A Danbury Overview #2
by Rodney J. Grega,
CAD Systems Operations Manager (Retired)
Instrument Division and Optical Technology Division were not the only Perkin -Elmer irons in the fire.
Electro-Optical Division was also present, and as an outgrowth an orphan child was created,
the Microlithography Division.
by Rodney J. Grega, CAD Systems Operations Manager (Retired)
Although I never met Richard S. Perkin or Charles W. Elmer, I am grateful for, and would like to thank them (albeit posthumously), for creating a great place to workâ€¦ a company that treated employees like family, fostered cooperation, creativity, and technical excellence, yielding products that made us all very proud.
I am fortunate to have spent 35 years at P-E starting as a Junior Designer in 1966, and retiring from SVGL/ASML as CAD Systems Operations Manager in 2001. A highlight of my career was my role as lead designer on the team that developed the "Micralign" projection mask aligner, the tool that played a major role in transforming the integrated circuit (microchip) from success-at-any-cost primarily military device to the ubiquitous commercial product that has literally changed the world.
The creation of an integrated circuit includes the printing of multiple layers of microscopic imagery on silicon wafers coated with photoresist, a material that changes properties when exposed to ultraviolet light. Each layer must be precisely aligned with the preceding layers. In the early days, the images were created by contact printing, bringing a patterned glass mask into physical contact with the wafer and exposing the combination to ultraviolet light. Developing each image removed the photoresist from selected areas, which were then subjected to other stepsâ€¦ diffusion, conductor plating, etc. This process yielded a wafer populated with multiple integrated circuits with features 3 to 5 microns in size produced by overlaying 5 or more imaging/processing steps. (NOTE: 100 microns is about the size of a human hair). This was a very low-yield (and hence expensive) process due to defects resulting from the physical contact and the motions required to align the mask to features in preceding layers.
It had long been evident that projecting an image of the mask onto the wafer would eliminate the defect problem and greatly increase the yield; however, devising an optical system with the required resolution, the required image placement precision, and a useful field size was a formidable challenge.
In 1967, recognizing this challenge, and in accordance with their charter to plant seed money in areas needing a technology boost, the United States Air Force awarded a contract to P-E's Electro-Optical Division for a tool capable of projecting an image onto a wafer at the requisite performance level. This project, managed by John Bossung, yielded a tool with a 16-element refractive lens capable of exposing a 1.5 inch diameter wafer. Note: John Bossung's name will live forever as it has been attached to the Bossung curve, a very powerful diagnostic device. The tool met the requirements of the contract, but clearly was not a commercial product. However, P-E was now fully aware of the problem, the need, and the opportunity. The creative juices started to flow! The timing was good inasmuch as the E-O division was looking for ways to apply their aerospace expertise to the commercial marketplace.
Sufficient field size was a major part of the challenge, especially since it was clear that larger wafers were coming. Dr. Rod Scott of P-E, with years of experience with scanning aerial camera systems, wondered if a wafer image could be created by a scanning system rather than with a system with a field large enough to cover the full wafer. That prompted optical designer Abe Offner to come up with an extremely simple optical systemâ€¦ two concentric spherical mirrors configured to yield extraordinary optical performance in a narrow annular fieldâ€¦ a "ring field". Abe suggested a lithography tool in which the mask and the wafer would be scanned in coordinated fashion across that narrow field.
A Perkin-Elmer engineering team headed by Dave Markle, Jere Buckley, and John Bossung was assigned the task of exploring that approach, proving the concept, and developing the idea into a commercial product. E-O Division Director Harold Hemstreet provided funding and steadfast support in spite of widespread skepticism. ("Why are you spending out precious Bid and Proposal funds on such an unlikely project?"
The team invented a curved mercury capillary lamp to illuminate the narrow curved region of good correction provided by the Offner optical design.
The team was able to circumvent what would have been the formidable challenge of separate precisely coordinated mask and wafer scanning mechanisms by adding three folding flats to Abe's 2-mirror system, allowing the simplest possible scanning geometryâ€¦ the mask and the wafer both carried by the same stage, with that stage rotating around a single fixed flexure axis.
One of those three mirrors had to be a beam-splitter to allow access by the viewing system required for image alignment. The relative positions of the three mirrors had to be held to arc-second tolerances. In order to irrevocably achieve that result, the team opted to make the three-mirror array a single optical element, created by assembling multiple pieces of glass using an optical contacting technique. This required a tour de force by the optical shop but avoided unending heartache on the assembly line and in the customer's facility.
The emphasis on practicality was further enhanced by the mounting system for the three-mirror array. It included three precision adjustment knobsâ€¦ one for focus and one each for scan-direction and cross-scan distortion adjustment. Unlike many optical systems, these adjustments were not interdependentâ€¦ a big plus for users unfamiliar with adjusting optical systems.
In the marketing department, Peter Moller initially had to convince skeptical customers that a scanning system could work. Little did he know how soon his job would become keeping customers in an orderly line!
The original Micralign 100 was a rapid success and was soon followed by the Model 200 and the Model 300, with added features including automatic alignment, automated wafer loading, and increased wafer capacity with enhanced throughput. Later came the Model 500 with a new optical system and 5 inch diameter wafer capacity but still using the ring-field scanning concept. The advantages of the Model 500 included the larger field size required for 5 inch wafers, elimination of the distortion problems inherent in the mirror array of the earlier models, a wider ring-field and hence greater throughput, and a magnification adjustment capability that previous models did not possess.
Moore's law kept demanding smaller features, and wafers kept getting larger. In response to those pressures, P-E invented the "Step and Scan" approach in which the machine steps from exposure field to exposure field but scans each of those fields, effectively combining the best features of scanning and stepping machines. This concept has enabled lithography tools with wafer diameter capacities of 8 inch and larger, e.g. the series 90 and 92 "Prime and Blue". The P-E invented step and scan technology required a carefully synchronized motion between mask and wafer with a 4:1 velocity ratio. A wafer stage with the capability to handle 8 inch plus diameter wafers was designed. Enhanced marketability was introduced by accommodating storage of up to 16 customer masks in an onboard library allowing quick selection of next needed reticle to the device.
When the Micralign development was in its formative stages, Harold Hemstreet predicted that we could sell 1,000 machines. History proved that estimate to be way too low! The program was a great adventure and a huge financial success. I'm proud to have been a part of it!
For more information on the early Micralign and how it transformed the integrated circuit industry, conduct an internet search for "The Near Impossibility of Making a Microchip", an Invention and Technology article authored by Daniel P. Burbank.
Many thanks, ( and a tip of the hat) to Messrs. Jere P. Buckley and David A. Markle for their technical and historical perspective and for their contributions to the writing of this article.
Authors note: Every digital device i.e. watches, cell phones, calculators, computers, laptops ,hearing aids, medical devices, TVs, automobiles, appliances et al: owes its existence to the invention of the integrated circuit and the "Micralign", "Micrascan", and subsequent integrated circuit production devices for enhanced delivery of these products to the consumer.