Cure-induced photoresist distortions in double patterning

Abstract

Many processes are under evaluation as simplifications to current double patterning methods. Reduction in process complexity and cost may be achieved by use of track-based photoresist stabilization methods that eliminate one etch step by allowing a second resist to be patterned over a first resist. Examples of stabilization methods using numerous curing processes have been reported. At least some resist shrinkage during stabilization appears to be generally observed for these methods. We evaluate the link between shrinkage and three-dimensional pattern distortions at line ends and elbow corners using experimental and simulation-based methods. A 172-nm UV resist curing process was used to produce controlled shrinkage ranging from 5% to 30%: shrinkage was correlated with resist distortions. At cure dose sufficient to stabilize the resist, shrinkage of approximately 23% results in measured line-end pullback and elbow displacement of approximately 16% and 13% of nominal linewidth respectively, when measured at resist half-height. Finite element analysis of resist beam structures produces shrinkage distortions that are in good qualitative and semiquantitative agreement with these measurements and thus appears to provide a provisionally general and useful method for predicting pattern distortions that arise during cure-based resist stabilization methods used in double imaging.

Correlation of experimental and simulated cure-induced photoresist distortions in double patterning

“…Increasing efforts in this area center on double imaging processes in which two or more independent lithographic patterns are generated in sequentially applied resist films, then transferred in a single etch step. [1][2][3][4][5][6][7][8][9][10][11][12][13][14] In previous reports, 15,16 we demonstrated that 172 nm cure stabilization, when properly implemented, can result in superior retention of photoresist profiles compared to broadband UV curing (Fig. 1).…”

“…19 Details of the simulation approach have been described in a previous report. 16 Fig. 2a illustrates the experimental correlation between 172 nm cure dose and volumetric shrinkage as measured using offsets between dose-linearity curves of uncured and cured resist patterns.…”

“…15,16,20 Simpler automated methods are desirable for process development, optimization, and monitoring.…”

Photoresist stabilization for double patterning using 172 nm photoresist curing

“…13,15,16,18 To assess shrinkage in XP-7600A, we have employed the exposure latitude curve offset approach described elsewhere. [14][15][16] This method allows estimation of shrinkage for all CDs within an exposure latitude range of interest for a given line-space target. Results for XP-7600A are summarized in Fig.…”

“…A parallel set of studies developed fundamental linkages between shrinkage and distortion, and highlighted the need for curable resist materials with very low cure shrinkage. 15,16 Here, we show that use of resist materials designed for cure can greatly improve the lithographic performance of the 172 nm based process. We have developed new curable resists with low shrinkage and found that these materials are readily capable of patterning cross-grid contact holes at lithographic k 1 = 0.28, well beyond the practical k 1 limit for singleexposure contact patterning.…”