That the glass would melt in heat,
That the water would freeze in cold,
Shows that this object is merely a state,
One of many, between two poles.
—Wallace Stevens, “The Glass of Water”
To someone who lives in the temperate north, the event seems simple enough. When the air turns cold, the water in ponds and brooks, lakes, rivers, ditches, and water glasses freezes on top. If the air stays cold, the ice on top gets thicker. It may get thick enough, in lakes and ponds at least, that as the Christian knight told the disbelieving Saracen in Sir Walter Scott’s novel The Talisman, a horse could carry you “over as wide a lake as thou seest yonder spread out behind us, yet not wet one hair above his hoof.” To the Saracen, this preposterous claim was evidence that the knight belonged to a nation that loves to laugh. “Neither the Dead Sea,” he scoffed, “nor any of the seven oceans which environ the earth will endure on the surface the pressure of a horse’s foot.” He was misinformed; several of the Earth’s seas and many thousands of its lakes and ponds can take the pressure of a horse’s foot in winter, and more besides. Yet there are things that happen when a body of water turns to ice that seem almost as improbable, when you find out about them, as the knight’s claim, made under a knot of palm trees by a plashing fountain, seemed to the Saracen.
The most improbable thing is that the ice is on top. Almost all other substances are denser and heavier as solids than as liquids, typically 10 to 20 percent heavier, but ice is about 9 percent lighter. Drop other solids in tubs filled with their liquid form, and they will sink; drop ice in (or shove it under, as Galileo did), and it will float (or bob up). As temperatures get colder, molecules generally move around less vigorously and pack more closely together than at warm temperatures, but when water gets cold enough to freeze, the molecules move apart. Lucky for us.
As the water freezes, molecules take up precise positions in a lattice, repeated in all directions and over many layers, to form an ice crystal. The pattern is hexagonal, with an oxygen atom and its two attached hydrogen atoms occupying each of the six corners.* A single layer of identical hexagons would look like tiles on a bathroom floor, or maybe a chicken-wire fence. Stacked layers produce rows of identical hexagonal cages. Because of the angle at which bonds hold the molecules apart, each cage has a large void in it. It is this openness, the hole at the heart of each submicroscopic hexagon, that accounts for the fact that ice floats. If it did not, we probably wouldn’t be here to notice. Lakes and seas would freeze from the bottom up, and once seas froze they would have a very hard time melting. Aquatic life over much of the world wouldn’t survive or wouldn’t have gotten started in the first place. Human life might not have evolved.
Another improbability related to water’s freezing is that we get to see it happen. If water were like most other substances, it would freeze at much lower temperatures than it does (by one estimate, at about -150F instead of 32F) and boil at much lower temperatures (by one estimate, at -112F instead of 212F). Ordinarily, molecules in a liquid move about randomly, bumping into one another and bouncing away, repelled on contact. But water molecules often stick to others because of the unusual way they are charged, positively at one end and negatively at the other.* Since opposites attract, water molecules link up readily in the liquid, an oxygen atom in one molecule bonding to a hydrogen atom in another.
These hydrogen-to-oxygen bonds between molecules are called “hydrogen bonds.” And although they are much weaker than the hydrogen-to-oxygen bonds within molecules, they are strong enough to account for most of water’s strange properties, including its high heat capacity (think how slowly an ocean warms in summer and cools in autumn; think how far north the Gulf Stream goes and how much heat it still has to give to Europe when it gets there); its high surface tension (think of the globular perfection of a cloud droplet); its absorption of infrared radiation (think of water vapor’s large contribution to the greenhouse effect); and its high freezing and boiling points. The reason water has a high heat capacity is that it takes a lot of energy to break the hydrogen bonds between molecules and free the molecules to move faster. The reason water has a relatively high boiling point is that an enormous amount of heat is needed to rupture enough hydrogen bonds to turn the liquid into vapor.
And the reason that water has a high freezing point is that, as heat is withdrawn, the hydrogen bonds grow so much more robust that they lock molecules into positions in ice crystals at a higher temperature than would be expected if water were an ordinary substance. It is the strength of the hydrogen bonds, then, that enables water to convert to ice at the relatively benign (to warm-blooded creatures) temperature of 32F. It is their assertiveness that allows us to walk on frozen water and chop up pieces of it to drop into our drinks. Hydrogen bonds put ice in our world. Then we were on the roof of the lake.
