond. Both elements can be romantically thought of as made in the heart of a star and delivered to us by supernova. Microscopic nuclei for diamonds grown on Earth may even have arrived in meteorites billions of years ago. The dispersed atoms of carbon that crystallized into visible diamonds presumably came from space, like the crystal nuclei, but they aggregated to macroscopic proportions only 1-2 billion years ago, so far as anyone can decipher today (Science 285, 851 (6 Aug 1999)). While this makes a nice symbol of Eternal Love, my iron crystal beats it several-fold for endurance, having formed completely in space before the Earth even cooled, 4-5 billion years ago. Nothing like this could originate or endure on Earth. So I think engagement rings should be meteorite iron. Time to start a new DeBeers cartel and organize the advertising industry! No, not quite time: we need first to corner the market low-nickel irons at the next Tucson Gem and Mineral Show...
Here I confess I got too excited and forgot the discipline of solo investigation. I got a book from a web site about meteorites: Norton, Rocks from Space. It tells that iron meteorites (a 6% minority of observed falls) are indeed commonly single crystals, or at least those irons with less than 6% Ni content are. This "kamacite" is called for its cubic crystal habit "hexahedrite" because a cube viewed at random angle or right along the body diagonal looks hexagonal in silhouette. Its strangely parallel lines were first reported in 1848: "Neumann lines" of crystal twinning, presumably along to slippage planes induced by mechanical shock. This typically happens in a few directions at the same time, just as you might imagine the shearing of a stack of balls. It might be of interest to accurately determine the angles observed between such sets and see if they correspond to ways a face-centered cubic stack can easily shear.
Meteorites with more Ni than about 7% crystallize more complexly as a mixture of kamacite and an "octahedrite" called "taenite". The mosaic presents "Widmanstatten figures" where taenite and kamacite domains are mixed. None of those appear here, consistent with my lowest estimate, 5% Ni. According to more professional observations reported in Norton's wonderful book, Sikhote-Alin samples are 6% Ni, 1/2% Co, and 1/2% P. Their crystal habit typically lies just at the border between the coarsest octahedrite and marginal hexahedrite. The sample Bob gave me maybe is exceptional in being solidly hexahedrite (kamacite).
And Norton does tell that meteoric kamacite single crystals as big as a brick have been found .. well, at least one: a 7 kg piece from Calico Springs, Arkansas, first recognized as such in 1964. Divide by density 7.86: if this were a cube it would be about 10 cm on edge.
This topic will continue and end next time with an examination of some maybe-less-exceptional samples.
Meanwhile don't forget the Rainbow Moon that you might be able to witness, even to capture to graphics, on Monday 25 March then again on Sunday 3 March.
Copyright 2002 by A.T.Winfree. All rights reserved. Used by permission.
