the thickness of the metal layers in 0.25/0.18um process

[zhonghan]

Can someone tell the approximate value of the thickness of the metal layers in 0.25/0.18um process? I want to make a comparison between it and that in 2um bipolar process to evaluate the maximum current density that the metal can tolerate.

 

[GDF]

The metal thickness depends on the technology of foundry. But it's often
less than 0.9um unless ultra thick metal is adopted.

 

[zhonghan]

0.9um seems a little bit thick than my evaluation. if 1.5um of the metal layer in bipolar process is reasonable, than 1 um wide metal can allow max 2mA current to flow through it. It also seems a little bit small.

 

[Colbhaidh]

The first 5 layers are typically 0.8um and the top metal 1.2um (thicker top metal for bonding efficiency. These thicknesses include the barrier metals (Ti/TiN at the bottom oif the stack and TiN on the top). For 0.18um, where all the interlevel dielectrics between the metals are planarised using chemical mechanical polishing and tungsten plugs are used to connect between layers, the elcromigration performance of these backend processes often out perform older technologies. This is because the older processes did not planarise the inter metal dielectrics and so the thicker metal thinned over steps or sharp discontinuities especially at contacts or vias. So although the process sepcifies 2.0um metal, it might actually thin down below 0.5um in places.