A brief history of ion traps (for quantum information processing)

Chiara Decaroli
9 min readJan 30, 2021

--

Springtime has come for the field of quantum computing. Researchers around the globe are hard at work, governments are announcing large-scale projects and new facilities, while start-ups and companies are popping up like mushrooms after the rain. One of the promising technologies for the implementation of a useful quantum computer is trapped ions. But what are these ions? And how are they trapped? Here I will try to give a brief (and by no means exhaustive) history of ion traps, and their evolution across the years toward becoming one of the building blocks for quantum computers.

The early history

Our story begins in the Netherlands, in the 1930s. A young physicist named Frans Michel Penning had been conducting his doctoral research on the thermodynamics properties of gases. He was fascinated by the behaviour of different gases and discharge at very low temperatures and under specific conditions. After completing his doctoral work, he joined the Phillips research laboratory, where he was tasked with continuing research on discharge of gases. The main application for his studies at the time was the development of new lamps! While investigating discharge, he invented a device, called the Penning gauge, which used magnetic fields to accelerate electrons through a tube, and allowed a precise measurement of the pressure within the tube, a vacuum gauge [2].

Fast forward to 1949, across the ocean, the physicist J. R. Pierce described in his book “Theory and Design of Electron Beams” an electron trap, which was able to confine electrons in a specific region of space, using a combination of electric and magnetic fields [1]. The work caught the attention of the german physicist Hans Dehmelt [5] , who was at the time working on his doctoral thesis and would move to the Unites States shortly after. He developed more in detail the mathematics describing the motion of the electrons when trapped, and was able to create the first “magnetron” trap and trap electrons in 1959! He named the device the Penning Trap, honouring the first efforts in studying the effects of magnetic fields on discharge by Penning. The Penning trap would later become the tool of choice to perform high precision measurements of the properties of particles such as electrons and protons.

From J.R.Pierce “Theory and Design of Electron Beams”, 1949

Meanwhile, back in Germany, another physicist was curious about techniques to confine charged particles. The field of mass spectroscopy, in which atoms would get ionised, and subsequently accelerated towards a detector, had been active for several years. A combination of electric and magnetic fields pushed and deflected ionised atoms on the detector according to their charge to mass ratio, allowing for a precise determination of the mass of specific substances. Wolfgang Paul [6], while at the university of Bonn, proposed in 1953 a “New mass spectrometer without a magnetic field”. Not only this mass spectrometer did not need a magnetic field, but for certain geometries it was able to confine charged particles, opening up the opportunity of trapping ions, much like in a Penning trap, but without the need for a strong magnetic field. The Radio-frequency ion trap, also called the Paul Trap, was born.

“A new mass spectrometer without a magnetic field”, the proposal by Paul in 1953

A few decades later, in 1989, Paul and Dehmelt were awarded the Nobel Prize in Physics “for the development of the ion trap technique” [3]. Their early discoveries has given scientists the chance to study in detail the properties and spectra of atoms, as the Nobel Prize committee writes:

“The properties of atoms are determined by laws of quantum mechanics that say they can have only fixed energy levels and that electromagnetic radiation with certain frequencies is emitted or absorbed when there are transitions among different energy levels. Opportunities to study the properties and spectrums of atoms are improved if individual atoms can be isolated under constant conditions for longer periods.” [3]

Back row, from the right: Dehmelt and Paul at the Nobel Prize ceremony in 1989, image from [2]

(A fun fact which amused me about Wolfgang Paul: he used to refer to Wolfgang Pauli as his imaginary part if their surnames were considered as complex numbers.)

Basic operation and initial geometries

Now that we know a little about the history, let’s take a closer look at ion traps! We already know there are two main kinds: the Penning trap and the Paul trap. Both kinds utilise static electric fields to confine the ions, however the static electric fields on their own cannot create a trapping region in all three dimensions. If the ions are not trapped in all directions, they can easily escape in the direction which is not confining. To ensure that the ions are trapped, one extra field is required. While the Penning trap uses a magnetic field for this, the Paul trap uses an oscillating electric field [4], [7].

The two original geometries for the ion trap. On the left, a realisation of Paul and Penning trap. On the right, a Paul linear 2D trap. Illustration by C. Decaroli

The original geometry for the Paul and Penning trap was the Ring 3D trap shown on the left in the illustration above [8]. The Paul ring trap had a static electric field applied to the endcaps and an oscillating electric field applied to the ring. The endcaps created confinement along the trap axis, while the ring trapped the ions in the radial plane, the plane perpendicular to the trap axis. For the Penning trap, a magnetic field was applied in the direction of the trap axis, and all ring and endcaps had static electric fields applied to them. Similarly to the Paul version, the magnetic field had the role to create confinement in the radial plane. In both cases, the ions would be trapped in the very middle of the ring.

The 2D linear trap was an extension of the 3D Paul trap, but did not confine the particles along the trap axis unless extra endcaps were included. This trap was used as a mass spectrometer in the 2D configuration [7], but became the basis for later design developments for 3D Paul traps.

The evolution of Paul & Penning traps

By now, you might be asking yourself “why is there a (for quantum information processing) in the title, and when is that stuff going to come up? It feels like we are still talking about mass spectrometers!”. Hang in there, we are getting to it!

