The Ultimate Clock

Keeping precise time on the universe’s scale

Time is an ancient and contrary mystery. Augustine of Hippo, writing his Confessions in a North African monastery, asked “Who can even in thought comprehend it, so as to utter a word about it? But what in discourse do we mention more familiarly and knowingly, than time?”

In what’s called an “optical lattice clock,” thousands of atoms are held in wells made of laser light. Oscillations of the atoms’ electrons keep the time. NIST

TIME OVER TIME | Timekeeping’s accuracy improved dramatically during the last millennium. A Chinese water clock could lose around 10 minutes a day. Within five years, scientists hope to have an atomic clock that loses less than one second over the universe’s lifetime. from left: New York Public Library/Photo Researchers, Inc; Paul Clements

The current best long-term timekeeper is the United Kingdom’s atomic clock. It loses or gains less than a second every 138 million years. U.K. National Physical Laboratory

More than 16 centuries later, many scholars share the feeling, if not the prospect of sainthood. “We don’t even know what time is. But we can measure it really, really well,” says Chris Oates, a physicist at the National Institute of Standards and Technology’s Boulder, Colo., campus.

His team operates a ytterbium optical lattice clock, one of the latest types of souped-up atomic timepieces. To track the passing seconds, such clocks rely on the fixed frequencies of photons absorbed and emitted by atoms’ electrons as they change energy levels. Recently, scientists have found ways to make these quantum counters even better, by switching from a reliance on microwave frequencies to the faster-paced optical regime and introducing a system of checks that relies on multiple atoms in levitated grids. In a remarkable recent development, the central atomic metronomes are protected from distortion by a method so powerful that physicists formally call it magic.

Oates is a member of a worldwide cadre working with such devices at the frontier of clockmaking. His team’s clock loses time at a rate of about one second every few hundred million years.

Speedy metrology

Such accuracy is why time is not just one dimension among several but a foundation for defining other fundamental units. The meter’s definition has been defined with increasing accuracy by such things as one ten-millionth the distance on a circular arc from the equator to the North Pole, and by a precision-made “prototype meter” bar of metal alloy kept in Paris. In 1983 the meter officially became the distance light will travel in a vacuum in 1/299,792,458 of a second. The better the stopwatch, the better such definitions can be applied.

The metrology of time is not holding still. In the April-June issue of Reviews of Modern Physics, experimental physicist Hidetoshi Katori of the University of Tokyo and theorist Andrei Derevianko of the University of Nevada, Reno declared dramatic ambitions for a record-breaking atomic clock based on emissions from mercury atoms.

“If someone built such a clock at the Big Bang and if such a timepiece survived the 14 billion years, then the clock would be off by no more than a mere second,” they note in the paper. That is actually conservative. The goal formally is to lose or gain no more than one out of every billion billion seconds. That is one second in about 32 billion years, and is 10 to 100 times better than any existing clocks.

In scientific shorthand, the proposed mercury clock would reach a fractional uncertainty of at most one part in 1018 — it would run for 1,000,000,000,000,000,000 seconds before being one second awry.

Already, atomic clocks have come a long way. While experimental clocks are moving ahead, a device called the NIST-F1 is the official U.S. timekeeper. It’s accurate to a few parts in 1016. It occupies a large first floor room in Building One at NIST’s Boulder campus. The dominant feature is a shiny steel vacuum chamber 8 feet high. Inside is a laser-controlled fountain of cesium atoms chilled to near absolute zero. The atoms rise in clumps about as large as a man’s thumb and, responding to gravity, fall back through a cavity in a tunable microwave generator.

Oscillations within the cesium atoms are akin to the to-and-fro of the balance wheel in an old wristwatch, but it is the microwave generator that communicates with the outside world. Just as the ticking of a watch arises from the escapement mechanism connected to the gears and hands, oscillations within the cavity are recorded electronically.

By itself, the microwave generator would drift off time. So with each passage of the atoms, the generator checks to be sure its pulsations exactly match the signal from a chosen energy transition in the atoms’ electron clouds — an electromagnetic wave that beats 9,192,631,770 times a second.

NIST is now working on a successor, called F2. With an improved cooling system and superior way of moving the atoms through the microwave chamber, it will be about four times better and will beat out the current record for a long-term timekeeper, a clock in the United Kingdom that is accurate to about two parts in 1016.

Such astonishing accuracy is no mere intellectual exercise. Recent advances in timekeeping have brought practical payoffs in the design of better global positioning systems that triangulate locations on Earth by measuring distance via radio-signal travel time, as measured by satellite-borne atomic clocks. Further progress should lead to instruments able, from the slowing or speeding of time’s passage due to shifts in gravity, to improve maps of the planet’s interior and to find mineral deposits or detect the movements of deep magma. Pure research on Earth and in space may gauge to almost unimaginable exactness the stability (or drift) of supposed constants of physics that not only affect nuclear decay, but also, some astronomers say, may have worked differently in distant eras. 

