Plasma-developed resist process

Figure 13. Schematic of plasma-developed resist film composed of polymer host (P) and volatile monomer (m). Processing steps are (a) exposure which locks monomer in place, (b) fixing which removes unlocked monomer, (c) plasma development. (Reproduced with permission from...

G.N. Taylor, L.E. Stillwagon, and T. Venkatesen, Gas phase functionalized plasma developed resists Initial concepts and results for electron beam exposure, J. Electrochem. Soc. 131, 1658 (1984) M.A. Hartney, R.R. Kunz, D.J. Ehrlich, and D.C. Shaver, Silylation processes for 193 nm excimer laser hthography, Proc. SPIE 1262, 119 (1990) M.A. Hartney and J.W. Thackeray, Sily lation processes for 193 nm lithography using acid catalyzed resists, Proc. SPIE 1672, 486 (1992) ... 

One important area of resist research in recent years is the development of plasma-developable resist systems. The aim of plasma developable resists is to use nonsolvent, all dry development methods to avoid the problems of swelling and consequent resolution limitation associated with conventional resists. Much of the semiconductor fabrication process now utilizes plasma techniques as they are capable of providing high resolution images. An important consideration in this is that the plasma-developable resist images should stand up well to the plasma etching treatments. 

In 1979, Smith and co-workers described the development of a system they called PDF (which presumably stands for Plasma Developable Photoresist) that is based on the use of a material, the structure of which has not yet been divulged 61). In this process the resist is coated in the usual fashion and exposed optically. The exposed film is then subjected to a baking cycle that produces a relief image of negative-tone that is, depressions are generated in unexposed areas (Figure 45). This relief structure is... 

Figure 44. A schematic representation of the plasma developed x-ray resist process. Exposure serves to covalenty bind the monomer (m) into the polymer matrix (p). Heating (fixing) drives out (volatilizes) the monomer except where it is "locked in place" by exposure. Plasma treatment converts the silicon to Si02 which retards the etch rate in the exposed areas through formation of a metallic oxide (MO) layer.

It has been found, however, that the etch rate of PBS can be reasonably controlled in both oxygen and CF4/O2 plasmas if the substrate temperature is kept below room temperature (9). This fact has been utilized to reduce the defect density in the manufacture of chrome photomasks by exposing the developed PBS pattern to a low-temperature oxygen plasma (descum) prior to wet-etching the chrome. We have now found that the plasma-etch resistance of PBS in a CF4/O2 plasma can be markedly enhanced at room temperature simply by exposing the resist to a short oxygen plasma pretreatment prior to exposure to the fluorinated plasma. This effect can be used in a variety of pattern transfer processes to controllably generate submicron features on wafers and masks. This paper examines the parameters associated with this effect, proposes a mechanism to account for the results and delineates some possible pattern transfer processes. 

Figure 3.35. Negative plasma-development process via selective silylation of a positive-negative resist.

The mass spectroscopic analysis of the gases formed in thermal development of the UV exposed poly(l-butene sulfone)/pyridine N-oxide revealed only 1-butene and S02 as the products, which indicated depolymerization of the polymer initiated by energy transfer form the sensitizer. These photosensitized poly(olefin sulfones) are not suitable for dry etching processes, and they are not reactive ion etching resistant. Resists made of poly(olefin sulfones) and novolac resins which will be described next are CF plasma etch resistant with reasonable photosensitivities. 

The swelling problem in the negative resist can be avoided if the differential solubility between exposed and unexposed area does not totally rely on the crosslinking of the exposed resist (4), or a dry development process, such as plasma development, is used. 

The development process converts the latent image in the polymer into the final 3-D relief image. This process is perhaps the most complex of resist technology. It can generally be achieved by either liquid development or dry (plasma) development. Numerous considerations are critical to either alternative. We will first focus on the wet development process. Plasma development will be discussed in a later section. 

Plasma development derives its advantage over its liquid counterpart mainly from the anisotropic nature of the process, except when loss from the unexposed area is vanishingly small, in which case the isotropy comes from that already in the latent image. The resist must be tough to avoid being eroded completely while the substrate underneath is etched. To improve the plasma etch resistance, aromatic compounds have been added to PMMA (50, 51). Hence, for... 

---