“Over decades, both military and space programs all around the world have known the negative impact of radiation on semiconductor-based electronics,” says Meyya Meyyappan, Chief Scientist for Exploration Technology at the Center for Nanotechnology, at NASA’s Ames Research Center. What has changed with the push towards nanoscale feature sizes is that terrestrial levels of radiation can now also cause problems that had previously primarily concerned applications in space and defence. Packaging contaminants can cause alpha radiation that create rogue electron-hole pairs, and even the ambient terrestrial neutron flux at sea level – around 20 cm−2 h−1 – can have adverse implications for nanoscale devices.
Fortunately work to produce radiation-hardy electronics has been underway for some time at NASA, where space mission electronics are particularly prone to radiation exposure and cumbersome radiation shielding comes with a particularly costly load penalty. Vacuum electronics systems, the precursors to today’s silicon world, are actually immune to radiation damage. Alongside Jin-Woo Han and colleagues Myeong-Lok Seol, Dong-Il Moon and Gary Hunter at Ames and NASA’s Glenn Research Centre, Meyyappan has been working towards a renaissance of the old technology with a nano makeover.
In a recent Nature Electronics article, they report how with device structure innovations and a new material platform they can demonstrate nanoscale vacuum channel transistors that compete with solid-state system responses while proving impervious to radiation exposure.
Standing up for better drive currents
The first nanoscale vacuum channel transistors (NVCTs) followed a traditional architecture with the emitter and collector in the plane of the substrate but with a vacuum channel between the two instead of a semiconductor. As well as radiation immunity, the vacuum channel offers additional advantages in terms of the electron velocity. However, while the 10 μA at 2 V for a channel 50 nm long demonstrated an impressive proof of concept, the drive current did not match up to metal oxide semiconductor field effect transistors (MOSFETs) of the same size.
In the vacuum channel devices, the source pad is shaped into a 10 nm radius tip, which as Meyyappan described it proved “really a waste of the source pad real estate”. Instead the researchers flipped the structure on its side and redesigned the device for vertical emitters. “The drive current increases with the number of emitters fabricated on the source pad and thus, can be used as a design parameter,” says Meyyappan.
The next compromise to tackle was balancing the thickness of the oxide layer separating source and gate. Avoiding direct tunnelling leakage current requires a thick oxide layer, but a thinner layer allows better gate controllability. By folding over the gate, the researchers designed their NVCT to have a thick oxide layer in the field region to prevent parasitic capacitance and a thin layer at the emitter edge for optimized control.
The final masterstroke in optimizing their NVCT was to deviate from the standard silicon substrate. Although silicon is the go to material for today’s electronics, it turns out that silicon carbide has a higher bonding energy, as well as an exponentially increasing emission current with decreasing electrode gap and no current saturation or quantum screening, suggesting it may provide a better substrate for radiation-resilient NVCTs. The researchers produced NVCTs on both silicon and silicon carbide substrates and were able to demonstrate the superior radiation robustness with the carbide.
“SiC is not the first material anyone thinks of,” Meyyappan tells Physics World. “The first attempt is always using silicon. Then come materials like SiC, GaN, diamond etc. It is also possible that 2D material enthusiasts could attempt graphene and others as well.”
Back to the future of vacuum technology
The researchers demonstrate the wafer-scale manufacturability of their devices on a 150 mm SiC wafer. In the near term these NVCTs could provide complementary devices to conventional solid-state electronics for specific applications where voltage and breakdown field endurance, reliability and noise immunity requirements exceed what solid-state technology can achieve. However, for vacuum nanoelectronics to permeate the vast proliferation of electronics in everyday life, will require a significant amount of work to enable an industry that has been geared towards silicon for decades to embrace the fabrication, design and support of an alternative technology.
Despite the hurdles Meyyappan is hopeful of overcoming them. “There seems to be a resurgence of vacuum electronics, as witnessed by papers appearing in the literature and funding calls,” he says. The next step is developing full circuits based on NVCTs.