The second before atoms
For most of history the second was a fraction of the day. A second was defined as 1/86,400 of an average solar day (24 hours times 60 minutes times 60 seconds). That worked until astronomers and engineers could measure the Earth’s rotation precisely enough to see that it is not perfectly steady. Tides, shifts in the Earth’s interior and seasonal changes in the atmosphere all nudge the length of a day by tiny amounts.
In 1960 the international definition moved to the Earth’s orbit around the Sun, which is steadier than its spin. That version of the second was quickly superseded, because laboratories had begun to build something better: clocks based on atoms.
The idea behind an atomic clock
Every clock needs something that repeats at a steady rate. A pendulum swings, a quartz crystal vibrates and an atom does something similar at a much higher frequency. When an atom changes between two particular energy states, it absorbs or emits electromagnetic radiation of a very specific frequency. Every atom of the same isotope behaves the same way, which means the frequency is a natural constant, not something a craftsperson has to adjust.
The physicist Isidor Rabi discussed the possibility of using atomic transitions for timekeeping in the 1940s. Early laboratory devices followed, including a clock at the US National Bureau of Standards in 1949 that used ammonia molecules. These early machines were more of a proof of concept than a standard.
The cesium clock
The first accurate cesium beam clock was built by Louis Essen and Jack Parry at the National Physical Laboratory in Britain in 1955. In simple terms, a beam of cesium atoms passes through a microwave cavity. A quartz oscillator drives the microwaves, and a detector looks for the frequency at which the largest number of atoms change state. The electronics then adjust the quartz oscillator to stay locked on that frequency, so the quartz oscillator is effectively disciplined by the atoms.
The cesium-133 atom turned out to be an excellent choice. Its transition lies in the microwave range, which electronics of the day could handle, and it was well understood. Within a few years cesium clocks were running at several national laboratories and comparisons between them showed agreement far better than any astronomical method.
How the second was redefined
In 1967 the General Conference on Weights and Measures defined the second as the duration of 9,192,631,770 periods of the radiation corresponding to the transition between two hyperfine levels of the cesium-133 atom’s ground state. In practice, that means a cesium clock counts roughly nine billion cycles and calls the result one second.
This was a change of principle. The second no longer depended on a planet or on a particular piece of equipment. Any properly equipped laboratory could, in principle, reproduce it.
Better atomic clocks
Cesium beam clocks were steadily refined. In the 1990s, cesium fountain clocks became important standards. These toss cooled atoms upward, let them fall back through the microwave field, and so give the atoms a longer time to be observed, which sharpens the measurement. Other designs, such as hydrogen masers, are valued for short-term stability and are widely used alongside cesium clocks.
More recently, optical clocks that use transitions at visible-light frequencies have shown even better precision in laboratories. Because the new devices tick far faster, they can be more precise. Metrologists have been discussing for some years whether the definition of the second should eventually be updated to use one of these transitions. Any such change would be decided internationally, so treat any timeline you read with caution.
Standard time before and after atomic clocks
Atomic clocks did not create standard time. That came earlier, from the problems of travel and communication. In the nineteenth century, towns kept their own local solar time, which was awkward for railroads running timetables across many places. In Britain, railway companies adopted Greenwich time in the 1840s, and Greenwich time was later spread by the electric telegraph. In North America, railroads introduced standard time zones in 1883. An international conference in Washington in 1884 recommended Greenwich as the prime meridian, though countries adopted time zones at different speeds over the following decades.
Atomic clocks gave this system a far more accurate foundation. Today, International Atomic Time is built by combining the readings of many atomic clocks around the world. Coordinated Universal Time (UTC) is based on that atomic time, but it is occasionally adjusted by a leap second so that it does not drift away from the Earth’s rotation. Leap seconds were added starting in 1972, and the most recent one was at the end of 2016. In 2022, the international metrology community agreed in principle to stop using leap seconds in the future, with implementation planned for later in the 2030s.
Where atomic time is used
Most people never see an atomic clock, but they depend on one. Satellites used for global positioning carry atomic clocks, and a receiver works out its position by comparing the timing of signals from several of them. Cellular networks and financial systems use precise time to coordinate operations, and many of those systems get it from satellite signals. Time services also distribute the time through radio broadcasting and internet protocols, from laboratories like those in Britain, the United States, Germany and many other countries.
What the clocks do not do
An atomic clock is not magic: the time shown on a phone or computer may differ slightly from UTC because of network delays, and a household “atomic” wall clock is typically a radio-controlled clock that receives a time signal rather than a miniature cesium device. The real devices are laboratory instruments, usually in controlled rooms, compared and averaged by international agencies.