[LEAPSECS] The next primary frequency standard?
Jonathan E. Hardis
jhardis at tcs.wap.org
Sun Feb 7 13:49:39 EST 2010
>There are some other misleading statements in that article.
>There's no need to change the definition of the second because the latest
>frequency standard is based on a different quantum transition. Magnesium,
>mercury, and ytterbium have been used in the past, though AFAIK all the
>current primary frequency standards are caesium.
The SI "second" is currently defined in terms of a microwave
transition in cesium. This allows the best atomic clocks to have
accuracies on the order of 1 part in 10^15. Atomic clocks based on
optical transitions are thought to have potential stability on the
order of 1 part in 10^18. However, as long as the definition of the
second is based on cesium, they can never achieve better _accuracy_
than a cesium clock.
Because of this, many believe that one day the definition of the
second will be changed to reference an optical transition. Which
system might be best for that purpose is still a subject of research
>The speed of light in a vacuum is a fixed constant that is used to define
>the metre in terms of the second, so I don't know what questions about it
>need to be resolved.
The speed of light was perhaps not the best example. However, it is
perhaps the best known "universal physical constant," which goes back
to the point that the author was making. The question being explored
is how "constant" the physical constants might be. At the top of the
list is the value of the fine structure constant, alpha, which equals
e^2/(h-bar c). Today, "c" is a defined constant, and if the CGPM
continues along its present path, in a few years "e" and "h-bar"
might be defined constants as well. Nonetheless, one can determine
experimentally an upper limit to the rate-of-change, if non-zero, of
alpha. There are physical theories that predict that alpha does, in
fact, change over time because of coupling between quantum physics
For many years the best bound to "alpha-dot," as the issue is
sometimes known, was based on astronomical observations of the early
universe. Today, the best bound comes from studies utilizing optical
atomic clocks. As the precision of these clocks increases (i.e.,
towards a part in 10^18), the bound on "alpha-dot" can be tightened.
If, in fact, a non-zero alpha-dot were to be observed, it would be
one of the great discoveries in physics in the 21st century.
Similarly, optical atomic clocks are being used to look for drifts in
the electron/proton mass ratio, and the light quark mass (the average
mass of up and down quarks).
More information about the LEAPSECS