to develop the standards that
serve as reference for time and frequency services and to
research advanced measurement systems.
INTENDED OUTCOME AND BACKGROUND
The Time and Frequency Division
maintains standards with the accuracy,
continuity, and stability essential for
supporting U.S. commerce and scientific
research; provides an official source of
time for U.S. civilian applications; and
supports coordination of international
time and frequency standards, including
realization of the SI second.
NIST time and frequency standards are
based on the NIST time scale and the
NIST primary frequency standard,
NIST-F1. The time scale is an ensemble
of five (soon, six) hydrogen masers and
four cesium-beam clocks. The stability
of the time scale is approximately
0.2 fs/s for thirty days of averaging, with
a long-term frequency drift of less than
3 (fHz/Hz)/year. The frequency of the
time scale is calibrated by periodic comparisons
to the NIST-F1 laser-cooled
cesium primary frequency standard
(9.2 GHz microwave frequency), with
a fractional frequency uncertainty Δf/f
approaching 6 × 10-16 (0.6 fHz/Hz,
as of June 2004).
The NIST time scale is the basis of
NIST's realization of Coordinated
Universal Time (UTC), the international
time scale. NIST is one of about 60
timing laboratories across the world
continuously contributing to the realization
of UTC. Through improvements to
the NIST time scale, NIST's realization
of UTC rarely differs from the international
average by more than 10 ns.
In addition, NIST is one of only
four laboratories worldwide (as of late
2004) operating the highest accuracy
primary frequency standards to determine the frequency (rate) of UTC.
The extraordinarily stable NIST time
scale, coupled with world-leading performance
of the NIST primary frequency
standard (as of June 2004), provides
U.S. industry and science with a unique
resource for the most demanding applications
of accurate time and frequency.
However, commercial and scientific
needs for even more accurate and stable
time and frequency standards drive a
vigorous NIST research program to
improve microwave frequency standards
and to develop new, optical frequency standards.
Since the first atomic clock was invented
at the National Bureau of Standards
(NIST's predecessor) in 1949, the performance
of primary frequency standards
has consistently improved by
about a factor of ten each decade--driven by, and enabling, advances such
as telecommunications synchronization
and the Global Positioning System
(GPS). NIST research on microwave
and optical frequency standards strives
to at least maintain this rate of performance
improvement.
Accomplishments
Improved Performance of
the NIST-F1 Primary Frequency Standard
The NIST-F1 laser-cooled, cesium
fountain primary frequency standard
(Fig. 1) is the U.S. national standard for
frequency and the realization of the SI
second. Since the first formal report of
NIST-F1 frequency to the International
Bureau of Weights and Measures
(BIPM) in 1999, the NIST-F1 uncertainty
has been reduced by about a factor of three.
|
Figure 1. The NIST-F1 cesium fountain primary
frequency standard, the Nation's standard for
frequency and the SI second. |
The NIST-F1 frequency evaluation
reported to BIPM in June 2004 included
an "in-house" fractional frequency
uncertainty of approximately 6 × 10-16
(0.6 fHz/Hz), increasing to about
9 × 10-16 (0.9 fHz/Hz) as received
at BIPM due to uncertainties in the
satellite-transfer process. Both of
these results were, at the time, the
best ever reported to BIPM--by a significant margin.
Ongoing NIST-F1 changes continue to
improve performance. New laser systems,
for generating the optical molasses
and repumping, maintain frequency
lock for weeks without intervention.
Since NIST-F1 data must be averaged
for about a month to achieve ultimate
uncertainties, this improvement has substantially
reduced downtime. The higher
power molasses-laser system also permits
reshaping of the atom cloud to reduce
atom density in the vertical (launch)
direction, decreasing the spin-exchange
shift without loss in detected atom
number. Systems that ensure that
mechanical shutters fully block the lasers
have reduced the light-shift uncertainty
by 50 %. Other improved systems
include 9.2 GHz microwave synthesis,
vacuum, temperature, computer
automation, and data transfer to the time scale.
We are carefully exploring the role of
deadtime in precision frequency measurement
and satellite transfer--that is,
time when the NIST maser ensemble
continues to operate, and potentially
drift, while the primary frequency reference
is not operating. In addition to
purposely building deadtime into formal
frequency evaluations, the Division is
collaborating with the NIST Statistical
Engineering Division to formally model deadtime effects.
Second-Generation Cesium
Fountain Primary Frequency Standard
While continuing to optimize NIST-F1,
the Division is actively developing the
next-generation primary frequency standard,
NIST-F2. The largest nonstatistical
contributions to NIST-F1 frequency
uncertainty are the spin-exchange shift,
due to cold atom collisions, and the
blackbody shift (AC Stark shift from
thermal radiation near 300 K). NIST-F2
is intended to substantially reduce those
uncertainties through two means.
First, NIST-F2 will use a multi-toss,
multiple-velocity system, in which
about ten low-density atom balls will be
launched to different heights in rapid
succession, all coalescing in the detection
zone without having crossed paths
in the Ramsey interrogation region.
