Advanced atomic clocks will go way beyond timekeeping
an interview with Dr. Jun Ye

NIST physicist Jun Ye adjusts the laser setup for a strontium atomic clock in his laboratory at JILA, a joint institute of NIST and the University of Colorado at Boulder.

Why should gravity affect time? While not an intuitive expectation, it does. Experiments have proven that time passes more slowly in stronger gravitational fields than it does in weaker fields, and the GPS location system does not work accurately unless the time signal from the satellites is adjusted for this effect, because their orbits have them operating in a weaker gravitational field than the one on Earth’s surface. If a person knew this, but was unfamiliar with the theory of General Relativity, they might be inclined to think that gravity acted to slow the mechanism of the clock, but that’s not the case. The clock works correctly, but time itself is slowed. This is called gravitational time dilation.

For a decade following his Special Theory of Relativity, Albert Einstein worked to ‘generalize’ it by also accounting for the effects of accelerated motion and gravity. In 1915 he published his General Theory of Relativity which expanded the idea of 3-dimensional space into a new 4-dimensional concept he called spacetime. This concept combined space and time into a singular, inseparable “fabric” that the Universe, as we know it, exists within and it revolutionized the thinking about gravity.

According to classical Newtonian physics, gravity acts as an attractive force, similar to magnetism, between bodies of mass such as the Earth and an apple. Sir Isaac Newton learned how to calculate the effects, as they acted on these bodies, but never understood the mechanism of action. In a letter he wrote to Richard Bentley, an English scholar, he stated, “Gravity must be caused by an Agent acting constantly according to certain laws, but whether this Agent be material or immaterial I have left to the consideration of my readers.”

General Relativity, in contrast, describes gravity as an effect of spacetime that’s been warped or curved by the presence of massive bodies. An example would be gravitational lensing, where light rays bend according to the curvature of spacetime around a massive body such as a star or black hole. In fact, the first observational confirmation of General Relativity came from a solar eclipse on May 29, 1919. During the brief period of totality astronomers measured the apparent dislocation of known star positions as they were very close to the Sun. This confirmed that the Sun’s mass was bending the light from those stars as it passed by, according to Einstein’s theory.

Because time is also warped by being integral with spacetime it allows us to measure gravitational effects by measuring differences in time, but the challenge is that the time differences are extremely small. To resolve them clocks of incredible precision are required. A 19^th century pendulum clock is far too crude. Current cesium atomic clocks can measure relativistic differences but not accurately enough for practical detection purposes. A new generation of atomic clocks, however, is bringing gravitational measurement into reality. Dr. Jun Ye is one of the physicists at NIST working on these. The following is a Q&A exchange we had about some of the possibilities:

(Q1) Gravity-related measurements, using clocks, have become possible because of the accuracy of this new generation of optical atomic clocks. I’ve seen it written that they have an accuracy of ~1 second in 14 billion years. How can you determine this accuracy without an equally accurate but different reference?

(A1) We usually build multiple clocks and then use them to evaluate each other’s overall systematic uncertainty. “Accuracy” is a term reserved for the primary standard, so until the definition of time is changed to that based on optical atomic clocks, we tend to use the phrase “systematic uncertainty” to describe the accuracy of our experimental clocks. This type of evaluations has been well documented in a series of our group’s publications, as well as from other research groups. For example, please refer to our article of 2014 [1] and 2015 Nature Communications paper [2].

Now, when the total systematic uncertainty is evaluated to say 2 parts in 10¹⁸, (this is where the quote of “~1 second in 14 billion years” came from), and knowing that the gravitational red shift is about 1 x 10^-18 for a 1 cm change in height on the surface of Earth, you can say that we can now determine in principle the absolute gravitational potential at the 2 cm level using our modern optical atomic clocks.

By the way, the work on redefining time based on the new generation of optical atomic clocks is also underway. In our recent publication (arXiv:1907.03184, which will come out in Physical Review Letters in the near future), we demonstrated for the first time an ALL-OPTICAL time scale, which represents the first step toward the employment of optical atomic clocks for the realization of universal time at the stability level defined by the optical clocks.

(Q2) Einstein’s General Theory of Relativity has been validated repeatedly over many decades but it’s believed to be problematic because it breaks down with quantum mechanics and inside black holes. For your purpose of measuring time dilation does that matter and do you need to make any adjustments or modifications to the original theory?

