Nikon Corporation began to develop ruling engines in 1961, and in that year Mr. Yoshida participated in its development project as an engineer in his fifth year at the company. Ruling engines are precision machines used for making diffraction gratings. They can draw 1,000 to 1,500 parallel lines per millimeter on the surfaces of 10cm x 10cm flat glass plates using diamond cutters. Through the ruling engine development project, which lasted for about 10 years, Nikon engineers developed a very accurate laser interferometer-based XY stage positioning mechanism and peripheral technologies. These technologies contributed greatly to Nikon's development of steppers, of which the company has become the world's number one manufacturer.

Mr. Yoshida says, "I tried very seriously to find any possible application of these technologies because it had taken the company 10 years to develop them." Fortunately, the Japanese semiconductor industry had just entered its growth stage at that time. Nikon developed mask coordinate measuring machines for semiconductor manufacturers in the early 1970s, making good use of these technologies. The machines were the world's first instruments that could measure the positions of masks with a precision of 100nm.

Mr. Yoshida says, "Nikon used to make special custom-made ruling engines and astronomical telescopes. But the company received repeat orders for its mask coordinate measuring machines from customers for the first time. This convinced me that semiconductor manufacturing- related machines would provide a good business opportunity for the company. So I started my career in the semiconductor-related industry."
When he started his career in this industry, semiconductors were being produced using the contact printing microlithography system, in which masks are placed directly on wafers to imprint microscopic circuitry patterns. Imprinting finer circuitry patterns on wafers requires the reduction of mask patterns through optical lenses. Mr. Yoshida thought reduced patterns could be imprinted accurately on wafers by moving the wafers very accurately, if his company were to make use of its mask coordinate measurement technology coupled with its high-resolution lenses. "But," he says, "no one in the industry considered it possible to develop such steppers, arguing that the basic concept of steppers at that time could not satisfy the needs of semiconductor manufacturers in terms of productivity."

In 1976, Nikon was asked to develop steppers by an organization that had conducted a VLSI R&D project. The organization was set up to carry out Japan's first national semiconductor development project.
Mr. Yoshida says, "We delivered our first stepper to the organization in 1978, and a commercial version was launched in 1981. Initially, our model was able to imprint circuit patterns with a line width of 1 micrometer using a 1/10 reduction projection system. After the first machine was introduced, a novel 1/5 reduction projection system, which can also imprint lines that are 1 micrometer wide, was released. This helped semiconductor device maker improve their productivity fourfold, and this is why Nikon has become the world's number one stepper supplier."

The key to the success of his company was setting clear project targets and thorough discussion among project members. Mr. Yoshida says, "I think setting clear targets in application research is very important, and this can be applied to the nanotechnology area as well."

When Nikon started developing steppers, most of the researchers and engineers were young. They discussed issues very frankly, and their honest discussions sometimes caused arguments. "I believe that how deeply staff who are involved in a project can discuss issues determines whether the project team can make most of its integrated ability," says Mr. Yoshida. These, for him, are the necessary factors for a successful project. As a project leader, he always wants to encourage participants to generate new ideas.

(Interviewer: Kuniko Ishiguro, Cosmopia Inc.)

For more information,
http://www.nanonet.go.jp/english/mailmag/2004/025a.html