W. R. C. Rowley

Frequency measurements on optically narrowed Rb-stabilised laser diodes at 780 nm and 795 nm

Laser diodes, optically narrowed using the technique of resonant optical feedback, have been frequency stabilised to hyperfine transitions of the two Rb D lines at 780 nm and 795 nm. The best frequency stability of the beat between two similar lasers was 1.5 kHz (4 parts in 1012 of the optical frequency) observed for an averaging time of 10 s. A frequency reproducibility of 44 kHz (one standard deviation) was observed on strong isolated hyperfine components, and possible causes of frequency shift were investigated. Values for the Rb hyperfine intervals were obtained, leading to an improved determination of the excited state hyperfine constants of 85Rb and 87Rb, and the isotope shift. The absolute frequencies of the hyperfine transitions of the two D lines were determined interferometrically by comparison with an 127I2-stabilised He-Ne laser at 633 nm. Measurements were made on component c′ at 795 nm and the d/f level crossing at 780 nm. The frequencies were found to be 377106271.6 MHz and 384227981.9 MHz respectively under the chosen conditions, with an uncertainty of ±0.4 MHz, limited by knowledge of the reference frequency. These results represent the most accurate and complete characterisation to date of laser diodes stabilised to Doppler-free Rb spectra.

Absolute frequency measurement of a 1.5-µm acetylene standard by use of a combined frequency chain and femtosecond comb

We have developed and characterized a pair of Doppler-free acetylene-stabilized diode laser frequency standards as optical communications references. The Allan deviation sigma/f of an individual system reaches a minimum of 4 x 10(-14) at a sampling time of 5000 s, and the long-term lock-point repeatability is found to be 0.4 kHz (one standard uncertainty), corresponding to a fractional uncertainty of 2 x 10(-12). Using a combination of a frequency chain and a self-referenced femtosecond comb, we have measured the frequency of line P(16) of the v1 + v3 overtone band of 13C2H2 to be 194,369,569,385.9 (3.0) kHz. The uncertainty of this number is limited solely by the reproducibility of the standards.

High-accuracy length metrology using multiple-stage swept-frequency interferometry with laser diodes

The optical length of a 1 m Fabry-Perot etalon has been determined by swept-frequency interferometry using laser diodes. The method involves progressively building up the measurement accuracy using frequency sweeps over increasing ranges, from 150 MHz (one optical fringe) to 19 GHz and 7 THz. The 7 THz sweep is referenced to the splitting of the rubidium D lines at 780 nm and 795 nm. The result from the 7 THz sweep is sufficiently accurate to use the known frequency of either end point of the scan to determine the length to a few parts in , without the need for any further measurement. The scope for further development of this technique to a range of interferometric systems is discussed.

Observation of the 5s 2 S 1/2 −4d 2 D 5/2 transition in a single laser-cooledtrapped Sr + ion by using an all-solid-state system oflasers

The first application, to our knowledge, of an all-solid-state system of lasers to the study of a single cooled trapped Sr(+) ion is described. Quantum jumps have been observed by driving the 674-nm 5s(2)S(1/2)-4d(2)D(5/2) transition, and preliminary observations of the line shape are reported. An upper limit for the temperature of a single ion, derived from the 674-nm linewidth, was 200 mK. If non-Doppler sources of broadening such as unresolved Zeeman structure dominate, then the temperature limit would be even lower.

Precision Measurement of the 2

The 4f 13 6s 2 2 F 7/2 –4f 14 5d 2 D 5/2 3•43 μm infrared transition in 172 Yb + has been driven for the first time and measured to be 87 360 087(4) MHz. The frequency was determined by probing a cloud of 172 Yb + ions held in a radiofrequency trap in the presence of helium buffer gas. The infrared radiation was generated by difference frequency mixing in LiNbO 3 . The frequency measurement is part of a programme to locate the 4f 14 6s 2 S 1/2 –4f 13 6s 2 2 F 7/2 467 nm ultra-narrow transition in laser-cooled Yb + .

