▎ 摘　　要

One of the prerequisites for accurate measurements of relative humidity using thin-film humidity sensors is to obtain a single calibration equation that can be effectively applicable to individual sensors even at different temperatures. Here, we report on a normalization methodology in order to reduce difference among sensor units and also temperature dependency of humidity sensors based on quartz crystal microbalance (QCM). Three independent QCM humidity sensors using graphene oxide (GO) as a humidity sensing material are fabricated from a same batch of GO solution. Frequency shifts of these sensors upon humidity changes from 10 %rh to 90 %rh at 20 degrees C are normalized to minimize measurement errors occurring due to fabrication tolerance among sensors. It is demonstrated that the maximum measurement error of three sensors is reduced from 4.83 %rh to 1.66 %rh after using a normalized calibration equation. The normalization methodology is then extended to humidity measurements at different temperatures (10 degrees C and 30 degrees C). It is found that the normalized frequency shifts of GO-based QCM humidity sensors are consistent at different temperatures. The maximum measurement error due to temperature dependence of these sensors is reduced from about 5.60 %rh to 2.53 %rh after using normalized frequency shifts for a calibration equation. The same normalized calibration equation is applicable to a faulty sensor with smaller frequency shifts within the same budget of the maximum measurement error (2.53 %rh). The normalization methodology using GO-based QCM humidity sensors will provide an economical calibration process because it requires only two humidity points.