Ticking gears

By historic standards, clock progress is now frenzied.

People have long kept track of time by monitoring processes that change measurably in a steady way. Early peoples monitored the seasons by the motion of the sun and moon. An 11th century Chinese water clock, its gears driven by a steady stream, might lose or gain 10 minutes a day — an accuracy of about a part in 100. Large, stable swinging pendulums in the 1600s were good to a few seconds a day. Eighteenth century navigation clocks that were the pride of the British Navy weren’t much better. They lost or gained a minute or less per month, an accuracy of about one in 10,000, but they did it while tossing about in ships at sea. Quartz clocks and watches, paced by electrically stimulated crystals vibrating at about 32,768 times per second, were developed in the late 1920s. They keep time to within a second per day, better than a part in 105.

Then along came atomic clocks, following the beat of quantum mechanics, the laws that govern the energies of electrons bound to nuclei. Every 10 years since the first one debuted in 1949, based on oscillations in the ammonia molecule, the accuracy has increased by about 10 times. Recently, things have gone even faster.

While devices like the NIST-F1 use atomic signals of microwave frequency with billions of cycles per second, newer clocks, including Oates’, rely on light waves beating a million times faster. The new approach “is like having a ruler with more divisions,” says Tom O’Brian, NIST’s chief of the divisions of Time and Frequency and of Quantum Physics. “The pace of improvement is skyrocketing.”

A further development, the lattice clock, has been imagined only in the last decade, with rapid progress in the last five years. For now, related devices called single ion optical clocks, which key in on solitary electrically charged atoms such as aluminum, are the most accurate. However, lattice clocks’ use of many atoms simultaneously, with a strong combined signal, appears to give these clocks the ultimate edge.

Katori says his team in Tokyo hopes to have the first clock with one part in 1018 accuracy working within five years. A look at how the record-setting mercury clock would work reveals the basics of all contemporary neutral-atom lattice clocks.

At a glance, the proposed clock is a bewildering laser beam jubilee — but there is underlying order.

The action starts with a system of cooling lasers that bathe a thin vapor of mercury atoms in what is called “optical molasses” to slow their motion. Temperatures hit a few millionths of a degree above absolute zero, a coolness at which each atom drifts roughly at the walking speed of ants. But even at that slow speed, the motion causes a slight blur in the atoms’ collective optical signals.

The cooling lasers propel the chilled atoms gently into a zone where another laser system’s beams cross one another. The interacting light waves, sometimes doubled up by mirror systems, form a tiny three-dimensional array of shimmering energy fields.

This is the lattice. Its standing waves rise and fall but do not propagate. When the fields’ energies are diagrammed, they take on a pattern that looks a bit like the hollows in egg cartons. These nodes trap and hold the atoms — ideally one atom per energy well — in perfectly aligned ranks. The entire array of atoms is levitated in a tiny near-vacuum about 100 micrometers across, roughly the thickness of a page in a glossy magazine like this one. Most important, the trapping lasers whose beams produce the lattice will be set to a “magic frequency” — a recent breakthrough in the field — to grip the atoms in place while not distorting the shapes of their electron clouds.

All that is preamble to the key step. A clock laser will, a bit faster than once per second, illuminate the atoms, adjusting itself as needed to match the frequency at which they most easily absorb and emit light. Lasers may be popularly considered the essence of precision optics and purity of color. But at the esoteric edge of the timekeeper’s craft, they are too wobbly to keep time by themselves. Thus the clock laser’s orderly light waves are paced by the atomic metronome — just as a drill sergeant keeps troops in precise cadence.

Just one more big step: reading the clock. Though this clock relies on visible light, optical waves flicker far too fast, nearly a million billion times per second, for electronic circuits to count one by one. So, the clock laser is keyed in turn to yet another recent invention, what’s called an optical comb laser. It is many thousands of lasers in one. Its multiple wavelengths, when plotted, look like the teeth of a comb, stretching across a vast frequency range. Optical combs were a big reason another scientist at NIST, John Hall, shared a Nobel Prize in 2005.

The comb’s function is akin to a transmission’s gears. By synchronizing one of the optical teeth with the clock laser, atomic clockmakers force the other teeth to become equally stable. That way one of those in the microwave domain can be selected to deliver a countable beat, a million times slower.

Now, a pause to ponder magic. By the 1990s it was clear that laser lattices would allow a probe laser to gaze at a throng of trapped atoms long enough to get a better reading on their signal than is possible with a fountain clock such as the F1. The lattice would also prevent collisions among the atoms, a key source of distortion in microwave fountain clocks.

