Building the World’s Smallest Atomic Clock: A Nitrogen Atom in a Carbon Cage | Hacker News Reader

Building the World’s Smallest Atomic Clock: A Nitrogen Atom in a Carbon Cage

spectrum.ieee.org

62 pointssohkamyung33 comments

They're shockingly ugly (and expensive) beasts, but you can now get atomic wristwatches. (Actual atomic watches, not the "atomic" kind that simply receives the radio time signal.)

https://www.hoptroff.com/collections/atomic-timepieces

Well the pocket watch isn't that bad looking......

Perhaps that's the reason why it's the only model that's sold out.

> Requires an iPhone or Android phone.

Oh my.

For the initial time sync, apparently. I guess if you have an atomic clock on your wrist, you want the time displayed to actually be correct.

I would definitely be willing to pay less for an atomic stopwatch.

That’s impressively small. I had a rubidium frequency standard which was several orders of magnitude larger I.e medium sized hardback book. This used a cesium beam approach. Haven’t seen one of them less than a 2U rack before.

Dubious as to the point though as the sources all have a finite operational life, usually a decade or so.

Nowadays wrist watches rarely have a point besides showing off your wealth.

This is true. I don't actually own one because it is of no use to me.

How do you know what time it is?

Look up a bit on your phone. ;)

Exactly that! :)

Ah so you take your phone out of your pocket to check the time? Like a pocket watch! Cool, now if only there was something you could keep on your wrist that told time.. Would be so much more convenient!

I actually have a Casio watch that is identical to one that I used to have 30 years ago. No frills or whistles. Roughly the same price now as it was back then, too. Under ten pounds.

(this obviously does not contradict your point - I mention it because I like talking)

I wore a Casio watch before I got a phone in the late 1990s. Really the best utility there is in the watch space. It's a done and dusted thing.

It's super interesting how even before phones, knowing the time was so important that we would put it on our wrists.

We have clocks in most places. You could put a clock into your bag. But no, this is so important that we would place this on our wrists for instant access.

Time is only important if you have to synchronise events with other people.

Or if you want to keep track of your work schedule.

One second per millennium accuracy, it says. I wonder how much empirical testing has been done to compare these to traditional quartz watches. A quick Google search puts quartz watches at about 15 seconds per month. I’m sure the atomic oscillator itself is more resistant to everyday variance (like temperature) than quartz, but I wonder if the detector is actually that much more accurate in practice, or if they’re just citing a theoretical number.

Oh well, at the end of the day your clock is still going to vary based on your altitude (distance from Earth’s gravitational field), unless you also correct for that. ;)

To save power the atomic resonator doesn't run all the time. The practical accuracy is apparently 1.5 seconds per thousand years. Still pretty good.

Could atomic clocks provide more accurate GPS positioning for GPS receivers? It’s my understanding that GPS works by measuring tiny differences in the time it takes equivalent signals to reach the receiver from different satellites. It may be the case that other factors (like noise and refraction) dwarf any inaccuracies in the quartz clocks in our everyday GPS receivers.

We don't need super accurate clocks in the GPS receivers; instead, the GPS satellites carry atomic clocks with them, and include timestamps in their broadcasts. So you get two signals at the same time, compare the time the satellite said it sent each signal, and you've got the difference in travel time.

Oh, interesting. What’s the time granularity transmitted by the satellites? I assumed that the satellites broadcasted a course stream of times and the receiver had to finely measure its local time difference between receiving the same time from multiple satellites.

My assumption, for example:

Receiver receives time “noon” from GPS satellite 1 at receiver-local time T=0.

Receiver receives time “noon” from GPS satellite 2 at receiver-local time T=2 nanoseconds.

Receiver receives a couple more “noon”s from other satellites, then uses its receiver-local time differences to trilateral its location.

So did some poking around; apparently, there are two timestamps in GPS messages, and the more precise one is packaged in a custom hash function and called the "ranging code". The timestep (called a "chip") is a bit under a microsecond.

But yes, the receiver does have to measure small time differences between received signals; but that just requires a local clock with high resolution, not one with low drift or low absolute error.

For example, if your local clock is a second off, but still measures time intervals to pretty high precision (say, 0.01% - losing a second every three years) that's probably not going to mess up your location fix very much. Especially when the problem is overconstrained as it usually is intentionally. Compare the theoretical minimum requirement of three visible satellites to make a measurement with the 4-12 satellites that the system guarantees will be in sight at any time.

