A linear cryogenic 16-pole wire ion trap has been developed and constructed for cryogenic ion spectroscopy at temperatures below 4 K. The trap is temperature-variable, can be operated with different buffer gases, and offers large optical access perpendicular to the ion beam direction. The housing geometry enables temperature measurement during radio frequency operation. The effective trapping potential of the wire-based radio frequency trap is described and compared to conventional multipole ion trap designs. Furthermore, time-of-flight mass spectra of multiple helium tagged protonated glycine ions that are extracted from the trap are presented, which prove very low ion temperatures and suitable conditions for sensitive spectroscopy.

Multipole radio frequency ion traps are important for various fields of research, in particular, for ion–molecule reaction1–3 and inelastic collision4,5 studies, laser-induced reactions,6–8 ion-tagging spectroscopy,9–13 laser-induced inhibition of cluster growth,7,14 terahertz-visible two-photon spectroscopy,15–17 and ion cluster research.18,19 Multipole ion traps are also used in studies of ion Coulomb crystals.20 

Multipole ion traps offer a large trapping volume that allows us to store a large number of ions with a range of different masses and with limited ion–ion interaction. Buffer gas cooling can be achieved down to a few Kelvin because radio frequency heating is suppressed by the multipole potential unlike the case for quadrupole traps.1,21 Several new designs and implementations of cryogenic multipole ion traps have been presented in recent years.22–28 

In this work, we present a cryogenic 16-pole radio frequency ion trap with thin, 0.1 mm diameter, wires as electrodes. The design of this trap is inspired by our recent room temperature octupole wire ion trap27,29 and by the cryogenic quadrupole ion trap that was designed by Jašik et al.26 The main goal was to design a trap that provides large optical access in the radial direction to allow very different types of radiation to be passed through the trapped ion cloud without being limited by electrodes, other obstacles, or long distances to viewports in the vacuum setup. This is particularly important for long-wavelength terahertz radiation but also helpful for ultraviolet radiation where contact with electrode surfaces may release unwanted electron pulses.

In order to enhance optical access to the trapped ions in the sub-4 Kelvin temperature regime, we have designed a radio frequency ion trap based on wire electrodes, which is capable of reaching these low temperatures. This trap also provides different pre-cooled buffer-gas inlets, allows for continuous temperature measurement, and is highly adaptable to the different experimental needs. The new instrument has been successfully employed in a vibrational predissociation spectroscopy experiment of the amino acid tryptophan, which was tagged with multiple hydrogen molecules in the cryogenic ion trap.30 

The components and the mechanical and thermal design of the trap are presented in Sec. II. A thorough description of the effective potential characteristics of the 16-pole wire trap is presented in Sec. III. In Sec. IV, the successful formation of protonated glycine ions tagged with multiple helium atoms is described, followed by conclusions.

The focus of this article is on the wire trap itself. The setup that includes this trap has been described by Spieler et al.30 A brief description of the setup is given in Subsection II A, followed by a description of the properties of the ion trap.

The cryogenic 16-pole wire trap is part of a setup that has been built in our laboratory for the purpose of preparing molecular ions at low temperature for spectroscopy and ion-neutral collision experiments. Ions are produced via electrospray ionization (ESI), allowing the gentle transfer of ions from the liquid to the gas phase. The ions are coupled into vacuum by a home-built interface well-suited for investigating non-volatile ions in the mass range from around ten to few hundred atomic mass units, such as small bare and doped water clusters, amino acids, or small peptides. Two differential pumping stages with a radio frequency octupole ion guide and a radio frequency quadrupole, which can be used as a mass filter but is presently operated in guiding mode, bring the ions to the cryogenic radio frequency ion trap that is at the heart of the experiment. Buffer gas cooling and action spectroscopy experiments are carried out inside the cryogenic trap by irradiating the ion cloud with different radiation sources. After interaction, the ions are extracted and passed into a perpendicular reflectron time-of-flight mass spectrometer. There, the signal is recorded on a microchannel plate detector with single ion counting ability and a mass resolution of up to about 400.

The linear cryogenic 16-pole wire trap has been developed to be able to operate at temperatures below 4 K while subsequently providing the ability to couple long-wavelength radiation to the trapped molecular ion and ionic complexes cloud. This low temperature is essential for forthcoming helium-tagging experiments. Consequently, the design process focused on the optimal thermal behavior in conjunction with as thin electrodes as possible.

Figure 1 shows an exploded view of the new trap. The main trap housing is built from oxygen-free, high-conductivity copper. Good cooling behavior has been achieved primarily by significantly reducing the number of different components and thereby minimizing thermal resistances between different parts of the trap31 and, second, by the choice of materials: oxygen-free copper is ensuring high thermal conductivity at cryogenic temperatures,32 and copper electrodes have been shown to enhance the buffer gas cooling efficiency.33 

FIG. 1.