The ice looked like a photograph of water.
—David Berman, “Snow”
One winter, I rented a cabin in Elkins, New Hampshire, so I could watch a lake freeze. Although I grew up in the temperate north (Ohio) and had seen some puddles and brooks ice over, I had never seen a larger body of water freeze up. The lake was Pleasant Lake, probably named in summer by someone with a hamper full of beer; in winter, temperatures around it run as low as -30F. The lake is a mile and a half long by three-quarters of a mile wide (narrow enough that at 10 o’clock each morning for several days straight I saw a deer appear on my side of the lake and swim across to the other side, which oddly looked exactly like my side) and over 100 feet deep at its deepest point. The long line of the lake lies on a northwest-to-southeast heading, the way the North American ice sheet was going when it scooped out the basin for it 20,000 years ago, and the way the prevailing winds blow now, toward the post office and town beach.
The lake is also only a short drive away from one of the world’s premier centers for the study of ice, snow, and frozen ground, the U.S. Army’s Cold Regions Research and Engineering Laboratory (CRREL), whose declared mission is to solve technical problems in cold regions, most of which “arise because water changes into ice.” The scientists and engineers at CRREL work on such problems as how to build a bridge over weak ice; what’s the thickest ice that submarine conning towers can break through; how to keep helicopter blades from icing up; and what composite makes the best gliding surface for skis. Along the way they also do basic work, on, for instance, how glaciers move over their beds and salt drains out of sea ice. I hoped they could answer some of my noodling questions about what was going on at Pleasant Lake. Ice is a fit subject for contemplation.
—Henry David Thoreau, Walden
It could be like watching paint dry, I am thinking, or maybe an orange grow old, the object of my attention just gradually getting stiff. I am warned that any major changes will probably take place in the middle of the night, when temperatures usually run lowest and I won’t be able to see much. I am prepared for only modest thrills. I read up.
I learn that before a lake—any small, freshwater, temperate lake—freezes, the water in it usually “turns over.” In summer, the top layer, heated by the sun and therefore lighter than the water below, stays on top, except as winds disturb it. But as the days grow shorter and the air cools, the top layer of water cools too, growing denser and heavier until it sinks and is replaced by another layer, which cools in contact with the air, sinks, and is followed by another layer, which upon chilling descends with the rest. Eventually the entire water column, or the entire upper part of it, reaches the same temperature, the one at which water is at its densest and heaviest, 39F—not, as would be expected if water were an ordinary substance, at 32F.
This improbability was recognized at least 300 years ago, although not the reason for it. In 1870, William Roentgen, the discoverer of X-rays, seeking a reason, proposed that liquid water contains two kinds of molecules, simple molecules of water and molecules of ice. As water temperatures fall, the proportion of ice molecules will rise, and since ice molecules take up more space than simple molecules, the water will expand in the cold. He was not the first to advance the idea that there’s ice in water. A dozen years earlier, H. A. Rowland, after noting the “remarkable” fact of water’s contraction at 4C (39F), observed that “the water hardly seems to have recovered from freezing.”
Some scientists still hold the view that water is a partially broken-down ice structure. In his book Meditations at 10,000 Feet, physicist James S. Trefil writes that he likes to express the view by saying that “water never quite forgets that it was once ice.” “The positions that water molecules take in ice crystals,” he explains, “are duplicated in the liquid.” Although the molecules hold these positions only briefly, with any individual molecule joining an icelike structure then “flitting” to another, there will be more of these structures to flit to as the water temperature gets closer to freezing.
Nobody can yet prove Trefil and Roentgen right or wrong. Even at this late date, scientists don’t know what the structure of water is. (The structure of ice, by contrast, is well known, all molecules in their assigned places.) “There’s a lot of dissension,” Stuart A. Rice, professor of chemistry at the University of Chicago, points out. “The volume of the literature on water is like an ocean in itself.” One reason for the uncertainty is that the current methods of studying water, X-ray and neutron diffraction, don’t give a precise picture. “It’s like looking at a statue where the arms are moving,” Sidney W. Benson, Distinguished Professor of Chemistry, Emeritus, of the University of Southern California in Los Angeles, states. “You can measure the distance between the shoulder and the elbow and between the fingers and the toes, but you get a blurry pattern. The molecules don’t keep their orientation in space. They break apart in nanoseconds.”
