An American company (which ASML subsequently bought) had already developed a unique means of producing this particular and peculiar type of EUV radiation. Some said the company’s method verged on the insane, and it is easy to see why.
Extremely pure metallic tin is heated until it becomes molten, and the hot liquid is then squirted out into a vacuum chamber in a tiny jet stream that looks continuous but is in fact composed of fifty thousand droplets moving past each second. The droplets themselves are then hit with light from a first laser, which pancakes each one, making a larger surface area for a second and very powerful carbon dioxide laser to irradiate each flat droplet—each of which turns instantly into a superheated plasma that emits a second jet stream of the wanted extreme ultraviolet radiation. (The bombarded droplets also produce fragments of waste tin, which might solidify were it not for a conveniently sited jet of hydrogen gas that casually brushes them out of the way.)
The EUV radiation that is born in this Hadean environment is then passed through the intricate masks on which the transistor arrays are drawn, that is, the new and ultra-tiny integrated circuit, after which it is moved down a staircase pathway of Bragg reflectors, each made to formidable optical precision, and onto the silicon wafer itself, to begin its work at mechanical tolerances of as little as seven-, maybe even five-billionths of a meter. If everything works properly—and at the time of this writing, it seems to be—then the first of these supercomplex chips, made in this bizarre manner, will be on offer from 2018 onward. And Moore’s law, by then fifty-three years old, will prove to have kept itself on target, again.
YET, ALL ARE asking, for how much longer? The use of EUV machines may allow the law’s continuance for a short while more, but then the buffers will surely be collided with, at full speed, and all will come to a shuddering halt. The jig, in other words, will soon be up. A Skylake transistor is only about one hundred atoms thick—and although the switching on and off that produces the ones and zeros that are the lifeblood of computing goes on as normal, the fact that such minute components contain so very few atoms makes the storage and usage of these digits increasingly difficult, steadily more elusive. There are plans for getting around the limits, for eking out a few more versions of what might be called “traditional” chips by, among other things, making the chips themselves increasingly three-dimensional—by stacking chip on top of chip and connecting each by forests of ultraprecisely aligned and very tiny wires. This would allow the number of transistors in a chip to keep on increasing for a while without our having to reduce the size of individual transistors.
And there are other materials, other architectures. There is talk of using the curious one-molecule-thick substance graphene, a filmlike, two-dimensional form of pure carbon, for the making of chips. Molybdenum disulfide, black phosphorus, and phosphorus-boron compounds are also being spoken of as alternatives to silicon, as means to keep the juggernaut of ever more miniaturization trundling along, to achieve whatever purpose is demanded of it. The ever-more-alluring field of quantum computing, which uses the weird ambiguities of the subatomic world, as described by Werner Heisenberg in 1927, as the basis for its abilities, is being touted as the next step.
Yet, down at this level, measurement becomes increasingly fluid, ambiguity transcends accuracy, precision wanders into the world of paradox, limits become meaningless, numbers vanish into a quantum-infused mist—except that there are some real numbers to be taken seriously. Perhaps most important, there is the so-called Planck length, the fixed and calculated dimension at which classical ideas of space-time begin to evaporate and the very idea of physical size becomes meaningless.
This length has an actual value—or, at least, it has a value if you believe that the two sure constants in our known universe, the speed of light and Newton’s gravitational constant, are immutably constant themselves. The Planck length has been worked out as 0.000 000 000 000 000 000 000 000 000 000 000 016 229 (38) meters, and so is about twenty decimal places smaller than the diameter of a hydrogen atom. And once you have that distance, you can work out an extent of time—if the same constants are similarly immutable, that is. And so the time it would take for a photon to journey through a Planck length can be calculated, and has been: the best estimate of this minute expanse of temporal extancy is 5.39 × 10−44 seconds.
