In-Situ Identification and Concentration Detection of Microfluidic Liquid Using a Non-Contact Photoacoustic Sensing Approach

Microfluidic liquid samples are crucial in both biomedical and chemical engineering, however, their in-situ identification and testing remain a major challenge. Here, we developed a non-contact photoacoustic (PA) sensor for in-situ and real-time identification and concentration detection of substances in a microfluidic environment. The ultrasonic vibration of the microfluidic samples was generated by a free-space pulsed laser pump and detected by a non-contact laser Doppler vibrometer. The time- and frequency-domain responses of the PA signals were employed to achieve substance identification and concentration detection. Principal component analysis (PCA) and adaptive boosting (AdaBoost) algorithms were employed to extract the frequency-domain characteristics, and an identification accuracy of over 98% was achieved. Chemical samples down to the nanomolar (nM) level can be detected by analyzing either the time-domain or frequency-domain response. The proposed sensor with non-contact and fast detection provides a promising platform for in-situ and real-time monitoring of samples in microfluidic systems.