PEM-CCD
In the process of IC failure analysis, EMMI (also known as PEM, Photon Emission Microscope) is a very basic and commonly used tool for locating failure points. Traditional EMMI uses cooled charge-coupled devices (C-CCDs) to detect photons, with a detection wavelength range between 400nm and 1100nm, which corresponds to visible light and infrared light.
When semiconductor devices generate an excess of electron-hole pairs, photons are produced due to the recombination of electron-hole pairs, or excess kinetic energy is released by thermal carriers in the device, which manifests as photons. These photons generated by both mechanisms can be detected by EMMI. Therefore, issues such as junction leakage, oxide breakdown, electrostatic discharge damage, latch-up effects, impact ionization, forward bias, and transistors operating in saturation regions can be precisely localized by EMMI, thereby inferring the defect locations in integrated circuits, which is greatly beneficial for subsequent circuit analysis and physical failure analysis.
-
defects that can trigger photon emission
- P-N junction leakage or breakdown
- transistors misoperation due to open or short circuits
- latch-up effects
- Gate oxide leakage
- Residual polysilicon from fine wires
- Damage to silicon substrates
- Burned componentsFalse defects (photon emission occurs during normal operation)
-
False defects (photon emission occurs during normal operation)
- Floating gate state
- BJT or MOSFET in saturation region operation
- Forward-biased diode
-
The situation of being unable to detect photons is trapped.
- The light excitation position is blocked by the upper layer.
- Embedded junction
- Leakage locations under large metal lines
- Resistance short circuits or bridging
- Metal connection shorts (sometimes still detectable by EMMI)
- Surface conduction paths
- Silicon conduction paths
- Leakage current is too small (<0.1uA)
- The light excitation position is blocked by the upper layer.
