- Deterministic Quantum Entanglement. Q. Turchette,
C. Monroe, D. Wineland, and other members of the quantum computer
project of the Ion Storage Group have recently demonstrated the ability to
entangle two quantum particles with high efficiency, advancing the
possibilities for reducing noise in stored-ion frequency standards and
demonstrating realistic quantum computation. Previously, entanglement of
particle states was obtained by post selection from a large number of trial
experiments, such as the production of two correlated photons that occasionally
occurs when a single photon passes through a special crystal. Such entanglement
has proven useful for tests of quantum nonlocality, but entangling a large
number of quantum particles--essential for noise reduction in atomic frequency
standards and for building a practical quantum computer--becomes much less
likely if it is dependent on a probabilistic process. In their "deterministic
entanglement" process, a pair of beryllium ions is confined in an ion trap and
laser cooled. Using a predetermined sequence of laser pulses, the internal spin
of one ion is entangled with the ions' shared external motion, and the motion is
entangled with the spin of the other atom. Entangled spins are therefore
obtained "on demand" in each trial. It should be possible to apply
the techniques used in these experiments to entangle larger numbers of ions.
(Q. Turchette)
- Lithographically Fabricated Micro-Traps. In collaboration with
J. Beall of EEEL, C. Myatt of the Ion Storage Group has designed and
fabricated the first-generation, lithographic, linear ion trap from a ceramic
substrate with gold-plated electrodes. Lithographic fabrication (see
Fig. 1) provides for more precise control of dimensions for small traps,
and allows construction of the much-more-complex trap arrangements needed for
future work on the entanglement of larger numbers of ions.
Figure 1. Schematic representation of the
lithographic trap. The four rods in the traditional linear trap are shown at
the top. In the new trap, shown with substantial separation between the two
substrates for clarity, metallization along edges of the slots in the two
substrates replaces the rods.
Small numbers of ions have been laser cooled and crystallized in the trap with
confinement provided by alternating electric fields at frequencies up to
200 MHz. Independent laser beams have been tightly focussed on each of the
two trapped ions, which were separated by about 5 µm, with the
off-focus ion receiving only 1/5 of the radiation intensity intended for the
other ion. This is important because many of the applications of these
entangled states require the separate addressing of individual ions. In
addition, the Group has recently shuttled ions along the axis of the trap and
separated them by applying pulsed voltages to the electrodes. This technique
may relax the laser-focusing constraints for quantum logic gates and
individual ion detection and suggests the possibility of multiplexing a more
complex array of trapped ions by moving ions between accumulators.
(C. Myatt)
- Laser Cooling to the Ground State for Two Ions B. King and
other members of the quantum computer project of the Ion Storage Group have
cooled two ions to the ground state of motion, an important step in reducing
noise in stored-ion frequency standards and in implementing quantum logic
operations on multiple ions. Of critical importance is the heating rate and
decoherence of the modes of two-ion motion. The Group found that the
center-of-mass modes of the ion pair are heated at a rate of 5 quanta/ms
to 10 quanta/ms, similar to previous single-ion results. However, the
three, internal, motional modes (stretch and rocking modes) are found not to
suffer from heating, up to the level of experimental uncertainty of about
0.1 quanta/ms. This is not unexpected, since the internal modes are immune
from the effects of noisy fields affecting both ions equally. The heating
results indicate that the (unknown) source is not differential heating, thus
ruling out sources such as atomic collisions, field gradients, and certain
types of rf-micromotion heating.
These results imply that internal modes are more suitable than any
center-of-mass mode for use in quantum-logic or noise-reduction schemes
following the general proposal of Cirac and Zoller. However, any logic
operation will be affected to higher order by motion in at least one of the
center-of-mass modes, so center-of-mass heating remains a concern.
(B. King)
- Narrowest Linewidth Laser. In the course of recent efforts to
develop an optical frequency standard, B. Young and J. Bergquist,
with F. Cruz from Brazil, have demonstrated the narrowest-linewidth
visible laser ever built. The ultimate goal of this project is to lock a
narrow-linewidth laser to a well defined (282 nm, 2 Hz linewidth)
transition in 199Hg+ ions. The resulting optical
frequency standard could be used directly in the optical region, or frequency
divided to the microwave region to serve as a traditional atomic clock.
In order to demonstrate the performance of the new laser system, the difference
frequency (beatnote) between two laser beams, locked to independent reference
cavities on independent isolation platforms, was shown to have a linewidth of
0.84 Hz for an averaging time of 40 s (see Fig. 2). This implies
that the linewidth of one of the lasers was less than 0.6 Hz,
corresponding to a fractional linewidth of about
1 × 10-15. The key to this result is the isolation
from seismic and acoustic noise and from pressure and temperature fluctuations
of the high-finesse, optical cavities used to stabilize the lasers.
