Interviewer: Can you describe where you work and what you do?
DAVID: I'm a physicist at the National Institute of Standards and Technology in Boulder, Colorado. At NIST, our primary business is to try to make accurate frequency standards. One of the really nice things about working at NIST—we, our group, Jim Bergquist, and I and the other people working in the group, is that we were blessed because our managers have been very encouraging to us. I think often in a university environment, the department wants individual stars, and I think NIST has been very supportive of our group environment. So, I think I can give an example of how well this has worked because Jim, myself, and two other guys, we put the group together in the late 70s, and basically we've been together this whole time. And as I say, NIST has been very supportive of this group environment. And I think it's worked very well for all of us. And for all of my career, I've been working on methods to try to improve atomic clocks. And basically that's the name of the game for us—we're just trying to make oscillators that oscillate back and forth but at a very precise frequency and one that doesn't change in time.
Interviewer: As a physicist, do you have a particular notion of time?
DAVID: Well, I think as a physicist, we tend to take maybe a low-brow approach—we don't get too philosophical about what time is. But I think our notion of time generally is the same as everybody else's. It's a way to mark a series of events and durations of time, of course, and it's just a way to mark the distance between these events. So, I would say certainly in my case, and I think for most physicists, our notion of time is very much that it is the same for everybody else: how to get to work and be there when everybody else is there. It's really the same measure.
Interviewer: Why is there a need for better clocks?
DAVID: Well, I would say in the business of clocks, I don't see it ending. And I think—it has always been true for the last ten centuries, at least, that when a better clock was built, use was immediately found for it. And this particularly applies to navigation. So, whenever a better clock was devised, it improves our ability for navigation. We're at the level now where, for example, the global positioning system, the GPS is a good example where routinely we can go to the store and buy a device that tells our position to roughly ten meters or so. And yet, as we make the better clocks in the lab this technology eventually finds its way into the public domain. So, you might ask, well, aren't we good enough now in navigating this precision? And I think the answer is no. You can always find applications where we like to have better precision. And the good example that is starting to be developed now is to locate any position on Earth to say to the millimeter or centimeter level, this is very useful. For example, earthquake prediction uses this kind of data. And I think this is unending. If we could make clocks even that much better, after that application has solidified, someone will come up with an application where we can use more precise clocks.
Interviewer: What are the principles behind an atomic clock and how are they different than a mechanical clock?
DAVID: Basically all clocks are made the same way. A common example and one that everybody is familiar with is a grandfather clock. And there, typically, the length of the pendulum is adjusted so that it oscillates back and forth maybe once a second or once every two seconds. And from that, of course, we can drive the second. And actually, an atomic clock works on exactly the same principle. Rather than an atom having a little pendulum, you can still think of an atom as having a characteristic oscillation and in fact a vibration is a good example—a good analogy that the electrons and the atom are vibrating back and forth. And they act very much then like a pendulum clock. And in the same way, we just count these vibrations of the atom, these oscillations, and when a certain number of oscillations have gone by, then we know a certain amount of time has elapsed. And so, one dramatic difference in an atomic clock is that typically nowadays the vibrations we use are occurring not at once a second or once every two seconds but more typically about a thousand trillion times a second. Nevertheless, we have the capabilities now to count each one of these oscillations and when a thousand trillion have gone by, then we know that a certain time has elapsed.
Interviewer: What might affect the accuracy of different kinds of clocks?
DAVID: So, in a quartz clock—say the one sitting on your nightstand or in your wrist watch—the vibrating component is a small piece of crystal often quartz or something closely allied to quartz. If we were to strike it, it will vibrate, and if we cut the crystal to a certain size, and then a certain shape, then we can predict pretty well what the frequency that these oscillations occur at. So, one of the things, though, that constrains the accuracy of crystal clocks is we have to know the dimensions of this quartz crystal very well. And in fact, we're limited on how well we can know dimensions of the crystal, so that quartz clocks all run at slightly different frequencies. But in the case of atoms, all atoms of a given species, they run at exactly the same frequency. And so, there are no dimensions that come into play. We don't have to engineer that. We don't have to cut the atoms in a certain way as we did for the crystal. They all run at exactly the same frequency. So, our task is to be able to measure these frequencies. And so part of the limitation that we have on atomic clocks is just perturbations and how well we can determine the frequency that they're oscillating in. We're always thinking of ways to try to get rid of the effects or suppress the effects that affect the frequency of the clock. And one nice thing about our jobs here is this process will never end. We're always able to think of new ways to suppress these effects and clocks continue to get better for that reason.
