Hydrides: the nemesis of high-quality SRF cavities?

Hydrogen has long been known as a possible enemy of superconducting radiofrequency (SRF) cavities – like those needed for the ILC – because of its potential to form non-superconducting hydrides that limit cavity quality factor (Q) and gradient. Researchers at Fermilab have made further progress in understanding the full physics behind hydrogen involvement, which is an important step towards improvements in cavity processing.

Hydrogen in the walls of niobium SRF cavities can bond to the niobium, forming compounds called hydrides that cause a known quality-factor limiting – and menacing-sounding – condition: hydrogen Q disease. While baking the cavities at the right temperatures was successfully used to “cure” the Q disease, scientists have never understood the full details of hydride formation. Over the past two years, Fermilab researchers Alex Romanenko and Fedor Barkov have developed a technique for directly observing how hydrides form. This technique allowed them to model another hydride-related limitation called high field Q slope, and simultaneous experiments shed light on how this condition is “cured” by a different bake.

“Fundamental understanding is the easiest and most natural path to improved performance,” said Romanenko, who is part of Fermilab’s Superconducting Materials Department and led this research. “As soon as you understand the physics behind the problem, then you can develop strategies on how to overcome it.”

ILC cavities require high Qs in order to efficiently reach the high gradient required by the ILC. Gradient refers to the energy transferred to a particle over a particular distance. Q, or the unloaded quality factor of a cavity, is related to the cavity’s power dissipation: the higher the Q of an SRF cavity, the lower the power required for the cavity to achieve a certain gradient. Therefore, higher Q cavities lead to more compact and cost-effective accelerators. ILC cavities must have a Q of at least 8 billion and a gradient of 31.5 megavolts per metre.

But hydrogen that enters a cavity’s niobium wall during cavity processing can limit the Q if allowed to clump together to form hydrides when the cavity is cooled to its cryogenic operating temperature.

Hydrogen Q disease refers to a large decrease in Q and gradient that occurs when large hydrides form on the inner surface of an SRF cavity. The decrease occurs because the hydrides aren’t superconducting like the rest of the cavity and cause extra surface resistance. Scientists learned in the past that baking the cavities at 600-800 °C for several hours can cure and prevent the disease; in fact, this bake is part of the ILC cavity processing procedure. However, researchers had never actually observed the hydrides that cause Q disease in cavity niobium, meaning they knew little about their structure and formation. In 2012, Romanenko and Barkov’s direct observation of hydrides literally shed light on the process, capturing images of the formation of hydrides in real time using a laser confocal scanning microscope coupled with a cryostage.

“We directly observed what was known to cause a limitation but was never actually seen,” Romanenko said.

But hydrogen Q disease may not be the only cavity problem caused by hydrides. Even after a cavity undergoes a 600-800 °C bake to prevent Q disease, hydrogen still exists in the 100-nanometre layer closest to the inner surface of the cavity. Romanenko’s group proposed a mechanism linking another condition called high field Q slope to the formation of the smaller hydrides. These hydrides only sustain superconductivity up to a certain magnetic field level, roughly 100 milliteslas. When the hydrides lose their superconductivity, they can limit cavities’ gradient and Q, just like they do in hydrogen Q disease. In 2013, Romanenko and his team published a model showing how the hydrides may cause this quality-limiting condition.

Like Q disease, high field Q slope can also be removed by a bake, this time at 120 °C for 48 hours. This bake is also a part of the ILC cavity processing procedure. Though researchers have known the bake works, they haven’t understood how it works. Using a technique called positron-annihilation spectroscopy, Romanenko and his team collaborated with the University of Bath, UK, and the University of Western Ontario, Canada, to discover that prior to the bake, hydrogen atoms are located in between the cavity’s niobium atoms and are free to move and form hydrides upon cooldown. During the bake, vacancies, or missing niobium atoms in the niobium crystal structure, form near the surface layer. These vacancies bind hydrogen so that it can’t form the small hydrides that may cause the high-field Q slope. The group published a paper last month detailing the findings.

According to Romanenko, understanding the mechanism of the 120 °C bake can help researchers think of methods for preventing Q disease or high field Q slope that might have additional benefits to cavity performance.

“This research really adds to our basic understanding of high-field Q slope,” said Camille Ginsburg, deputy department head of Fermilab’s SRF Development Department and Fermilab SRF cavity coordinator, in regard to Romanenko’s hydride research. “If we have that basic understanding, then we can more efficiently plan our cavity processing, which may reduce costs.”

Despite all the new information about hydrides and their role, researchers’ search to understand them is not over. According to Romanenko, the next step is to visualise the small hydrides at cryogenic temperatures, just as they did with the large hydrides.

“Understanding the underlying physics of SRF materials for accelerator applications is a must if we want to enable further technological advances, such as even higher gradient cavities for ILC,” said Slava Yakovlev, head of Fermilab’s SRF Development Department. “This is yet another demonstration of how important the basic SRF R&D is.”