Figure 2. Beatnote observed in mixing the outputs of two
separate 563 nm sources. The inset shows the simple experimental
arrangement. The data are for a measurement time of 40 s.
The performance of this laser is now better than that needed for a local
oscillator for the optical, mercury-ion, frequency standard. Since systematic
frequency shifts for the optical-frequency standard are anticipated to be very
small, this new standard should perform better than all previous frequency
standards. (J. Bergquist)
- A Diode-Laser, Optical Frequency Standard Using Trapped Calcium Atoms.
C. Oates, L. Hollberg, and guest researcher F. Bondu have
developed and tested an all-diode-laser optical frequency standard based on
trapped, laser-cooled, calcium atoms. Measurements of the intercombination line
at 657 nm demonstrate linewidths as narrow as 400 Hz (line
Q = 1012), the natural linewidth for this transition. The
fractional frequency stability of a laser locked to this resonance (in an
unoptimized system) is already 5 × 10-14
τ
-1/2, where
τ is the averaging time. Ramsey fringes for
typical operating conditions are shown in Fig. 3. The narrow linewidth of
the transition along with the convenient wavelengths for probing and cooling
(allowing use of diode lasers for interrogation and cooling) makes this an
especially attractive optical-frequency standard.
Figure 3. Ramsey fringes for the cold-calcium optical
frequency standard.
The trapping and cooling light at 423 nm is generated by frequency
doubling 846 nm light from a semiconductor master-oscillator power
amplifier. This provides 50 mW of 423 nm light, which is used to cool
and trap 107 atoms in 20 ms. A "shelving-detection"
scheme, similar to that used for spectroscopy of trapped ions, has been used to
interrogate the calcium line and stabilize the frequency of the 657 nm
laser (fast linewidth of about 50 Hz). A phase-coherent,
frequency-measurement chain for connecting this transition to the 282 nm
transition of the mercury ion is now under development. (C. Oates).
- Subsystems for Optical Frequency Measurements. Staff members of
several Groups in the Division have developed laser-frequency measurement
technologies for application to the construction of a frequency synthesis
system capable of measuring optical frequencies with an accuracy limited by
atomic frequency standards. The system will be used to interconnect and compare
new optical-frequency standards such as the calcium and mercury-ion standards,
and eventually to connect these references to the cesium primary frequency
standard. The design concept involves three successive subdivisions of
optical-frequency intervals plus one frequency doubling to arrive at a point at
which ultrashort laser pulses can be used to measure the smallest frequency
interval relative to the cesium primary standard. The ultrashort-pulse
technique (developed at the Max Planck Institute) to be used in this last step
involves a comb of closely spaced frequencies generated by a mode-locked,
titanium-sapphire laser.
The larger steps of optical-frequency subdivision are achieved using mixing
crystals of periodically poled lithium niobate. These crystals have been custom
fabricated (in collaboration with staff of EEEL) to obtain efficient
second-harmonic generation and sum/ difference frequency generation. With these
devices a preliminary measurement was made of the mercury ion transition
(532.360 800 THz) relative to the accurately known calcium transition
at 657 nm. Further development of the system this year will provide an
improved measurement of the mercury-ion transition relative to the calcium
transition. Completion of the connection to the cesium standard is expected
within approximately two years. (L. Hollberg and J. Bergquist)
- Direct Observations of Spatial Structure of Crystallized Ion
Plasmas. Large numbers of beryllium ions (up to 106 or more) can
be stored in Penning traps by a combination of static electric and magnetic
fields and laser-cooled temperatures so low that the ions freeze into a rigid
lattice. Previously, Bragg scattering with the same laser light used to cool
the ions was used to determine general features of the spatial structure. It
was found that, for approximately spherical plasmas having 200,000 or more
ions, the Bragg scattering pattern was consistent with a body-centered cubic
(bcc) lattice, the theoretically predicted structure for the infinite-volume
limit.
More recently, T. Mitchell, J. Bollinger, and W. Itano of the
Division observed direct images of the ions fluorescing in large, spherical
plasmas. These images showed the central regions of the plasmas to have bcc
structure, or, more rarely, face-centered-cubic structure. In order to obtain
these images, it was necessary to phase-lock the rotation of the plasma to an
external electric field. The imaging camera could then be gated synchronously
with the rotation, which has a frequency of about 1 MHz in this case. The
ability to control the state of the plasma with this high a degree of accuracy
would be of great importance to a frequency standard based on ions in a Penning
trap.