Exploded view of the main parts of the 16-pole radio frequency wire trap. The trap’s copper housing consists of a main component that provides the base and four side walls, a copper cover, and a bottom part that covers the inlets and a labyrinth structure for the buffer gas. Temperature measuring diodes are mounted in the bottom of the main part and in the top cover. Additional parts are the 16 copper wires, the printed circuit boards, two electrode stacks, two windows, and the buffer gas inlets. The outer dimensions of the copper cube are 46.5 mm (height) by 49 mm (axial length) by 52 mm (transverse width). Reprinted with permission from Spieler et al., J. Phys. Chem. A 122, 8037 (2018) Copyright (2018) American Chemical Society.

FIG. 1.

Exploded view of the main parts of the 16-pole radio frequency wire trap. The trap’s copper housing consists of a main component that provides the base and four side walls, a copper cover, and a bottom part that covers the inlets and a labyrinth structure for the buffer gas. Temperature measuring diodes are mounted in the bottom of the main part and in the top cover. Additional parts are the 16 copper wires, the printed circuit boards, two electrode stacks, two windows, and the buffer gas inlets. The outer dimensions of the copper cube are 46.5 mm (height) by 49 mm (axial length) by 52 mm (transverse width). Reprinted with permission from Spieler et al., J. Phys. Chem. A 122, 8037 (2018) Copyright (2018) American Chemical Society.

Close modal

The main component of the trap provides the four side walls of the trap housing. The inside walls of the trap housing have a flat design, ensuring that no screws or mounting features may perturb the radio frequency field. 16 oxygen-free-high-conductivity copper wires of d = 100 μm in diameter serve as electrodes. Vented drill holes in the straight solid design ensure fast evacuation time, therefore avoiding virtual leaks. The mounting of the wires is realized by soldering the wires individually to two printed circuit boards. The first connection is made without applying force, and the second end is soldered under a tension force of about 1 N to the second circuit board. The circuit boards introduce a small amount of differential thermal expansion between the copper wires and the trap housing, but this was found to be small enough to keep the wires straight at room temperature and at cryogenic temperature. Outside of the circuit boards, three cylindrical “endcap” electrodes are mounted on either side of the trap, which serve as electrostatic lenses for guiding the ions and to close the trapping volume in the axial direction.

An optimal heat flow between the individual components of the trap is achieved via large, flat, and polished contact areas. Thread inserts of steel in the copper material allow for good mechanical connection via screws with up to M5 metric screw threads.31 With these large screws, a pressing force between the contact surfaces of several thousand Newtons can be achieved. Additionally, Apiezon N is applied to fill small surface roughnesses in order to further enhance the heat conductance between the individual parts.

The trap housing possesses two lateral openings through which optical access can be provided. Windows, mirrors, or closed lids may be mounted on these openings; at present, one laser window and one lid is installed. Behind the openings, the small diameter of the wires allows for a large open area that ensures good optical access even for long-wavelength radiation up to the terahertz (or sub-millimeter) regime (i.e., down to less than 1 THz). In case a laser beam with a diameter of 6 mm is passed to the trapped ions, only three of the wires are in front of the ion cloud resulting in 95% transmission.

The trap is mounted on a closed-cycle pulse tube cooler (Sumitomo SRP-082B2) with a nominal cooling power of 1.0 W at 4.2 K on the second stage and 40 W at 45 K on the first stage. Blackbody radiation reaching the trap is the main source of heat and subsequently to the buffer gas and the trapped ion cloud. Consequently, the copper housing of the trap is protected by an aluminum blackbody shield that is cooled by the first stage of the cryo cooler. It only leaves the ion entrance and exit to the trap and the two side windows open. The electrical connections to the trap are cooled on both the first and the second stage of the cryo cooler using small copper bobbins that serve as heat anchors. For the parts at low temperature, manganin and copper wires were chosen, while steel wires were installed to connect to the feedthroughs through the vacuum vessel.

The trap is equipped with two different gas inlets allowing mixtures of different buffer gases and for creating tagged ions or for investigating ion–molecule reactions. One of the inlets is connected to a self-built piezo-electric valve to allow for pulsed buffer gas input and thus higher peak densities. The buffer gas lines are thermalized on the first and on the second stage of the coldhead. On the first stage, a large copper tube is connected to the coldhead, and two independent buffer gas lines made of smaller copper tubes are bent spirally inside it, resulting in a long precooled path. The heat flow via warm buffer gas is diminished by using teflon tubes in the warmer region between the first and the second stage. A self-designed, gas tight connection to the trap housing is mounted on the second stage of the cryo cooler stage. Thermalization of the buffer gas is achieved in the bottom of the trap housing by guiding the gas into a labyrinth structure that features several 180° bendings and ensures that the gas particles have multiple inelastic collisions with the cryogenic walls before entering the trapping region.

