Vol.4 No.2 2011

Research paper : Thermoelectric hydrogen gas sensor (W. Shin et al.)−104−Synthesiology - English edition Vol.4 No.2 (2011) and fast response could be determined by specific numeric target values requested from the user's needs. A prototype sensor is fabricated and its performance is evaluated, verifying the achievement against the targets. During this process, the elemental components are selected and integrated for the synthesis.When we select the elemental technology by considering the boundary conditions, the knowledge of the past is sometimes insufficient, and it is necessary to confirm and to recognize the details by fabricating and testing the actual prototype device. At the start of the development of the sensors, we had a catalyst problem. The composition of the catalyst combustor followed the previously reported data in the papers, which said that a few nanometer thick catalyst film was selected for higher activity. However, after trial and error, thick film catalyst of much higher platinum content was found to be better by our experiment. This thick film catalyst made hydrogen combustion possible at room temperature[7]. The researchers in the field of catalyst science have wondered and were surprised by the very high amount of 20 ~ 40 wt% of noble metal, platinum, in the catalysts of our sensor devices. This was because their common knowledge of the catalyst was that several wt% of metal content was better to prohibit unwanted sintering of the metal during the operation. However, the catalyst of our design was found to exhibit a high activity and stability to detect the hydrogen gas in air[8]. Unlike the typical catalytic reactions, the catalyst for the sensors is required to be able to burn gas at extremely low concentrations of flammable gases, so that thick film type of high metal contents was integrated into the sensor. As a result, boundary conditions of gas sensor application stimulated the elemental technology of the catalyst to be something of a totally different quality.4.2 Scenario for commercializationAfter investment for research such as capital, facilities, and human resources, papers and patents are always desired as results of research. Of course commercialization of the new technology is an ultimate goal to complete the R&D. Especially for AIST, a research institute without any commercial product vending department, to complete R&D is to transfer technology obtained in the development to private companies as intellectual property such as in the form of patents. In mass production technology of our prototype, it was easy to scale up the fabrication of sensors, as a result of the strategy described above. It was necessary to make technology transfer possible even with some companies that do not have their own semiconductor process facilities. We performed 4 inches or 6 inches wafer processes not only in the laboratory prototype manufacturing but also in a commercial foundry. In 2007, the foundry services in Japan still had little business experience, including the issue of acceptance of service and other contractual issues of technical problems, and it was very difficult to work out our first batch. The experience of the difficulty of the foundry process, and confirming the process yield was a big step toward commercialization. Fortunately, the Seebeck coefficient of the SiGe thermoelectric film pattern in our device which is the most important parameter of the sensor, was relatively process independent. Therefore, we established a manufacturing technology sensor device with reduced discrepancies in sensor performance.4.3 Results over the scenarioUnexpected sensor performance triggered a start of new development which was not drawn in the scenario. A leak detector is one example in our research. The very promising thermoelectric hydrogen sensor capability detecting the gas leak of several-ppm level was over-specification in the application of safety sensors. However, this performance is necessary in new industrial application of leak detection, which is the technology to check the air tightness or sealing of the products, such as in fuel tanks, battery packs, water proof housings, etc. In this gas leakage test, currently helium gas is used, but hydrogen is now also used in place of helium because of recent instability of He import from the US. Recent testing equipment with 95 % nitrogen 5 % hydrogen gas as a gas alternative is now widely used, and with the high cost of helium, there is a strong need for cheap, sensitive, and selective hydrogen sensors. Another development is a medical device to support human life, which aims to expand the application of the sensor which can measure hydrogen concentration in the breath[9]. Current technology used in medical centers is Fig. 5 In the synthesis of thermoelectric gas sensor, the elemental technologies have been configured and integrated with the boundary conditions of social demands.CatalystThermoelectricStudy・applicationBreath gas analysisLeak testerHydrogen detectorIntegration of technologiesMicrofabricationSelection・specificationComposition・integrationBoundary condition for technologiesInternational standardCommon test methodSelectivityFast responseStabilitySensitivitySocial demands /applicationsSensingperformanceTest methodObservation・analysisSensor deviceThermoelectricCatalystHeat transferElectronic stateGas combustionChemical/physical phenomena


page 33