Molecular Scale E-Beam Resist Development Simulation for Pattern Fluctuation Analysis

Abstract

To better understand nanometer scale pattern fluctuation in lithography, we present a new model for electron-beam exposed (poly)methyl-methacrylate resist development, which considers the molecular scale material structure. The polymer chains in the resist are represented as spheres with a radius proportional to the square root of the chain length. Monte Carlo electron scattering simulation in the resist gives the local absorbed energy, which in turn sets the local molecular weight distribution. The development model uses the Poisson probability for removing each surface polymer chain in a small time step. The Poisson removal rate is derived using a mass equivalence with macroscopic etch rate models. The simulations predict the macroscopic shape of patterned resisi features as well as microscopic surface roughness and nanometer scale line-edge variation. Observations with atomic force microscopy confirm the simulation results, including 10 nm edge variation and rough developed surfaces at exposures doses near the transition from low etch rate to high etch rate.

Surface and line-edge roughness in solution and plasma developed negative tone resists: Experiment and simulation

“…Because of its importance in the last few years, several publications discuss the problem of roughness. [1][2][3][4][5][6][7][8][9] Surface roughness ͑SR͒ appears after the spin coating of the photoresist film, while line edge roughness ͑LER͒ appears after the development of the microstructure. However, the two phenomena are closely related.…”

Effect of developer molecular size on roughness of dissolution front in electron-beam resist

“…The LER of nonchemically amplified resists has been studied with regard to molecular weight, polydispersity, and the structure of base polymers, 1,12 and by simulations of the development process. 13 Though the relationship between these factors and LER were demonstrated, the origin of LER has not yet been definitely clarified. It is thought that LER is generated through inhomogeneous dissolution behavior at the nanometer scale.…”

Linewidth fluctuations caused by polymer aggregates in resist films.

“…They are mainly evaluations of the linewidth fluctuation of resist patterns in connection with the resist molecular size and its molecular weight distribution [3,4]. Recently, we found that the linewidth fluctuation is caused by granular structures in the resist films [5].…”