With the ability to confine charged particles in a stable manner for extended period of times came the opportunity to start studying in detail the properties of the particles, and to study their interactions with their environment and external fields. Several techniques were developed to allow scientists to control the trapped particles and manipulate them, reaching regimes which were previously inaccessible, with David Wineland being one of the pioneers of the field. One such technique was laser cooling, thanks to which the energy of the trapped ions is lowered by its interaction with laser light.

The field was very exciting and offered very interesting avenues not only in the field of spectroscopy and precision measurements but also towards studying quantum mechanical effects. In 1995, Ignatio Cirac and Peter Zoller, proposed a way to implement the quantum equivalent of a classical gate, a quantum gate, using cold trapped ions [9]. Their proposal ignited a whole new research direction: quantum computation with trapped ions. The field has been growing ever since, and has focused not only on trapped ions but on a variety of other platforms such as superconducting circuits, solid state-based quantum systems, neutral atoms, photons etc.

As the needs and expectations from the humble trapped ions changed and grew, it was time to rethink the original Paul trap design. The linear Paul trap with endcaps was transformed into the rod trap, made up of cylindrical segmented rods, and later into the blade trap, to allow for a larger optical access angle, to send in and collect light emitted by the ions. These traps were still macroscopic, with dimensions in the order of centimeters in size.

A blade trap used at the University of Oxford. Photo credit: David Nadlinger

As trapped ions showed better and better performance as qubits, or quantum bits, the long term goal of building a quantum computer out of them imposed requirements on the size of the traps. It became necessary to be able to shrink down the devices and work on making them more scalable. The macroscopic Paul trap was miniaturised, and two main microfabricated designs emerged: the 3D microfabricated trap and the surface (or planar) trap. Microfabricated traps went from centimeters to millimeters in size. The ions were now trapped tens to hundreds of micrometers away from the electrodes.

The evolution of the ion trap: blade trap, microscopic 3D trap and surface trap. Illustration by C. Decaroli

Microfabricated 3D traps are made by stacking a number of individual wafers on top of each other. The wafers are either made of conductive materials, or insulating materials that are subsequently coated in metal to create electrodes. The ion sits at the centre of the stack, and sees a field landscape very similar to that generated by the original linear trap. These traps create deep trap for the ions, and can be operated at room temperature down to cryogenic temperatures of a few degrees Kelvin. However, their fabrication and assembly is fairly troublesome and the reproducibility is low.

On the other hand, surface traps are made of a single wafer which is patterned using commercial technologies and lithography techniques. The ion(s) is trapped above the surface, and can be as close as a few tens of micrometers to the surface of the electrodes. Due to their relatively easier fabrication, these traps are a promising candidate when it comes to scaling up the devices to be able to trap many ions.

The design of innovative Paul traps is an active field of research, and as new fabrication technologies are developed, more opportunities present themselves. Lots of exciting developments are unravelling, such as the integration of optical elements in the trap wafers themselves to deliver the necessary laser light to perform quantum gates.

A surface trap with integrated waveguides for the delivery of laser light. Illustration by C. Decaroli for ETH news article [10]

And what about Penning traps? While being extensively used in precision measurements, they have also found applications in quantum simulations and many-body physics such as in [11], and are following a similar trajectory to Paul traps in the realm of quantum computations.

An ion crystal trapped ion a Penning trap [11] in NIST, Colorado

Lots of engineering efforts, research collaborations and innovative design will be necessary to transition from trapping tens of ions to trapping thousands of them, the coming years will be full of really exciting research!

About the author

Chiara Decaroli is a physicist working towards her PhD in the field of trapped-ion quantum information processing. She focuses on designing and fabricating ion traps. She is also a freelance scientific illustrator and loves all things sci comm. You can find her on twitter: @DecaroliChiara.

References

[1] “A brief history in time of ion traps and their achievements in science”, Michael H Holzscheiter 1995 Phys. Scr. 1995 69

[2] “Die glimmentladung bei niedrigem druck zwischen koaxialen zylindern in einem axialen magnetfeld”, F. M. Penning, Physica, Volume 3, Issue 9, November 1936, Pages 873–894

[3] www.nobelprize.com

[4] “Physics with Trapped Charged Particles”, Martina Knoop , Niels Madsen and Richard C. Thompson, https://arxiv.org/pdf/1311.7220.pdf

[5] https://en.wikipedia.org/wiki/Hans_Georg_Dehmelt

[6] https://en.wikipedia.org/wiki/Wolfgang_Paul

[7] “Electromagnetic traps for charged and neutral particles”, Wolfrang Paul, Reviews of Modern Physics, Vol. 62, №3, July 1990

[8] “Ein neues Massenspektrometer ohne Magnetfeld”, Wolfgang Paul and Helmut Steinwedel, Zeitschrift für Naturforschung A, Volume 8: Issue 7, 1953

[9] “Quantum Computations with Cold Trapped Ions” J. I. Cirac and P. Zoller, Phys. Rev. Lett.74, 4091–4094 (1995).

[10] https://ethz.ch/en/news-and-events/eth-news/news/2020/10/optische-verdrahtung-fuer-grosse-quantencomputer.html

[11] “Quantum simulation and many-body physics with hundreds of trapped ions”, John J. Bollinger, Joseph W. Britton, and Brian C. Sawyer, CLEO 2013 Technical Digest, OSA 2013

[12] https://aquadrupauliontrap.wordpress.com/

--

--

Chiara Decaroli

Quantum physics researcher, Yoga teacher and occasional illustrator based in Zurich.