Prospects for Ultra-stable Timekeeping with Sealed Vacuum Operation in Multi-pole Linear Ion Trap Standards

A recent long-term comparison between the compensated multi-pole Linear Ion Trap Standard (LITS) and the laser-cooled primary standards via GPS carrier-phase time transfer showed a deviation of less than 2.7×10/day. A subsequent evaluation of potential drift contributors in the LITS showed that the leading candidates are fluctuations in background gases and the neon buffer gas. The current vacuum system employs a “flow-through” turbomolecular pump and a diaphragm fore-pump. Here, we consider the viability of a “sealed” vacuum system pumped by a non-evaporable getter for long-term ultra-stable clock operation. Initial tests suggest that both further stability improvement and longer mean-time-betweenmaintenance can be achieved using this approach.* INTRODUCTION A compensated multi-pole Linear Ion Trap Standard (LITS) developed at the Jet Propulsion Laboratory (JPL) has recently demonstrated extremely good long-term stability over a 9-month period [1]. The short-term stability of this clock has been measured at 5×10/τ and all contributions to systematic stability have been measured to be less than 5×10. An upper limit on fractional frequency drift <2.7×10/day was measured in comparison to the laser-cooled primary standards using GPS carrierphase time transfer. This clock is designed for continuous operation and is able to run uninterrupted for up to 2 years, limited primarily by the need to perform maintenance on the mechanical fore-pump used in conjunction with a turbo pump. This “flow-through” vacuum system is used because of the need to pump unwanted background gases in the presence of a relatively large buffer gas pressure. Whereas an ion pump would quickly become saturated with the noble buffer gas, the turbo pump provides a much longer operating life and provides a stable pumping speed. Sealed pump buffer-gas-cooled ion trap clocks using a getter/ion pump combination have operated for up to 5 years in prior timekeeping applications, but exhibited significant drift rates [2]. For spacecraft applications, a vacuum system loaded with a getterable buffer 309 * The research described in this paper was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. 39 Annual Precise Time and Time Interval (PTTI) Meeting gas and pumped by only a small ion pump was operated for several years. The measured buffer gas pressure variation in an open loop system corresponds to a long-term stability limit in a mercury ion clock due to the pressure shift of 6×10/day [3]. An alternative vacuum pumping strategy uses only a getter pump, which does not pump inert gases such as the current neon buffer gas. With a fixed quantity of buffer gas, a getter pump could provide decades of operation with no interruption provided that the outgas rates of both getterable and non-getterable gases are sufficiently low. Operation of LITS technology in a small sealed vacuum system baked out to very high temperatures and pumped using only a getter pump has shown encouraging results towards achieving long operational life [4]. In addition to greatly extending the time between maintenance interruptions, this approach simplifies needed electronics and may possibly lead to an even more stable vacuum environment. However, at present, details of the long-term evolution of gases not pumped by the getter are poorly understood. In this paper, we present initial tests on the long-term behavior of such gases in the presence of a getter pump and discuss the viability of this approach for ultra-stable timekeeping applications. THE COMPENSATED MULTI-POLE LITS The multi-pole Linear Ion Trap Standard (multi-pole LITS) consists of a conventional linear quadrupole trap for ion loading and state preparation/readout and a second 12-pole trap to hold ions during the interrogation of the microwave clock transition [5]. The multi-pole trap significantly reduces the ionnumber-dependent second-order Doppler shift. The compensated multi-pole LITS further reduces this sensitivity by using a small stable inhomogeneous magnetic field to compensate any remaining secondorder Doppler shift [6]. This frequency standard also uses a neon buffer gas to minimize the buffer gas pressure shift and has achieved a short-term stability of 5×10/τ. LONG-TERM PERFORMANCE To demonstrate the new level of performance achieved by the compensated multi-pole LITS, this standard (known as LITS9) was compared to several UTC (k), UTC, and to the laser-cooled primary standards using GPS carrier-phase time transfer [1]. Fig. 1 shows the Allan deviation of fractional frequency offsets between LITS9 and UTC (NIST). The measurement is limited at the 10 level by GPS noise, but shows no significant drift over the 9-month run. Also shown in the graph is the stability of an uncompensated multi-pole LITS compared to the USNO Master Clock [7] (the short-term stability is masked by measurement noise). The improved long-term stability achieved by applying the magnetic compensation is readily apparent. By double-differencing these data with UTC (NIST) – UTC data, we are able to obtain LITS9 – UTC. Frequency offsets for LITS9 – UTC obtained in this way are shown in Fig. 2. A straight-line fit to these data gives a relative drift rate of 3.3×10/day. This can be viewed as an upper bound on LITS9 performance, as the comparison also includes GPS noise and steers to UTC. A further double-difference with all laser-cooled primary frequency standards reporting accuracy evaluations (frequency offsets from TAI) over the same time period gives a relative deviation between LITS9 and those standards of 2.7×10/day.