A common technique used for multiple patterning is double patterning. MOSFET’s are fabricated using 193nm wavelength light in a method known as optical lithography. At 40nm half-pitch, 193nm wavelength single-exposure lithography achieved its physical limit. As we progress to lower technological nodes, i.e. channel lengths less than 30nm, the process’s accuracy may suffer. Because the features are too small in comparison to the wavelength of light. This is where the concept of double patterning in VLSI is useful. Chipmakers can image IC designs at 20nm and smaller using multiple patterning.
What exactly is double patterning in VLSI?
Double patterning is a common multiple patterning technique. At advanced process nodes, the lithographic process uses the double patterning technique to specify the characteristics of integrated circuits. By utilizing typical optical lithography methods, it will allow designers to manufacture chips in sub-nanometer process nodes. Double patterning often refers to the litho-etch-litho-etch (LELE) pitch-splitting process in the fab. Self-aligned double patterning (SADP) is a spacer method that is part of double patterning.
How does double patterning in VLSI work?
Double patterning necessitates novel layouts, physical verification, and debugging for the designer. On the design side, for example, the mask layers are allocated with colors depending on spacing requirements. The mask layers are divided from the originally created layout into two new layers. Foundries use a variety of double patterning design processes at 20nm. One of the most popular flows does not necessitate the design team to decompose their layers into two colors. However, there are several circumstances in which the designer might need to be aware of the colour schemes. Even though it makes sense, seeing the double patterning colours will probably degrade debug productivity.
Also read, What is FGPA in VLSI?
Why do we need double patterning in VLSI design?
Double patterning mitigates the impact of diffraction in optical lithography. These diffraction patterns make it difficult to produce accurately defined deep sub-micron patterns using existing illumination sources and traditional masks. As the diffraction problem became more acute with each successive process node. Several novel reticle enhancement techniques were introduced to counter.
In the 180nm process node, phase-shift masks were introduced. They change the phase of the light as it passes through certain sections of the mask. This alters how it is diffracted and decreases the defocusing effect of mask dimensions smaller than the wavelength of the illuminating light. The disadvantage of employing phase-shift methods is that masks are more complex and costly to create.
Optical-proximity correction (OPC) approaches determine how to distort patterns on a mask to compensate for diffraction effects. The approach includes layout constraints, has a computational cost in design, and implies that making the corrected masks takes longer and costs more.
Conclusion
If you want to know more about the semiconductor industry or are interested in making a career in VLSI, Chipedge is the right place for you. It is the best VLSI training institute in Bangalore that offers a variety of VLSI courses online including Physical Design, DFT courses, Design verification courses, RTL courses, etc. Enroll yourself now.
Explore our VLSI courses online here.