The temperature of the trap is monitored on the hottest and the coldest part of the trap using silicon diodes (Lakeshore DT-670) as temperature sensors. Mechanical shielding of the sensors allows for temperature monitoring during radio frequency operation. Special heat anchors and the chosen materials for both the signal wiring and the gas lines ensure minimized heat flow through the connections to the ambient room temperature.

By the interplay of all these technical details, cryogenic temperatures as low as 2.97 K are achieved without radio frequency and buffer gas, 3.05 K with the radio frequency voltage applied, and 3.09 K when the radio frequency is applied, helium buffer gas is passed into the trap, and ions are trapped. This was tested with trapped protonated glycine ions with a radio frequency of 5.05 MHz at an amplitude of 400 V. Higher temperatures of 3 K up to about 40 K can be realized using electric heating elements inserted into the body of the main part of the trap.

In linear radio frequency multipole ion traps, the confinement in the radial direction is achieved by radio frequency potentials of opposite phase alternatingly to the trap electrodes (see Fig. 2). In the axial direction, static voltages are applied to the two sets of endcap electrodes to create an axially confining potential. In the following, we summarize a few key properties of multipole ion traps.

FIG. 2.

Cut through the central perpendicular plane of a linear 16-pole radio frequency ion trap. The blue and red circles depict the cross sections of the electrodes with a diameter of 2r0 to which radio frequency voltages +V0 and −V0 with opposite phase are applied. R0 is the radius of the inscribed cylinder. The lines A and B are used to describe the electric field along the two most different radial directions, with line A going straight through an electrode and line B moving with the maximum distance in between two neighboring electrodes.

FIG. 2.

Cut through the central perpendicular plane of a linear 16-pole radio frequency ion trap. The blue and red circles depict the cross sections of the electrodes with a diameter of 2r0 to which radio frequency voltages +V0 and −V0 with opposite phase are applied. R0 is the radius of the inscribed cylinder. The lines A and B are used to describe the electric field along the two most different radial directions, with line A going straight through an electrode and line B moving with the maximum distance in between two neighboring electrodes.

Close modal

Assuming the motion of the trapped ions can be separated into a rapid oscillation at the radio frequency termed micromotion, and a slow drift motion, termed macromotion, one finds that the macromotion is subject to an effective trapping potential34 

VeffR=q2E2R4mΩ2+ΦDC,
(1)

where m is the mass of the trapped particle with charge q, E(R) is the electric field at the point R, and Ω is the frequency of the oscillatory electric field. ΦDC denotes an electrostatic offset potential that is not relevant for trapping.

To estimate when the effective potential approximation holds, Gerlich34 introduced the adiabaticity parameter

η=2|q|||E||mΩ2,
(2)

which is a generalization of the q-parameter in quadrupole ion traps. Stable trapping conditions were found if the adiabaticity parameter is limited to η < 0.38.35 

Equation (1) can be applied to various electrode geometries. In the case of an ideal, infinitely long linear multipole, the effective potential takes the explicit form

Veff(R)=q2n2V024mΩ2R02RR0(2n2),
(3)

with R being the radial distance from the trap center and R0 being the inscribed radius. n denotes the multipole order (n = 8 for the 16-pole trap), and V0 denotes the voltage amplitude of the radio frequency field.

The ideal multipole field can be approximated with 2n cylindrical electrodes of diameter 2r0 that are equally arranged on a cylindrical surface with inscribed radius R0. In this case, the optimal ratio of radii is given by34 

r0=R0n1,
(4)

which has been used as a guideline for many radio frequency multipole ion traps in the past.1,21

The new wire trap breaks away from this guideline. It consists of 16 copper wires of diameter d = 100 µm arranged on a cylinder with an inscribed radius R0 of 5.15 mm. For a multipole order of n = 8, the above criterion would result in an inscribed radius of only 0.35 mm. Alternatively, maintaining the inscribed and the electrode radii would lead to an impractically high order of n = 207. Both of these options are not suited for the goal to maximize the radial optical access to the ion cloud. Therefore, we chose to keep n = 8 and R0 = 5.15 mm while using d = 100 μm diameter wires as electrodes, which provides an open area of about 95% (see Sec. II B). The consequences of this design for the effective potential are presented in Sec. III.

