Solvent-shifting nanoprecipitation is a simple route to spontaneously generate ultra-dispersed organic systems—nanodroplets or polymer nanoparticles—by adding a poor solvent to a solute dissolved in a good solvent. Also known as the Ouzo (or “Pastis”) effect, it applies broadly to multicomponent formulations. Despite its apparent simplicity, laboratory protocols face two major bottlenecks: the need to remove solvents afterward and a strong dilution limit in solute loading (typically ~0.1%), historically identified as the “Ouzo boundary”.
We performed a systematic study of how mixing quality competes with precipitation kinetics in these systems. Because nanoprecipitation can occur faster than macroscopic mixing, local supersaturation hotspots can dominate the formation pathway. We showed that the Ouzo boundary vanishes when mixing is fast enough compared to precipitation, enabling solute volume fractions up to ~20% while still producing Brownian-sized particles. A simple two-stage millifluidic mixer with an integrated quench allowed us to measure characteristic formation times on the order of ~10 ms in a lab-scale setup. Finally, our results resolve a long-standing debate on mechanism: Brownian objects can form beyond the spinodal if mixing is sufficiently improved, supporting a decomposition–aggregation scenario rather than classical nucleation-and-growth.
The formation of metallic nano-objects is a major engineering challenge due to the extreme speed, multiscale nature, and strong parameter coupling of the underlying kinetics. Conventional syntheses in semi-closed reactors are often poorly reproducible and difficult to scale. Rather than exploring reaction space through incremental recipe changes, we developed an approach based on controlling nanoprécipitation pathways themselves, by imposing composition changes on timescales comparable to nucleation.
This strategy relies on a high-throughput, sequential fast-mixing process enabling reagent additions on the millisecond timescale. Custom-made tangential vortex mixers ensure efficient mixing at Reynolds numbers of a few thousand while maintaining flow rates above 10 L/h. This architecture makes accessible precipitation pathways that are impossible in batch reactors, such as delaying stabilizer addition relative to nucleation. As a proof of concept, seven distinct temporal pathways leading to the same final composition produced seven silver nanoparticle dispersions with different colors, sizes, and shapes. This demonstrates both the necessity of pathway control for reproducible synthesis and its power to explore new structural spaces. The approach opens strong prospects for complex nanosystems, including nanoalloys and functional lipid nanoparticles, and naturally integrates with active-learning strategies for systematic pathway exploration.