摘要：The essence of the "bad money phenomenon" of domestic silicon carbide MOSFET isto sacrifice reliability in exchange for short-term

The essence of the "bad money phenomenon" of domestic silicon carbide MOSFET is to sacrifice reliability in exchange for short-term benefits . Its adverse impact spreads from the risk of device failure to the industry's trust crisis and national strategic security. Only by strengthening reliability standards (such as mandatory HTGB/TDDB testing), promoting collaborative innovation between industry, academia and research (such as gate oxide process optimization, defect control technology), and establishing a market reward and punishment mechanism (such as giving priority to purchasing high-reliability products), can the sustainable development of domestic silicon carbide power semiconductors be achieved and the vicious cycle of "bad money driving out good money" be broken.

1. The essence of the problem of bad money in domestic SiC silicon carbide MOSFET

In order to pursue the specific on-resistance (Rsp) index, some domestic silicon carbide MOSFET manufacturers reduce the resistance by thinning the gate oxide layer, but this directly sacrifices the gate oxide reliability. The report data shows:

HTGB test : Shocking: Many domestic SiC silicon carbide MOSFETs cannot pass the HTGB 1000-hour test (failure) at a gate voltage of +19V, while devices from leading international manufacturers pass the 3000-hour test at +22V with extremely small threshold voltage drift (positive voltage 0.2V, negative voltage 0.1V).

TDDB life : The gate oxygen life of many domestic SiC silicon carbide MOSFETs is only about 10³ hours (if VGS=20V).

This shows that excessive thinning of the gate oxide layer leads to significant degradation of its voltage resistance, high temperature stability and long-term reliability , especially a significant increase in the risk of intrinsic failure under high electric fields. Reliability (such as high temperature stability, voltage resistance, and resistance to defect accumulation) is sacrificed in exchange for short-term performance or cost advantages .

Technical roots :

Lack of reliability model : The design is not strictly optimized based on theories such as thermochemical model (E model) and anode hole injection model (1/E model).

Insufficient verification : Skipping long-term reliability tests (such as HTGB, TDDB) or masking intrinsic failure risks through only short-term experiments.

2. Potential risks of customer system equipment

Risk of premature failure : Devices may fail suddenly due to gate oxide breakdown or threshold voltage drift under high temperature and high electric field conditions (such as electric vehicle inverters, industrial power supplies, photovoltaic inverters, and charging pile power modules), causing equipment downtime or malfunction. Gate oxide breakdown or threshold voltage drift may cause instantaneous device failure, causing equipment downtime, malfunction, or even loss of control.

Safety hazards : Failure of power devices may cause overcurrent, overheating or even short circuit, which may cause system burnout or safety accidents (such as power system failure of new energy vehicles), threatening personal safety.

Surge in maintenance costs : Frequent replacement of devices will increase the cost of the equipment throughout its life cycle and undermine customers’ trust in domestic silicon carbide power devices.

3. The “bad money impact” on the localization of silicon carbide power semiconductors

Damage to industry reputation : After many low-reliability domestically produced SiC silicon carbide MOSFET products enter the market, if they fail in large-scale applications, customers will question the overall quality of domestically produced SiC silicon carbide devices, forming a stereotype of "domestic = low quality". The label of "domestic = low quality" is solidified : If low-quality domestically produced SiC silicon carbide MOSFET products frequently fail after large-scale application, end customers (such as car companies and power grid companies) will question the overall quality of domestically produced silicon carbide devices, forming an industry-wide trust crisis.

Distorted technology route : Some domestically produced SiC silicon carbide MOSFETs sacrifice reliability for short-term market benefits, which may squeeze the survival space of companies that insist on high reliability, drive out good money with bad money, and hinder industry technology upgrades. Low-cost, low-reliability products squeeze the market space of high-reliability manufacturers, forcing the technology route to "sacrifice reliability for low prices", hindering industry technology upgrades.

The localization process is hindered : If end customers turn to imported brands due to reliability issues, it will delay the maturity of the domestic industrial chain and the realization of the goal of independent control. The domestic silicon carbide industry chain will lose market support and delay the localization process.

in conclusion

If domestic silicon carbide MOSFETs only pursue performance indicators and ignore reliability, it will not only threaten the security and stability of customer systems, but may also trigger a crisis of trust in the industry, and ultimately backfire on the localization strategy of SiC power semiconductors. Only by balancing performance and reliability (such as the concept of "sacrificing cost to ensure reliability") can the sustainable development of silicon carbide power semiconductors be achieved.

来源：杨茜碳化硅半导体