Very flat, radially extended plasmas have been observed to have a structure
like a stack of planes. As the density of the ions is increased, the plasma
goes through a series of structures having square, rhombic, or hexagonal
ordering within a plane. At certain values of the density, another plane is
formed. The observed structures are in good agreement with calculations made by
D.H.E. Dubin of the U. California, San Diego. (W. Itano)
- NIST-F1: A Cesium-Fountain Primary Frequency Standard. Over the last
half year, S. Jefferts, D. Meekhof, and D. Lee, along with
F. Levi of the IEN in Italy, have brought NIST's new cesium-fountain
frequency standard into operation, and they recently completed a preliminary
evaluation of its uncertainty at a level commensurate with that of NIST-7
(5 × 10-15). This evaluation was limited entirely by
statistical noise (not systematic effects), since the stability of the standard
is still more than an order of magnitude worse than we expected. A Ramsey
pattern for the standard is shown in Fig. 4. The narrowest Ramsey fringe
observed to date has a width of about 0.6 Hz. The magnetic field applied
to separate the Zeeman lines is 0.1 µT, and the Ramsey fringes
observed on the first Zeeman line indicate that the field along the flight path
of the atoms is uniform to about 10 pT. This indicates that the magnetic
shielding is quite good.
|
This standard differs from other fountain standards in that the microwave
cavity and atom drift tube are an integrated structure that serve as the vacuum
chamber for the standard. This provides exceptional immunity to microwave
leakage fields. For all other fountain standards, the microwave cavity and
drift tube are contained within a vacuum chamber, and microwave leakage can
cause difficulty. The laser system used to generate the multiple beams that
cool, trap and launch the atoms involves a single master-oscillator power
amplifier that provides sufficient power for all of the beams. Other fountain
standards typically employ an array of independent diode lasers that are
injection locked to a lower-power master oscillator. Future efforts on this
project will focus on improving stability so that better evaluations can be
made of systematic frequency shifts. (D. Meekhof and S. Jefferts)
|
|
Figure 4. Ramsey fringes for the NIST cesium-fountain frequency
standard. The dots show the actual data, while the lines simply connect data
points in sequence, so the amplitude noise apparent in the envelope of the
curve is not real.
|
- A Laser-Cooled Primary Frequency Standard in Space. Several Groups
within the Division, in collaboration with staff members of the Atomic Physics
Division and a faculty member at the U. Colorado, are now engaged in a
flight definition study for a laser-cooled-cesium clock in space. The project,
called Primary Atomic Reference Clock in Space (PARCS), is aimed at an improved
realization of the definition of the second, improved coordination of
time/frequency standards on earth, and tests of several aspects of special and
general relativity. The microgravity environment of space allows us to use
slower atoms and increase interrogation time, thus improving clock performance.
Assuming that the project can remain on an ambitious development schedule,
flight should occur aboard the International Space Station (ISS) in 2003.
Figure 5 shows a schematic diagram of the proposed clock. The core of the clock
(the physics package) is made up of: 1) the atom-preparation region where
atoms are laser cooled, trapped, and launched; 2) a microwave cavity where
atoms are subjected to microwave radiation at the cesium resonance frequency;
and 3) a detection region where laser fluorescence is used to determine
whether the microwaves have caused a transition. The objective is to achieve a
stability of 3 × 10-14 τ-1/2 and an absolute uncertainty of
1 × 10-16. (D. Sullivan)
Figure 5. Diagram of the proposed space clock. Details of the
microwave cavity have been omitted to more clearly show the expansion of the
atom balls as they proceed through the cavity.
- Joint U.S.-Japan Development of a Frequency Standard. A joint
project between the Time and Frequency Division and the Communications Research
Laboratory (CRL) in Japan to develop an improved version of NIST-7, the U.S.
primary frequency standard, has recently been completed. The objectives of this
project, funded by CRL, were to construct an optically pumped standard with an
uncertainty comparable to that of NIST-7, to compare this new standard with
NIST-7, and to improve a number of subsystems allowing for more-rapid,
automated evaluation of systematic frequency offsets. B. Drullinger and
D. Lee led the project, with major contributions from D. Jennings,
L. Mullen, C. Nelson, J. Shirley and F. Walls. In addition,
during the entire three-year course of development, at least one staff member
from CRL was always engaged in the project.