Pronouncing the idea of water as a “disordered” ice structure “dead as a doornail,” Benson worked out his own theoretical model. In it, water consists of cubes and rings “at all temperatures up to and beyond the boiling point.” Some of the cubes join together, forcing each other into loose assemblies; “two cubes might be joined at a corner or an edge.” The assemblies have big holes in them, so they take up more room than cubes. At 39F, “the magic number for water,” he says, the open structures “start to dominate.”
Other models of water include one where it contains a mix of five different species of molecules—having anywhere from zero to six hydrogen bonds each—and one in which simple water molecules inhabit the cavities of hydrogen-bonded clusters of molecules. Rice and Mark G. Sceats, now at the University of Sydney in Australia, have developed a “random network model,” in which the hydrogen bonds are continuous (they don’t break) but distorted (they bend and twist). Bonds that are strong and taut and linear in ice are relaxed and floppy in water, Rice explains. The twisting of the bonds creates strain, and the strain causes energies and frequencies in the network to shift, so that many different configurations form, with many different angles between them. Yet the molecules are connected, Rice notes, “like a bedspring.”
Whatever the structure of water turns out to be—a Hungarian scientist is reportedly working on an electron diffraction system that might sharpen the picture—there is general agreement that the hydrogen-bonded association of molecules in the liquid is responsible for its being densest and heaviest at 39F. Below 39F, water expands because the greater clustering effect from increased hydrogen bonding between slowed-down molecules overrides the more-efficient-packing effect of the slowed-down molecules.* Above 39F, the water expands because molecules are moving faster and taking up more room, and hydrogen bonds are being broken so there are fewer clusters. The water keeps on expanding as it is heated, all the way to the boiling point. thou the waters warp
—William Shakespeare, As You Like It
Once the temperature of the water column in a lake reaches 39F, any water on top that’s chilled by the air will be lighter than the water underneath and won’t sink but will stay on top, in excellent position to be further chilled by the air. It can take months for a lake to turn over, but once it has, the surface layer, in constant contact with the wintry air, can give up a lot of heat in a hurry. In a small lake or pond, all it may take for the top layer—thinner than a birch leaf floating on water, a few ten-thousandths of an inch deep—to change into ice is a single calm, cold, clear night. The brief sun flames the ice, on pond and ditches,
In windless cold
—T. S. Eliot, “Little Gidding”
The first time I go out to check on the lake is Thanksgiving Day, when every sensible person is in. Air temperature is 14F and the winds are gusting to 35 miles an hour, but the water temperature is 44F; the lake hasn’t turned over yet. Still, I find some ice, not on the lake but next to it. The winds have blown water off the lake surface onto rocks and low-hanging branches along the shore. Cold air, cold rocks: the water probably froze in a flash. On the lake, with the large storage tank of warmer water beneath it, the top won’t freeze so easily. The ice on the rocks is clear and shiny and looks like floor sealant. If the wind spraying goes on and on, I read, and the splash-ice builds in layers, it’s called “candle dipping.”
Water on the branches has frozen into icicles, which are as regularly spaced as rake tines or fringe on parlor lamp shades. When talking about ice, you can block that metaphor but not those similes; the material takes so many evocative forms that the images keep on coming, a few even making it into the scientific lexicon: pancake, grease, candle, bullet, plate, slob, honeycomb. Some of the icicles are two feet long and swinging in the gusts, like wind chimes, but without sound. I spot a twig with a blob of ice at the end, built around a bud. I break it off and suck on the ice like a lollipop, wondering: how can ice be so refreshing on a day so bitter?
The following day, air temperature is up to 37F, and the innocent prettiness of the splash-ice is gone. The slim tines, fringes, and chimes have fattened and coarsened into root vegetables, teats, and mittens. The sealant on rocks is cloudy now, like . . . like skim milk, or a blind man’s eye. The heat of the sun’s rays has opened up countless, minuscule cracks in the ice, which scatter later rays and turn the ice opaque.From the Hardcover edition.
Excerpted from Ice by Mariana Gosnell. Copyright © 2005 by Mariana Gosnell. Excerpted by permission of Knopf, a division of Random House, Inc. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.