1 Introduction

Although often considered a technology of the distant future, fusion devices have been researched, designed, and constructed for decades, with electricity-generating fusion plants anticipated as early as the 2030s. Despite the popularity of the traditional tokamak design—using a magnetic field to confine fusion plasma within a toroidal vessel—fusion research encompasses many different combinations of fuel, confinement technology, and concepts. There are three main types of confinement concepts today: magnetic confinement fusion (MCF), magneto-inertial fusion (MIF), and inertial confinement fusion (ICF). Despite the development of these alternative concepts, magnetic confinement fusion remains the most popular choice for eventual power production. However, steady power production with MCF can only be achieved using powerful superconducting magnets.

Superconductivity refers to the property of some materials to conduct electricity without resistance, typically at very low temperatures nearing 0 K. Ordinary conductors display a gradual decrease in resistance as their temperature decreases, while superconductors have a characteristic critical temperature, below which the resistance drops abruptly to zero [1]. Considering that the magnetic field is proportional to the electrical current carried by the electromagnet, superconducting magnets can fulfill the extreme field requirements for confining plasma in an MCF device, such as tokamaks, stellarators, or spheromaks. Despite their great potential in nuclear fusion, superconducting magnets still present challenges that must be addressed before their widespread usage. Some of the obstacles throughout history and being researched today are manufacturing difficulties, radiation damage, reliability with long-term use, and cost. Future applied superconductivity research aims to address these challenges, paving the way for fusion power plants. This paper reviews the historical context of superconductivity within the scope of fusion research, addresses the status of worldwide fusion projects using superconductors, and finally, identifies areas of ongoing and future research addressing these challenges.

2 History of Fusion and Superconducting Magnets

In the 1950s, claims presented by the German scientist Ronald Richter in his Huemel project catalyzed early fusion research [2]. Despite failing to achieve nuclear fusion, it prompted the US, UK, and USSR to begin their own endeavors into fusion energy production. In the US, the scientist Lyman Spitzer began research on his stellarator concept, while USSR counterparts Andrei Sakharov and Igor Tamm built their linear pinch fusion machine. By 1955, the USSR had built the first toroidal fusion device. Simultaneously, research into superconducting materials began to gain traction. In 1954, Nb3Sn was discovered as a superconductor, and the following year, the US scientist George Yntema coiled the first electromagnet with superconducting Nb wire, producing a field of 0.7 tesla [3].

In 1958, the USSR began operation of the first tokamak, T-1, while Spitzer’s stellarator model became the focal point of the Atoms for Peace convention in Geneva [4]. Despite growing interest in both tokamaks and stellarators, around this time, all fusion devices showed high rates of plasma leakage. The US physicist David Bohm had theorized that plasma diffused along lines of force at a rate inversely proportional to the magnetic field, whereas classical diffusion stated that plasma diffusion was inversely proportional to the square of the magnetic field [5]:

$$D_{{classical}} \propto \frac{1}{{B^{2} }}\quad \quad \quad \quad \quad D_{{Bohm}} \propto \frac{1}{B}$$
(1)

where D represents the diffusion rate of a plasma across a magnetic field in m2/s and B is the magnetic flux density in tesla.

For the same magnetic field, Bohm’s diffusion coefficient suggested higher rates of particle loss from the plasma and reduced plasma confinement times when compared to classical diffusion, indicating that magnetic confinement fusion would not be practical. Although his theory was later replaced by neoclassical diffusion, Bohm diffusion plunged fusion research into a period of stagnation, believing that, among other obstacles, the current magnets available could never supply the fields required for confinement.

Contrasting the pessimism surrounding fusion, superconductor research rapidly progressed with niobium-tin sustaining superconductivity at large currents and strong magnetic fields (nearly 9 T) in 1961, bridging the gap between the high magnetic field required and the current materials available for MCF [3]. The following year, comparatively less brittle NbTi alloys produced fields up to 10 T, prompting the commercial production of NbTi superconducting wire at Westinghouse.

