A hybrid ion-atom trap with integrated high resolution mass spectrometer

TL;DR

This paper presents a novel hybrid ion-atom trapping system integrated with a high resolution TOF mass spectrometer, enabling detailed study of ion-atom interactions and potential molecular ion cooling.

Contribution

The work introduces a new hybrid trapping setup combining MOT, linear Paul trap, and TOFMS, allowing precise detection of reaction products in ion-atom experiments.

Findings

Observation of interactions between potassium atoms and calcium ions.

Detection of reaction products involving different ion-atom pairs.

Discussion on cooling prospects for molecular ions in the system.

Abstract

In this article we describe the design, construction and implementation of our ion-atom hybrid system incorporating a high resolution time of flight mass spectrometer (TOFMS). Potassium atoms ($^{39}$K) in a Magneto Optical Trap (MOT) and laser cooled calcium ions ($^{40}$Ca$^+$) in a linear Paul trap are spatially overlapped and the combined trap is integrated with a TOFMS for radial extraction and detection of reaction products. We also present some experimental results showing interactions between $^{39}$K$^+$ and $^{39}$K, $^{40}$Ca$^+$ and $^{39}$K$^+$ as well as $^{40}$Ca$^+$ and $^{39}$K pairs. Finally, we discuss prospects for cooling CaH$^+$ molecular ions in the hybrid ion-atom system.

Figures

Full text

††preprint: AIP/123-QED

I Introduction

In the past decade, hybrid traps were employed to study elastic and inelastic collisions as well as rich chemical interactions between cold atoms and ions Grier et al. (2009); Zipkes et al. (2010); Schmid, Härter, and Denschlag (2010); Hall et al. (2011); Rellergert et al. (2011); Härter et al. (2012); Ravi et al. (2012a); Sivarajah et al. (2012); Ray et al. (2014); Haze et al. (2015); Meir et al. (2016); Hall and Willitsch (2012); Deiglmayr et al. (2012); Rellergert et al. (2013); Kwolek et al. (2019). The very recent research in this field focuses on controlling the chemical reactions in cold ion-atom collisions. These include controlling the electronic state Joger et al. (2017); Saito et al. (2017), the spin state Sikorsky et al. (2018) or the collision energy Saito et al. (2017); Puri et al. (2018). The long interrogation time and precise internal state control make these systems an ideal platform to explore quantum state dependent interactions.

Ion-atom hybrid traps are also proposed to attain the rovibrational ground state of molecular ions Hudson (2016). The rich internal structure of molecules makes them well suited for precision experiments such as time variation of fundamental constants Schiller and Korobov (2005), precise measurement of the electron’s electric dipole moment Cairncross et al. (2017) and study of parity violation Kozlov and Labzowsky (1995). Laser cooled atomic ions have proved to efficiently cool the translational degrees of freedom of co-trapped molecular ions Kimura et al. (2011); Rugango et al. (2015). The idea is to cool the internal states of the molecules sympathetically using ultra cold atoms. Recent experiments to achieve this goal gave promising results. Vibrational cooling of BaCl+ ions Rellergert et al. (2013) was demonstrated using interaction with neutral Ca atoms in a MOT. Rotational cooling of MgH+ down to 7.5 K was achieved with helium buffer gas cooling Hansen et al. (2014). One can expect more efficient rotational cooling using interactions with laser cooled atoms. This technique is independent of the molecular ion species as it does not involve any resonance phenomena Hudson (2016). One of the goals of our ion-atom hybrid trap presented in this manuscript involves realizing CaH+ molecular ions in their internal and external ground state.

In collision experiments involving different species, a reliable detection scheme to identify the reaction products is essential. In the past, indirect detection either by secular excitation of the trapped ions Baba and Waki (2002) or structural changes in the ion coulomb crystal, in conjunction with molecular dynamics simulations Roth et al. (2006), were used to identify the reaction products. Even though these indirect methods simplify the equipmental design, they lack high resolution and require more sophisticated data analysis. A time of flight mass spectrometer (TOFMS) can be coupled to an ion trap for axial Ravi et al. (2012b); Seck et al. (2014) or radial Schowalter et al. (2012); Meyer et al. (2015); Jyothi, Ray, and Rangwala (2015); Deb et al. (2015); Rösch et al. (2016); Schmid et al. (2017) extraction and detection of the trapped ions. The radial extraction has proven to be advantageous in improving the mass resolution because the trapped ion spatial extent is smaller along the radial direction. In our ion-atom hybrid trap, we have incorporated a TOFMS to radially extract the trapped ions with high mass resolution.

