September 1, 2020.
Back in the old days, humans used simple observations of natural phenomena to tell time. For example, the movement of the Sun is used to tell the time in the day, and the positions of constellations are used to time the time in the year. It did not take to long for us to realize that those methods are probably not very accurate. It is very difficult to tell the difference between 8 AM and 9 AM based solely on the position of the Sun. When technology becomes more advanced, there is a need for a better time keeping. This let scientists to come up with a better and more reliable clocks, such as pendulum clocks or quartz clocks. The latter can be made small enough to fit around our wrists.
In some applications, the precision of pendulum clocks or wristwatch is simply not enough. Nowadays, we use the global positioning system (GPS) for navigation. It is based on a time an electromagnetic wave takes to travel between a GPS receiver and a satellite. A mistake in only 1 millisecond will cause hundreds of kilometers mistake in the distance. Clearly, a better clock is needed.
Currently, we use so-called atomic clocks to tell time. This is nothing new. We have been using this piece of technology for about 60 years now. We will first describe how it works and will introduce what is new in clock technology.
The first generation of atomic clocks use cesium atoms as a reference. The structure of any atom is uniquely determined by its number of electrons. Once the number of electrons is fixed, the structure of the atom, in principle, can be calculated. The structure of an atom indicates possible energy levels that electrons can occupy. We can pick any pair of energy levels and measure the energy difference. This difference in energy will be the same independent of the location of the measurement. This is because atoms, provided that they are the same element, are indistinguishable.
This simple fact is actually one of the greatest things about atomic clocks. The fact that cesium atoms in Thailand behave exactly the same way as cesium atoms in Spain allows us to make sure that atomic clocks (based on the same type of element) give the same standard everywhere in the world, or even in the universe.
In order to understand how atomic clocks work, it is crucial to understand that for any clock, there are two main components: the oscillator and the frequency (or time) standard.
For cesium atomic clocks, the oscillator is an electrical circuit that oscillates at a frequency close to 9 GHz. This is in the range of microwave frequency. Scientists have been working with microwave technology for a very long time, so people know how to work with it. However, the microwave oscillator is not stable. This means that any external parameter can influence the frequency of the oscillator. For example, usually the oscillator will have a bit higher frequency if the ambient temperature is higher than usual. Not only that, how do we know if our oscillator is oscillating at the exact frequency of 9.00000 GHz or 9.00001 GHz. This is where the frequency reference comes in.
The lowest energy level of a cesium atom consists of two so-called hyperfine levels, and these two levels differ by a fixed amount of energy. There is a standard technique called Ramsey interferometry, which allows us to measure the frequency of an electromagnetic wave relative to the energy difference between any two atomic states. (It should be noted here that energy and frequency are related simply by a Planck’s constant, and in atomic physics sometimes we use them interchangeably.) Interaction between electrons in an atom and an electromagnetic wave causes a transition from one atomic state to another atomic state. When the frequency of the electromagnetic wave matches exactly the energy difference between the states, the transition probability is the greatest. This means that we can tell if the electromagnetic wave frequency matches the atomic energy or not. Not only that, sophisticated techniques also allow us to know if the frequency of the wave is too high or too low. This lets us correct the frequency of the electromagnetic to always match the atomic energy.
Once this is done, we say that the frequency of the electromagnetic wave (our oscillator) is locked to the atomic transition. If this is the case for cesium atoms, we define the locked frequency of the electromagnetic wave to be exactly 9,192,631,770 Hz.
We need to constantly measure the frequency of the oscillator against the atomic transition, because the frequency of any oscillator will drift or change in an unpredictable way. Once the frequency is locked, we can split off the signal from the oscillator and use that signal for any timing or frequency reference application.
For a clock, it is probably of great interest to define a second from the atomic clock. Since the frequency is defined exactly, one second is simply the time duration for this particular oscillator to complete 9,192,631,770 cycles of oscillation.
Atomic clocks based on cesium (or rubidium) atoms are what we call the first generation of the atomic clock. Current best stability of this kind of clock is about 1 part in 1015 . While this is good enough for most of the application we can think of, actually physicists realized that it is quite difficult to improve the performance of this kind of clock any further. People started to look for a better solution.
The second generation is to use a different kind of oscillators. Instead of a microwave oscillator, a laser might be a better option. Laser is a term that actually stands for light amplification by stimulated emission of radiation. It is a way to produce light wave with extremely high spectral purity. In other words, it is as close to being monochromatic as possible. Actually, physicists have built a laser light which is stable to more than 1 part in 1018.
Why laser is a better option? It is a question that requires quite a lot of technical discussions, but we can simply say that it is easier to build a better oscillator if the oscillation frequency is higher.
So, we have replaced a laser light as an oscillator. What about the frequency reference? We still would like to use atoms as a reference for the very same reasons. However, cesium atoms are not suitable anymore. It turns out that atoms like ytterbium (Yb) or strontium (Sr) atoms or even mercury (Hg) are better to be used as an optical clock. (The term optical clock is used because it is a clock with a visible light being the oscillator.) The diagram below shows a simple operation of such a clock. Note that we have a so-called frequency comb. This device allows us to link the laser light frequency (which is extremely high) to other frequency in the lower range like radio frequency or microwave frequency. Without a frequency comb, an optical clock is probably useless.
While the operation principle is the same as in the cesium clock, the technology involved are quite more demanding. For the past 10-15 years, a lot of progress in vacuum technology and optics technology have been made to realize optical clocks that surpass the first-generation clock in terms of stability and accuracy. Currently the best optical clock is accurate to a part in 1018 [2,3]. This means that this clock will not be wrong by more than a second even we wait for the age of the universe.
Currently, in Thailand, by the support of Thailand Center of Excellence in Physics (ThEP), a team of researcher from Mahidol University and National Institute of Metrology (Thailand), or NIMT, are building a second-generation atomic clock based on ytterbium ions. The ions will be trapped in an ultra-high vacuum environment, then a highly stable laser light (our oscillator) will illuminate this ion. The resulting interaction between the laser light and the ion will allow us to determine precisely the frequency of our laser light. At the time of the writing of this article, we are in a design stage of the ion trap. We are hopeful to have a prototype of this optical clock in Thailand very soon.
 Jefferts S R et al 2015 Proc. 2015 Symp. Freq. Stds. Metrology 723, 2778.
 Rosenband T et al 2007 Phys. Rev. Lett. 98, 220801.
 Nicholson T et al 2015 Nature Communication 6, 6896.
Thaned Pruttivarasin (Mahidol University)
Tara Chalermsongsak (Mahidol University)
Jirakarn Nunkeaw (ThEP)
Piyaphat Poonthong (NIMT)
Nakarin Jayjong (Mahidol University)
Asst. Prof. Dr. Thaned Pruttivarasin, Dept. of Physics, Fac. of Science, Mahidol University, Bangkok 10400, Thailand