Vol.4 No.2 2011

Research paper : Thermoelectric hydrogen gas sensor (W. Shin et al.)−102−Synthesiology - English edition Vol.4 No.2 (2011) performance, microfabrication was essential.A five-year scenario of microfabrication of the sensor was drawn in 2002. One study cycle consisted of fiscal years of 2003-07. By integrating the elemental components, the necessary sensor performances were achieved, and as a result, we could harvest the output of research, such as papers, patents, transfer of technology one after another, and start a new cycle. The cycle of the development can be divided into four steps as follows, among which the third is the synthesis:1) idea - to discover new ideas or to search social demands,2) integration of knowledge - to quantify it in such experiments that embody it,3) synthesis - to set the necessary properties (goals), to further development,4) completion - summarize research results leading to the following research.It takes a long time to organize the different elemental technologies. Based on a new idea for sensor performance improvement, new batch fabrication is processed, and then the performance of the manufactured sensor device is investigated, in relation to the improvement of the idea. By searching for new methods of development from the results obtained in this process, the leader of the research team could direct engineers who were responsible for each elemental technology to new directions. This feedback sometimes took a few months or one year. It is very similar to the practice of an orchestra. The most highly agile methodology practices in the development of thermoelectric sensors are the following two: ・ laboratory design aimed at real application from the beginning ( the full-automated process equipment and sensor testing systems)・ sharing of the whole development scenario and discussions within the laboratory with all the engineers. In the process of the integration of knowledge, it is important to speed up the feedback of the fabrication and of the results of sensor tests. Our strategy was the improvement of both the process and analysis tools with a compact lab design. Though all the details of tools or lab layout cannot be described in this paper, we can explain a unique idea of our laboratory for manufacturing micro-sensor devices. We have succeeded in minimizing the clean room space very efficiently, and to integrate most of the process equipment within 5 meters radius. The sensor test was carried out next door. This high quick fabrication and tests realized the commercialization of the sensor after the 5 years research period. Introducing full-automatic process equipment has brought good results in this research. Typically, especially at an early research stage, people tend to introduce an experimental facility which is designed specially for scientific researches. We, however, were able to develop a practical-looking production system by introducing a semi-generic one. Such equipment does not require a high professional staff with knowledge and is easy to maintain, and this led to a significant reduction in development costs.The people who put out the results by running the facility are the members of the team. It is important to position the members in the right places, supplying an environment not Fig. 3 Three test methods for sensor elementThey are developed for checking the working principle, social receptivity and mass production, respectively. (℃)16010040DCpowerDCpower30L TestchamberData LoggerTemp. & HumidityControl cabinet200LPCGas injectContinuous heatingBlowerCatalystThermoelectric507070Diffusion chamberDevice holderSensor deviceRubber filmCutterOperating at 100 ℃1 vol.% H2 in airIR cameraGas tankTest chamberTest system for long‐term stabilityVerification of operating principleSocial receptivityPallet(20 devices)Mass‐production techniquePractical test method for gassensor response characterizationGas sensor test systemDiffusion chamberfor response time measurement


page 31