In the following, the three dimensional electric potential simulations are described, and the results that were obtained for the wire trap are presented and discussed.

The electric potential of the trap is simulated by solving the Poisson equation in three dimensions with the software package COMSOL Multiphysics.36 The simulated configuration consists of 16 electrodes (multipole order n = 8) with diameter 2r0 and length 35 mm that are arranged on a cylinder with inscribed radius. This inscribed radius is fixed to R0 = 5.15 mm in accord with the experiment. No additional electrodes are included in the simulation to focus on the core properties of a wire-based ion trap with diameters much smaller than recommended by Eq. (4). The simulation volume is chosen to be a cylinder of radius 26 mm and height 35 mm [see Fig. 3(a)], and the outer boundary conditions are selected to be non-conducting. The focus of the simulation is on the central plane of the ion trap, and it has been tested that the details of the boundary conditions or the simulation volume did not affect the results for this plane.

FIG. 3.

(a) Schematic view of the simulation volume and the trap electrodes. The electric potential is shown for electrode amplitudes of V0 = ±175 V by a color code ranging from blue to red. The data are extracted from the central plane depicted in green. (b) Cross section of the mesh used in the simulation for an electrode diameter of 2r0 = 0.1 mm. The mesh is finer around the electrodes and coarser in the center as well as outside of the trap. The color code reflects the quality of the mesh elements, more specifically the local growth rate of the mesh, with green denoting high quality and red denoting low quality.

FIG. 3.

(a) Schematic view of the simulation volume and the trap electrodes. The electric potential is shown for electrode amplitudes of V0 = ±175 V by a color code ranging from blue to red. The data are extracted from the central plane depicted in green. (b) Cross section of the mesh used in the simulation for an electrode diameter of 2r0 = 0.1 mm. The mesh is finer around the electrodes and coarser in the center as well as outside of the trap. The color code reflects the quality of the mesh elements, more specifically the local growth rate of the mesh, with green denoting high quality and red denoting low quality.

Close modal

The mesh that is used to solve the partial differential equations is adapted to the simulation and offers different areas of precision for different radii as depicted in Fig. 3(b). Overall, the radial mesh size varies between 0.05 and 0.8 mm. In the axial direction, an equidistant mesh size of 0.78 mm has been chosen. The simulations have been repeated for a set of different electrode diameters 2r0, with mesh sizes adapted to the selected diameter.

From the simulated electric potential, the electric field vectors E(R) are computed using discrete derivatives. With the electric field, the effective potential Veff(R) is then numerically calculated using Eq. (1). It depends on the frequency Ω of the radio frequency potential, as well as on the charge q and the mass m of the ions to be trapped. In this work, the effective potential is calculated for the protonated glycine molecular ion with a mass-to-charge ratio of m/z = 76. The radio frequency is chosen to be Ω = 2π · 5 MHz.

The effective potential obtained from the COMSOL simulations for the central radial plane of the wire trap is shown in Fig. 4. The effective potential is shown on a logarithmic scale. The white lines represent equipotential lines (two lines per decade) starting in the center at 10−5 eV and 3 · 10−5 eV.

FIG. 4.

Simulated effective potential in the central plane of the trap for an electrode voltage of 125 V and electrode radius of 0.05 mm. The white lines are equipotential lines with 2 lines per decade. The innermost line represents the equipotential line for the 10−5 eV effective potential. The color code on the right is from 0 eV to 5 eV. Values above 5 eV appear as dark red.

FIG. 4.

Simulated effective potential in the central plane of the trap for an electrode voltage of 125 V and electrode radius of 0.05 mm. The white lines are equipotential lines with 2 lines per decade. The innermost line represents the equipotential line for the 10−5 eV effective potential. The color code on the right is from 0 eV to 5 eV. Values above 5 eV appear as dark red.

Close modal

A more detailed view, which shows the effective potential as a function of distance from the trap center along lines A and B, is presented in Fig. 5. The dashed lines follow line A, and the full lines follow line B (see Fig. 2). The trace of the ideal multipole potential, which does not depend on the angular direction and is thus identical for line A and B, is shown as red dashed line in the figure. By comparing the potentials along lines A and B with the ideal multipole potential, one notes that the effective potentials are increasingly reduced for electrode radii much smaller than the optimal radius of 0.735 mm given by Eq. (4). For all electrode radii, the potential along lines A and B is almost indistinguishable for small distances to the trap center. After a certain distance, the potentials then diverge significantly. The smaller the electrode radius, the closer this divergence is located to the trap center. For an electrode radius of r0 = 0.05 mm, the radius where the absolute difference between lines A and B becomes more than 10 meV is found at a distance of 4.1 mm to the trap center, while if the near-optimal electrode radius of r0 = 0.735 mm is chosen, this distance is 4.3 mm.

