Silicon carbide power devices are attractive for applications that demand high voltage, high temperature or fast switching. In space systems, those electrical advantages must be weighed against exposure to energetic ions that can cause device failure or lasting performance changes. NASA’s evaluations of SiC components highlight an important engineering point: passing a destructive-event test does not, by itself, establish that a device will remain reliable over its operating life.
For engineers and technical decision-makers, the challenge is to understand both catastrophic effects and less obvious degradation, then use that knowledge to guide component selection, qualification and radiation hardness assurance (RHA).
Why silicon carbide power devices draw interest
NASA technology roadmaps identify needs for compact, efficient power electronics, including high-voltage switches for power processing and electric propulsion, and integrated control and power electronics for motors and actuators. SiC is a wide-bandgap semiconductor technology being evaluated for these demanding applications. The source material notes that SiC power devices are already flying on space missions, while also identifying needs that existing silicon power devices do not meet, such as radiation-hardened, high-current MOSFETs above 250 V.
These potential benefits do not remove the need to assess performance in the intended environment. For space hardware, single-event effects (SEEs) caused by energetic particles are a key concern. Testing must establish how a device responds under relevant electrical bias and radiation conditions, not simply confirm its nominal operating characteristics.
Radiation-induced leakage complicates device assessment
One significant finding in the evaluations is that ion exposure can increase leakage current in SiC diodes and MOSFETs. In diodes, the change in reverse leakage current may not follow a simple, steadily accumulating pattern: degradation can saturate as ion fluence increases. The voltage at which the effect begins can also saturate quickly with linear energy transfer (LET). Reported observations found similar saturation onset voltages across Schottky diodes rated from 650 V to 1700 V.
That behavior matters because a test result at one fluence or voltage may not fully describe how a component will behave in service. In the cited work, leakage-current effects saturated before the high-flux iron region of the galactic cosmic-ray spectrum. The authors therefore cautioned that mission orbit may have less influence on this particular risk than a straightforward assumption might suggest. This does not make the effect negligible; it makes careful interpretation of test conditions especially important.
MOSFETs can exhibit several leakage paths, including drain-source, gate-drain and gate-source leakage. Drain-source leakage can resemble the reverse leakage behavior observed in diodes, while gate-related leakage adds further ways for irradiation to change device behavior. Some gate-source leakage may require a specific measurement approach to detect, and increased drain-source voltage can affect whether a device passes that test.
Why a safe operating area may be difficult to define
Single-event burnout (SEB) is a destructive failure mode, but leakage degradation can occur without immediate burnout. This creates a qualification dilemma: an apparent absence of catastrophic failure at a particular test fluence does not necessarily establish that the device is safe from SEB. The reported work describes leakage changes and nonlinear behavior occurring before a test reached a fluence of 104 xenon ions per square centimeter, and explicitly cautions against treating that exposure as sufficient to rule out burnout.
As a result, a single “safe operating area” should not be accepted without understanding how it was established. Radiation testing needs to account for the voltage conditions at which different responses occur, and should record changes in leakage as well as catastrophic events. A useful characterization dataset also shows how susceptibility varies with ion species and LET.
Radiation hardening involves trade-offs
Radiation-hardened-by-process (RHBP) design changes can reduce some risks, but the results described are not a complete solution. For a 1200 V MOSFET, a thicker epitaxial layer was associated with reduced SEB susceptibility, and reducing drain-neck width eliminated a particular gate-drain leakage degradation effect. However, other degradation effects changed little. The work also notes that latent gate damage may remain and that the long-term reliability impact of residual mechanisms needs to be established.
Hardening can also affect electrical performance. As with silicon power MOSFETs, design measures intended to improve radiation tolerance may involve performance costs. Component decisions therefore need to consider the balance between radiation response and the electrical requirements of the application, rather than treating radiation hardness as an isolated attribute.
Building a more informative test program
Radiation test planning should reflect the device’s intended operating conditions and the questions the qualification effort must answer. The NASA work points to several practical considerations:
- Test at relevant bias conditions. Record the voltage conditions associated with different leakage effects and failure responses.
- Include lower-LET exposures. Lighter-ion testing can reveal differences among device designs and support on-orbit degradation predictions, even when higher-LET responses have saturated.
- Specify LET and penetration range appropriately. The work recommends expressing LET in terms of LET(Si), consistent with space radiation environment requirements, while calculating penetration range for SiC.
- Evaluate application-specific temperature effects. Dopants may not be fully ionized at room temperature, and the effects of temperature on SEB susceptibility need to be established. The reported MOSFET leakage degradation rate increased with temperature.
- Assess reliability beyond the irradiation event. Determine whether non-catastrophic degradation persists and what it may mean for long-term operation.
These steps support more useful comparisons among parts and designs. They also help distinguish a device that avoids immediate catastrophic failure from one whose electrical behavior remains acceptable for the required service life.
Implications for power electronics engineering
The broader evaluation included related work on gate drivers and SiC integrated circuits. A gate-drive optocoupler was assessed across an extended temperature range and through thermal cycling, while NASA-developed SiC integrated circuits intended for Venus conditions were planned for total ionizing dose and SEE testing. These activities illustrate that qualification extends beyond the power switch itself: supporting components and integrated electronics also need evaluation against their own environmental and reliability requirements.
For engineers selecting SiC components, the practical lesson is to treat radiation response as a set of possible outcomes—not just a pass-or-fail burnout result. Measure leakage, examine bias and temperature dependence, consider residual damage, and document test conditions in enough detail to inform system-level risk assessment. Silicon carbide power devices can address demanding power-electronics needs, but their suitability depends on understanding how the specific device behaves in its intended environment.