No abstract

The successful implementation of extreme ultraviolet lithography ͑EUVL͒ requires the use of an electrostatic chuck to both support and flatten the mask during scanning exposure. The EUVL Mask and Chucking Standards, SEMI P37 and P40, specify the nonflatness of the mask frontside and backside, as well as the chucking surface, to be on the order of 50 nm peak-to-valley. Thus, characterizing and predicting the capability of the electrostatic chuck to reduce mask nonflatness to meet this specification are critical issues. Details describing the performance of the Coulomb electrostatic chuck have been presented in earlier publications. In this paper, the governing equation identifying the force-gap relationship for a Johnsen-Rahbek ͑J-R͒ chuck is described and compared to the Coulomb response. Using finite element techniques, numerical models of Coulomb and J-R electrostatic chucks have been constructed and evaluated for their clamping performance. The models include the effects of reticle and chuck nonflatness, surface friction, and the finite stiffness of the chuck. Modeling predictions are presented for the two types of chucks. The simulations indicate that using a reticle and chuck ͑Coulomb or J-R style͒ that meet the SEMI standards for flatness can result in a clamped reticle flatness of less than 100 nm. However, there may be a need to increase the chuck stiffness specified in SEMI P40. These results, which provide the first comprehensive comparison of Coulomb and J-R chucks, are currently being used to establish specifications for chuck geometry and to identify the range of flatness variations that can be accommodated with electrostatic chucking.

The purpose of this research is to assess the effectiveness of electrostatic chucks in reducing low-spatial frequency mask ͑or reticle͒ flatness variations and to validate finite element ͑FE͒ models of the chuck-mask interaction. The flatness of a sample extreme ultraviolet lithography reticle and an electrostatic pin chuck were measured using a Zygo interferometer. The measured flatness data were entered into the FE models, and electrostatic chucking was simulated by applying an area-weighted average pressure on the reticle. The shape of the mask when clamped by the electrostatic chuck was then predicted using the FE model. To validate these predictions, experiments were conducted in which the previously measured reticle was electrostatically clamped using the pin chuck. These experiments were conducted in a vacuum chamber to minimize the effects of humidity. Interferometric plots of the chucked reticle surface were obtained and compared with the FE predictions. It was found that the measured and predicted shapes were in good agreement. Results from the experiments and simulations will be used to establish specifications for the chuck design and to identify the range of flatness variations that can be accommodated with electrostatic chucking.

No abstract

No abstract