After physicists in the USSR began reporting impressive 1000 eV electron temperatures in their T-3 tokamak, the tokamak concept spiked in popularity with a flood of new proposals in 1969 [6]. Although Spitzer’s stellarator concept was originally popular in the US, MIT’s Alcator, General Atomics’ Doublet, UT-Austin’s Texas Turbulent Tokamak, and PPPL’s Symmetric Tokamak were all designed in the two years following T-3’s results.

In 1978, superconductivity and fusion research finally merged with the Kurchatov Institute’s T-7 tokamak, which used superconducting NbTi toroidal field (TF) coils [4]. Figure 1 shows the NbTi wires placed between the grooves of the nine copper cooling channels in the T-7 tokamak. Of the two available superconducting wires at the time, NbTi offered higher ductility, hence increased manufacturability and affordability, whereas NbSn provided a stronger magnetic field at a maximum of 30 T (compared to NbTi Bmax = 15 T) and a slightly higher critical temperature of 18.3 K (NbTi Tc = 10 K).

Fig. 1
figure 1

Cross-section diagram of T-7 NbTi wire [7]

Full size image

In the 1970s, Europe, Japan, and the US began designing large-scale tokamak experiments TFTR, JET, and JT-60, which each began operation in 1982, 1983, and 1985, respectively. Throughout the 1980s, the plasma operations of these large experiments revealed new plasma instability issues, requiring higher triple products than each machine was able to produce.

There had been some discussion between the USSR and US about an International Tokamak Reactor (dubbed INTOR), which was only revived during the Geneva Summit of 1985 where Reagan and Gorbachev proposed the new ITER project. In 1986, fusion research was consolidated under one multinational massive tokamak project [8].

The next few years also brought three major discoveries in superconductivity [3]:

  • 1986: Bednorz and Müller at IBM Zürich discover the first high temperature superconductor (HTS) lanthanum barium copper oxide (LBCO) at Tc = 35 K

  • 1987: University of Houston and University of Alabama discover yttrium barium copper oxide (YBCO) at Tc = 93 K

  • 1988: Bismuth strontium calcium copper oxide (BSCCO) discovered at Tc = 96 K in Japan

The ground-breaking discoveries of high-temperature superconductors in the 1980s marked a pivotal moment since these materials exhibited superconducting properties at temperatures significantly higher than those previously achievable. However, the discovery of HTS did not immediately translate into practical use for fusion magnets, as large-scale wires and tapes required for such applications were not immediately available.

Following through with the promises made at the Geneva Summit, design work on ITER began in 1988. Generally, ITER has three main magnet systems in their design: toroidal field (TF) coils, poloidal field (PF) coils, and the central solenoid (CS). Figure 2 shows the general design and location of each type of coil in ITER’s magnet system.

Fig. 2
figure 2

ITER’s magnet system [9]

Full size image

Reported in the 2002 design, the 18 TF coils are Nb3Sn strands wound with copper into a cable inside a stainless-steel jacket (shown in Fig. 3) and held in place by grooved steel plates. To form one large TF magnet, the cables are pancake wound in a welded SS case, treated with wind-and-react technique.

Fig. 3
figure 3

Schematic of ITER’s TF Nb3Sn cable-in-conduit conductor (CICC) wire [10]

Full size image

The central solenoid is made of Nb3Sn in a jacket, and split into six individual modules, built from multiple pancake layers, treated with the wind-and-react technique. Finally, the poloidal field coils are made of NbTi in a square SS conduit, constructed into double pancakes [11].

3 Present Status

In 2022, ITER’s magnet systems had the same basic materials and structure, mainly due to the limitations of ITER’s design phase schedule. Despite research advances in HTS materials, low temperature superconductors (LTS), like Nb3Sn and NbTi, are still common choices for tokamaks due to their lower cost and availability, which can also be partially attributed to the market demand created by ITER. One major change to ITER’s magnet design is the replacement of conventional Cu leads for HTS current leads. Despite most suppliers switching from Generation 1 HTS materials (e.g., BSCCO) to G2 HTS materials based on rare-earth barium copper oxides (REBCO), ITER made a final design choice in 2006 to use Bi2223 current leads, minimizing risk by avoiding the newer G2 HTS materials [12].

