Hans-Joachim Merrem

Modified polyhydroxystyrenes as matrix resins for dissolution inhibition type photoresists

It is generally accepted that the production of shrink versions of the 16 MB DRAM and the 64 MB DRAM generations will be patterned using deep UV radiation. This provides a new challenge to the photoresist suppliers, as the standard photoresist formulations are not suitable for this technology, mainly because the presently used novolak resins are highly opaque in the 200 - 300 nm region. This is especially true for the 248 nm wavelength of KrF eximer lasers. Poly 4- hydroxystyrene [PHS] has several advantages in transmission and thermal stability; however, its dissolution rate in commercial grade developers is unacceptably high. We report some recent results on modified, alkyl-substituted PHS derivatives. These polymers combine reduced alkaline solubiity with adequate optical and thermal properties, making them acceptable for future deep UV based production processes. Selected data of these new (co)polymers are discussed.

A novel two component positive photoresist for deep UV lithography

Abstract Multifunctional a - diazo - β - ketoesters (DKE) 6 provide an attractive source of deep UV sensitive photoactive compounds with a broad potential for modification of their chemical, physical and lithographic properties. In combination with highly transparent resins, they offer insights into the development of a positive working deep UV sensitive two - component photoresist.

Percolation view of novolak dissolution: 3. dissolution inhibition

The dissolution of novolak films in aqueous alkali is controlled by the diffusion of base through a thin penetration zone that forms at the interface between the developer solution and the solid. Base diffusion is a percolation process in which the ions of the base migrate through the zone by stepping from one hydrophilic site (phenol or phenolate) to the next. Dissolution inhibitors function by blocking some of the hydrophilic sites and thereby interrupting the diffusional pathways. Percolation theory suggests a relation between the strength of inhibition and the percolation characteristics of the resin. The two are linked together by the hydrophobic displacement volume of the inhibitor, which is that volume which the inhibitor occupies in the penetration zone. The hydrophobic displacement volume determines the effectiveness of an inhibitor; it depends not only on the molecular volume of the inhibitor, but also on the mobility of the hydrophilic sites in the zone; it is much smaller above the glass transition temperature of the zone than below it. It is also smaller in systems where some degree of motional freedom persists even below the glass transition of the zone.

Statistical experimental design on the optimization of high-performance photoresist

The ultimate photolithographic performance of a photoresist is essentially determined by the nature and relative concentration of the chemical components in the formulation and the process conditions selected. To meet the stringent performance requirements demanded by the advanced microlithography technology, each individual constituent of the photoresist process described must be optimized simultaneously. This study presents an effective and time saving approach using experimental design techniques to address the complicated multi-variable of resist formulation and process condition optimization. The responses for the design experiment in terms of lithographic performance were: exposure and focus latitude, photosensitivity and resist pattern thermal flow temperature are correlated with the resin dissolution characteristics, relative photosensitizer concentration, pre- and postexposure bake temperatures and development time.

Lithographic evaluation and characterization of a negative deep-UV resist system for the next generation of DRAMs

We report here the initial lithographic evaluation of AZ DN-21, a commercial, negative tone, aqueous alkali developable, chemically amplified resist. The resist was exposed with 248 nm light from a KrF laser on a Canon deep UV stepper with NA 0.37. Feature sizes down to 0.35 microns were printed with good focus and exposure latitude. The resist profiles are nearly vertical with a slight undercutting at the bottom of the feature and a slight rounding of the top. We also report some initial results from a study of the effects of delays in resist processing. For a given dose, delays in the processing increased the measured linewidths. Results from a calculation of the effective activation energy for crosslinking are also presented. For the PEB temperatures investigated, the effective energy was found to be a function of the PEB temperature.