(A2) Indeed, the missing connection between general relativity and quantum mechanics, at least the lack of any experimental observations so far, has been an important issue in the framework of modern physics. And this is actually a major inspiration for my work. I will come back to this point later. To answer your question here, however, the time dilation effect arises purely from the general relativity, and hence there is no tension with quantum mechanics, for now.

But imagine that we can continue to advance our clock, say by another three orders of magnitude (to the level of 1 x 10^-21), then we can get into a very intriguing place where quantum mechanics and gravity effect can both come into play in our experimental observations. This is a big goal, and it will take years of research to continue to advance our clocks. Our most recent work of using quantum gas of Sr atoms in a three-dimensional optical lattice clocks (Science 2017 [3]) provides a possible path toward this goal.

(Q3) One idea has been to use these clocks in an early warning system that would measure changes in the mass of a volcanic magma chamber. These changes could indicate the chamber is filling and may lead to an eruption. What would that arrangement of clocks look like?

(A3) To do this, we first need to advance our clock precision (stability) a little further. Our most recent work [4] has demonstrated a record clock precision at 3.4 x 10^-17 in 1 second. This means that to detect a 1 cm of gravitational potential change, we need to average our clock signal for about 1000 seconds. The better the clock stability, the less time it will take to detect such gravitational potential changes. Our near term research goal is to advance our clock stability to be better than 1 x 10^-17 at 1 second, which would mean we need only 100 s (less than 2 minutes) to detect absolute gravitational potential change of 1 cm. This is where we can start dreaming up to perform real time relativistic geodesy.

We will need to establish a clock network where clocks placed at geo-active regions will be linked to other clocks at stable reference points. We have existing technologies of using phase-stabilized fiber networks to link clocks, some of the early work was pioneered in Boulder (Journal of Opt. Soc. Am. B, 2003 [5]). Another technology that is being developed by researchers is to establish phase-stable optical links in the air, and eventually allowing clocks to be linked between satellites and ground.

The LIGO interferometer facility in Hanford, Washington, US. It uses two 4 km (2.5 mile) long arms arranged at a right angle. A laser beam is split and shines down both arms, then recombined to check for interference in the light waves. If the arms are exactly the same in length the waves cancel (destructive interference). But if the arm lengths change, relative to each other, because of spatial distortion from gravitational waves, then a signal is generated and analyzed.

(Q4) Another idea proposed is to use these clocks for detecting and measuring gravitational waves caused by cosmic events such as the merger of massive objects like black holes and neutron stars. Currently there are a number of detectors for this around the world that use the principle of a Michelson Interferometer. Two instruments in the US are called LIGO (Laser Interferometer Gravitational-Wave Observatory) and have been renamed Advanced LIGO in 2015 after modifications to increase sensitivity. Would the clocks you’re working on now have an accuracy that would allow for a detector of similar or even better accuracy?

(A4) This is an excellent question. We actually wrote a paper on this very topic. Please see Physical Review D 2016. At the current performance level, optical clocks are not yet competitive for gravitational wave detection. But if the clock performance gets improved further, say into the 10^-21 stability level (see my comment to your question (2)), then we could get into the business of detecting gravitational waves using optical atomic clocks.

(Q5) Because of the faintness of the gravitational signal, and the sensitivity needed to resolve it, the current interferometer detectors are affected by terrestrial and systemic noises, such as seismic, thermal, electronic, etc. What sort of noises would you be concerned with in using clocks?

(A5) Let me try to explain in simple terms how we would measure gravitational waves using clocks. Imagine we have two satellites, each housing an optical atomic clock, that are connected via a laser link. The atoms act as very precise frequency references, and the atoms in the two satellites are identical, and so their internal transitions will have the same frequency. A gravitational wave will induce small frequency shifts in the laser light that is sent from one satellite to the other. These small differences in the frequency of the laser light at the two satellites can then be detected by comparing the light to the atoms in each satellite.

The comparison between the laser light and the atoms in each satellite is made using the same techniques that are used in all atomic clocks. The frequency of the laser light can be compared to the frequency of a very narrow atomic transition between two internal atomic states. By shining the laser light directly on the atoms, which are initially all prepared in the same state, and then afterwards counting how many atoms are in one internal state and how many are in the other, it is possible to very precisely measure how close in frequency the laser light is to the atomic resonance.