The final comparisons between the new standard and NIST-7 indicate agreement
within the nominal uncertainty (5 × 10-15) of the two
standards. Major improvements made during the project included a more robust
diode-laser system for optical-state preparation and detection, new
servo-control and monitor software using a more-flexible object-oriented
approach, identification of a number of smaller sources of systematic offset,
and improved modeling of several of the larger systematic frequency shifts.
Improvements made to the new standard during this development project will now
be incorporated in NIST-7. Aside from these improvements, the key benefit of
this project was the demonstration of agreement between these independent
standards. It would have been difficult for NIST to justify construction of a
second standard for this purpose. (R. Drullinger)
- Frequency Synthesizer for Laser-Cooled Atomic Clocks. F. Walls,
with guest researchers A.S. Gupta and D. Popovic, has recently
developed an improved microwave frequency synthesizer with a performance
sufficient to support laser-cooled atomic clocks being developed as new primary
frequency standards and for advanced space applications. This new synthesizer
makes use of simple and rugged digital technology, some key components of which
are already space qualified. The phase stability, temperature coefficient, and
frequency agility should be more than adequate for every standard now under
active development and might well serve generations of standards beyond these.
It should also find application in standards for phase-noise and
amplitude-noise measurements.
|
The key design advances have been the use of digital technology and the removal
of narrow-band filters, which typically produce temperature-stability and
phase-stability problems. Measurements between a pair of these synthesizers are
shown in Fig. 6. For each of the pair of synthesizers, these data indicate
a fractional-frequency stability of better than
7 × 10-16 at 10 s, averaging down to
1 × 10-18 at 1 day. The measured temperature
coefficient is 0.12 ps/K. This synthesizer is small, and the circuitry is
easier and less costly to assemble since there are many fewer critical
adjustments involved. (F. Walls) |
|
Figure 6. Measured fractional-frequency stability between two of
the new microwave frequency synthesizers as a function of averaging time
τ
. |
- Improved Time-Scale Reference for Division Programs. T. Parker,
F. Walls, and J. Levine have collaborated to improve both the
performance of the NIST time scale and distribution of time-scale signals to
key programs within the Division. Drift and noise in signal distribution to
clock research laboratories and time-transfer stations have been reduced
through installation of improved distribution amplifiers and much
higher-quality coaxial cables. This has been needed particularly for studying
and evaluating the new generation of laser-cooled frequency standards and for
improving the reference used at satellite time comparison stations where the
NIST time scale is compared with those of other world standards laboratories.
The temperature coefficient of delay of the new cable is dramatically better
than conventional cable, resulting in a more stable transmission delay despite
the fact that the cable runs through areas that experience rather large
temperature excursions.
The real-time output of the time scale is now generated by steering the output
of a special synthesizer (driven by a good clock in the time scale) to the
ensemble average of the clocks in the time scale. Of course, steering
corrections to international UTC are also interjected into this system. The RMS
error in this generation is now below 10 ps, an improvement of about a
factor of 30. Finally, all five of the NIST masers are now contributing to the
time scale. This has improved the time-scale stability to σ
y(τ) · 3 × 10-16
at τ = 5 days.
(T. Parker)
- GPS Carrier-Phase Time Transfer. In collaborative work between
J. Levine of the Division and K. Larson of the University of
Colorado, GPS signals were used to achieve a time transfer resolution between
Washington and Boulder of 100 ps for an averaging time of 1 day. The
traditional approach to high-accuracy GPS time transfer involves two observers
making observations of the same code-based timing signal from a satellite that
can be viewed simultaneously by both. This approach is limited in resolution to
2 ns to 3 ns when averaging for 1 day.
The method employed in these experiments uses the phase of the GPS microwave
carrier (rather than the code) for the common-view time transfer. This process
involves identifying the same cycle of the carrier (or cycles of the carrier
separated by a constant number of cycles). This is a substantial problem
because the frequency is high, each satellite is in common-view for only a
short period, and there is no reference available to help identify a particular
cycle. The process employed is patterned after that used by geodesists, wherein
the observations from a large number of other sites are compared and adjusted
to obtain consistency and arrive at the appropriate cycle identification. They
were able to achieve this for periods lasting many weeks.
This work is critical to the comparison of the frequency accuracy of new
generations of laser-cooled, atomic frequency standards that are now being
developed at laboratories around the world. The method should provide for an
order-of-magnitude improvement in the precision of frequency comparison. This
first experiment already allows frequency comparisons at a level of
1 × 10-15 over about 1 day. (J. Levine)
- GPS Common-View Timing Receiver with Multiple Channels.
