Contactless VLSI Measurement and Testing Techniques by Selahattin Sayil

Author:Selahattin Sayil
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

6.2 Charge-Density Probing

Electrical signals in an integrated circuit cause charge-density modulation within devices and parasitic p-n junctions. This modulation effect changes the local refractive index. Charge-density or the plasma-optical probing senses these local refractive index changes from the backside of an IC, which can be related to either a current or a voltage in the circuit through the charge-signal relationship of a specific device.

Because of its centrosymmetric crystal structure, silicon does not show the electro-optic effect or the Pockels effect [2]. In the Pockels effect, the optical properties of a crystal change according to an applied electric field applied across it. The effect commonly changes the refractive index of crystal. Because of the fact that silicon is a symmetric crystal having the same random arrangement of its atoms in every direction, silicon does not exhibit the Pockels effect and therefore cannot be probed using electro-optic sampling [9]. However, most semiconductor devices function by modulating charge density within a control region, which contributes to material index of refraction. Hence, the carrier density plasma-optical relation can be exploited for detecting electrical signals in silicon.

Since charge-density modulations occur in all semiconductor materials and it relies on free carriers, this technique is applicable to GaAs ICs as well or any semiconductor material [2, 9].

For silicon, the index of perturbation in the plasma-optical effect is isotropic. This means that the index of perturbation in silicon does not depend on the direction of light propagation in the silicon medium [10]. Therefore, one single beam cannot detect the free carrier index perturbation in silicon. However, by using two probe beams, it is possible to detect charge-density modulation in an active device [8].

Figure 6.4 shows the block diagram of such a detection system. The Nomarski prism divides the incoming polarized light into two beams, separating them by a small angle. The backside of the silicon IC should be polished before these two beams pass through. One of the beams, the probe beam goes through an active device, such as a transistor or a p-n junction. The reference beam is focused onto a region without an active device. Charge in the active device varies the probe beam’s optical path length. Finally, both beams are reflected back from the front surface metallization. The Nomarski prism then recombines them. Charge-density modulation in the active device changes the phase of the probe beam relative to the reference beam. “Because of the finite effective mass of free carriers the probe beam experiences a measurable phase delay as it propagates through active region” [8]. The phase delay changes the polarization of the combined return beam. The polarizing beam splitter senses the polarization modulation and converts it to an intensity modulation as in electro-optic sampling.

Fig. 6.4Block diagram for charge-density probing scheme

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.