In this article we describe the design and implementation of our ion-atom hybrid trap which incorporates a MOT for potassium atoms (39K) and a linear Paul trap for ions. In section II, the simultaneous cooling and trapping of 39K and 40Ca+ as well as the details of TOFMS are discussed. The details of experimental control is also described in this section. In section III, we present measurements to demonstrate the spatial overlap of ion-atom trap centers and a few experimental results to demonstrate the potential of the experimental system to perform cold interactions studies. These include elastic as well as resonant charge exchange collisions between 39K+ and 39K, long range Coulomb interaction between 40Ca+ and 39K+, and charge transfer collisions between 40Ca+ and 39K. We conclude by discussing future goals for the hybrid system.

II Experimental Design

The experimental setup consists of an 8*′′* octagon vacuum chamber (Kimball Physics MCF800-SphOct-G2C8) maintained at a pressure of 1$\times$10*-9* Torr. A schematic of the experimental system is illustrated in Fig. 1 (a). A segmented electrode linear quadrupole trap (LQT) is mounted at the center of the chamber. The ion trap consists of four rods in a quadrupole configuration. Each rod is broken into three electrodes. A radio frequency (rf) electric field is applied on the two central diagonal electrodes to confine the ions in the radial directions (YZ). The other two central electrodes function as the rf ground. The eight end-cap electrodes are used to apply direct current (dc) voltages to confine the ions in the trap axial direction (X). The rods are assembled such that the neighboring rod separation is 10 mm in order to accommodate the 10 mm MOT beams. The experimental chamber is integrated with a high-resolution TOFMS for the identification of the ionic species. The TOFMS is mounted to one of the 2.75*′′* ports perpendicular to the ion trap axis for radial ejection of the ions. The ion trap voltages and fast rf switching for the TOFMS are performed using the electronics developed by Hudson Lab at UCLA Schneider et al. (2014). The chamber is also equipped with a leak valve to input gases such as H2 for creating molecular ions for future experiments.

The 39K MOT is formed using three mutually orthogonal retro-reflected laser beams. The ground state hyperfine splitting of 42S1/2 of 39K isotope is 461.7 MHz and the excited state manifold is just 33.5 MHz. This hyperfine structure makes it possible to derive both the cooling and the repumper laser beams using a single laser at 766.7 nm (external cavity diode laser and tapered amplifier, MOGLabs). The laser is locked to the crossover peak of the saturation absorption spectrum. The laser output is split into two beams and are sent through acousto-optic modulators (AOM) to get the desired detunings of 20 MHz from the F=2 (cooling) and F=1 (repumper) to the excited state manifold. Potassium vapor is produced by heating a potassium ampule (Sigma Aldrich 244856) placed in a stainless steel bellow. The glass ampule is connected to the chamber via an all metal right angle valve that can be used to isolate the ampule if required, and is broken during the initial pump down of the chamber. The magnetic field gradient required for the MOT is produced using a pair of coils in anti-Helmholtz configuration, with a field gradient of about 10 G$/$cm in the axial direction. The MOT anti-Helmholtz magnetic coils are symmetrically mounted outside the chamber on a two dimentional translation stage to precisely adjust the center of the magnetic field. An additional pair of coils in Helmholtz configuration is mounted concentric with the MOT coils. These coils are used to shift the magnetic field center in the vertical direction to shift the MOT atoms from the ion trap center when required.

In our hybrid trap experiments we use laser cooled 40Ca+ ions. Neutral calcium atoms are produced by evaporating a piece of calcium inside a stainless steel tube by passing a current through a steel wire inserted into the stainless steel tube. The calcium ions are produced by resonant two photon ionization using 423 nm and 379 nm laser beams Lucas et al. (2004). The ions are trapped with an rf field of 400 V at 1.7 MHz and a dc voltage of 7 V and are cooled by 397 nm laser with 866 nm laser as the repumper. All wavelengths are derived from a multi diode laser module (Toptica Photonics).

