1. IntroductionDew condensation causes plant disease, metal corrosion, glass fogging and so on. In order to prevent dew condensation, its detection in early stages and preferably in advance should be followed by lowering relative humidity on the target. However, it is difficult to realize early and advanced detection of dew condensation accurately and in a short time by using current technologies such as hygrometer. Therefore, we have developed a sensor to detect fine water molecules directly. It has different metal arrays which are alternately arranged at fixed intervals (minimum distance: 100 nm). When water droplet or molecules cross over the adjacent arrays of the sensor, electric current flows spontaneously by the galvanic effect [1,2]. In addition, wettability of this sensor’s surface was changed [3]. Furthermore, the surface condition of the sensor could be made closer to that of the target for dew condensation if its surface temperature was controlled. So far, the temperature of the sensor surface was able to be changed from the back of silicon chip by using a Peltier device [2]. Taking into account actual application of this sensor, however, such Peltier device should not be adopted because the electric circuit for measurement and control becomes more complicated and the power consumption increases. Instead, it is suggested that heat capacity of the sensor can be close to the target by attaching a heat sink on the back of the silicon chip. Therefore, the purpose of this study was to clarify the influence of the material and weight of the heat sink on changing behaviour of the temperature of the sensor surface.2. Experimental methodSilicon wafer covered with a silica layer was used as the substrate of a sensor chip. Arrays of Al and Au were alternately arranged with 1 μm in width and 0.2 μm in thickness on the substrate to form an opposing interdigit structure and used as the electrodes. After fabrication of the electrodes, the wafer was cut into a 5mm square chip after Pt wires was also placed on the chip surface near the edge of the chip. The sensor chip was adhered on an aluminum lead frame. Except areas for the arrays, the rest surface of the chip was covered with the resin to make a sensor package. Electric current between two electrodes of Al and Au were measured by the hand-made device with precise amperemeter. The temperature of the sensor chip was estimated from after calibrated straight line of electric resistance of Pt as a function of changing temperature. Heat capacity of the sensor was varied by attaching a heat sink directly to the backside of a sensor package through a window which was open in the resin of the package. In this study, aluminum and copper were used as the heat sink and their volume was varied; 109 (cylindrical), 378 and 818 (cylindrical with rectangular) mm3. Temperature of a sensor was changed by putting it into a chamber and its internal temperature and water vapor pressure were controlled for 840 sec at 1 kPa from 25℃, respectively.3. Results and discussionTemperature changing rate of the sensor was calculated from the slope of temperature with the square root of time. It was plotted as a function of additional heat capacity of the sensor with and without heat sink and results at cooling process were shown in Fig. 1. The temperature changing rate of the sensor was decreased by adding heat sinks to the sensor and it seems to be related almost linearly with the additional heat capacity. This means that it is possible to make temperature change of the sensor with close to that of the target when it becomes important on application such as dew condensation detection.4. ConclusionThis study clarified that the heat capacity of moisture sensor could be controlled by attaching heat sink of metals directly on a part of the sensor. Since the wettability of the sensor had been already controlled successfully, the surface status of the sensor could be made closer to the target on application that this sensor is needed even when the temperature of the target is changed.5. Reference[1] J Kawakita et al., ECS trans., 75 (2017) 51.[2] Y. Kubota et al., Sensors and Actuators A: Phys., 303 (2020) 111838.[3] R.G. Shrestha et al., Sensors, 19 (2019) 4500. Figure 1