However, lattices come with a price. Their oscillating electromagnetic fields typically and severely distort atomic energy levels. One can easily guess how badly a violin would go off tune if someone squeezed its belly and twisted its scroll just as the violinist was striving for a delicate note.

In 2001, Katori and colleagues began publishing proposals that there might be a cure. Perhaps there could be certain frequencies of trapping fields that would displace the boundaries of one key energy transition by exactly the same degree. Katori’s insight, bolstered by calculations by Derevianko and others, offered a way to grab that violin roughly yet leave one selected note unwavering.

In conversations with Katori, Nobelist Hall said such frequencies sounded like magic. Some journal editors initially resisted when formal papers referred to magic. But the name stuck.

Magic frequencies give lattice clocks accuracy a billion-fold better than they would otherwise manage. “That is nine orders of magnitude,” Derevianko says. “That is magic happening.”

Katori calls it “a gift from nature.”

With magic frequencies soon discovered for several atoms suitable to trapping in lattices, physicists now had an army of atoms, says Jun Ye of JILA, an institute that NIST and the University of Colorado operate jointly. “A million-man army is better than a one-man army.”

Ye’s group operates a strontium-based lattice clock. In 2007 he interlinked, via optical fiber, his clock with a calcium-based lattice clock in Oates’ lab at NIST. The two researchers were able to cross-check the time to an accuracy of one in 1016, an exercise proving that such clocks can be interlinked over electronic and optical circuits.

Beyond the tock

Such advances are the latest step in humankind’s ancient drive to measure how long it takes things to happen, whether for a season to pass or an automobile equipped with GPS navigation to drive 10 feet as monitored by the changed time for a radio signal to reach and return from two distant satellites whose locations are known. A one in 1018 clock would allow location to a matter of inches — on Mars using satellite transceivers at Earth.

The European Space Agency aims to make the International Space Station a platform for fundamental discovery with lattice clocks. That agency, along with the French space agency, has already begun a mission called ACES, or Atomic Clock Ensemble in Space. Its aim is to install on the space station a laser-cooled cesium microwave clock, with accuracy of about one in 1016.

Linking to ground clocks based on different atoms and atomic transitions in a worldwide network, the timekeeper will probe fundamental laws of physics to high accuracy, says Luigi Cacciapuoti of ESA’s astrophysics and fundamental physics division. Under scrutiny will be such esoterica as the constancy of the speed of light in all directions, and Einstein’s equivalence principle declaring that gravity and acceleration have the same effects on time and physical processes. With shuttles retired, the first such clock could be delivered by a Japanese or other automated transfer vehicle as early as 2014.

Stephan Schiller of Heinrich-Heine-Universität in Düsseldorf, Germany, and colleagues, in a program called Space Optical Clocks, hope to put aboard the station around 2020 an optical lattice clock 10 times as accurate. They look forward to comparing the relativistic effect of Earth’s gravitational field on time with that of the sun. Some theories suggest that Earth, with its iron core and other heavy, neutron-rich matter, may very slightly affect time’s passage differently than the neutron-poor hydrogen that dominates the sun.

Can timekeeping possibly continue a tenfold improvement every decade? There are many orders of magnitude before timekeeping would reach the smallest increment allowed by known physics — called the Planck time, 10-43 seconds. To even approach it seems out of reach.

As for boosting precision relatively modestly by bringing today’s atomic lattice clocks into ultra­violet or even X-ray domains, several technical hurdles stand in the way.

While atomic clocks depend on the activity of electrons in an atom’s outer boundaries, there may be other processes to tap — nuclear processes. Within an atom’s nucleus, neutrons and protons jostle and change energy according to quantum mechanics, too. Calculations suggest, for instance, that in the nucleus of the isotope thorium-229 is a transition that should emit ultraviolet rays. If that signal can be confirmed, and stimulated, a laser frequency comb might lock on to it and produce a lower frequency beat that could be fed into an electronic counter.

“This is a very important development, as the nuclei are very well isolated from external perturbations,” says Derevianko. It may open a path “toward the ultimate clock.”

Only time will tell.


A laser jubilee
From the outside, a lattice clock appears to be a dizzying array of lasers, but (as shown in one type of strontium clock at right) each laser has a role in reading the time from atomic oscillations.

Thousands of strontium atoms are cooled by a system of blue lasers.

Red lasers further cool and shrink the cloud of atoms.

An infrared laser system traps the atoms, locking them into pancake-shaped wells.

Another red laser, the “clock” laser, bathes the atoms, causing many of them to become excited.

A blue laser probes the cloud; emitted light reveals how many atoms became excited.

The clock laser adjusts to properly keep pace based on the atoms’ excitations. Light from the clock laser is passed through a “frequency comb,” allowing the light’s ticks to be read.