For the simultaneous cooling and trapping of 39K and 40Ca+, the 767 nm laser beams for 39K and the 397 nm cooling beam for 40Ca+ need to be alternately switched at high frequency. This is because the ionization energy of the potassium atoms from the 4P3/2 excited state is only 2.7 eV and any light with wavelength lower than 455 nm can ionize the excited state 39K atoms (the relevant energy levels for 39K and 40Ca+ are shown in Fig. 1 (b)). Hence, while the MOT is operational, the 767 nm and 397 nm lasers are alternately switched at 2 kHz and the 423 nm and 379 nm lasers are blocked using a mechanical shutter to avoid ionization of excited state 39K atoms. Hence in all experiments, we first load the 40Ca+ ions, block the 423 nm and 379 nm laser beams, start alternate switching of the 767 nm and 397 nm laser beams using AOMs and start loading the 39K MOT.

The fluorescence from the 39K atoms and 40Ca+ ions are collected using two calibrated photo multiplier tubes (PMT) and an EMCCD camera. A schematic of the fluorescence collection is shown in Fig. 2. The camera images are used to confirm the ion-atom spatial overlap in one plane (XY) along the ion trap axis. The spatial overlap in the third direction (Z) cannot be imaged due to design constrains and is ensured by another method described in the result section. The overlap between the atoms and the ions can be optimized by tuning the trapped ion position using the dc voltages on the ion trap electrodes or by tuning the MOT position by shifting the magnetic field center in the horizontal plane using a 2D translation stage and in the vertical axis by adding a constant magnetic field using a Helmholtz coil pair.

The TOFMS is an essential part of the experimental set up to identify the reaction products of the cold collision experiments, especially the nonfluorescent ions. A schematic of the TOFMS is shown in the Fig. 3. It consists of a grounded skimmer which determines the entrance aperture of the TOFMS (5.6 mm in diameter), two Einzel lenses for ion collimation, a micro channel plate (MCP) for ion detection and a grounded wire mesh in front of the MCP detector to shield the MCP from the drift tube. The entire TOFMS parts are mounted on a 267 mm long drift tube and is attached to the 2.75*′′* ports perpendicular to the ion trap axis for the radial ejection of the ions. The MCP (Photonis) consists of a pair of impedance matched microchannel plates with an active area of 18 mm. The two plates are assembled in a chevron configuration allowing maximum signal multiplication. A metal anode at the back of the MCP detector collects the electrons resulting from charged particle detection. The timing signal is extracted from the back plate of the MCP pair through an RC decoupling circuit and is collected on a digital storage oscilloscope.

The MCP is maintained at a high voltage of -2.1 kV. To extract the ions onto the TOFMS the trapping rf voltages are turned off and appropriate dc high voltages (approximately 1000 V and 700 V) are applied to the 4 central rods to create a potential gradient towards the MCP. To improve the resolution of the time of flight spectra of the ions, the residual oscillation of the rf field after it has been turned off is damped. To facilitate this, the electronics system is equipped with damping circuitry to remove residual rf oscillations. After the ions are ejected, the ions travel through a field free region. In order to collimate the ions in the two planes perpendicular to the spectrometer axis, high voltages are applied to the two Einzel lenses. The construction of the Einzel lens is similar to that used in Schneider et al. (2014). The lens voltages are adjusted to optimize the TOFMS spectrum. In the present experiment a resolution, m/$\Delta$m of about 208 was achieved. This resolution is sufficient to resolve the 39K+ and 40Ca+ ions in our experiments, whose masses differ by only one atomic unit.

The real-time control of the experiment was implemented using hardware and software solutions provided by the Advanced Real-Time Infrastructure for Quantum physics (ARTIQ) framework Bourdeauducq et al. (2018) from m-labs. Using ARTIQ, experiment code is written in Python, then compiled and executed on dedicated Field Programmable Gate Array (FPGA) hardware with nanosecond timing resolution and sub-microsecond latency. The central controller used in the experiment was the Kasli FPGA carrier, which was accompanied by several Eurocard Extension Modules (EEMs) to complete the setup. Two digital input/output EEMs, each with four channels, were used to count PMT inputs as well as to send trigger signals to other electronics such as the mechanical shutter and TOFMS trigger. A third EEM contained four DDS output channels from an AD9910 DDS chip and was used to drive the AOMs. An example timing sequence for the measurement of sympathetic cooling of 39K+ ions by laser cooled 40Ca+ ions as discussed in section III (B) is shown in Fig. 4.

