Fig. 1. Evanescent waveguide lab-on-chip basic structure, with two sectioned insights on the sample–waveguide interaction site (left side) and on the waveguide–photosensor detection site (right site).

Fig. 2. (a) Modeled optical power absorption in the longitudinal section of the interaction site by a biological sample, with a 4-mm-long sample–waveguide overlap. (b) Evaluated optical power absorption versus samples’ complex refractive indices (n and k). (c) Modeled coupling efficiency in the longitudinal section of the detection site, with a 150-μm-long waveguide–photodiode overlapping length. (d) Evaluated waveguide–detector coupling efficiency versus ITO top contact thickness.

Fig. 3. (a) Fabricated system. (b) Enlarged top view of the fabricated detector.

Fig. 4. Current density versus bias voltage curve in dark conditions for reference sensor and sensor under test.

Fig. 5. Investigation on milk’s fat content sensing with the proposed micro-device. (a) Sensor’s expected photocurrents (from an initial reference state corresponding to 40 nW coupled to the diode) at 589 nm, 633 nm, and 660 nm. (b) Calculated sensitivity (S), minimum detecting signal (MDS), and limit of detection (LoD) at 589 nm, 633 nm, and 660 nm.

Fig. 6. Time evolution of the sensor’s photocurrent (using a milk sample to fill the bridge channel), with pictures on top illustrating the three phases (using blue-dyed deionized water as demonstrative liquid).

Fig. 7. (a) Mean and standard error of the sensor’s measured photocurrent drops, with respect to its initial reference current, at different milk fat concentrations. (b) Sensor’s photocurrents obtained for milk and cream samples, and linear regression.