Gravitational waves stretch space in a periodic fashion, bringing two objects closer together and then farther apart, then closer together again, and so on. Optical interferometric detectors like LIGO measure these periodic changes in distance (hence LIGO employs an optical interferometer to measure the difference between two arms). The change in distance between the two objects means there will also be a change in their relative velocity. In other words, at first the gravitational wave will cause the two objects (in our case satellites,) to move towards each other, and then after half a period of the wave, the two objects will be moving away from each other. The changes in relative velocity between the two satellites will result in an oscillating Doppler shift of the laser light sent from one satellite to the other. Just as a Doppler shift of sound waves causes the pitch of an ambulance siren to rise as it approaches you, then drop as it drives away from you, the color, or frequency, of the laser light sent from satellite A will be different when it is compared to the atoms in satellite B. When the two satellites are moving towards each other the light frequency will be shifted slightly higher, and when they are moving further apart it will be shifted slightly lower. It's these small periodic shifts in the laser frequency that we hope to detect by comparing the laser light to the atoms in satellite A, where there should be no shift because the light hasn't traveled anywhere yet, then sending the light over to satellite B, at which point it may have been Doppler shifted by a gravitational wave, and then comparing it to the atoms in satellite B. If everything is working properly, any differences between the laser/atom comparisons in the two satellites would be due to the gravitational wave.

Obviously, technical noise contributing to the relative motion between the satellites needs to be minimized.

(Q6) It seems like a network of many clocks, spread in a wide pattern, could have advantages over a 2-arm interferometer. Has such a network been discussed?

(A6) The main difference between an atomic approach is that one does not need two arms. We need to simply compare the frequencies of the two clocks, as stated in Response 5. However, in order to be able to sense the direction of the gravitational wave, then indeed a network of clocks will be useful. For a futuristic picture of clock networks in space, please see our publication in Nature Physics 2014 [6].

(Q7) Dark Matter is a popular subject. Its existence is based on the idea that the behavior of galaxies, as we see them, is not possible with only the gravitational force that observable matter would exert. I’ve read clocks might be used to explore that as well. So far no Dark Matter has been found, however, despite several decades of effort and there’s an alternate idea that gravity may have been miscalculated instead. Do you think clocks could be useful in helping to solve this question?

(A7) Yes I think as clock technology continues to advance, and the clock performance continues to improve, it will play an important role in sensing of the space-time variations and in search of dark matter. With the tension between astrophysics observations and our current basic framework of physics, something has got to give. There are many alternative theories out there trying to explain the origin of dark matter, and already the current clock stability can set limits to some of the proposed couplings between dark matter and ordinary matter. There is no question that we need to continue to improve our detection sensitivity to explore this intriguing mystery. I believe clock is one of the most fundamental tools we can improve to provide an answer.

(Q8) Currently the clocks you’re working on are experimental laboratory arrangements, not yet practical for use in the field. What needs to be done to make them portable and robust, and what work lies ahead for you?

(A8) Again, this is a very important question, and a timely one. Quantum information science is catching a lot of people’s attention, and quantum sensing, such as a modern optical atomic clock, is part of this new quantum revolution. A key thrust is to develop robust tools to engineer systems that can be portable and field deployable. I can name quite a few items that the whole community needs to work on: an optical atomic clock employs multiple lasers, including very stable clock lasers. How do we integrate these lasers into company systems? We need cold atom technology to provide suitable atoms used in atomic clocks, so how do we provide robust cold atom sources for a new generation of quantum sensors? Other important technologies include cryogenics, optical resonators, optical frequency combs, opto-electronic components. □

Referenced articles/papers:
[1] JILA Strontium Atomic Clock Sets New Records in Both Precision and Stability
[2] Systematic evaluation of an atomic clock at 2 × 10^-18 total uncertainty
[3] A Fermi-degenerate three-dimensional optical lattice clock
[4] Using a silicon silencer, the Ye Group has reached record stability for optical atomic clocks
[5] Delivery of high-stability optical and microwave frequency standards over an optical fiber network
[6] A quantum network of clocks