Several conventional methods for absolute sphericity testing of optical surfaces are known. One method has been used previously for real time interferometry, but a detailed investigation has not been carried out so far. In this work the known conventional methods are reviewed and assessed for suitability in real time interferometry. One method has been applied to phase stepping interferometry and examined. Adjustment peculiarities, experimental results of deviation measurements on normal surfaces, and application of normals are reported. Measuring errors, especially coherent noise, are analyzed and a special subsequent digital spatial filtering technique for diminution of noise is described.

Absolute sphericity measurement

A method of approximating the homogeneous second derivatives of a merit function with respect to a set of design variables is presented. If these derivatives are used in an automatic lens design program, the familiar problem of stagnation may usually be avoided.

Pseudo-second-derivative matrix and its application to automatic lens design.

The special properties of lens design as an optimization problem are discussed and are related to the properties of the Levenberg-Marquardt (damped least-squares) optimization method. Several variations of the basic method are discussed and compared. Some examples are given to illustrate the effectiveness and limitations of the method.

Use of the Levenberg-Marquardt (damped least-squares) optimization method in lens design

A microlithographic projection lens is expected to give constant results from one lens to another, anywhere within the field, over the full Rayleigh depth of focus, and with a finite variation of exposure dose. In particular, the requirements for image dimension and position control are an order of magnitude smaller than the features being printed. These demand that the paper design have wavefront correction significantly better than the Marechal criterion, and almost zero distortion. Furthermore, impressive theoretical performance is of no use if the manufactured lens cannot always approach the ideal as closely as possible. Lens designers traditionally bridge the gap between customer requirements and production realities. Compensators -- parameters that are adjusted to compensate for construction imperfections -- are a well-known way of relaxing tolerances. For a lens on the edge of possibility, they come to the aid of the designer caught between two vociferous groups of people with contradictory needs -- the users and the manufacturers. A lens will be shown that is used in a Wafer Stepper capable of printing features in the 0.7- to 0.6-micron range, and the rationale given for an extensive scheme of compensation, both before and after assembly. A selected number of elements are adjustable after assembly, while the lens is tested both interferometrically and lithographically, after which they are fixed in place.

Compensator Selection In The Tolerancing Of A Microlithographic Lens

A lens mount including a lens cell wherein a lens is mounted through flexures. The lens seat in the lens cell has a radius of curvature match-mated to the lens surface which rests upon it. The rotational symmetry axis of the lens seat is coincident with the axis of rotation of the lens cell. The optical axis of the second lens surface is coincident with the axis of rotation of the lens cell.

Mounting for high resolution projection lenses

Centring was done by interferometric measurement of surface height and the lens cells were assembled in plain diamond turned barrels; improvements in symmetry of about an order of magnitude over results previously reported were found.

Some Experiments on Precision Lens Centring and Mounting

Some experiments have been carried out to investigate the efficiency of various lens optimization procedures, such as the method of steepest descent, the method of least squares and the method of damped least squares The results of these experiments demonstrate the efficiency of the damped least squares method Methods of damping the least squares equations and controlling the boundary conditions are also described and the results of some experiments on these are given

M.J. Kidger

Papers

Some Experiments on Precision Lens Centring and Mounting

Experiments with Lens Optimization Procedures