Interviewer: What steps have you taken to improve atomic clocks at NIST?
DAVID: So, one thing that's happened in our development of atomic clocks is that the frequency of the vibrations that we use in the atoms or molecules has increased. So, for example, in the Cesium atom, the characteristic oscillations occur at about 10 gigahertz, which means about ten billion cycles per second. Now in the more modern clocks that we're pursuing now, the vibrations that are important to us are the vibrations typically of electrons in their orbits. And these occur at about a thousand trillion times a second. In other words, about a hundred thousand times faster than the oscillations that occur in the Cesium atom. And fundamentally, it isn't so important that it's oscillating faster, but in practice it turns out that using higher frequency vibrations gives us an advantage. And the simple way to think about that is, that in any given unit of time, the oscillations essentially divide that unit of time up by finer and finer steps. So, by going from Cesium to optical clocks, and we've divided up any unit of time, say the second, in that case by a factor of a hundred thousand. We just define these increments of times into finer and finer units. In fact, that's one of the main reasons we get a higher precision.
Interviewer: Describe the basic quantum mechanical principles behind the mercury ion clock.
DAVID: In our mercury clock—the same idea would apply for any other optical clock. What we do is we prepare the mercury ion—the mercury atom in its ground state. And then we apply radiation at these roughly thousand billion cycles per second. And when the radiation is tuned to a particular value, it causes the energy state of the mercury ion to go from a lower energy state to a higher energy state. And that occurs with maximum probability when this incoming radiation is tuned to the exact vibration frequency of this transition on the mercury atom. So, what we do in the lab is we make an apparatus that allows us to tell when the atom has gone from its lower energy state to its higher energy state. Now, in principle, the way we could detect it is we could wait for the atom to decay back to the lower energy state—what we call the ground state. And in principle, we could catch that photon coming out and measure it. It turns out that's not a very efficient process for detection. So, what we do in the case of mercury is we play a trick that involves the same lower energy state of the atom—the ground state—and we have another laser that when we shine that laser on the atom, if it's on the lower energy state, it will scatter light at a very high rate. So, let's say the clock radiation is not tuned to the right frequency, then the atom remains in its ground state. It can't make this transition to the excited level of the clock transition that we're using. So, when that happens—when the radiation is detuned, then when we shine the second laser in on the atom, we see scattering. And in fact, the atom lights up. It's a little dot—that if the radiation weren't in the ultraviolet that we're using—we could actually see it with our eye—a single atom. So, the idea is that then we use the second laser as kind of a discriminator. It tells us when the atom is in the ground state. On the other hand, when the clock radiation is tuned exactly right—when the atom goes from this ground state to the upper state we're using for the clock, and we shine the second laser in, the atom doesn't scatter anymore. So, we can clearly see this effect. The atom is either bright, or it's dark. And as I say, in the lab, we're not able to see this with our eye. It's actually bright enough to see with our eye, but it's in the ultraviolet. We typically just use a simple video camera that's very similar to the one you could buy at the store, but it's sensitive to ultraviolet light, and we see the atom blink on and off, and we know when it's either made the transition, or it hasn't.
Interviewer: Just how precise are your single-ion clocks?
DAVID: Often what's said about our advanced optical clocks is if we made a clock—and we could keep it running long enough—it would neither lose or gain a second in roughly the age of the universe. Well, that's a tough task to keep this clock running that long. But nevertheless, that gives an idea of the performance.
Interviewer: Are there effects that limit the precision of these advanced clocks?
