Âé¶ą´«Ă˝

The epic quest to redefine the second using the world’s best clocks

A more precise definition of the second is crucial to all sorts of physical measurements – but to get there, scientists have to pack up their extraordinarily fragile optical clocks and take them on tour

On a large table draped with hundreds of cables, a maze of mirrors and lenses bounces and guides a thin beam of laser light. It culminates at a silvery capsule, which holds 40,000 strontium atoms cooled to within a whisker of absolute zero. This delicate edifice is an optical clock, one of the world’s most accurate timepieces.

Instruments like this aren’t exactly designed to be portable – which makes it more than a little surprising that the operators of one such device at the German national metrology institute packed it into a trailer and sent it hurtling down a motorway. It was the start of a perilous journey: a bad jolt could disrupt the beat of its precise ticks. But it was necessary.

That was because, in 2022, scientists globally agreed that we should start work on redefining the second based on our latest and greatest timekeeping technology: optical clocks. However, this meant bringing together several of the world’s best specimens for comparison.

Doing so proved a huge challenge, but it will surely be worth the trouble. A new definition of the second will be profoundly consequential for nearly every other measurement that scientists use to describe nature, from speeds to masses and more. Our efforts to define it more precisely, then, will ripple out across our entire view of the world. “This was the first global comparison of optical clocks. It’s such an important and impressive achievement that it really brought together labs from across the world,” says at the University of Colorado Boulder, who helped build the world’s best optical clock.

This story begins long before anyone was using lasers and vibrating electrons to tell time. In order to keep time accurately, we have always had to set our clocks by the ticks of a better one. Until the 1950s, the best were those set by astronomers based on the position of the sun in the sky. Back then, the most reliable clock was the one at the Royal Observatory in Greenwich, London.

Not everyone could bring their clocks to the observatory to set them, though – and that is where Ruth Belville, the “Greenwich Time Lady”, came in. Every day, come rain or shine, she would take her stopwatch to the Royal Observatory, then travel all over London so her hundreds of clients could synchronise their timepieces with hers.

Atomic clocks

It is now 85 years since Belville retired, and in that time this synchronisation process has become both easier and harder. For the most part, nobody has to trek through the London drizzle to bring anyone else the time – it can be transmitted through fibre-optic cables or satellites with enough precision for nearly any purpose. As things stand today, the official definition of a second is based not on the motion of the entire planet, but on the quantum vibrations of electrons in caesium atoms in what is known as an atomic clock. These vibrations happen with a fixed frequency. Measure this, and a simple calculation will yield the length of a second. “Frequency is the thing that humans have measured best,” says Aeppli, so this is an unquestionably better way to measure time than anything that came before.

Caesium clocks aren’t quite the best way to measure time we have, though. That accolade now belongs to optical clocks, which are also based on electrons jumping between quantum states. But the atoms used in optical clocks hop between these states much more quickly, making the clocks hundreds of times more accurate and precise.

Belville’s watch was accurate to about one-tenth of a second, precision to a single-decimal place that seems almost laughable in comparison with the 15 or 16 decimal places offered by today’s best optical clocks. Even if you waited for four times the current age of the universe, the optical clock riding down the motorway from Germany would be off by less than a minute. How times have changed.

“I’ve been involved with atomic clocks for over 20 years now, and by 2015 it became clear that optical clocks were getting better than the caesium clocks that are used to define the second currently,” says at the Paris Observatory in France. That renders the current definition obsolete, he says. And the International Bureau of Weights and Measures, which controls the world’s clocks by setting the global standard for them – Coordinated Universal Time (UTC) – agrees: in 2022, it put forward a resolution to officially redefine the second using optical clocks.

Optical clock at the National Physical Lab
Optical clocks like this one at the National Physical Lab in London are a long way from the watch once used to synchronise the city’s clocks
Guido Wilpers/Andrew Brookes AB Still LTD

Travelling timepieces

So why is our global timekeeping system still based on clocks that are second best? That comes down to the difficulty of building and operating optical clocks, plus the even more vexing problem of comparing them with each other. There are fewer than 100 optical clocks in the world, and before we can use those timepieces to officially redefine the second, researchers must be absolutely certain that each one is functioning at its peak.