J. Levine, V. Zhang, and A. Gifford have developed a Common-View
Timing Receiver based on a commercially available, general-purpose,
multi-channel GPS receiver. The system functionality is similar to that of
earlier receivers developed at NIST except that the current receivers are much
simpler and less expensive and track up to eight satellites simultaneously. One
objective of this development was the replacement of older, NIST-developed
receivers that have been used for many years for international time
coordination. Many parts for these older receivers are no longer available, so
maintenance is becoming progressively more difficult. Aside from Boulder, the
new receivers are now located at the US Naval Observatory and the BIPM. Data
from these receivers are automatically transmitted to NIST once each day.
The receivers are also being used in other applications. For example, data from
one receiver located at NIST are downloaded every morning to a web site
providing the means for achieving NIST traceability using GPS signals. These
receivers are also being installed at a number of DOD sites where they will be
used to achieve high-level synchronization for DOD programs. (J. Levine)
- Multipath Effects in Time Transfer. Improved primary frequency
standards and time scales have put increasing demands on the performance of
international time comparisons using GPS common-view and two-way time transfer.
In an effort to respond to this demand, F. Ascarrunz and T. Parker
have been studying methods for improving the performance of Division
time-transfer systems. Of particular significance is their development of an
understanding of the impact of multipath signals on time transfer with
pseudo-random-phase codes used with both two-way satellite time transfer and
with common-view, GPS time transfer. Their analysis shows that multipath
effects exacerbate the difference between the observed phase delay and the
group delay. This understanding should allow improvement of two-way time
transfer through more appropriate choice of the code chip rate. It focuses more
attention on minimizing multipath effects in common-view time transfer, where
it is not feasible to make chip-rate changes.
They have also developed a calibration system for two-way time transfer,
providing a means for evaluating (and thus controlling) the delays through the
entire satellite ground station. Improved cables have cut the delay variations
in the calibration system from 200 ps to 50 ps. (T. Parker)
- Frequency Traceability to NIST Using GPS. M. Lombardi and
J. Levine have developed an on-line database of comparisons between the
NIST time scale and the Global Positioning System (GPS) satellite signals.
Calibration laboratories using GPS signals as a frequency reference can access
the database to complete their chain of traceability to NIST. The database is
automatically updated each morning and past data are archived. The archive
allows users to retrieve past data and retroactively confirm the traceability
of their measurements. This service was developed in response to requests from
calibration and standards laboratories and from GPS receiver manufacturers who
develop products for the time and frequency marketplace. (M. Lombardi).
- Year 2000 Time/Date Service. The Time and Frequency Division has
established a time server to assist users in testing the performance of
time-setting software after the year 2000 (Y2K). The transmitted time of day is
correct and is directly traceable to the NIST time scale, but the date portion
of the message is exactly 2 years in the future. Access to the Y2K service is
by telephone or through the Internet. All of the common digital time formats
are supported.
The service, developed by J. Levine was inaugurated at the end of October
1998, and will stay in operation through the end of 1999. To facilitate using
the Internet-based test system, NIST has also modified its client software to
allow users to select either the normal servers or this special test system.
This modified software is available on the Internet. (J. Levine)
- Completion of the Second Phase of Upgrade of WWVB. Staff members
at the NIST radio-station site north of Fort Collins, Colorado, working with
staff members of the Time and Frequency Group, have completed the second phase
of upgrade of WWVB. This phase of upgrade of the 60 kHz broadcast service
involved development of a second, identical, transmitting system including a
reconditioned low-frequency antenna and associated helix house, new
transmission line, and high-power transmitter. The two transmitting systems,
operated together in phase, will produce 50 kW of radiated power. This
should be compared with the initial starting point of 10 kW and the most
recent output of 23 kW achieved following completion of the first phase of
the upgrade. In emergencies, one transmitting system can radiate the full
50 kW of power. However, this mode of operation reduces tube life
significantly. Modifications of the WWVB building, now in progress, will allow
the installation of a third transmitter to back up the present transmitters. At
50 kW of radiated power, the broadcasts will more completely cover the
continental United States as shown by the modeled, field-intensity contour in
Fig. 7. This should allow for commercial development of a broader range of
simple clocks and frequency standards based on these broadcast signals.
(W. Hanson).
|
|
Figure 7. Electric-field-intensity contours (100 µV/m)
projected for operation of WWVB at 50 kW radiated power during night-time
hours. The nulls in the pattern are caused by interference of the ground wave
and the sky (reflected).
|
Figure 8. Photograph of the new diode-laser source for
wavelength-scanned interferometry.
Pages designed and maintained by the Office of ECSED.