Frequency-stabilised diode lasers in the visible region using Doppler-free iodine spectra

Abstract Hyperfine components of the 7-4, R(39) transition of the B-X system of 127 I 2 at 637 nm have been used to frequency-stabilise a pair of extended-cavity diode lasers. The variation of component frequency with laser operating parameters has been investigated. A frequency stability slope (square root of Allan variance) of 9.3 × 10 −11 τ −1 2 and a long term reproducibility of 28 kHz (5.9 × 10 −11 , 1 σ ) have been measured by analysing the beat-note between these lasers locked to components a 4 and a 13 . In addition, the absolute frequency of component a 4 has been measured to an uncertainty of 0.10 MHz (corresponding to 2.2 × 10 −10 , 1 σ ) by interferometric comparison with an I 2 -stabilised HeNe laser. The frequency separations of all hyperfine components with respect to a 4 are reported.

Clearly resolved secular sidebands on the /sup 2/S/sub 1/2/-/sup 2/D/sub 5/2/ 674-nm clock transition in a single trapped Sr/sup +/ ion

A prestabilized 674-nm diode laser has been narrowed by locking it to a high-finesse ultra-low-expansion cavity and kilohertz error signals have been observed. Lamb-Dicke confinement of single laser cooled Sr/sup +/ ions in a miniature rf trap has been demonstrated and micromotion reduction achieved. RF trap sidebands have been observed on the Sr/sup + 2/S/sub 1/2/-/sup 2/D/sub 5/2/ 674 mm clock transition. The transition frequency has been measured to be (444779043.98/spl plusmn/0.12) MHz (2/spl sigma/).

A 633 nm iodine-stabilized diode-laser frequency standard

An extended-cavity diode laser has been stabilized to hyperfine components of the 6-3, P(33) transition of the B-X system of 127I2 at 633 nm, and its performance evaluated by comparison with an iodine-stabilized He-Ne laser. The Allan standard deviation follows a slope of 3.8 × 10-11 τ-1/2, reaching a minimum of 1.7 × 10-12 at 500 s, and the reproducibility of the diode laser locked to component b21 has been determined to be 7 kHz (standard uncertainty) corresponding to a fractional uncertainty of 1.5 × 10-11. In addition, the influence of modulation depth, iodine pressure and axial power density on laser frequency has been investigated. Complete hyperfine interval sets and absolute frequency determinations of component b21 are presented for modulation depths of 2 MHz and 6 MHz.

Sub-kHz “clock” transition linewidths in a cold trapped 88Sr+ ion in low magnetic fields using 1092-nm polarisation switching

Abstract A small RF (Paul) trap has been developed, which allows Lamb–Dicke confinement of a single strontium ion. This paper describes the trap and the diagnostics used to verify Lamb–Dicke confinement. In particular, it is necessary to operate in a near zero magnetic field, so that the Zeeman splitting is minimised. In such low magnetic fields, the ion fluorescence at the 2 S 1/2 – 2 P 1/2 cooling wavelength at 422 nm normally falls to zero. This arises from optical pumping into some of the Zeeman sub-levels of the 2 D 3/2 state, which occurs because the laser which drives the 1092 nm transition ( 2 P 1/2 – 2 D 3/2 ) has a fixed polarisation state. However, if the polarisation state of this laser is rapidly switched, the fluorescence is recovered. A scan over the strontium ion 2 S 1/2 – 2 D 5/2 quadrupole transition shows only the carrier and sidebands corresponding to the radial and axial secular frequencies. Short range high resolution scans over a part of the carrier reveal Zeeman structure with ≤1 kHz-linewidth features.

Development and absolute frequency measurement of a pair of 778 nm two-photon rubidium standards

A pair of diode laser frequency standards stabilized to 778 nm two-photon rubidium transitions have been developed and characterized. Their sensitivity to vapour pressure, modulation depth and optical power have been investigated, along with their relative stability and day-to-day reproducibility. The Allan deviation of the standards locked to the 5 2S1/2(Fg = 3)–5 2D5/2(Fe = 5) transition of 85Rb is 9.3 × 10−13τ−1/2 for 1 s < τ < 100 s, reaching a floor of 1 × 10−13 at 100 s. The long-term repeatability of an individual system locked to this line is 1.2 kHz, corresponding to 3.1 parts in 1012. The system-to-system reproducibility is limited by cell impurities to 4.9 kHz and 4.5 kHz, respectively, for this line and the 5 2S1/2(Fg = 2)–5 2D5/2(Fe = 4) transition in 87Rb. The mean observed frequencies for these transitions, averaged for the two systems and corrected for light shift and second-order relativistic Doppler effect, are 385 285 142 375.1 (4.9) kHz and 385 284 566 374.2 (4.5) kHz, respectively.