Scientists have seen this coming for a good while and they knew that the only thing that can test one exceptional clock is another exceptional clock. One of them is at the Physikalisch-Technische Bundesanstalt (PTB), the German national metrology institute. So, more than a decade ago, he and his team started hatching a plan to put their clock on the road. The idea was to pack it into a trailer and take it to the National Physical Laboratory (NPL) in Teddington, London, to meet another of the world’s best optical clocks.

They recognised that such a journey, while anxiety-inducing, was absolutely necessary: there is no other way to meaningfully compare optical clocks across the globe without introducing errors that would drown out the precision we are trying to measure. Researchers have tried using satellites and fibre optics, among other potential methods to send signals between clocks, but everything they tested introduced far too much uncertainty. “We’re really back to the situation that we had in the olden times, where, to do a comparison of clocks, you had to carry around a clock,” says at the National Measurement Institute in Australia. It is time for a modern-day Greenwich Time Lady.

But while Belville’s pocket watch was a small, hardy thing, an optical clock is anything but. “These are quite complex apparatuses with a lot of vacuum systems and lasers and frequency stability,” says Lisdat. “If you kick this stuff around too hard, there’s a big chance that something breaks.”

For the whole thing to work, the laser light in the clock’s guts must be patterned just right for each strontium atom to sit undisturbed in its own spot in a neat grid, similar to eggs resting in an egg carton. Taking it on the road was like loading these fragile quantum “eggs” onto a trailer and hoping that their carton, made from something as ethereal and finicky as light, will never crease, fracture or dent. “I had knots in my stomach when I worked on similar measurements,” says at PTB.

Despite their apprehension, in 2023 the PTB researchers loaded their 800-kilogram clock into an air-conditioned trailer, hired a professional driver to shepherd it through the European countryside, then hit the road. “Seeing the clock leaving PTB was quite a thing. [At that point] you know that now things will start, no matter if you are well prepared or not,” says Lisdat.

Elizabeth Ruth Belville receives a timekeeping certificate from an official at the Royal Greenwich Observatory
The Greenwich Time Lady was a key part of timekeeping in London until her retirement in 1940
Fox Photos/Hulton Archive/Getty Images

The German clock wasn’t the only one making the trip to London. At the same time, a team at the Japanese scientific institute RIKEN was preparing its clock for an even longer journey, this one by air (see “A clock caper”, below). The researchers spent years miniaturising their equipment, making sure the lasers could be kept stable and the whole thing carefully packed and shielded to stay intact despite any disturbances to its environment during its long flight. “It arrived packaged into big wooden crates and was moved about with forklift trucks and transported on lorries from the airport,” says . “Care was taken, but it got thrown around a little bit.”

When we compare clocks, there are always surprises

When the two visiting clocks arrived, they were hooked up to NPL’s optical clock. This was a more standard device, its many components neatly arranged on a table in a laboratory. Its stationary and stable nature made it an ideal comparison for the potentially more fickle travelling clocks.

At NPL, each one was hooked up to a laser with the frequency of its light carefully locked to that of the main laser for the stationary clock, and those coupled lasers were used to measure the performance of the clocks. For three weeks, they ticked silently away, their frequencies monitored by laser light.

At the end of the experiment, the researchers couldn’t immediately declare success – much of the data they had collected from the clocks still had to be analysed – but morale was high. They all went to a local curry house for a celebratory meal and immediately started talking about doing it all again.

For some of the team, the next trial was imminent. After meeting in Teddington, the German and Japanese clocks made another journey, this time across the English Channel and through northern Europe to Braunschweig, Germany, for another comparison with a different laboratory-bound optical clock.

Another three weeks later, with all the comparisons done, it was time for the researchers to finally analyse their hard-won data. For the most part, the results were positive. The measurements taken in Germany were effectively identical to those collected in the UK, which was the researchers’ main concern. This shows that optical clocks can travel widely and still tick along at the same rate as when they started – the largest hurdle we have to leap if we want to use them to redefine the second.