FIG. 5.

Simulated effective potential following lines A (dashed line) and B (solid line) (compare Fig. 2) for various electrode radii from 0.05 mm up to 0.735 mm for the electrode voltage V0 of 125 V. The effective potential for the ideal multipole is depicted as a red dashed line. The inset shows the effective potential in the region of the turn-around radius, which is ∼3.7 mm for an electrode radius of 0.05 mm at room temperature.

FIG. 5.

Simulated effective potential following lines A (dashed line) and B (solid line) (compare Fig. 2) for various electrode radii from 0.05 mm up to 0.735 mm for the electrode voltage V0 of 125 V. The effective potential for the ideal multipole is depicted as a red dashed line. The inset shows the effective potential in the region of the turn-around radius, which is ∼3.7 mm for an electrode radius of 0.05 mm at room temperature.

Close modal

To characterize the radial dependence of the effective potential, a polynomial of the type Veff = aRb has been fitted to the effective potential as a function of distance to the trap center. For distances smaller than the divergence distance, the fit to the potential along line A and line B yields an exponent b ≈ 14. This is the expected (2n − 2)-scaling for a 16-pole ion trap [see Eq. (3)]. This scaling even holds for the smallest studied electrode radii for distances of less than 4 mm to the trap center. Upon approaching the electrodes, the scaling changes and differs for positions along lines A and B (see Fig. 5).

The distance where the potential along lines A and B begins to differ significantly may be compared to the classical turn-around distance for trapped ions that are thermalized to a given buffer gas temperature. For a room temperature ion cloud, the average turn-around distance is estimated by assuming Veff(Rt) = 3/2kBT, where kB is the Boltzmann constant and T is the absolute temperature. The effective potential is taken as the average along lines A and B. For protonated glycine ions at room temperature, this yields a turn-around distance Rt ≈ 3.7 mm for our wire trap with r0 = 0.05 mm with V0 = 125 V. This distance is substantially smaller than the distance where the effective potential deviates between lines A and B and from the R14 scaling.

The change in the effective potential for different angles is shown in Fig. 6 for three different distances from the trap center for a wire trap with an electrode radius of 0.05 mm. It focuses on the angular range from 0 to 90°. 11.25° corresponds to the direction along line A and 22.5° to line B. It can be seen that the amplitude variation is very small for average values of the effective potential up to about 40 meV and then grows to less than 10% at 80 meV average effective potential. Thus, within the classical turn-around distance, the angular modulation of the effective potential is not significant.

FIG. 6.

Angular dependence of the effective potential for three different distances to the trap center for a wire trap with 0.05 mm electrode radius.

FIG. 6.

Angular dependence of the effective potential for three different distances to the trap center for a wire trap with 0.05 mm electrode radius.

Close modal

It is instructive to plot the effective potential also as a function of the radio frequency amplitude. In Fig. 7, the mean of the effective potential for lines A and B at a distance of 4 mm to the trap center is plotted as a function of the radio frequency amplitude V0. It shows the expected quadratic increase in the effective potential with amplitude. This figure allows one to compare the effective potential for different electrode radii. For example, along line A, a trap with an electrode radius of 0.1 mm delivers the same effective potential of 0.16 eV (at a radial distance of 4 mm) with an amplitude of V0 = 125 V as the ideal multipole geometry would offer with V0 = 85 V. Using electrodes with a radius of 0.05 mm requires an amplitude of about 150 V. It can be concluded that electrode voltages need to be about a factor of 1.75 times larger for the present wire trap (2r0 = 0.1 mm) compared to the ideal 16-pole trap geometry.

FIG. 7.

The mean effective potential at a distance of 4 mm from the trap center is shown for different radii and different electrode voltages for the mean of lines A and B.

FIG. 7.

The mean effective potential at a distance of 4 mm from the trap center is shown for different radii and different electrode voltages for the mean of lines A and B.

Close modal

A different view on the scaling of the effective potential for different electrode radii is provided in Fig. 8. For each electrode radius, the mean effective potential along lines A and B is computed for different distances from the trap center and then divided by the potential for the ideal multipole at that position. Six different positions ranging from 2.0 mm to 4.5 mm are shown. One observes a clear decrease in the effective potential for small electrode radii down to about 30% of the ideal multipole potential for r0 = 0.05 mm.

FIG. 8.

Mean effective potential, along lines A and B, divided by the ideal multipole potential as a function of the electrode radius. Calculations are shown for six distances from the trap center.

FIG. 8.

