Line pattern collapse (LPC) becomes a critical concern as integrated circuit fabrication continues to advance towards
the 22 nm node and below. Tokyo Electron Limited (TEL) has been investigating LPC mitigation methods for many
years <sup></sup>. These mitigation methods include surfactant rinses to help reduce surface tension and Laplace pressures
forces that accompany traditional DIW rinses. However, the ability to explore LPC mitigation techniques at EUV
dimensions is experimentally limited by the cost and availability of EUV exposures. With this in mind, TEL has
adopted a combined experimental and simulation approach to further explore LPC mitigation methods.
Several analytical models have been proposed <sup>[2, 3, 4]</sup> for a LPC simulation approach. However, the analytical models
based on Euler beam theory are limited in the complexity of profile and material assumptions. Euler beam based
models are also now questionable because they are outside the beam theory's intended aspect ratio regime <sup></sup>. The
authors explore the use of finite element models in addition to Euler beam theory based models to understand resist
collapse under typical EUV patterning conditions. The versatility of current finite element techniques allows for
exploration of resist material property effects, profile and geometry effects, surface versus bulk modulus effects, and
rinse and surfactant rinse effects. This paper will discuss pattern-collapse trends and offers critical learning from this
simulation approach combined with experimental results from an EUV exposure system and TEL CLEAN TRACK
ACT<sup>TM</sup> 12 platform, utilizing state of the art collapse mitigation methods.
In order to quickly and cheaply test candidate fluids and coatings for immersion lithography, we have devised a fluid handling scheme that we call drag-a-drop. We have constructed a prototype tool in order to test materials using this fluid scheme, and conducted several experiments with it. From these tests, we have determined that a hydrophobic topcoat with low contact angle hysteresis on the substrate increases the maximum stable scanning velocity by at least a factor of 2 over a standard 193 nm photoresist. We observed that instabilities on the receding contact line are unaffected by height, but the onset of instability on the advancing contact line occurs when the height of the lens is low. We also examined the drag-a-drop technique for possible use in laser mask writing, and found that by means of a hydrophobic topcoat, the lens can be completely removed from the substrate while keeping the immersion droplet affixed to the lens.