III Results

The main requirement to effectively use the potential of an ion-atom hybrid trap is the spatial overlap between the trap centers to initiate the interactions. Overlap can be visually observed using the CCD camera if the species are fluorescing. In our set up we have the camera set up along only one axis and hence we can see overlap only in one plane. We measure the overlap in the third direction by using the fact that the 39K atoms in excited state get ionized by the cooling laser for the 40Ca+ ions. We use this effect to ensure the ion-atom spatial overlap in the following way. The MOT image analysis shows that the MOT full width is less than approximately 3 mm. The 40Ca+ cooling laser beam radius at its focus is approximately 150 $\mu$m and the beam position can be scanned spatially across the MOT in the YZ plane using a translation stage. First, the 40Ca+ axial cooling beam is optimally aligned with respect to the ion trap center by observing the fluorescence of 40Ca+ ions (denoted as position Y=Z=0). The ions are removed and after loading the MOT to saturation, the 397 nm laser is turned on and the MOT decay rate is measured at Y=Z=0. The 397 nm laser beam is scanned across Y(at Z=0) and Z(at Y=0) directions and the MOT decay rate is measured. The rates as functions of Y and Z as given in Fig. 5 show that the center of the MOT gaussian profile matches with the ion trap center. The measurement reveals good overlap between the trap centers along the Z axis, which is not observable using the camera. These measurements are done with the ion trap voltages off. For all other experiments, the 767 nm and 397 nm lasers are switched alternatively to avoid the MOT atom ionization.

In the following three subsections we describe the experimental results of collision experiments performed in the hybrid trap.

III.1 Sympathetic cooling of 39K+ by 39K MOT

Like every alkali metal ion, 39K+ ions have closed electronic structure and cannot be laser cooled due to unavailability of suitable lasers. The established way of cooling such ions is via sympathetic cooling by laser cooled atoms or laser cooled ions. The elastic collisions are more probable in an ion-atom hybrid system, leading to transfer of energy from the hot ions to the cold atoms. However, if the mass of the atoms and ions are the same, resonant charge exchange interaction plays an important role in accelerating the cooling process Ravi et al. (2012a); Sivarajah et al. (2012); Ray et al. (2014). In a charge exchange collision between 39K and 39K+, the electron hopes from 39K atoms to 39K+ ions in a glancing collision, giving rise to cold 39K+ ions and hot 39K atoms. The hot 39K atoms are lost from the MOT and the MOT is immediately refilled by another 39K atom as the MOT is loaded continuously.

Fig. 6 demonstrates the cooling of 39K+ ions by 39K MOT atoms. In this experiment, after loading the MOT to saturation, 39K+ ions are produced by photoionizing 39K atoms in the MOT for 10 ms by the 397 nm laser. The 39K+ ions are held in the ion trap for fixed amount of hold times either in the absence or in the presence of 39K atoms in the MOT and then extracted onto the TOFMS. The life time of the 39K+ ions in the absence of cooling is only 2.2(4) s due to heating caused by trap imperfections, rf heating as well as background collisions. However when held with the 39K MOT, the lifetime of 39K+ ions increases to 25.3(4) s. The cooling caused by resonant charge exchange as well as elastic collisions increased the ion lifetime approximately by an order of magnitude.

III.2 Sympathetic cooling of 39K+ by 40Ca+

The long range Coulomb interaction between co-trapped ions leads to the sympathetic cooling of hot atomic or molecular ions by laser cooled ions Larson et al. (1986). Here we demonstrate the sympathetic cooling of 39K+ ions by laser cooled 40Ca+ ions.

