Browse Prior Art Database

In-situ Exposure Tool Monitor Disclosure Number: IPCOM000014231D
Original Publication Date: 1999-Oct-01
Included in the Prior Art Database: 2003-Jun-19

Publishing Venue


Related People

Joseph Fatula Robert Fabinski Kathryn Gutberlet


Post Disclosure Text Drawings This idea addresses the problem of knowing when enough energy has been used to expose photoresist. Current techniques are to monitor (the) tool with a meter to establish stability, then optimize process settings and run product adjusting to achieve the required critical dimension or resist clearnance; a procedure which incurs many misprocessed wafers. What we suggest is to monitor the reflection of the expose light from the surface of the wafer/resist as it is being exposed and monitor the change in the amplitude of the peaks which represent the spectral composition of said light. By comparing with a reference spectrum (e.g. that of the wafer/resist at the start of the exposure, or taken at another time), and monitoring the rate of change in the amplitude, one can determine the point at which the resist has been exposed to the correct amount of light required to make the image in the photoresist. The change in the amplitude of the peaks which represent the spectrum of the reflected light is known to be related to the resist absorption, and the reaction of the photo-active-compound (PAC) reacting with the photons in the incident light. The attached drawing shows the relative positioning of the fiber optic bundles in the optical path of an Ultratech Stepper for example. One bundle samples the light after it passes through the mask but before it exposes the resist this fiber optic leads to the reference spectrophotometer. The other fiber optic bundle samples the light after it has exposed the wafer and been reflected back into the optics of the stepper this fiber optic leads to the 1024 channel spectrophotometer. Both spectrophotometers would sample their respective light paths in a periodic fashion, wiht a period onthe order of micro-seconds. Note that different expose optics would require varying the actual positions of the inputend of each fiber optic bundle to sample the correct spectra. The expose tool could be controlled by the software that calculates either the optimum rate of change of amplitude of any combination of peaks, or a relative amplitude when compared to the reference peaks. (In a stepper, one could use this technique to control the wafer to wafer mean resist image size which is a critical problem for magnetic head production P2 width sometimes called P2B a first order effect on the density/performance of a hard disk drive. One could do this either dynamically if the signal processing was fast enough or by exposing a non-product area of the wafer to determine the characteristics of the resist/exposure system.