DAVID: Yes. Well, we often use the term frequency jitter. There are myriad effects that can cause the frequency to change. A common one that we think about and have to worry about is that if the local magnetic field changes slightly, it will cause the vibrations to run at a slightly different rate. If the field fluctuates, then the frequency is going to change and fluctuate as well. It might be—well, a good example is the earth's magnetic field, which in part depends on the local environment and temperature. And so, for example, when the temperature changes, the earth's field can change slightly, and we have to worry about those effects. And another thing in the lab, where it's sort of a nightmare we have to worry about all the time, is that our electronic equipment that derives power from the 60 hertz electricity that we get out of the wall causes the frequency of the clock to move around back and forth at 60 hertz. And this you might classify as a form of jitter. It's something that's happening fairly rapidly. That nevertheless causes a frequency error in our clocks. Our main job then is to try to control these environmental effects we call systematic effects as well as we can. And in fact, that's fine. In the end, that's what limits the accuracy or the precision of the clocks is just how well we can control these environmental effects—or more to the point—our inability to control them as precisely as we'd like.
Interviewer: What other things might affect the frequency of the clock?
DAVID: In very simplistic terms—what we do is we go through every known force of nature and try to understand how these forces of nature might affect the clock. Now in the end, the dominant ones are electric and magnetic fields—electromagnetic fields. And they come from various sources. I mentioned, for example, the ambient magnetic fields that might be in the lab. But there are some more interesting ones. For example, gravity affects the rate that clocks run. One of the effects of gravity comes from Einstein's theory of general relativity. And one of the consequences of Einstein's theory of general relativity was that clocks, if they're placed near a gravitational mass, say the Earth—will run at a slower rate than if they're removed from the source—say clocks on a satellite. But nowadays the precision of the clocks is such that we have to worry, when we compare clocks, if one clock in one lab is 30 centimeters higher than the clock in the other lab, we can see the difference in the rates they run at. And this is an extremely small effect that we haven't had to worry about before.
Interviewer: Will increasing the accuracy of clocks allow the possibility of discovering new physics?
DAVID: Yes, well, one of the things that is interesting is, of course, our job as clock makers is to try to identify everything that might affect the frequency of the clock, and then we compensate or correct for that. But on the more fundamental side, what we're always hoping for is that we might discover something that nobody has seen before. And one of the things that's surfaced over the last decade or so is that clocks that drive their basic frequencies from physical principles might see the ratio of the frequencies that they run at starting to evolve over time. So, in fact, as a result of making better clocks, we're able to search for these fundamental effects. And in our recent comparisons of two optical clocks in our labs here, we have been able to put some of the most stringent limits on the ratios that the frequency of these clocks run at. We haven't seen anything yet, but if we saw something—first of all, we'd have to expect that—well, we just did something wrong. We didn't account for say a magnetic field in the right way. But if we could somehow rule out all the effects that we know about—we might be able to say something about that—well, there's a new fundamental effect there. I think, any physicists dream is to discover some new effect—and of course, we aren't banking on this. But we certainly want to keep our eye out. And so, we continue to perform these checks as the performance of the clocks gets better. And before we would say anything very loudly, we would have to check for a very long time to make sure it wasn't some simple environmental effect.
Interviewer: What parts of your work have you found most rewarding?
DAVID: I think you'll probably get this similar answer from Jim. I think when we've compared notes about our careers here, and maybe there's some paperwork we'd rather not do, but I would say we're pretty lucky with our jobs because we get to play with fancy toys and yet, it has a purpose. And it's like a really fancy video game where you can do amazing things. And the only difference, but an exciting difference is that we don't know the rules in the game exactly. So, part of the game is to find out the rules—how does nature work. So, I think that's one of the things that continues to make it exciting for us.
So, the other part of this physics is that it does involve mathematics. And I think one of the things I always liked was how the things we work on can be described by mathematics, quantum mechanics, for example. It's maybe more complicated than simple algebra, but on the scale of what the mathematician thinks about, it's not really that complicated. And yet, I never cease to be satisfied by how we can use this relatively simple math to describe what seem to be fairly complicated situations, including how our atomic clock works.