Pushing photolithography to the limit

Francis Sedgemore, Wednesday 13 February 2008 at 15:19 UTC

Scientists at the National Institute of Standards and Technology (NIST) in Gaithersburg, US, have made the first direct measurements of the swelling and collapse of thin polymer films used in photolithography. The results of the experiments provide important new insight into the complex chemistry involved in the production of nanoscale electronic circuits.

In order to reach the industry goal of 22 nm transistor gates in microprocessors, we need to understand what happens at the interface between the exposed and unexposed photoresist in the lithographic process. Swelling of the resist can induce the polymer formulation to separate, resulting in a roughening of the edges which must be kept below 2 nm in size.

Schematic of the photolithography process shows the formation of a gradient extending from the photoresist material to be removed (centre), into the unexposed portions of the resist on the sides

Research leader Vivek Prabhu and his NIST colleagues found that at the edges of exposed areas the developer fluid can penetrate several nanometres into the unexposed resist. This interface region swells and remains swollen during the rinsing process, collapsing only when the surface is dried. The magnitude of the expansion, while only a couple of nanometres, is greater than the size of the molecules in the resist, and this can limit the ability of the photoresist to achieve the desired edge resolution.

“The balance between hydrophilic and hydrophobic interactions (like oil and water) forms a swollen, sponge-like layer only a few nanometres thick at the line edge,” says Prabhu. “Since every nanometre counts on the road to 22 nm, we envisage that these measurements, which clarify the nanoscale mechanisms of photoresist development, will inspire alternative photoresist architectures and developer strategies that push photolithography to the limit by attempting to control this residual swelling fraction.”

“Prabhu’s team has been doing an excellent job in the fundamental investigation of complex imaging processes of chemical amplification resists,” says IBM Research scientist Hiroshi Ito, who co-discovered chemically-amplified photoresists. “The measurements provide critical insights into line edge roughness and developer control as the feature dimensions shrink to below 50 nm.”

Further reading:

“Direct measurement of the spatial extent of the in situ developed latent image by neutron reflectivity”, Prabhu et al., J. Vacuum Sci. Tech. 25, 2514 (2007).

Figure: Schematic of the photolithography process shows the formation of a gradient extending from the photoresist material to be removed (centre), into the unexposed portions of the resist on the sides. NIST measurements document the residual swelling fraction caused by the developer that can contribute to roughness in the final developed image (source: NIST).

Article first published in Nanomaterials News.