NIST-F2 Atomic Clock News Briefing: Statement from Steve Jefferts

[full news release]

So as Tom said, we're announcing that we've built another atomic clock, NIST-F2. We're terribly imaginative about the names around here. But F2 is substantially more accurate than our current official frequency standard, NIST-F1.

In order to make NIST-F2 an official time standard or frequency standard for the U.S. we've had to basically get it validated by the International Bureau of Weights and Measures. It's been accepted by them. They've got the data from the clock, there's a technical paper in to the journal Metrologia, which should be published next month, I guess. It's been accepted in any case. So that all clears the path for NIST-F2 to be the official frequency standard for U.S. civil time.

F2 is really a refinement on NIST F-1. They are very similar clocks in a lot of ways if you look at the big picture of it. But when you look at the inaccuracies inherent in the NIST-F1 design, you come to the realization that it's dominated by a frequency shift called the blackbody radiation shift, which really is a result of the walls of the experiment glowing, if you will—emitting blackbody radiation which shifts the energy levels of the cesium atom a little bit and causes an inaccuracy. And it turns out if you cool the walls of the experiment you can make the shift get smaller and therefore the inaccuracy associated with the shift gets much smaller. And it's one of the few places that Mother Nature is nice to you, because the shift gets smaller in a big hurry as you cool things off.

And so by cooling NIST F-2 down to about 80 degrees above absolute zero, from 300 degrees above absolute zero where NIST-F1 runs, we've made the shift more than 100 times smaller, and the inaccuracy associated with the shift gets much, much smaller as well.

Having more or less eliminated that shift from the inaccuracy budget of the clock, we're left with a number of shifts that depend mostly on the microwave cavity associated with the clock, and it turns out that the microwave cavity gets much, much better when you cool it off as well, and so those shifts also got much smaller. So at the end of the day, NIST F-2 today is about a factor of three more accurate than NIST F-1 is. And that means that roughly speaking if you could make either one of them run for 100 million years, NIST F-1 would lose a second and NIST F-2 might lose a third of a second in that period.

Why is that important?

As Tom was saying, not so far in the future we're going to end up redefining the second in terms of these optical clocks that are now research devices. And when we do that we need to be able to make the transition from the definition of cesium, from the use of cesium as the definition of the second, to the use of some optical standard that hasn't been determined yet. And in order to do that seamlessly you need the best measurement you can of the relative frequency of those two clocks.

I should say that that redefinition is not NIST's job, that's a thing that happens at the international metrological community level, and so it will require a whole bunch of people agreeing for that to happen.

Let's see, what else do we have to say about this ...Well, there's a picture of the clock up there. We also have a demonstration video—that's available online as well—about how the clock works, sort of a cartoon, and maybe we'll play that and we can give you the quick description of how this all works.

So we start off in this clock by gathering a sample of cesium atoms out of a background cesium vapor using lasers and laser cooling. And we cool off this sample of maybe 10 million cesium atoms to a temperature which is well below 1 millionth of a degree kelvin. You can sort of see this happen here. So now we have this sample, and with a little trick with the laser beams we launch the sample. The sample goes up, interacts with a microwave field in this cavity, which is going to try to make the atoms change states.

If the frequency is right, the atoms will change states. And you see here a few of them went from blue to red, which means the frequency wasn't quite right. And we count how many change states by looking at the atomic fluorescence. Now we'll do it again with a slightly higher microwave frequency, the atoms will come back and if we're closer to right more of them will change state. You see most of them change to red, you got more photons out as a result, and we get a larger peak. So this is basically . . . you're now seeing how this works. Now the microwave frequency is much higher, and the atoms come back, interact with the microwaves a second time, only a few of them change states. So the intermediate frequency was closer to right.

And we just keep doing this over and over to optimize our knowledge of the frequency required to make the transition. So that's more or less how it works.

Why is it important? Well, I think Tom covered that pretty well already.

Nothing here is going to change the way we live tomorrow in terms of having a three times more accurate clock. But these technologies keep getting adopted for use in our society and so we have to keep inventing things to make them better because we keep using them. I would say that's more or less what I have to say.