8 Reasons Why Silicon Carbide Diodes Outperform their Silicon Peers

Wide bandgap semiconductors are making possible a host of high-power applications previously unfeasible using silicon (Si). This blog compares the characteristics of both materials and shows why silicon carbide diodes (SiC) have a clear advantage across multiple metrics.

1. SiC diodes take up less space than Si for the same voltage rating

SiC has ~10x better dielectric breakdown field strength and, for a given blocking voltage, a thinner and a higher doped drift layer than silicon, giving it lower resistivity and better conduction performance. This means a SiC die can be smaller than its silicon equivalent for the same voltage rating. An extra benefit of having a smaller die size is lower device self-capacitance and lower associated charge for a given current and voltage rating. This, combined with the higher electron saturation velocity of SiC, enables faster switching speeds with lower losses than Si.

2. SiC diodes demonstrate better thermal performance

The thermal conductivity of SiC is nearly 3.5 times better than Si, allowing it to dissipate more power (heat) per unit area. While packaging can be a limiting factor during continuous operation, the significant extra margin offered by SiC brings added confidence in applications susceptible to transient thermal events. In addition, the ability to withstand higher temperatures means SiC diodes offer more rugged performance and better reliability without risking thermal runaway.

3. SiC diodes have magnitudes lower reverse recovery loss which drastically improves power converter efficiency

SiC diodes are unipolar Schottky metal-semiconductor devices in which conduction only takes via majority carriers (electrons). This means hardly any charge is stored in the junction depletion layer when a diode is forward-biased. By contrast, P-N junction silicon diodes are bipolar and store charges that must be removed during the transition to reverse bias. This causes a spike in reverse current which means higher power losses in the diode (and any associated switching transistors and snubbers), with the power losses worsening as the switching frequency increases. A SiC diode exhibits a reverse current spike due to the discharge of its self-capacitance under reverse bias, but this can be up to an order of magnitude lower than in a P-N junction diode, meaning less power is dissipated not only in the diode itself but also in the corresponding switching transistor.

4. Forward voltage drop and reverse leakage current of SiC diodes matches Si

The maximum forward voltage drop of SiC diodes is comparable with the best ultra-fast Si types and is constantly improving (with minor differences for higher blocking voltage ratings). Despite being a Schottky type, the reverse leakage current and resulting power dissipation in a high-voltage SiC diode under reverse bias are relatively low and similar to an ultra-fast Si diode of the same voltage and current class. Any slight difference in dissipation caused by variations in forward voltage drop and reverse leakage current between SiC diodes and ultra-fast Si diodes is more than offset by the improvement in dynamic losses with SiC due to the absence of reverse charge recovery effects.

5. SiC diode recovery current is stable across temperature, lowering power losses

A silicon diode's recovery current and time vary widely with temperature changes, making circuit optimization difficult, but significantly, with SiC, there is no such variation. In some circuits, like ‘hard-switched’ power factor correction stages, a silicon diode acting as the boost rectifier can dominate losses, going from forward bias at high current to reverse bias (typically around 400 V DC link voltage) for a typical single-phase AC input. The characteristics of SiC diodes can bring significant efficiency gains to this application and eases the design considerations for hardware designers.

6. SiC diodes can be connected in parallel without the danger of thermal runaway

An additional advantage of SiC diodes over Si is that they can be connected in parallel because their forward voltage drop has a positive temperature coefficient (in the application-relevant area of the I-V curve), which helps correct any current imbalances. By comparison, the negative temperature coefficient of Si P-N diodes can result in thermal runaway when devices are connected in parallel, heavy derating or additional active circuitry is required to force devices to current share. 

7. SiC diodes have better Electromagnetic Compatibility (EMI) performance than Si

A further benefit of SiC diodes’ soft switching behaviour is a significant reduction in EMI. When Si-diodes are used as switching rectifiers, the potentially snappy spike in reverse recovery current (with its broad frequency spectrum) can cause conducted and radiated emissions. These create system disturbances (via various coupling paths) that may result in system EMI limits being exceeded. At these frequencies, filtering can be complex due to this stray coupling. In addition, EMI filters designed to attenuate switching fundamental and low harmonic frequencies (often below 1 MHz) typically have high self-capacitance, making them less effective at higher frequencies. A snubber can be used across a fast recovery silicon diode to limit edge rates and damp oscillations, reducing stress on other components and lowering EMI. However, the snubber has to dissipate significant energy, thus reducing system efficiency.

8. SiC diodes have lower forward recovery power loss than Si

Forward recovery is an often-overlooked source of power loss in Si diodes. During the transition from off- to on-state, the diode voltage drop temporarily increases, producing overshoot, ringing and additional loss relating to the initial lower conductivity of the P-N junction. However, this effect is absent in SiC diodes, meaning forward recovery loss is not a concern.

To find out more about Nexperia’s Silicon Carbide Schottky Diodes portfolio and to discover all the products, head over to nexperia.com/sic.