This approach will minimize spin-exchange
shifts while still providing
sufficient atom numbers at detection
for good stability.
The general principles of this approach
have been successfully demonstrated in
NIST-F1. To make the system practical,
NIST-F2 will use a low-velocity intense
source (LVIS) of atoms for the required
rapid cooling, capture, and launch of
many atom balls in succession. Early
results demonstrate molasses atom fill rates of about
4 × 109 s-1.
In addition, the fountain team has demonstrated
cooling to approximately 1 µK using the
required (1,1,1) NIST-F2 fountain
geometry. Both these results are commensurate
with NIST-F2 performance requirements.
The second major improvement in NIST-F2 will be to cool the drift tube
and interrogation regions to cryogenic
temperatures, vastly reducing the blackbody
shift. Use of different cryogens,
and/or pumping on the cryogens, will
also enable accurate measurement of the
blackbody shift, the value of which has
been the subject of intense debate.
The ultimate goal for NIST-F2 is to
approach an "in-house" fractional frequency
uncertainty of 1 × 10-16 in the next few years.
Research on Optical Frequency Standards
While the ultimate accuracy limit of
cesium microwave standards operating
near 1010 Hz is not yet clear, optical frequency
standards operating on the order
of 1015 Hz have the potential for substantially
greater stability and accuracy.
Optical frequency standards also have a
potential for dissemination through
optical fiber, which may be advantageous in many applications.
The Division conducts a vigorous
research program on prospective optical
frequency standards, simultaneously
pursuing several different approaches.
These include cold, trapped single ions; cold, neutral atom clouds;
and "logic clocks," using techniques of quantum
computing. A crucial part of optical
clock research is optical frequency synthesis
using femtosecond laser frequency
combs, described in a later section.
Mercury-Ion Optical Frequency Standard
A frequency standard based on optical
transitions (282 nm, 1064.7 THz) in a
single, laser-cooled, trapped mercury
ion has potential for better accuracy
than cesium fountain standards by a factor
of 100 or more. With a Q factor
> 1014, and a transition that is relatively
insensitive to environmental factors, the
potential fractional frequency uncertainty
Δf/f for a mercury ion standard is as
small as 10-18.
The reproducibility of the mercury ion
standard relative to the NIST-F1 primary
frequency standard was studied
over a two-year period. Measurements
were referenced to NIST-F1 through a
hydrogen maser, using a femtosecond
laser frequency comb to compare optical
and microwave frequencies. The mercury-ion optical frequency standard has
better short-term stability than NIST-F1,
so the stability of the maser was key to the intercomparison.
The experiment set a limit on the
fractional frequency variation of the
NIST-F1 cesium microwave transition,
referenced to the mercury ion optical
transition, as no greater than ±7 × 10-15
per year. From the point of view of the
cesium transition being defined as the
international frequency standard, this
means the mercury ion clock frequency is stable to this level.
This experiment also set a limit on possible
variations of fundamental constants
related to the fine-structure constant a,
in particular the possible temporal variation
of the quantity gCs
(me/mp) α6.
Assuming any variation is due to a only,
this result sets an upper bound for the
fractional change in the fine-structure
constant as no greater than 1.2 × 10-15
per year. It is a tighter bound by about a
factor of 30 than astronomical observations
suggesting variations in the finestructure
constant over periods comparable to the age of the universe.
Calcium-Atom Optical Frequency Standard
|
Figure 2. The calcium-atom optical frequency
standard, being adjusted by Chris Oates. |
The calcium optical clock transition at
about 657 nm (456 THz) has several
advantages. It is spectrally sharp
(470 Hz natural linewidth), can be
detected with excellent signal-to-noise
using shelving-detection techniques, and
is largely insensitive to perturbations
from external electric and magnetic
fields and from cold collisions.
The frequency of the calcium optical
clock transition has been measured to a
fractional uncertainty approaching
1 × 10-14, making it one of the bestknown
optical frequencies. The 423 nm
resonance line permits rapid first-stage
laser cooling to the Doppler limit of
about 2 mK, and we have demonstrated
three-dimensional, quenched narrowline
second- and third-stage laser cooling
to about 10 µK. Experiments to slow
calcium atoms further using pulsed,
one-dimensional, quenched narrow-line
laser cooling have reached temperatures
as low as 300 nK. (See Fig. 2.)
A calcium optical frequency standard
should ultimately demonstrate fractional
stability approaching 2 × 10-16 with
one second of averaging. Performance
comparable to cesium fountain microwave
standards could be attained with
averaging over seconds, rather than weeks.
It is likely that ion and neutral atom
optical standards will each display ultimate
strengths and limitations. They
may find uses in different applications
in much the same way that different
microwave standards are optimal for
different applications--such as cesium
standards for absolute accuracy and
hydrogen masers for greatest stability.
First strategic focus |
Second strategic focus |
Third strategic focus |
Fourth strategic focus
"Technical Activities 2004" - Table of Contents |