Process for reducing edge roughness in patterned photoresist

Details Process for reducing edge roughness in patterned photoresist Process for reducing edge roughness in patterned photoresist

2000-11-14

2003-06-24

Huff, Mark F. (Department: 1756)

Radiation imagery chemistry: process, composition, or product th

Imaging affecting physical property of radiation sensitive...

Including heating

C430S317000, C430S322000, C430S313000, C430S315000, C430S328000, C430S311000, C427S336000, C427S335000, C427S337000

Reexamination Certificate

active

06582891

ABSTRACT:

FIELD OF THE INVENTION
The present invention provides a process for reducing the roughness from the surfaces formed in photoresists after a relief image is patterned and developed in the photoresist. In particular, the process involves exposing the patterned photoresist to a vapor to reduce roughness from the exposed surfaces of the photoresists.
BACKGROUND OF THE INVENTION
In the course of manufacturing integrated circuits, a semiconductor wafer is typically coated with a photoresist layer. Generally, the photoresists used for this layer are organic photosensitive films used for the transfer of images and are typically exposed through a patterned photomask to a source of activating radiation such as ultraviolet light to form a latent image. The photomask has areas opaque and transparent to the activating radiation such that the photomask image is transferred to the photoresist layer. The latent image is then developed to yield a positive or negative relief image that is identical to or very similar to the photomask pattern. The photoresist layer may be positive or negative acting depending on the photoresist chemistry and developers chosen. The wafer is then further processed until the desired integrated circuit device is fabricated. The processing may be carried out according to a number of processing steps, including etching, doping and the like. Each module of the process typically requires the use of at least one patterned photoresist layer to transfer the desired properties and designs to the integrated circuit.
The relentless drive towards smaller sub-micron feature sizes in order that more and faster components may be included in a chip of a given size has created numerous technical challenges for microlithographic processing. The technical challenges include developing new materials and processes capable of accurately and reproducibly creating the desired small lithographic structures. Utilizing smaller wavelengths of radiation have made it possible to resolve smaller features. However, in order to transfer those small features into the integrated circuit it is necessary to make modifications to the photoresist chemistry.
Traditional I-line resists, based on novolac resins, have functioned well for critical dimension feature sizes down to 0.35 um. However, novolac resins are not suitable for use at the smaller wavelengths since these materials strongly absorb radiation from the wavelengths used below 365 nm. As such, in order to ensure the requisite optical transparency, new materials, such as poly(hydroxystyrene) and polyacrylates, are generally used as resins for the shorter wavelengths.
DUV photoresists resists are optimized for exposure to radiation having a wavelength at about 248 nm. DUV resists are generally based on a chemical amplification mechanism because it offers high contrast, good sensitivity and demonstrated ability to resolve patterns at 0.25 um and smaller. DUV photoresist formulations typically include a resin and a photoacid generator (PAG). Commonly used resins include modified polyvinyl phenol polymers or polyvinylphenol/acrylate copolymers in which the phenol or carboxylate groups are partially “blocked” or protected by moieties that can be chemically cleaved. As previously mentioned, these systems are typically based on an acid catalyzed mechanism wherein a strong acid is generated upon exposure to a photoacid generator compound present in the photoresist formulation to activating energy. The acid catalytically and chemically cleaves the protected group. As a result, a dissolution differential then exists between exposed (deprotected polymer) and unexposed (protected polymer) regions. It is this catalytic mechanism that is primarily responsible for the high sensitivity of these systems.
Photoresists optimized for exposure to radiation at 193 nm employ a variety of other protected polymers that necessarily contain few or no aromatic groups due to the strong absorption of 193 nm light to aromatic compounds. For example, some of these photoresists are based on acrylate polymers having various non-aromatic functional groups. Similar to DUV photoresists, these resists also utilize an amplification mechanism. The PAG chemistry of 193 nm resists is essentially the same as that of DUV resists.
At feature sizes below about 0.10 um it is anticipated that 157 nm wavelength light will be used for imaging. Photoimaging of these types of resists will likely require a vacuum since oxygen is known to strongly absorb at this wavelength. As such, a new generation of chemically amplified photoresist polymers will have to be designed to meet the challenging demands of lithography at this wavelength.
Inherent to all photoresists lithographically patterned is a phenomenon known as edge roughness, nanoedge roughness, or line edge roughness. Edge roughness refers to the irregularities or coarseness present on the surfaces of the photoresist after patterning and development. The greatest severity of edge roughness is typically observed on the sidewalls of the patterned photoresist and has a significant impact on the critical dimension budget. Edge roughness becomes increasingly important as one transitions to smaller feature sizes. For example, the critical dimension error budget targeted by lithographers for 100 nm lines is ±7 nm. If sidewall roughness for a 100 nm line causes variations of only 4 nm on either side of the targeted edge, then over half the critical dimension (CD) error budget has been used up. Edge roughness thus consumes an increasingly larger portion of the CD budget as the critical dimension of the feature shrinks. Accurate transfer of the roughness into the subsequently etched pattern will contribute to unacceptable variations in electrical performance.
There has been a great deal of effort directed towards understanding the root causes of edge roughness. The cause of edge roughness can be attributed to numerous factors. For example, edge roughness can be attributed to inadequate mask quality, poor resist performance, poor aerial image contrast near the limits of optical resolution, the plasma etch process and numerous other sources recognized by those skilled in the art. One particular class of edge roughness is associated with top surface imaging (TSI) processes based on selective silylation of the exposed resist film. Well-known problems with this approach include poor silylation contrast of the resists and low glass transition temperature, and hence higher mobility of the silylated polymer.
However, in most cases other than TSI systems, edge roughness can be attributed to characteristics of the resist itself or to the lithographic processing of the image, i.e., post-exposure bake and developing conditions.
Another causal factor contributing to edge roughness is excessive acid diffusion generated upon exposure of the photoacid generator to the activating radiation at the edges of the latent image pattern, which may cause partial deprotection of the resist polymer beyond the edges. This is the main reason that so much effort has been directed at PAG design and understanding the kinetics of the deprotection reaction. Post-exposure bake conditions which drive diffusion of the photogenerated acid, can therefore have a very significant effect on edge roughness.
Another cause of edge roughness is believed to be uneven dissolution of the resist during development. Developer process conditions have been found to have a large influence on edge roughness with different effects on isolated and dense features often resulting. Several researchers have attributed this type of edge roughness to the resist polymer composition and non-homogeneous regions formed during lithographic patterning and development. For example, researchers have determined that the formation of polymer clusters during exposure and post expose bake (PEB) leads to edge roughness due to the differing solubility of the clusters and non-clustered polymers (T. Ha et al., Proc. SPIE, vol. 3999, 66 (2000)). Similarly, research has indicated that for some resists, phase separation of pro

Affiliated with

Also associated with

Rating

Process for reducing edge roughness in patterned photoresist does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Process for reducing edge roughness in patterned photoresist, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for reducing edge roughness in patterned photoresist will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3123512