From the article:

> GPS receivers will benefit from onboard clocks. This is perhaps surprising, given that the GPS signal itself carries time information, but it happens because the signal from the satellite is weak—comparable to the power of a lightbulb transmitted across a continent. Landscape features, buildings, and interference make it harder to detect. To track this weak signal, the receiver must precisely lock onto the broadcast frequency. The more stable the local frequency reference, the faster and more reliable this tracking can be.

> In hostile environments, such as battlefields, this becomes even more important. The GPS signal is vulnerable to jamming, and effective (but illegal) jammers are widely available and likely to be encountered in future wars. With precise timing information, GPS receivers could isolate the true signal above the noise of the jammer. The receivers could even allow navigation to survive partial destruction of the satellite network.

> Present-day receivers must determine their positions by using signals from four or more satellites simultaneously, but a sufficiently accurate clock could instead use successive signals from a single satellite. Other defense applications include frequency-hopping communication, bistatic radar (in which an attacker stealthily acquires a radar signal from a target illuminated by a distant transmitter), and sensitive monitoring of enemy communications. For these reasons, portable clocks are of great interest to militaries in several countries.

Hm; tried to revive some primary school-level physics/chemistry knowledge to calculate price-per-molecule or #-of-molecules-per-£. Is my calculation correct, or have I botched it? :)

Given atomic mass of C is 12.011, of N is 14.007, the molar mass [1] of the fullerene should be:

M(N@C60) = 14.007 + 60×12.011 [g/mol] = ~734.667 [g/mol]

Then, given:

N_A = 6.022140857e23 [1/mol] — https://en.wikipedia.org/wiki/Avogadro_constant (number of molecules in 1 mol)

p_g = 200e6 [£/g] — price per gram given in article

n = ? [1/£] — number of molecules one would get for £1

I devised:

n = N_A / (p_g × M(N@C60)) = [1/mol] / ([£/g] × [g/mol]) = [1/mol] / [£/mol] = [1/£] = 6.022140857e23 / (200e6 × 734.667) [1/£] = ~ 4e12 [1/£] ?

In other words, it seems you could buy ~4'000'000'000'000 molecules for £1 ?

—

[1] https://en.wikipedia.org/wiki/Molar_mass

The description of their "high-pressure liquid chromatography" technique sounds an awful lot like how one would enrich uranium (you know, as one does.) Did they accidentally invent a better uranium enrichment process (or independently reinvent a classified one?) Or is there some difference between {C60, N@C60} and {U235, U238} that makes using HPLC to enrich it impossible?

HPLC separates molecules based on some chemical property. In this case, the N@C60 fullerenes are probably very slightly larger, distorted from pure icosahedral symmetry, or have different electronic structure to the plain C60. Chemically speaking, U235 is probably very much like U238 and the chemical compounds of the isotopes are probably chemically almost identical (since isotopic chemical reactivity depends basically on the so-called kinetic isotope effect, which is larger if there is a large difference in mass between isotopes). HPLC is basically a standard synthetic laboratory procedure and has no relevance (that I know of) to uranium separation aside from purifying bulk chemicals.

It seem to me one could apply an RF field at the resonant frequency of the nitrogen to selectively heat the N@C60 molecules. Do this near their boiling point to distill them. Or maybe they'd break down before boiling off. This of course would not work for isotope separation.

My understanding was that HPLC was just a way to do gaseous-diffusion enrichment (i.e. "standard chromatography") without the really long tubes. From the article:

> In standard chromatography, substances having different chemical characteristics are separated by making them run a kind of gauntlet—an obstacle course that blocks the passage of one thing more than the other. HPLC works by using a pumped solvent (hence the term “high pressure”) to strip off the laid-down film of carbon fullerenes in such a way that the desired molecules—the fullerenes encasing nitrogen—are carried away preferentially.

My interpretation of that statement is that HPLC uses a solvent that manages to detach the slightly lighter (and/or less electrostatically-attracted-to-the-film) molecule first, giving it a head start down the diffusion tube, letting the tubes be really short, and thus giving you a tabletop device instead of capital equipment in a dedicated facility.

Maybe HPLC of N@C60 really leans on that differential-electrostatic-attraction property, but it seems like the process would still work to separate isotopes that differ only by molecular weight.

I'm not a chemist of any sort, though, so I'd be glad to "get schooled" here.

So how precise would this make our location?

Could I replace my 3D-mouse by a cellphone using this technique?

I’m not sure that is the goal. Having cheap and small atomic clocks simplifies time sensitive applications that can’t always rely on satellites or ntp. Clock synchronization is currently not a solved problem for the masses.

Reply on news.ycombinator.com