Polyimide-Based Capacitive Humidity Sensor

Abstract

The development of humidity sensors with simple transduction principles attracts considerable interest by both scientific researchers and industrial companies. Capacitive humidity sensors, based on polyimide sensing material with different thickness and surface morphologies, are prepared. The surface morphology of the sensing layer is varied from flat to rough and then to nanostructure called nanograss by using an oxygen plasma etch process. The relative humidity (RH) sensor selectively responds to the presence of water vapor by a capacitance change. The interaction between polyimide and water molecules is studied by FTIR spectroscopy. The complete characterization of the prepared capacitive humidity sensor performance is realized using a gas mixing setup and an evaluation kit. A linear correlation is found between the measured capacitance and the RH level in the range of 5 to 85%. The morphology of the humidity sensing layer is revealed as an important parameter influencing the sensor performance. It is proved that a nanograss-like structure is the most effective for detecting RH, due to its rapid response and recovery times, which are comparable to or even better than the ones of commercial polymer-based sensors. This work demonstrates the readiness of the developed RH sensor technology for industrialization.

Keywords: flat polyimide; humidity sensor; nanograss polyimide; rough polyimide.

Figure 1

Layout of studied interdigitated transducer. The electrodes are surrounded by a temperature sensor and a heater.

Figure 2

Photograph of (a) sensor chamber for the impedance analysis and (b) evaluation kit with a PEEK lid. Both can be connected to the gas mixing apparatus.

Figure 3

Two infrared spectra covering a range from 3000 and 800 cm⁻¹ measured at zero RH (grey) and 50% RH (purple). The difference spectrum of both spectra is shown for clarity (blue). Between 3000 and 1800 cm⁻¹, reflections on the flat polyimide surface cause interference. The small peaks at 2300 cm⁻¹ result from remaining gaseous CO₂ inside the FTIR chamber. RH, relative humidity.

Figure 4

Frequency-dependent capacitance for different RH = 5% (black), 39% (red), and 85% (blue) of a 11 µm thick polyimide layer. The Error bars represent the uncertainty given by the impedance analyser.

Figure 5

(a) Temporal evolution of capacitance of 4.6 µm and 11 µm polyimide layers at different RH measured by SHT25 commercial sensor. (b) Calibration curve of 4.6 (blue) and 11 µm (red) polyimide layers. The capacitance change is traced versus relative humidity at a frequency of 40 kHz.

Figure 6

(a) Stream lines distribution between two adjacent electrodes on glass a substrate. The gold electrode width and the gap between two adjacent electrodes are fixed to 6 µm. (b) Electric field strength along a perpendicular axis to the IDT’s planar surface and situated in the middle between two electrodes adjacent electrodes.

Figure 7

SEM cross-section pictures of plasma etched polyimide layers using different pressures: (a) p = 720 mTorr and (b) p = 50 mTorr. The thickness of the resulted grass structure is indicated by arrows. SEM top view pictures of etched polyimide layers at (c) p = 720 mTorr resulting in a roughened surface with a surface coverage compared to (d) p = 50 mTorr.

Figure 8

(a) Normalized capacitance of humidity sensors measured at 40 kHz versus time under different RH levels and for different thicknesses and surface morphologies. (b) Calibration curves of developed humidity sensors with different thicknesses and surface morphologies as indicated.

Figure 9

Response and recovery times of relative humidity sensor compared to a commercial one (SHT25) measured inside the chamber of the (a) gas mixing setup at 40 kHz (b) evaluation kit.