Successful development of a micro-total-analysis system (µTAS, lab-on-a-chip) is strictly related to the degree of miniaturization, integration, autonomy, sensitivity, selectivity, and repeatability of its detector. Fluorescence sensing is an optical detection method used for a large variety of biological and chemical assays, and its full integration within lab-on-a-chip devices remains a challenge. Important achievements were reported during the last few years, including improvements of previously reported methodologies, as well as new integration strategies. However, a universal paradigm remains elusive. This review considers achievements in the field of fluorescence sensing miniaturization, starting from off-chip approaches, representing miniaturized versions of their lab counter-parts, continuing gradually with strategies that aim to fully integrate fluorescence detection on-chip, and reporting the results around integration strategies based on optical-fiber-based designs, optical layer integrated designs, CMOS-based fluorescence sensing, and organic electronics. Further successful development in this field would enable the implementation of sensing networks in specific environments that, when coupled to Internet-of-Things (IoT) and artificial intelligence (AI), could provide real-time data collection and, therefore, revolutionize fields like health, environmental, and industrial sensing.

Government regulations and environmental conditions are pushing the development of improved miniaturized gas analyzers for volatile organic compounds. One of the many detectors used for gas analysis is the photoionization detector (PID). This paper presents the design and characterization of a microfluidic photoionization detector (or µPID) fabricated using micro milling and electrical discharge machining techniques. This device has no glue and facilitates easy replacement of components. Two materials and fabrication techniques are proposed to produce a layer on the electrodes to protect from ultraviolet (UV) light and possible signal noise generation. Three different microchannels are tested experimentally and their results are compared. The channel with highest electrode area (31.17 mm²) and higher volume (6.47 µL) produces the highest raw signal and the corresponding estimated detection limit is 0.6 ppm for toluene without any amplification unit.

SABRE is emerging as a fast, simple and low-cost hyperpolarization method because of its ability to regenerate enhanced NMR signals. Generally, SABRE hyperpolarization has been performed predominantly manually, leading to variations in reproducibility and efficiency. Recent advances in SABRE include the development of automated shuttling systems to address previous inconsistencies. However, the operational complexity of such systems and the challenges of integration with existing workflows hinder their widespread adoption. This work presents a fully automated lab workstation based on a benchtop NMR spectrometer, specifically designed to facilitate SABRE of different nuclei across different polarization fields. We demonstrated the capability of this system through a series of routine SABRE experimental protocols, including consecutive SABRE hyperpolarization with high reproducibility (average standard deviation of 1.03%), optimization polarization of ¹³C nuclei respect to the polarization transfer field, and measurement of polarization buildup rate or decay time across a wide range of magnetic fields. Furthermore, we have iteratively optimized the durations for pulsed SABRE-SHEATH ¹³C pyruvate. The constructed SABRE workstation offers full automation, high reproducibility, and functional diversification, making it a practical tool for conducting routine SABRE hyperpolarization experiments. It provides a robust platform for high-throughput and reliable SABRE and X-SABRE hyperpolarization studies.