A lipid nanoparticle (LNP) packages genetic material such as mRNA or siRNA inside a particle built from a few lipids. On its own, that cargo is fragile and goes nowhere useful, because naked RNA is broken down by nucleases within minutes and carries the wrong charge to cross a cell membrane. The LNP solves both problems at once, shielding the cargo in circulation and carrying it into the cell, where the RNA is released to do its work.

That delivery job is why the particle, and not the sequence alone, decides whether an RNA therapy works. Two teams can start from the same mRNA and get very different results depending on the LNP they put it in, which is why so much of RNA drug development is really LNP development.

An LNP is assembled from four lipid components, and each one carries a defined role. The ionizable lipid is the functional core, taking on a positive charge at low pH to bind the RNA during mixing and returning to near neutral at physiological pH so the particle can release its cargo inside the cell. A helper phospholipid and cholesterol give the particle its structure and stability, while a PEG-lipid sits on the surface to control particle size and limit aggregation.

What makes an LNP a design problem rather than a fixed recipe is that these four lipids work as a system. The molar ratio among them, the choice of ionizable lipid, and the ratio of ionizable lipid to RNA all shape particle size, encapsulation, and how efficiently the cargo reaches the cell. A small change to any one of them can move those properties in ways that are hard to predict from the structures alone, which is why teams test compositions empirically instead of calculating a single best recipe.

Composition sets what an LNP can be, but the mixing step decides what it actually becomes. LNPs self-assemble within milliseconds of the ethanolic lipid phase meeting the aqueous phase that carries the RNA, and as the ethanol is diluted past a critical point the lipids drop out of solution and condense around the cargo into particles. Because that window is so short, every particle in a batch has to pass through the same drop in ethanol at the same moment to come out the same size.

This is why how fast and how evenly the two phases combine sets both the size and the uniformity of the result. When mixing is faster than self-assembly, particles form in a consistent environment and come out small and tightly distributed, and when mixing lags, they form across a range of local environments and the distribution broadens. The mixing method is the main lever a team has over particle quality, so it shapes the formulation decisions that follow more than any single lipid choice.

Once particles form, a short set of readouts decides whether a formulation is worth advancing. Particle size, reported as a Z-average diameter, and polydispersity index (PDI) describe how large the particles are and how tightly the batch is distributed, both measured quickly by dynamic light scattering. Encapsulation efficiency measures how much of the RNA is actually loaded inside the particles rather than left free in solution, which is what separates a particle that can deliver from one that only looks right by size.

These readouts, measured on every formulation, let a team compare candidates on the same footing and decide which ones move forward to cell-based testing or further morphological characterization such as SAXS. No single number is a target on its own, because the size and PDI that matter depend on the cargo and the program, so a screen is really a search for the composition that fits the profile a program needs. Reading the readouts together, rather than chasing any one in isolation, is what keeps that comparison meaningful.

Because the effect of changing a lipid or a ratio is hard to predict, teams do not design one formulation and hope it works. They build a library of candidate compositions and run them together so the results can be compared side by side, a process known as LNP formulation screening. Each round tests many compositions, reads their size, PDI, and encapsulation, and carries only the few that meet the program’s criteria into cell-based work.

How a team makes those candidates bounds what the screen can actually learn. A screen produces a fair comparison only when every formulation is made the same way, so the differences in the data reflect the compositions and not the handling. The method a team formulates with therefore sets both how many compositions it can test in a round and how far it can trust the ranking that comes out, which is why the formulation step, more than the assays, is usually what limits a screen.

The LNP Screening Array brings microfluidic mixing onto a standard liquid handler, so every formulation in a screen forms through the same controlled mixing. Each one runs in seconds and collects on deck, which lets a full library move from mixing straight into dilution and processing inside a single method.

Because the mixing is consistent from the first formulation to the last, the differences a team sees across a screen reflect the compositions it set out to test rather than variation in how each one was made. A screen can then cover more lipid compositions in a round, at 100+ formulations per hour, without adding a separate formulation instrument or the manual transfers that come with one.

Screen 100+ LNP formulations an hour on the liquid handler you already own.