† Corresponding author. E-mail:
‡ Corresponding author. E-mail:
Project supported by the National Natural Science Foundation of China (Grant Nos. 11074248 and 11474320).
In the microwave 199Hg+ trapped-ion clock, the frequency instability degradation caused by the Dick effect is unavoidable because of the periodical interrogating field. In this paper, the general expression of the sensitivity function g(t) to the frequency fluctuation of the interrogating field with Nπ-pulse (N is odd) is derived. According to the measured phase noise of the 40.5-GHz microwave synthesizer, the Dick-effect limited Allan deviation of our 199Hg+ trapped-ion clock is worked out. The results indicate that the limited Allan deviations are about
The first frequency standards based on trapped ions capable of continuously operating were reported separately by the Laboratoire de l’Horloge Atomique (LHA) Laboratory in 1980 and Hewlett–Packard (HP) in 1981.[1] Since 1987, scientists at the Jet Propulsion Laboratory (JPL) have been studying the microwave 199Hg+ trapped-ion clock. They have developed the microwave 199Hg+ trapped-ion clock based on the single linear ion trap and two-segment extended linear ion trap successively.[2] JPL’s results indicate that the performance of the microwave 199Hg+ trapped-ion clock is very excellent in microwave frequency standard.[3–7] More noticeably, the demonstration of a mercury ion clock for space at JPL is realized with the Allan deviation about
During the dead time, the local oscillator’s frequency is not sensed by the atomic interrogation process, so local oscillator frequency fluctuations during those times are not appropriately corrected, which will cause degradation of frequency instability. This problem was first investigated and analyzed by Dick[11] and Dick et al.[12] at the JPL in microwave 199Hg+ trapped-ion frequency standard, which is now called the Dick effect. In the microwave 199Hg+ trapped-ion clock, the Dick effect causes the degradation of the frequency instability. Many other frequency standards, such as optical lattice clock,[13] cesium fountain,[14] 113Cd+ trapped-ion clock,[15] etc., also have this effect. In our group, we are developing a small, compact microwave 199Hg+ trapped-ion clock, in which the Dick effect is a significant limitation of the frequency instability.
The rest of this paper is structured as follows. In Section 2, the general formulation for the frequency instability degradation induced by the Dick effect in the microwave 199Hg+ trapped-ion clock is presented, and the general expression of the sensitivity function g(t) to the frequency fluctuations of the interrogating field with Nπ-pulse is derived. In Section 3, the phase noise of the frequency synthesizer for the microwave 199Hg+ trapped -ion clock is measured, and the experiment parameters of our experiments both in a linear ion trap and in a two-segment extended linear ion trap are described, then the Allan deviations limited by the Dick effect are calculated both in the Ramsey interrogation mode and in the Rabi interrogation mode due to the different types of traps. Brief conclusions are drawn from the present studies in Section 4.
In the 199Hg+ trapped-ion microwave clock, the frequency variation due to the phase fluctuations will cause the transition probability to change during the microwave interrogation time. The sensitivity function g(t) is used to describe the relation between the frequency variation and the variation of transition probability.[11,16]
When the Ramsey interrogation scheme is used, g(t) is given by[11,12,16]
If the π-pulse Rabi interrogation is used, the details of g(t) can be found in Ref. [12]. However, in the microwave 199Hg+ trapped-ion clock, in order to acquire a narrow linewidth, the 3π-pulse, 5π-pulse or other long time microwave pulse should be used for interrogation. In order to obtain the sensitivity function for this case, we can follow the rotation transformation method in Ref. [11]. If the time for interrogation is τN, and
Depending on the harmonic content of sensitivity function g(t), the noise of the oscillator at harmonics frequencies of the operation frequency 1/Tc will be down-converted to low frequency fluctuations. In a general way, the Allan variance limited by the Dick effect is given by[11,12,17]
In the 199Hg+ trapped-ion clock, the local oscillator is a 10-MHz ultra-stable oven controlled crystal oscillator (OCXO), which is used as a reference to generate the 40.507-GHz clock transition signal. As the time of one cycle operation Tc is several seconds, the power spectral density of frequency fluctuations from 1/Tc Hz to several hundred Hz accounts mainly for the Dick effect. However, the phase noise of the synthesizer from 1/Tc Hz to several hundred Hz is too small to measure directly by the commercial equipment. There is a method by which two identical microwave synthesizers are mixed as listed in Ref. [15].
In order to reduce the noise, the passive mixer is usually used. However, in the 199Hg+ trapped-ion clock, the power of the synthesizer is about −13 dBm, which is too small to drive a passive mixer. So in our experiment, the following setup of phase noise measurement is used as shown in the following Fig.
The 40.507-GHz microwave synthesizer is mixed with a 20.2-GHz phase locked dielectric resonator oscillator (PDRO) and the mixer is a sub-harmonic mixer. The power of the PDRO is about 13 dBm. At this power level, the conversion loss of the sub-harmonic mixer is about 13 dB. The output signal of the sub-harmonic mixer is about 107 MHz and it is amplified by an ultra-low noise amplifier, then the phase noise is measured by the E5505A. In the microwave synthesizer, there is a similar PDRO that is mixed with a direct digital synthesizer (DDS) to produce a 40.507-GHz signal, the phase noise of the single PDRO should be smaller than that of the microwave synthesizer, it will still introduce some noises to the 107 MHz. Furthermore, the sub-harmonic mixer and the ultra-low noise amplifier may also add some noises to the 107 MHz, all these additive noises are very small. So the phase noise of the 107 MHz is the upper limitation of noise of the microwave synthesizer. The result is shown in Fig.
In our group, the main experiments are carried out in the linear ion trap (shown in Fig.
In this type of trap, the ion-loading, the optical pumping, and the microwave interrogation are in the same region. The Ramsey interrogation scheme is used in our experiment, and the time sequence is shown in Fig.
Two microwave pulses (π/2-pulse), each with a duration τπ/2, interact with the trapped ions, and the two pulses are separated by a time of Ts, then the 202Hg discharged lamp is used for detection and optical pumping, and the time for detection, optical pumping and ion-loading is τd which is the dead time. So the total time of one cycle of operation is Tc = 2 * τπ/2 + Ts + τd. The realistic experimental parameters are τπ/2 = 0.3 s, Ts = 2 s, and τd = 3 s then the total time Tc = 5.6 s. and the Allan deviation is
In order to reduce the second order Doppler shift, our group is also studying the two-segment extended linear ion trap which is shown in Fig.
In this trap, the Rabi interrogation scheme is usually used in our experiment, and the time sequence is shown in Fig.
In the two-segment extended linear ion trap, the realistic experimental parameters are as follows: τπ = 0.6 s (N = 1), τm1 = τm2 = 0.5 s, and τo = 3 s, then Tc = 4.6 s and the Allan deviation is
When the dead time is the same, the Allan deviation for 3π-pulse interrogation is a little worse than that for π-pulse interrogation, however, the linewidth for 3π-pulse interrogation is one third of that for π-pulse interrogation, which is an advantage in improving the stability.
In this work, the sensitivity function for an Nπ-pulse (N is odd) is derived, and the Allan deviations limited by the Dick effect are calculated in the microwave 199Hg+ trapped-ion clock. In our experiment with the realistic experimental parameters, the Allan deviations are
1 | |
2 | |
3 | |
4 | |
5 | |
6 | |
7 | |
8 | |
9 | |
10 | |
11 | |
12 | |
13 | |
14 | |
15 | |
16 | |
17 | |
18 |