Ernst Haugeneder

Influence of Silicon on Insulator Wafer Stress Properties on Placement Accuracy of Stencil Masks.

The issue of placement control is one of the key challenges of stencil mask technology. A high placement accuracy can only be achieved with a precise control of mechanical stress on a global and local scale. For this reason, the stress properties of the mask blank material -typically silicon on insulator (SOI) wafers- have to be known and adjusted properly. A systematic investigation of initial stress properties of various SOI materials is presented. The study covers bonded and non-bonded wafers and demonstrates a global membrane stress accuracy of 0.1 MPa. Initial SOI-stress properties on a global and local scale are discussed. With the precise control of layer stress and pattern correction methods based on finite-element calculations, a placement accuracy of 12 nm (3σ) is achieved. A sample-to-sample error of 12 nm (3σ) indicates the high stress-uniformity of non-bonded material.

PN and SOI wafer flow process for stencil mask fabrication

Two process flows for the fabrication of stencil masks have been developed. The PN Wafer Flow- and the SOI Wafer Flow Process. Membranes and stencil masks out of different 6 inch Si base wafers with 3 micrometers membrane thickness and a membrane diameter between 120 mm and 126 mm were fabricated. The membrane stress depending on the material property and doping level has been determined. First metrology measurements have been carried out.

IPL stencil mask distortions: experimental and theoretical analysis

Ion Projection Lithography (IPL) requires stencil masks. These masks are manufactured in a SOI wafer flow process. This means that e-beam patterning and the pattern transfer in silicon is done on the bulk mask-wafer blank before the membrane is formed. The last steps are deposition of a protective carbonic layer and removal of carbon from the stencil openings by etching. The internal stress control of the finally remaining silicon and carbon layers is decisive for the pattern placement accuracy of the stencil mask. The surface geometry and pattern placement are measured with a LEICA LMS IPRO system at different process steps. The initial bow and warp of the SOI mask-wafer blank is measured. Then, the pattern placement is measured after e-beam writing. After membrane formation the samples are measured a third time followed by a final measurement after carbon layer deposition and etch. These results are to be compared with FE (Fenite Elements) modeling calculations. Compared to previous investigations the effect of wafer warp will be included. Furthermore, LMS IPRO measurements will be done with improved tool accuracy on stencil mask membranes as achieved recently. Thus, the claimed functional dependence between stress and pattern distortion is to be verified experimentally.

Layout postprocessing in ion projection lithography (IPL)

Several post-optical lithography technologies are under development. Ion proj ecti on lithography (IPL) is one choice. For the industrial usage of the ion projection lithography a powerful software tool for layout post processing is necessary to achieve the required (TI) uniformity and pattern fidelity. The application of stencil masks in the IPL exposure step requires the concept of complementary masks. The IPL software has to provide the mask pattern split and the complementary mask layout as well as a mask pattern pre-distortion in order to compensate ion optical and membrane stress induced pattern distortions. The IPL specific pattern transfer process shows fig. I. The goal of this figure is to achieve a correspondence between chip layout on design level and wafer layout.

Measures to achieve 20nm IPL stencil mask distortion

From detailed comparisons of stencil mask distortion measurements with Finite Element (FE) analyses the parameters of influence are well known. Most of them are under control of the mask manufacturer, such as the membrane stress level and the etching process. By means of FE analysis the different contributions may be classified. Some of the errors can be reduced if more stringent specifications of the SOI wafer are fulfilled, some of them may be reduced after pre-calculation. Reduction of the remaining placement errors can be achieved if specific means of an Ion Projection Lithography (IPL) tool are applied. These are mainly magnification and anamorphic corrections removing so-called global distortions. The remaining local distortions can be further reduced by applying the concept of thermal mask adjustment (THEMA).

Improvements of the membrane bulging method for stress determination of silicon open stencil masks for ion projection lithography

Ion Projection Lithography is one promising candidate for a next generation IC technology. Within this field of research one of the most critical aspects is the development of open stencil masks. The stress formation during the mask fabrication process affects the critical dimensions of the structures to be formed. The stress measurement in the mask blanks is performed by the well known bulging method. The accuracy of the membrane bulging method depends on the correct description of the bulged membrane shape. The accuracy of the method can be improved by including the deviations from spherical approximation by correcting the membrane bulging formula. Additionally, the shape of the membrane deformation and the improved bulging formula are compared with FE simulations. The commonly used interferometer technique can not be used to determine the absolute zero point of the membrane (e.g. plane membrane). With the diffraction image technique the zero point is determined by the geometry. Additionally, by bulging of the membrane in two directions, the zero point can be deduced from the anti-symmetry of the pressure vs. deflection curve.

Stencil mask key parameter measurement and control

Stencil masks for Ion Projection Lithography (IPL) are manufactured in a SOI wafer flow process. They consist of a 3 micrometer thick stencil membrane coated by a 0.5 micrometer thick carbonic protection layer. For mask manufacturing, the key parameters which have to be kept under tight control in order to have a high yield are critical dimensions (CD), image placement and defect density. In order to control critical dimensions, the parameters determining CD have to be known in detail. E-beam writing, resist processing, silicon and carbon etching are main contributors. Their impact will be discussed. For CD measurement, different alternatives of tools, optical CD microscopes, AFM and SEM are discussed. Image placement is one of the most critical parameters for IPL stencil masks, as process-induced distortions occur and are to be corrected by a software using FE calculations. Masks usually are specified to 0 defects. Defect inspection results of IPL stencil masks of optical tools are presented, as well as results from e-beam inspection. In addition, defect management for stencil masks in general and cleaning techniques are discussed.

Placement measurement and FE modeling results for distortion control of stencil masks

Distortion control is one of the key issues to solve for IPL stencil mask development. Placement is measured by a LEICA LMS IPRO system. Registration as well as overlay results and the error contributions of the measurement will be presented. The production flow of IPL stencil masks is marked by the fact, that e-beam patterning is done on the bulk wafer, whereas the removal of the bulk silicon and the creation of the free membrane takes place at the end of the process, after silicon trench etching. Therefore, distortions appear at the release of the membrane after bulk silicon etching and oxide removal. At e-beam patterning, the mask wafer blank is pre-stressed by the sum of the stresses of the different layers as bulk silicon, silicon oxide, the silicon of the latter membrane and resist. Additionally, the initial warp and bow of the mask wafer blank have to be considered. The analysis of the finite element modeling compares the placement at e-beam writing to the situation after membrane completion. With this information, the efficiency of a FE-supported software correction before mask patterning can be improved. Measurements of masks with different stress values are to be discussed in order to deduce the optimum stress values for IPL stencil masks.