Mean effective potential, along lines A and B, divided by the ideal multipole potential as a function of the electrode radius. Calculations are shown for six distances from the trap center.

Close modal

The presented numerical simulations show how the effective potential of a 16-pole radio frequency ion trap changes when the electrode radius is reduced from the optimal value to that of thin wires. The dominant changes are the occurrence of a significant angular modulation at large distances to the trap center and a significantly reduced magnitude of the effective potential. Higher radio frequency amplitudes are therefore required to operate wire ion traps, but the needed increase of less than a factor of two is actually moderate. The angular modulation turns out to be strong only in the region of high effective potential close to the trap electrodes, far outside the classical turn-around distance of thermal trapped ions. It is therefore expected that this modulation does not have a strong influence on the trapping and buffer gas cooling conditions in the trap.

For the test of the wire ion trap, protonated glycine cations are used. The ions are produced by electrospray from a 1 mM glycine solution in a 50:50 water:methanol mixture with 1% acetic acid. The ions are passed through the guiding octupole and quadrupole (see also Sec. II A) into the ion trap. For loading the trap, the ions have to overcome the axial potential of the inner entrance endcap electrode, which is set to +1.7 V above the dc potential of the trap. Once inside the trap, any ion that loses kinetic energy due to buffer gas collisions then stays trapped. In this way, ions are loaded for 180 ms after which an electrode between the octupole and quadrupole ion guides blocks further ions from the source. At the beginning of the loading period, a pulse of helium buffer gas is passed into the trap, which is used both for translational cooling and trapping and for creating helium-tagged ions by three-body collisions. The ions are then kept trapped for 800 ms. During this time, the helium gas is pumped out of the trap so that it cannot lead to collision-induced dissociation of the weakly bound tagged molecular ions during extraction. For unloading the ions, the dc potential on the inner exit endcap electrode is switched from typically +5 V relative to the trap dc potential to a negative voltage. The trapped ions are then passed to the reflectron time-of-flight mass spectrometer, which is pulsed at the proper delay to detect the protonated glycine ions and neighboring masses, and are detected in single-ion counting mode in order to be sensitive to small ion signals. Time-of-flight spectra are obtained by recording the individual arrival times with a constant fraction discriminator and time-to-digital-converter electronics suite. A usual mass spectrum is acquired in 300 s, which corresponds to 300 trap fillings. For a better signal to noise ratio, three such spectra have been added in Fig. 9 for each shown mass spectrum.

FIG. 9.

Time-of-flight spectra of protonated and helium-tagged glycine at trap temperatures of 3.3 K in blue and 30 K in orange with protonated glycine at m/z = 76 and attached with 1, 2, 3, and 4 helium tags at m/z = 80, 84, 88, and 92. Each spectrum is an accumulation of 900 trap fillings.

FIG. 9.

Time-of-flight spectra of protonated and helium-tagged glycine at trap temperatures of 3.3 K in blue and 30 K in orange with protonated glycine at m/z = 76 and attached with 1, 2, 3, and 4 helium tags at m/z = 80, 84, 88, and 92. Each spectrum is an accumulation of 900 trap fillings.

Close modal

Representative time-of-flight mass spectra for trapped protonated glycine are plotted in Fig. 9 for two trap temperatures, 3.3 K and 30 K. The largest peak in the spectrum at m/z = 76 stems from protonated glycine. It is saturated due to overlapping detection pulses in the employed single-ion counting mode, which also broadens the peak toward larger masses. All other peaks appear unsaturated and allow us to estimate a mass resolution of about mm ∼ 275 in this case. From mass spectra taken for the ion beam from the source without trapping, it is estimated that the protonated glycine peak is actually factors of about 10 and 15, respectively, larger than the two second largest peaks at mass 78 and 93, which are attributed to the solvent but are presently unassigned.

At m/z = 80, 84, 88, and 92, signal appears only for the 3.3 K ion temperature, but not for 30 K. These peaks are assigned to protonated glycine ions that are tagged with between one and four helium atoms. The rate for three-body collisions that are required to form these complexes with helium decreases strongly with temperature. In addition, the stability of these clusters is lower at higher temperature due to thermally activated collision-induced dissociation. Hence, they can only be formed and trapped at low temperatures. A strong sensitivity of the helium tagging efficiency to the trap and buffer gas temperature has also been observed in other cryogenic ion trap experiments.23,37 This shows that the present wire multipole ion trap is well capable of achieving low enough temperatures to allow for efficient and multiple helium tagging.

We have presented the mechanical design for a new linear cryogenic 16-pole wire trap. Careful consideration of an optimal thermal conductance led to a wire ion trap that can be cooled to 3.1 K under ion trapping operation.