Currently, public tokamak research projects like ITER, EAST in China, and KSTAR in Korea still utilize LTS, however, several private and public projects have begun designing tokamaks with HTS. Today, HTS are generally defined as materials that achieve superconductivity above 77 K, with the ability to be cooled with liquid nitrogen rather than the more expensive, difficult-to-handle liquid helium. By increasing the magnetic field, MCF devices with HTS can be significantly smaller than low-field, large tokamaks like ITER. In recent years, REBCO HTS wires have become commercially available, supplied by companies like Bruker, AMSC, Fujikura, SuperOx, SuperPower, and Theva, among others. YBCO has become particularly popular for fusion applications due to the lower neutron cross section of Y (1.28 b) compared to other rare-earth elements.

Commonwealth Fusion Systems (CFS), formed in the US in 2018, has begun construction on their SPARC project, a compact tokamak utilizing the high field generated from YBCO HTS magnets. For their proposed next stage fusion power plant, ARC, the major radius will be 3.3 m, almost half the size of ITER’s R = 6.2 m. CFS’s HTS magnets were made possible thanks to breakthroughs in production of YBa2Cu3O7 with uniformly distributed Y2O3 nanoparticles, creating a simpler reproducible microstructure. In 2021, SuperOx manufactured 300 km of 4-mm-wide YBCO tape in 9 months, making history as the largest completed order of 2G HTS tape [13].

In the UK, the private company Tokamak Energy is constructing their Demo4 magnet system, comprising of 14 TF and 2 PF coils with REBCO HTS at 20 K and expected to reach a field over 18 T. The system will demonstrate the interaction of the magnetic coils and test control and protection systems of the coils in a simple tokamak configuration, informing the design of their ST80-HTS spherical tokamak. At the UKAEA, the spherical tokamak for energy production (STEP) project is in its conceptual design phase, predicted to also utilize REBCO HTS to achieve the high-field requirements.

One interesting alternative to using either HTS or LTS is the proposal of hybrid HTS–LTS magnets for EU-DEMO and the China Fusion Engineering Test Reactor (CFETR). Using hybrid electromagnets maintains the high-field advantage of HTS while balancing their high cost with the comparatively cheaper LTS materials. For CFETR, Nb3Sn and NbTi strands are combined with Bi2212 and REBCO, with their placement dependent on which areas need a higher, more reliable magnetic field [14].

Renaissance Fusion, a private company based in France, has proposed direct deposition and patterning of HTS onto large cylinders to form the magnets needed for their stellarator. Multiple layers of HTS are deposited onto the cylinder to imitate multiple windings of coils. A laser then ablates the layers, forming grooves on the surface of the cylinder that act as boundaries for the current to follow. The complicated patterns will allow the HTS covering the cylinder surfaces to produce the unusual magnetic fields required for stellarators [15]. However, there is little public information on this process.

Although scientific breakeven (greater energy produced than required) was achieved by the National Ignition Facility in California in December 2022, breakeven with MCF has not yet been accomplished. Currently, there are MCF devices under construction which theoretically will produce more energy than they consume, but greater materials R&D, with emphasis on superconductivity, is required for the deployment of fusion power plants.

4 Areas of Future Research

Fortunately, the high-field requirements for MCF have bolstered research into superconducting magnets, with a large focus on commercialization of HTS tape. However, there are still obstacles to applied superconductivity for fusion. Ongoing research on superconducting magnets for fusion aims to address these main challenges:

  • Manufacturing difficulties and mechanical strength

  • Irradiation and long-term availability

  • Quench protection issues

  • Cost of superconductors and cryogenic cooling

One of the main obstacles to mass-producing superconducting magnets is its characteristic brittleness, a property unideal for forming flexible wires and tapes in lengths of kilometers required for fusion. The majority of high temperature superconductors are also ceramics, compared to the metallic low temperature superconductors. To combat this, HTS tapes are manufactured on a metallic substrate and stacked between layers, leading to better mechanical properties (see Fig. 4).