And it is at this point that the story of precision becomes, quite literally, topsy-turvy. It becomes wholly impossible to go down, down, down beyond a certain point. Though techniques being studied at some of the national metrology centers and in a few high-octane national and university laboratories around the world allow for some penetration of the atomic limits—a technique known as light squeezing, for example, allows some actual measurement (rather than calculation, which is the basis of those two immensely small numbers just given) of subatomic-level dimension—there is a near-universally acknowledged limit below which things are unmeasurable, and therefore unmakeable.
GOING DOWN INTO the world of the near-atomically minute may have real and proven limitations, but at the other end of the spectrum there are still possibilities. The making of ultraprecisely finished devices and instruments still has validity at the other ends of this topsy-turvydom of extremes. It has value when it comes to the examining of the faraway—as with the James Webb Space Telescope, precisely honed to gaze into the edge of the universe. It has use and validity, too, in the examination of the big cosmological questions that haunt our modern imaginations.
This is why the most exacting limits of precision engineering are now being tested in the construction of the giant instruments at the LIGO sites in Washington State and in Louisiana, and that are about to be built in the plains of western India. LIGO may be the scalar opposite of the integrated circuit, being vast in every sense, the one extending across kilometers and the other covering mere nanometers, but much the same purity and exactitude obtain in the making of both, and perhaps even more dramatically so in the lonely outposts where LIGO is based, and from where it examines one of the enduring and fundamental questions of our cosmos.
EINSTEIN PREDICTED MORE than a century ago that faraway cosmic events could trigger ripples in the fabric of space-time—gravitational waves, he called them—and that they would change the shape of planet Earth if they ever passed by or through us. The LIGO sites were established to see if such infinitesimal changes in the world’s shape were measurable, whether they did in fact exist.
To demonstrate and prove such a tiny change in the shape of the planet required the building of an enormous and ultrasensitive interferometer. Hence, in 1991, the birth (or, more accurately, the government’s funding approval) of LIGO, the Laser Interferometer Gravitational-Wave Observatory, within which are components that lay claim to being the most precise objects ever made by humankind, and that demonstrate just why the utmost precision is needed to examine or create not only at the near-atomic limits of minuteness, but also at the massive scale and near-endless distance of the objects far off in the outer universe.
A classic interferometer uses a powerful light of a pure and known color of which one knows the wavelength. That light is shone through a lens toward a device, basically a half-silvered mirror, that splits the beam in two, exactly. These two tubular beams of pure red light are then directed along paths that are exactly ninety degrees from one another, and toward mirrors that will reflect the beams back toward the first splitting mirror, where they now recombine and are superimposed upon each other as they are directed toward a detector.
If the beams are of exactly the same length, the circular image of the recombined red light will be amplified; the light will be as bright as it was before its beams were split in two. On the other hand, if the two beams differ in length, they will destructively interfere with one another, and the detector will register rings of color that will tell the observers and analysts by how much the difference is.
LIGO is very basically an experiment that employs a pair (soon to be a trio) of enormous interferometers of th
is quite simple design. Anyone who has used an interferometer would easily recognize, if flying five miles high over the central desert of Washington State, or over the lush forests of south central Louisiana, the two LIGO instruments for what they are: the two long arms at precisely ninety degrees, the building where the two arms meet and where the splitting mirror must be, the extensions and smaller structures where the laser light source is housed and where the detectors and analytical devices are situated, desert scrub up north, beech-magnolia woodlands deep down in Dixie, each suggesting placid and undisturbed nature. The long die-straight pathways cut across their landscapes look Nazca Line–like, stunning in their incongruity.
The purpose of the LIGO experiments is to determine if those two long arms at each observatory change their lengths relative to one another—for, if they do, to the tiniest degree imaginable, then there is a chance that it was the passage through the planet of a gravitational wave that made them do so.
