Microfluidic reactors have enabled the continuous conduction of chemical reactions with exceptional control of fluids and operating conditions. To enhance reagent mixing in laminar flow conditions, the effective strategy of adding cylindrical obstacles (i.e., pillars) in the mixing channel is appealing. Indeed, they can passively extend the contact surface of the incoming fluid streams, where diffusion and reaction occur. From our previous work, optimized sequences of pillars in a T-microchannel – that differ in diameter and position – can significantly improve the mixing for single-phase flow (i.e., mixing of miscible fluids of similar density and viscosity). However, real-world mixing applications often require the merging of fluids with significantly different fluid properties before, during, and after mixing. Hence, further investigations are required to unveil efficiency and robustness in these more complex practical mixing conditions. We jointly carried out experiments and numerical simulations on miscible fluids in an optimized sequence of pillars called IAF. Our analysis shows how the fluid velocity (i.e., the Reynolds number, Re) and the pillar configuration affect the formation of horseshoe vortices, which we consider crucial for water mixing when 1<Re<100. Then, we analyzed water–water and water–ethanol mixtures at Re=40 to observe how fluid properties alter the flow passively controlled by the pillar sequence. Eventually, the analysis extends within Re=10–100, showing efficiency and pressure drops of optimized designs that depend on the evolution of physical properties. These findings provide insights into the predictions of confined fluid flow around obstacles and its optimization for engineering processes, e.g., nanoparticle precipitation.