By simulating the effective potential of different 16-pole traps, we have shown that even with electrode radii far off the ideal multipole configuration, the effective potential scales with the distance to the trap center as R2n−2 with n = 8. The angular modulation of the effective potential becomes important for thin wire electrodes, but this modulation is found to still remain below about 2 meV near the average turn-around distance for a thermal ion ensemble at room temperature. To achieve the same effective potential depth, a higher amplitude for the radio frequency potential is required, but for the given wire electrodes, an increase by less than a factor of two turns out to be sufficient.

Successful operation at temperatures near 3 K has been achieved by tagging protonated glycine ions with up to four helium atoms. The modulation of the effective potential near the classical turn-around distance is therefore not found to be a limiting factor for buffer gas cooling and helium tagging. The strong reduction of the electrode surface area in the vicinity of the trapped ions may actually be beneficial, as it suppresses the chance for electric field distortions in the trapping volume due to surface patch potentials.

The presented wire ion trap is useful for a range of spectroscopy experiments with weakly bound ion complexes.30 With the large optical access the trap offers in the radial direction, it is suitable for a wide range of radiation frequencies from the terahertz to the ultraviolet regime.

We thank Engelbert Portenkirchner, Markus Nötzold, and Robert Wild for many helpful discussions. We are grateful to Florian Zweiker and Simon Mayregger-Kasseroler for the fabrication of the ion trap. This work was supported by the Austrian Science Fund (FWF) through the Doctoral Programme Atoms, Light, and Molecules, Project No. W1259-N27 and by the European Research Council (ERC), Grant Agreement No. 279898.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