Down at ground level, the instruments are industrial-scale behemoths, with the arms (basically subway-size tubes that stretch into the invisible distance) connected where they join to congregations of humming engine work and electronics of bewildering complexity. Technologies that employ engine oil and technologies that employ silicon perform here in perfect symbiosis. So vacuum pumps pump, laser generators generate laser light, servomotors make microscopic tweaks, and computers in control rooms work through numberless days and nights to interpret the data that stream in as the beams race hither and yon, back and forth, hundreds of times each second, between the mirrors, all in the faintest imaginable hope that, once in a while, the tubes down which the laser beams are pulsing will change length relative to each other.
And this they did, on September 14, 2015, when observers made their first-ever detection of the phenomenon Einstein had predicted almost exactly a century before. The computers in the Livingston control room noticed it: an aberration, an oddity, a variance in the signal, at 05:51 on that Thursday morning, half an hour before local Louisiana sunrise, and with the bayou alligators still asleep. The observers there may have been weary, but this being part of a vast network of participants in what is known as the LIGO Scientific Collaboration, others around the world more bright-eyed and bushy-tailed noticed it, too, at more propitious hours. Back in Hanford, Washington, it would have been 03:51, the dead of night; but in Leibnitz, it was 12:51; in Delhi, 17:21; in Tokyo, 20:51; and at Monash University, in Melbourne, 22:51 in the late evening.
And in every squirrel hole out in the wide beyond, people noticed it. A sudden uptick in a signal was noted in Livingston and was duplicated exactly by the detectors at Hanford. Not that all the detectors were switched on: the observatories were in the middle of an engineering run, when for many months at a time the various components are sedulously checked for precision and accuracy. Normally—not that there is much normal in the world of gravitational waves—observers look out only during observing runs. Yet because nothing had been seen or heard during all the runs of the previous thirteen years—the first basic LIGO was built in the late 1990s and started looking for waves in 2002—and with hundreds of millions of dollars of taxpayer treasure having been spent, with nothing to show for it, there was a sense of, if not quiet desperation, then at least institutionalized eagerness for a result.
So, when the first message came in from a middle-of-the-night observer in Pasadena, headed “Very Interesting Event on E[ngineering] R[un] 8,” the community pricked up its ears and, as one, shifted into skeptical overdrive.
This could not possibly be, they said. The equipment was in mid-shakedown mode, they said, with machines certain to throw up spurious data from time to time. Besides, part of the system set up to avoid jumping the gun had observers and machines firing off what are called injections, that is, anonymous false results injected, as it were, into the system to keep all the astrophysicists on their intellectual toes.
LIGO has two observatories in the United States, one in Louisiana and this one, seen from the air, in the desert of central Washington State. A third is being constructed in an arid region in western India.
Days went by, then weeks and months, during which time people around the planet were canvassed. Did you send out an injection? each was asked. And as each responded in a cascade of negatives, and as the results from the two observatories and from other, smaller centers were parsed over and over by analysts and mathematicians of ever-increasing skill and learning and wisdom, the skepticism gradually fell away. The LIGO meisters realized they had a story on their hands. They presented a scientific paper in Physical Review Letters, and then, at a crowded press conference in Washington, DC, on February 11, 2016, made an announcement that would shake, or at least stir, the scientific world—and much of the lay world besides.
After a courteous introduction by the director of the National Science Foundation (which took the greatest series of financial risks in its history by committing some $1.1 billion over the forty years since the project commenced), it fell to LIGO’s then-director, David Reitze of Caltech, with his astrophysicist colleague Kip Thorne an avuncular presence beside him, to make the formal announcement: that by using the most precise measuring equipment ever built, gravitational waves had now been discovered, or more accurately, their presence had been inferred.
“We have done it,” he said, and the room erupted in applause. A new era in astronomy had commenced, a new means of discovering the magical complexities of the universe. And a peaceful new era, to boot. It was a moment, someone said, akin to that of Galileo’s first looking through his telescope four hundred years before. There were tears of pleasure and pride.
THERE IS AN irony immediately apparent to anyone who has been up close and personal with, on the one hand, the 160-ton ASML machines in Holland, which allow for the placing of seven billion transistors onto a wafer of silicon no larger than a fingernail, and the airline hanger–cum–train station vastness of the LIGO machinery that has been established to detect what one author has called “gravity’s whispers.”