The experimental sequence is shown in Fig. 4. First, the 40Ca+ ions are loaded to the ion trap. The MOT is then loaded a few millimeters above the ion trap center and are photoionized to create 39K+ ions. The trap centers are shifted to avoid any ion-atom charge exchange interaction between 39K atoms and 40Ca+ ions. 40Ca+ and 39K+ masses are very close and can be trapped with the same optimal trapping parameters. The ions are then extracted onto the TOFMS. Fig.7 shows an average of 10 TOFMS traces of 39K+ in the absence and presence of laser cooled 40Ca+ ions. The effect of the sympathetic cooling can be observed in the time of flight spectrum as an increase in height and a decrease in width of the 39K+ peak in the presence of laser cooled 40Ca+ ions Schneider et al. (2014). The mass resolution increases from 26 to 208 in the presence 40Ca+ ions. The 39K+ peak height also is increased by approximately an order of magnitude from 0.7 mV to 6.5 mV in the presence of the laser cooled 40Ca+ ions.

III.3 Charge exchange collision between 40Ca+ and 39K

The long range interaction between neutral atoms and ions Côté and Dalgarno (2000)scales as $1/r^{4}$. The details of the long range potential and hence the charge exchange probabilities varies with the combination of atoms and ions Joger et al. (2017); Sikorsky et al. (2018); Kwolek et al. (2019); Saito et al. (2017). We observe for the first time the charge exchange interaction between 39K atoms and 40Ca+ ions. The charge exchange rate can be measured by collecting the 40Ca+ fluorescence in the absence and presence of 39K atoms as a function of time and is shown in Fig.8. Alternately, we can observe the reaction as a decrease in the 40Ca+ ion count and a simultaneous increase in 39K+ ion count using the time of flight measurements. The lifetime of laser cooled 40Ca+ ions is a few minutes in the absence of 39K atoms, but when spatially overlapped with the 39K MOT, the 40Ca+ ion number decreases due to charge exchange interaction. The 40Ca+ ions decay in 8.7(5) s with a MOT of density of 6.7(6)$\times 10^{8}$ atoms/cm3. The details of this interaction will be presented in a subsequent publication Jyothi et al. .

IV Summary and Discussion

We have presented the design details of our ion-atom hybrid trap integrated with a high resolution mass spectrometer, the experimental control details and the detection schemes. We have also presented the ion-ion and ion-atom interactions measured in the hybrid trap. The next step ahead involves control of these interaction using state preparation techniques.

In addition, it would be possible to induce strong interaction between the first and the second electronic potentials of KCa+ using an additional high intensity laser. This interaction can lead to non-adiabatic transition between these potentials and even make it possible to observe a light-induced conical intersection in the charge exchange interaction Kotochigova .

Attaining the ultimate ground state of molecular ions using interactions with laser cooled atoms and ions can be attempted in ion-atom hybrid traps. This would be an alternative approach to coherent internal state preparation by projection into a pure rotational state Chou et al. (2017). We are planning to produce CaH+ molecular ions by reacting 40Ca+ ions in 4P1/2 state by H2 molecules leaked into the chamber using a leak valve. Translational cooling of CaH+ by laser cooled 40Ca+ Rugango et al. (2015) and resolved rovibrational spectroscopy of CaH+ Calvin et al. (2018); Calvin and Brown (2018) have already been established in our lab. Our goal is to get the rovibrational ground state of CaH+ using cold 39K atoms. The resolved rovibrational spectrum of CaH+ can be used for the detection of the internal cooling of CaH+. CaH+ prepared in the ground state is a candidate for measuring the time variation of the proton to electron mass ratio Schiller and Korobov (2005).

The K-Ca+ hybrid system described here can be used to create a wide range of cold molecular ions beyond CaH+. Molecular ions with mass-ratios within a factor of 5 of Ca+ can be efficiently sympathetically cooled translationally Wübbena et al. (2012). The internal motion of molecular ions can be generically cooled by the 39K MOT as long as the reactions between 39K and molecules are relatively slow. The hybrid system is ideal for exploring the physics and chemistry of inelastic collisions between 39K atoms and polyatomic molecular ions.

Acknowledgements.

The authors acknowledge Eric Hudson’s group at UCLA for the technical support on the ion trap electronics. This work was supported by the MURI Army Research Office Grant W911NF-14-1-0378-P00008 and ARO Grant W911NF-17-1-0071.