Fig. 4
figure 4

General stacking structure of YBCO HTS tape [16]

Full size image

Since heat treatment is required for Nb and Sn to react and form Nb3Sn, there is ongoing research on whether to wind the cable in a steel conduit first and then perform heat treatment (wind-and-react, WR) or to perform heat treatment prior to winding (react-and-wind, RW). ITER currently uses WR technique, aiming to reduce the close strain monitoring of the brittle tape as it is wound and handled. However, for EU-DEMO, the Swiss Plasma Center has proposed revisiting RW technique, indicating that the critical current of the superconductor is 30% higher than when treated with WR [17]. Since the wire jacket and welds are not exposed to heat treatment with RW method, there is also greater freedom in the jacket thickness and shape [14]. After performing a comparison, however, the Korean Institute for Fusion Energy argued that there are not clear advantages to RW technique [18], indicating there is further room for research in this topic.

Superconducting magnets are also subject to internal loads from the current and magnetic field during tokamak operation, causing local bending of strands and eventual filament breakage. The performance degradation of the superconducting tapes will require maintenance and replacement of components that already contribute significantly to the overall cost of the tokamak. Another important factor contributing to material degradation is the irradiation of superconducting components. One of the products of the D-T fusion reaction is 14.1 MeV fast neutrons which can escape the vacuum vessel and interact with surrounding materials, especially in compact tokamak systems. The neutron radiation can affect the superconductor by creating defects in the lattice, consequently lowering the critical temperature and inducing changes in the critical current density. Annealing processes may recover the critical temperature and current in the superconductor after irradiation and future research is dedicated to finding magnet treatment and shielding methods which ensure long-term availability [14]. Additionally, continuous research into irradiation-resistant materials and magnet designs is needed to prevent degradation and costly maintenance.

Quench is the sudden loss of the superconducting state, propagating heat throughout the magnet and bringing it to a normal discharged state. If either the field or the rate of change of the field is too large within the magnet, quenches can occur, endangering the entire magnet system. The design of quench protection systems includes two goals: to rapidly detect the quench and to immediately discharge the energy in the magnets. ITER currently tackles this by monitoring the resistive voltage of the coils, sending a signal if it begins to rise, opening switches connecting the coil to large resistors, and discharging the whole system in roughly 1.5 min. However, quench in HTS magnets displays a comparatively low propagation velocity and highly peaked temperature profile in the normal zone, making the voltage tap detection method insufficient when applied to HTS. Optical fibers can provide higher spatial resolution and more reliable quench detection when applied specifically to HTS, but the practicality of such quench detection systems when applied to fusion devices is still an area of investigation.

Currently, the superconducting magnet systems make up 30% of ITER’s overall cost [14]. Eventually, with demonstration power plants creating greater market demand for superconductors and increasing competition among superconductor suppliers, this cost will decrease. However, at this point, the industry is still lacking in readily available, cost-effective superconducting tape options. Further research into mass-manufacturing methods and novel superconducting materials would address this glaring obstacle to economically feasible fusion power plants.

5 Conclusions

Historically, breakthroughs in superconductivity have aided and spurred fusion research, cementing superconductors as an enabling technology for magnetic confinement fusion power plants. Today, applied superconductivity research is even more critical. By addressing current obstacles like manufacturing challenges, long-term availability, quench issues, and high cost superconductivity research will play an important role in the realization and deployment of fusion electricity production. Multiple solutions to existing issues for fusion magnets have been proposed, but greater effort in international coordination between researchers will allow for more efficient and focused development.