2.
R.
Otto
,
J.
Mikosch
,
S.
Trippel
,
M.
Weidemüller
, and
R.
Wester
,
Phys. Rev. Lett.
101
,
063201
(
2008
).
3.
D.
Gerlich
,
P.
Jusko
,
Š.
Roučka
,
I.
Zymak
,
R.
Plašil
, and
J.
Glosík
,
Astrophys. J.
749
,
22
(
2012
).
4.
D.
Hauser
,
S.
Lee
,
F.
Carelli
,
S.
Spieler
,
O.
Lakhmanskaya
,
E. S.
Endres
,
S. S.
Kumar
,
F.
Gianturco
, and
R.
Wester
,
Nat. Phys.
11
,
467
(
2015
).
5.
F. A.
Gianturco
,
O. Y.
Lakhmanskaya
,
M. H.
Vera
,
E.
Yurtsever
, and
R.
Wester
,
Faraday Discuss.
212
,
117
(
2018
).
6.
S.
Schlemmer
,
T.
Kuhn
,
E.
Lescop
, and
D.
Gerlich
,
Int. J. Mass Spectrom.
185-187
,
589
(
1999
).
7.
O.
Asvany
,
K. M. T.
Yamada
,
S.
Brünken
,
A.
Potapov
, and
S.
Schlemmer
,
Science
347
,
1346
(
2015
).
8.
S. E.
Waller
,
Y.
Yang
,
E.
Castracane
,
E. E.
Racow
,
J. J.
Kreinbihl
,
K. A.
Nickson
, and
C. J.
Johnson
,
J. Phys. Chem. Lett.
9
,
1216
(
2018
).
9.
M.
Brümmer
,
C.
Kaposta
,
G.
Santambrogio
, and
K. R.
Asmis
,
J. Chem. Phys.
119
,
12700
(
2003
).
10.
D. J.
Goebbert
,
T.
Wende
,
R.
Bergmann
,
G.
Meijer
, and
K. R.
Asmis
,
J. Phys. Chem. A
113
,
5874
(
2009
).
11.
T. K.
Esser
,
H.
Knorke
,
F.
Siro-Brigiano
,
D. R.
Galimberti
,
K. R.
Asmis
,
M.-P.
Gaigeot
, and
J. M.
Lisy
,
Phys. Chem. Chem. Phys.
20
,
28476
(
2018
).
12.
A.
Masson
,
E. R.
Williams
, and
T. R.
Rizzo
,
J. Chem. Phys.
143
,
104313
(
2015
).
13.
F. S.
Menges
,
E. H.
Perez
,
S. C.
Edington
,
C. H.
Duong
,
N.
Yang
, and
M. A.
Johnson
,
J. Am. Soc. Mass. Spectrom.
30
,
1551
(
2019
).
14.
S.
Chakrabarty
,
M.
Holz
,
E. K.
Campbell
,
A.
Banerjee
,
D.
Gerlich
, and
J. P.
Maier
,
J. Phys. Chem. Lett.
4
,
4051
(
2013
).
15.
P.
Jusko
,
O.
Asvany
,
A.-C.
Wallerstein
,
S.
Brünken
, and
S.
Schlemmer
,
Phys. Rev. Lett.
112
,
253005
(
2014
).
16.
S.
Lee
,
D.
Hauser
,
O.
Lakhmanskaya
,
S.
Spieler
,
E. S.
Endres
,
K.
Geistlinger
,
S. S.
Kumar
, and
R.
Wester
,
Phys. Rev. A
93
,
032513
(
2016
).
17.
O.
Lakhmanskaya
,
M.
Simpson
,
S.
Murauer
,
M.
Nötzold
,
E.
Endres
,
V.
Kokoouline
, and
R.
Wester
,
Phys. Rev. Lett.
120
,
253003
(
2018
).
18.
C.
Hock
,
M.
Schmidt
,
R.
Kuhnen
,
C.
Bartels
,
L.
Ma
,
H.
Haberland
, and
B. v.
Issendorff
,
Phys. Rev. Lett.
103
,
073401
(
2009
).
19.
N.
Heine
and
K. R.
Asmis
,
Int. Rev. Phys. Chem.
34
,
1
(
2015
).
20.
J.
Pedregosa-Gutierrez
,
C.
Champenois
,
M. R.
Kamsap
,
G.
Hagel
,
M.
Houssin
, and
M.
Knoop
,
J. Mod. Opt.
65
,
529
(
2018
).
21.
R.
Wester
,
J. Phys. B: At., Mol. Opt. Phys.
42
,
154001
(
2009
).
22.
O.
Asvany
,
F.
Bielau
,
D.
Moratschke
,
J.
Krause
, and
S.
Schlemmer
,
Rev. Sci. Instrum.
81
,
076102
(
2010
).
23.
O.
Asvany
,
S.
Brünken
,
L.
Kluge
, and
S.
Schlemmer
,
Appl. Phys. B
114
,
203
(
2014
).
24.
A.
Günther
,
P.
Nieto
,
D.
Müller
,
A.
Sheldrick
,
D.
Gerlich
, and
O.
Dopfer
,
J. Mol. Spectrosc.
332
,
8
(
2017
).
25.
C.
Hock
,
J. B.
Kim
,
M. L.
Weichman
,
T. I.
Yacovitch
, and
D. M.
Neumark
,
J. Chem. Phys.
137
,
244201
(
2012
).
26.
J.
Jašik
,
J.
Zabka
,
J.
Roithová
, and
D.
Gerlich
,
Int. J. Mass Spectrom.
354
,
204
(
2013
).
27.
J.
Deiglmayr
,
A.
Göritz
,
T.
Best
,
M.
Weidemüller
, and
R.
Wester
,
Phys. Rev. A
86
,
043438
(
2012
).
28.
O. V.
Boyarkin
and
V.
Kopysov
,
Rev. Sci. Instrum.
85
,
033105
(
2014
).
29.
M.
Nötzold
,
S. Z.
Hassan
,
J.
Tauch
,
E.
Endres
,
R.
Wester
, and
M.
Weidemüller
,
Appl. Sci.
10
,
5264
(
2020
).
30.
S.
Spieler
,
C. H.
Duong
,
A.
Kaiser
,
F.
Duensing
,
K.
Geistlinger
,
M.
Fischer
,
N.
Yang
,
S. S.
Kumar
,
M. A.
Johnson
, and
R.
Wester
,
J. Phys. Chem. A
122
,
8037
(
2018
).
31.
E.
Gmelin
,
M.
Asen-Palmer
,
M.
Reuther
, and
R.
Villar
,
J. Phys. D: Appl. Phys.
32
,
R19
(
1999
).
32.
G.
Ventura
and
L.
Risegari
,
The Art of Cryogenics: Low-Temperature Experimental Techniques
(
Elsevier, Ltd.
,
2008
).
33.
S.-i.
Ishiuchi
,
H.
Wako
,
D.
Kato
, and
M.
Fujii
,
J. Mol. Spectrosc.
332
,
45
(
2017
).
34.
D.
Gerlich
,
Adv. Chem. Phys.
82
,
1
(
1992
).
35.
J.
Mikosch
,
U.
Frühling
,
S.
Trippel
,
D.
Schwalm
,
M.
Weidemüller
, and
R.
Wester
,
Phys. Rev. Lett.
98
,
223001
(
2007
).
36.
COMSOL Multiphysics v. 5.5, www.comsol.com, COMSOL AB, Stockholm, Sweden.
37.
D.
Gerlich
,
J.
Jašík
,
E.
Andris
,
R.
Navrátil
, and
J.
Roithová
,
ChemPhysChem
17
,
3723
(
2016
).