Proceedings Article | 5 April 2021
KEYWORDS: Opacity, Molecules, Line edge roughness, Stochastic processes, Photoresist materials, Lithography, Extreme ultraviolet lithography, Critical dimension metrology, Chemistry, Chemically amplified resists
Whilst traditional chemically amplified resists (CAR) support the initial insertion of EUV lithography, a wide range of other resist materials are being examined for future nodes, aiming to identify a photoresist that simultaneously meets RLS and defectivity requirements. It is becoming increasingly clear that this should involve a novel mechanism—a new chemistry that can be tuned to allow for improvement of the RLS requirements. One potential approach is the multi-trigger concept wherein a reaction will only occur when multiple elements of the resist are initiated concurrently and in close spatial proximity. At the centre of exposed features, where the exposure dose is sufficient the resist reaction is thus catalytic as in a CAR, but at the edge of the features the reaction is secondorder in nature, and thus the chemical gradient is increased. In effect the resist features an intrinsic, inversely dose dependent, quenching of the catalysis, enhancing chemical contrast and thus resolution, and reducing roughness, whilst eliminating the materials stochastics impact of a separate quencher. The multi-trigger material previously presented consists of a base molecule and a crosslinker, which represent the resist matrix, together with a photoacid generator (PAG). Research has been undertaken to improve this resist, in particular focusing on improving resist opacity and crosslinking density. Our work on high-Z cross-linker molecules mark I and mark II has previously been reported and LER figures below 3 nm for lines and spaces patterned at 14 nm half pitch using the high opacity MTR resist on the EUV-IL exposure tool at PSI were shown. Here we present results from further work focused on the enhancement of the high-opacity MTR resist. A new high- Z crosslinker molecule, mark III, has been synthesized and introduced in the MTR resist to make the high opacity MTR compatible with the ethyl lactate and PGMEA casting solvents. We report results obtained using the new MTR system containing the high-Z cross-linker mark III, with a variation of process conditions and formulation variations. The lithographic performance, of a formulation containing this crosslinker, at pitch 32nm patterned on an NXE3350 is presented A biased LWR of 4.2 nm for a line width of 15.1 nm is shown. Introducing a PEB induces performance changes for the MTR4L3Y(2) resist. The sensitivity improves by over 20% with 80 °C PEB. However, the PEB does lead to a 12% increase of the LWR. Overall, the lowest Z factor (using biased LWR) occurs with a 60 °C PEB temperature. The Z factor is also significantly lower with a film thickness of 22.5 nm compared to 20 nm. The thickest film thickness tested using the NXE3350 is 22.5 nm. However, at PSI, 12 nm lines on a 28 nm pitch were patterned with an LWR of 2.07 nm using a film thickness of 25 nm. In addition to varying the opacity of the resist, we have also investigated increasing the activation energy of the selfquenching aspect of the MTR system. In this case, MTR8 has a higher activation energy than MTR4. Having a higher activation energy should allow introduction of PEB to increase crosslinking and reduce pattern collapse, whilst simultaneously preserving the self-quenching behaviour. We present results which show a decrease in dose and Zfactor using MTR8 at this formulation ratio compared to MTR4, when tested at PSI. The results also show a Z factor improvement when using a 60 °C PEB. A standard opacity multi trigger resist was patterned on the MET5 at the Lawrence Berkeley National Laboratory, and the effect of different development processes was studied using 1:1 dense line features at pitch 32nm. Reducing the development time in nBA had an adverse effect on pattern collapse and LWR. However, changing to on track development process using 2-heptanone gives a 10% LWR improvement at the 25 nm film thickness studied. Additionally, a pillar pattern was studied when using a film thickness of 28 nm. Here we present pillars with a LCDU of 1.85 nm with a CD of 21.4 nm patterned at a 40 nm pitch.