Map of optical clocks travelling to Teddington in the UK

“This marks an important milestone for transportable clocks, demonstrating their potential as practical, real-world devices,” says at RIKEN. The German travelling clock and the British stationary clock agreed remarkably well, as did the Japanese travelling clock and the German stationary one, showing statistical differences on the order of just 1 part in 1 million trillion. In fact, the level of agreement between the latter pair was the best ever achieved among independent clocks of the same kind. All of the optical clocks emerged from as battle-tested champions of timekeeping.

But there were complications too. “When we compare clocks, there are always surprises that they don’t quite agree or behave as expected – always,” says Hill. While some pairs of clocks agreed exceedingly well, the whole group of four showed some surprising discrepancies.

Testing gravity

The researchers haven’t yet figured out where these discrepancies in frequency come from – whether there is a problem with the temperature of the atoms fluctuating or perhaps an unexpected interaction between the electrons in those atoms and the laser beams. “That’s really the point, to find out these discrepancies and to understand where they come from,” says Lodewyck.

Even if there is still some work left before we can use optical clocks to redefine the second, there is another problem that they could start solving now: measuring the minuscule changes in Earth’s gravity field that can be caused by shifting sea levels or seismic activity.

According to Einstein’s general relativity, clocks that are closer to the ground tick slightly more slowly than those that are higher up because of the pull of gravity. “Depending where you are in the gravity field of Earth or a black hole or whatever, your time passes at a different speed,” says Lisdat. “That’s what you see if you compare clocks at different height.” The effect is so minuscule that regular clocks simply cannot measure it. For example, a clock raised 1 centimetre above another one will tick 0.0000000000000001 per cent more quickly.

Luckily, this is exactly how precise optical clocks typically are, so carting them around all sorts of terrain could help map Earth’s surface and the details of its gravitational field like never before. The researchers tried using their clocks to determine the height difference between the two labs they visited and found that they performed as well as the best methods that experts currently use, showcasing precision to within less than 4 centimetres – and this was after they had been lugged around the world.

“There have been some proposals to put a network of [optical] clocks around a volcano to determine when it’s going to erupt. Or put it near some subducting tectonic plates and determine when the next earthquake will happen,” says Aeppli. These clocks would detect the tiniest of changes in the vertical position of the ground, which as the early rumblings of an earthquake or a volcanic eruption. Such a network could even prove useful for testing general relativity itself through careful examinations of how gravity affects their ticking.

But the most important step forward from the clocks’ journey is towards redefining the second – a goal that researchers are aiming to hit by 2030. That achievement will echo throughout science because of how fundamental the second is to many types of measurements, whether it be evaluating the brightness of light, the temperatures of celestial objects or even the amount of current in electronics. If you think of these measurements as our way to paint a picture of the world, how far you can zoom in to that picture before encountering fuzzy, uncertain edges is more often than not determined by how well we can define the second. A more precise definition of the second will mean a more precise picture of the world. Japan has had a head start on this: since June 2021, Japanese Standard Time has been calculated based on both a caesium clock and an optical clock, albeit with far less stringent accuracy requirements than those in place for doing such a thing globally – it is easier to redefine one time zone than to reset all of them.

“This quest [to redefine the second] is epic. We did something towards that,” says Hill. “We perhaps also revealed that a large amount of work is still required to reach the criteria to redefine. And so we continue.” As for the PTB researchers, Lisdat says they are getting their clock ready to travel to Italy for another comparison right now. It will take many more journeys like this one for us to be sure that optical clocks are reliable enough to form the foundation of global timekeeping.

This is the thing that all the researchers agree on the most: more optical clocks ought to be driving on highways, flying through cloudy skies and ultimately getting to sit next to the only other devices on Earth that keep time as well as they do. The first step on the road was certainly a success – the Greenwich Time Lady would no doubt be proud.

Topics: Quantum mechanics / Time