Access Official Technical PDF

Share Summary:LinkedInEmail

Forming low-resistance 4H-SiC ohmic contacts is a persistent process challenge for power-device manufacturing. The material’s wide bandgap contributes to a substantial metal–semiconductor barrier, while effective p-type doping is difficult because aluminum and boron acceptors have relatively deep energy levels. A reported aluminum-film laser-doping approach addresses the contact problem by introducing a very high aluminum concentration near the surface of p-type 4H-SiC, without relying on the high-temperature activation anneal commonly associated with ion implantation.

Why p-type 4H-SiC contacts are difficult

Silicon carbide is used in power-device development because of properties that include high breakdown voltage, strong thermal conductivity and high electron saturation drift velocity. However, those material advantages do not eliminate the need for low-resistance electrical connections. In a metal–semiconductor contact, a potential barrier can impede carrier flow. For p-type 4H-SiC, the barrier and the relatively low activation of acceptor dopants make it challenging to achieve an ohmic contact with suitably low resistance.

One route to reducing contact resistance is to increase the semiconductor’s doping concentration close to the interface. At sufficiently high concentrations, carriers can tunnel through the narrowed barrier, rather than relying primarily on thermionic transport over it. Ion implantation is widely used to create doped regions, but implanted dopants typically require high-temperature annealing for activation. The source study identifies annealing near 1700°C as a potential process concern, including the risk of unwanted crystal changes and reduced yield. High-temperature processing can also be undesirable when forming contacts on a wafer’s backside after device structures are present on the front.

Aluminum-film laser doping for 4H-SiC ohmic contacts

The investigated method uses a thin aluminum film as the dopant source. Researchers deposited a 120 nm Al film on the carbon face (C-face) of a non-doped, semi-insulating 4H-SiC substrate. They then irradiated the coated surface with a 248 nm KrF excimer laser while scanning the sample. The work focused on the C-face because it corresponds to the backside of conventional MOS-gated devices.

The laser pulses were stretched using an optical pulse stretcher. This lengthened pulse was intended to suppress ablation and increase the time during which heat was retained at the surface. The reported irradiation conditions included a fluence of 2–4 J/cm² and one to 36 pulses. The sample was held at room temperature during irradiation, and argon gas was directed over the surface to limit reactions during processing.

After irradiation, the Al film was removed. The researchers also used CF4 and O2 plasma etching, followed by buffered hydrofluoric acid, to remove surface products and oxide. They repeated the cleaning sequence several times before depositing the contact metal. The resulting electrodes used a Ti/Al stack: 100 nm of titanium followed by 400 nm of aluminum. The study reports no additional post-deposition heat treatment for these contacts.

Heavy doping and measured contact performance

Secondary ion mass spectrometry (SIMS) confirmed that aluminum had entered the SiC. Under a reported condition of 3.4 J/cm² and four laser shots, the measured Al concentration reached as high as 5 × 10²¹ cm⁻³. The profile showed particularly heavy doping near the surface, where the contact interface is formed.

Electrical measurements also indicated a substantial difference between irradiated and non-irradiated regions. Current between electrodes on the untreated sample remained at only a few picoamps at the maximum applied voltage. With Al-film laser irradiation, current reached several milliamps, and the current–voltage response was linear. A linear I–V characteristic is consistent with ohmic behavior in the measured structure.

To quantify contact performance, the researchers used the transfer length method (TLM). This approach evaluates total resistance across electrode patterns with different spacings, allowing contact resistance and transfer length to be estimated. From those values, the study calculated a specific contact resistance of approximately 4.0 × 10⁻⁶ Ωcm².

The reported value was lower than the cited results for several ion-implanted p-type 4H-SiC contact structures. Those comparisons involved different contact stacks and doping conditions, so they should be understood in that context rather than as a universal ranking of contact processes. The key result is that the laser-doped structure achieved the stated resistance without the additional contact anneal used in many conventional approaches.

How doping concentration supports low resistance

The study’s theoretical analysis relates specific contact resistance to barrier height and acceptor concentration. As doping increases, the depletion region at the interface becomes narrower, making field-emission tunneling more significant. In this regime, contact resistance can fall rapidly with increasing doping concentration.

Using a barrier-height estimate based on titanium’s work function and the electron affinity of 4H-SiC, the analysis indicated that a concentration around 10²¹ cm⁻³ would be needed to approach the measured resistance. That estimate is consistent with the high near-surface aluminum concentration observed by SIMS. The calculation therefore supports the interpretation that heavy doping is central to the low-resistance result; it does not establish that the same performance will follow under every device geometry or process condition.

Practical implications for SiC process development

For engineers and process developers, Al-film laser doping is a potential alternative to implant-and-anneal approaches when a heavily doped surface region and reduced thermal processing are priorities. The reported experiment demonstrates room-temperature laser irradiation of the substrate and Ti/Al contact formation without additional heat treatment. It does not, by itself, establish manufacturing readiness or provide a complete process window for production.

Evaluation for a specific process would need to account for the targeted wafer face, laser conditions, surface cleaning, electrode geometry and the device’s thermal budget. The result also highlights the importance of verifying both the dopant depth profile and the electrical contact behavior: a high surface concentration alone is not a substitute for contact-resistance measurements.

Conclusion

Al-film laser doping offers a route to forming heavily doped p-type 4H-SiC at the surface while avoiding the high-temperature activation step associated with ion implantation. In the reported study, this approach enabled Ti/Al contacts with a specific contact resistance of about 4.0 × 10⁻⁶ Ωcm² and a linear I–V response, without additional heat treatment. For SiC power-device development, the practical significance is a demonstrated method worth evaluating where low-resistance backside contacts and reduced thermal exposure are important process objectives.

Access Official Technical PDF

About The Author