Light emitting diode fault detection using p-n junction photovoltaic effect

Area of Science:

• Semiconductor Physics
• Optoelectronics
• Manufacturing Technology

Background:

• Current methods for light-emitting diode (LED) chip fault detection often rely on contact-based electroluminescence, which can be time-consuming and potentially damage the chips.
• Ensuring the quality of LED chips during the packaging process is critical for device performance and reliability.
• Identifying defects such as chip quality issues, contamination, and bonding deficiencies early is essential for reducing manufacturing costs.

Purpose of the Study:

• To propose and validate a novel online, noncontact fault detection method for LED chips utilizing the photovoltaic effect.
• To develop a high-sensitivity instrument capable of detecting various faults in different types and sizes of LED chips.
• To demonstrate the correlation between photovoltaic behavior and LED electroluminescent characteristics for comprehensive quality assessment.

Main Methods:

• Illuminating bonded LED chips on a lead frame to generate a measurable photocurrent via the p-n junction's photovoltaic effect.
• Characterizing the weak photovoltaic response to identify five specific LED faults: chip defects (quality, contamination) and bonding deficiencies (disconnection, debonding, rebonding).
• Developing a specialized instrument with a tunable magnetic core, coil, and detection system to optimize photocurrent measurement for diverse LED chips (color, size 9-15 mil).

Main Results:

• The developed instrument successfully detected various LED faults, including chip defects and bonding issues, in different colored and sized LED chips.
• Experimental results confirmed a direct relationship between the photovoltaic behavior of LED p-n junctions and their electroluminescent characteristics.
• The method allows for the derivation of internal optoelectronic characteristics and external Ohmic contact performance by analyzing the photocurrent.

Conclusions:

• The proposed online noncontact fault detection method based on the photovoltaic effect is effective for identifying critical LED chip faults before packaging.
• The developed high-sensitivity instrument demonstrates the feasibility of this technique on an LED assembly line for diverse chip types.
• This photovoltaic-based approach offers a viable, noncontact alternative to traditional electroluminescence-based online fault detection methods, potentially improving efficiency and reducing damage.