Both sets of machinery have been designed to deal with the tiny, the faint, the microscopic, the atomic, the cosmic—yet both sets of machinery are so Victorian-grand in design and so magisterial in scale, far bigger than the great machines of the past, those that dealt with steam and iron and lathes and screws and governor wheels and flywheels and heat and incessant noise and shuddering vibrations, back when precision was at its vague beginnings. Where precision once employed small machines to construct big things, it now employs big machines to create, or to detect, tiny ones.
THERE IS A further irony.
The first-ever device to call itself precise was a cylinder, bored from a block of solid metal by a Cumberland ironmaster in 1776, specially made for use in James Watt’s steam engine, and at the start of the Industrial Revolution. Now, the component at the heart of what LIGO’s David Reitze publicly described as “the most precise measuring instrument ever built” is a cylinder, too. Unlike Wilkinson’s, this one is solid, a forty-kilogram cylinder known as a test mass and made of fused silica that reflects all but one of every 3.3 million photons that hit it. The silica is tooled and lapped and polished to an immaculate flatness. It is suspended in a cradle by a network of 400-micron-thick silica filaments, and is balanced by an array of weights of glass and metal and magnets and coils that will allow it to be tested and measured by the laser that will hit it 280 times each fraction of a second, in order to measure the distance of the length of the tube at the end of which it lives, and which thereby detects whether a gravitational wave has passed through—as has happened so far four historic times.
LIGO’s precisely crafted fused-silica “test mass” (very basically, an ultraprecise mirror suspended inside a sophisticated damping system) reflects beams of high-intensity laser light that have been shot at it down a 4-km pure-vacuum tunnel in such a manner that it can detect microscopic changes in the tunnel’s length, and thus prove the existence of gravitational waves. At the time o
f this writing, LIGO has proved the existence of four such waves.
Photograph courtesy of Caltech/MIT/LIGO Lab.
John Wilkinson’s cylinder fit inside James Watt’s steam engine with a degree of precision amounting to the thickness of an English shilling, about one-tenth of an imperial inch. Such precision had never been achieved before, but after that, the world never once looked back.
Two and a half centuries on, and the engineers at LIGO have also made their test mass as a cylinder. This one was constructed out of fused silica—a pure form, effectively, of sand, of as elemental a substance, literally and metaphorically, as the iron that was used by John Wilkinson.
The test masses on the LIGO devices in Washington State and Louisiana are so exact in their making that the light reflected by them can be measured to one ten-thousandth of the diameter of a proton. They can also compute with great precision the distance between this planet and our neighbor star Alpha Centauri A, which lies 4.3 light-years away.
The distance in miles of 4.3 light-years is 26 trillion miles, or, in full, 26,000,000,000,000 miles. It is now known with absolute certainty that the cylindrical masses on LIGO can help to measure that vast distance to within the width of a single human hair.
And that’s precision.
Chapter 10
On the Necessity for Equipoise
The test of a first-rate intelligence is the ability to hold two opposing ideas in mind at the same time and still retain the ability to function.
—F. SCOTT FITZGERALD, THE CRACK-UP (1936)
And yet. The ever-increasing degree of precision that defines so many of the perfectly ordinary items that now surround us—and which is supposedly of such vital importance to the pursuers of today’s scientific truths—prompts a cascade of philosophical questions. Is such a wish for perfection truly an essential to modern health and happiness, a necessary component of our very being? Do the benefits it provides clearly outweigh the shortcomings that so clearly accompany its stealthy recent insertion into human life and society? Are we a happier and more contented collective of souls for possessing it and employing it in our everyday? Should we worship and revere and give thanks to all those of the past—Wilkinson, Bramah, Maudslay, Shockley, and their like—for blessing us with their notion of the need for endlessly improving exactitude?