First Component Sales to Multi-Beam Mask Writer and Multi-Beam Inspection System Manufacturers
In the past, JEOL never sold components to other companies, although it purchased components from other suppliers. However, with the appearance of the open innovation era, JEOL has changed its business attitude and begun engaging in business of other company’s components in a positive manner. In a collaboration with IMS (a subsidiary of Intel Corporation, USA), JEOL takes care of development and sales of necessary components for IMS Multi-Beam Mask Writer. New knowledge from outside the company also gives favorable influences to the development works of JEOL Ltd.
Multi-Beam Mask Writer with 260,000 electron beams
Now that technology advances greatly and speedily, the self-sufficiency policy will only result in a closed situation and people and companies will be left behind. Therefore, open innovation has attracted attention. Cooperation or collaboration with other companies, even competitors, has become necessary. JEOL has never developed or sold components to other companies, although JEOL has purchased components from them, adhering to the self-development and self marketing policy.
The culture has changed in 2012. JEOL had a request from the IMS Corporation, a research company for multi-electron and ion beam technologies in Austria. The request was for JEOL to develop components for the IMS multi-electron beam mask writer. Keisuke Sato, in the design department responsible for the electron beam lithography system, was assigned to this request.
“They asked us to cooperate with them for their multi-beam mask writer under development at that time. Normal mask writer uses a single electron beam. A multi-beam mask writer manipulates 260,000 electron beams at once, thus it can finish writing with the writing time two times or even three times faster. The IMS Corporation then was not a product manufacturer, but a research and development-based company. They had an extremely high technology of multi-beam. Therefore, they wanted to tie-up with JEOL who had the technology of holding the stage to place the mask blank on.” says Sato.
260,000 beams were to be illuminated onto the mask blank on the stage. The stage must be precisely controlled in the ten nanometer order. Technical collaboration of companies was required for the realization of such stage control.
Resolution of antinomy: air-bearing and vacuum
JEOL then had the know-how to control the stage but did not have such a device that can achieve the required accuracy by IMS Corporation. Sato put his eye on the “air-bearing vacuum stage” which was then at the prototype level, made by a Japanese leading heavy machinery manufacturer and decided to use it. The mechanism enabled noncontact, fine and smooth stage movement even in the vacuum, by making the stage floating by air.
It was a mechanism that never existed before. Sato’s team had discussions with the said Japanese manufacturer and started the development including the vacuum chamber. The development period was until the autumn of 2013, the next year. With such a short development period, the team members felt the pressures.
“The most difficult was how to overcome the antimony of outputting air from a few micrometre gap in order to float the stage, and of keeping the vacuum at the same time. We thought about the number of evacuation pumps, air output power, and the capacity of the chamber. To be concrete, the number of the turbo molecular pump for evacuation was increased to five from one or two, in order to evacuate the air all at once to keep the vacuum. “
Another problem was the position instability by vibration when the stage moved. Through trial and error, we completed the development on schedule and could ship it to IMS Corporation.
“Currently, the IMS Corporation is marketing this multi-beam mask writer. The sales records show good results. In particular, the big semiconductor factories have purchased this mask writer. Around 2012, the IMS factory was small, but it has become decent with a fair investment. IMS became a dedicated subsidiary of Intel Corporation in 2016.” says Sato.
Sense of speed in semiconductor industry
A major semiconductor equipment manufacturer inquired for the stage that JEOL developed, knowing the successful partnership between JEOL and IMS.
In 2017, the next project had been launched.
“The stage is still in the development. The basic mechanism has not changed. However, along with the semiconductor device node advancement, the tighter specification for the particle generation is requested. What is meant as the particle here is the out-gas from the stage materials. We have solved the particle issue by proceeding the stage baking in a vacuum. The stage is scheduled to be shipped in 2018’s fiscal year.”
“In the past one year, we have had TV conferencing with the responsible people of the company we provide to, almost every week. Every time additional items are made to our list of homework. So, it is tough.” continues Sato with a sparkle in his eyes.
As for developing a component per request from the world-leading company, Sato says, “the requirements from the user are severe, but they will help improve the performance of JEOL instruments.”
“The required development speed is much faster than JEOL’s and I truly feel how fast the semiconductor industry moves. Unless we move faster as requested, we cannot satisfy the customer.“
JEOL’ｓ technology admitted by the world
The most difficult thing in a series of development was “to develop the component and JEOL instruments concurrently.” says Sato.
“Since there are always several projects running in parallel, I had difficulty allotting my time and energy properly. I finally made it, supported by colleagues and friends. When I see the mechanism that I have designed works properly, I feel happy. If this stage sells well, as in the case of IMS, I will feel much happier. Good sales of the stage mean our technology is admitted by the world, that makes me feel rewarded.”
As for the future perspective, Sato says, “first of all, it is necessary to complete the components of on-going projects.”
“The user companies request a different shape of components even though the required mechanism is the same. In the future, I would like to make many parts in common to make the total development time shorter. Furthermore, I would like to develop our own mask writer of the highest performance better than IMS’s. “
JEOL will receive many more inquiries for component development from around the world.