The first two articles in this series ended at the same wall. Density is one number, so it can pin down one unknown. A binary solution has one unknown, and density handles it. But add a second dissolved thing — sugar and acid in a drink, alcohol and extract in a beer, two solvents in a formulation — and density alone cannot say how much of each. Many different pairs of amounts give the same density. The way past that wall is standard: measure a second property that also depends on composition, then solve for both amounts at once. That second property is often refractive index.
Two unknowns need two equations
This is the algebra you already know. One equation cannot fix two unknowns; two independent equations can. A ternary mixture — two solutes in a solvent — has two composition unknowns. Density gives one equation relating those unknowns. Measure a second property that also depends on composition, and you have the second equation. Solve the two together and both amounts fall out.
That second property has to be genuinely independent. If it rose and fell with composition the way density does, it would add no new information — the same equation twice. The pairing works only because the second property responds to composition differently.
Why refractive index is a good partner
Refractive index measures how much light slows and bends when it enters a liquid — conventionally yellow sodium light near 589 nm, the same wavelength for the standards and the unknown. Snell's law fixes the bend from the ratio of light's speed in the two media, and that ratio is the refractive index, n.1 What matters here is what it responds to. Density responds to how much mass is packed into a volume. Refractive index responds to how the material's electrons interact with light — its polarizability. Those are different physical things. A dissolved species that adds a lot of mass but few easily-polarized electrons moves density more than n; a species with loosely-held electrons moves n more than density. Because the two respond to different features of the same molecules, their equations are usually independent, and the pair is solvable. The Lorentz–Lorenz relation makes the link between refractive index, density, and polarizability explicit.1
Plot the two unknown amounts on a graph — solute A across, solute B up. A measured density is satisfied by a whole line of (A, B) pairs. A measured refractive index is satisfied by a different line. Where the two lines cross is the one composition that satisfies both. That crossing point is your answer. If the two lines meet at a healthy angle, the answer is sharp. If they run nearly parallel, the crossing is smeared and a small measurement error moves it a long way — the pairing is then ill-conditioned.
How it is done in practice
You do not derive the two equations from theory any more than you did for density alone. You calibrate both. Prepare standards that span the two-dimensional composition range, and measure density and refractive index on each. That gives two calibration surfaces: density as a function of the two amounts, and refractive index as a function of the two amounts. For an unknown, measure its density and its refractive index, then solve the two surfaces for the pair of amounts that produces both readings. In narrow ranges the surfaces are nearly flat and the solve is quick linear algebra; over wider ranges the standards and the fit carry the curvature. In practice the two surfaces are usually fitted by regression rather than read off a literal graph, and many density-and-refractometer instruments do the fit and the solve in their own firmware.
Beer and wine labs have done this for a long time. A fermented drink is mostly water, alcohol, and dissolved solids (the extract). Alcohol lowers density and raises refractive index; extract raises both. Because alcohol and extract pull density and refractive index in different combinations, one density and one refractive-index reading solve for both — the drink's alcohol content and its real extract — without distilling anything. Brewing and enology methods are built on exactly this pairing.2
Where it breaks
The failures follow from the setup, the same way they did for density alone.
You must know the three components. The method solves for the amounts of a known set. If a fourth unknown is present, two equations cannot fix three amounts, and the answer is wrong while still looking plausible — the ternary version of the contaminated-caustic problem from the first article.
The two responses can be too alike. When density and refractive index respond to the composition in nearly the same proportion, the two lines run close to parallel and the solve is ill-conditioned: small errors in either reading throw the result off badly. Some systems are simply poor candidates for this pair, and a different second property (conductivity, sound speed) resolves them better.
Temperature moves both. Density and refractive index each drift with temperature, and now you have two instruments to hold at a known temperature, not one. Both readings must be at the temperature the calibration assumed — and both instruments calibrated against the same set of standards. A mismatch there is a common, quiet source of error.
Where it is used
The pairing is common wherever a product is essentially three components and both amounts matter. Brewing and winemaking read alcohol and extract this way. The sugar industry reads dissolved solids and purity from density and refractometry together.3 Formulators use it for two-solvent systems and for a solute in a mixed solvent. The common thread is a known ternary and two bulk properties that disagree enough to be solved.
Any second property that is independent of density does the same job — refractive index, speed of sound, conductivity — each adds one equation and lets you name one more unknown. Refractive index is the most common partner because it is fast, needs a drop, and disagrees with density in a useful way. A later piece takes up speed of sound, the second number many density meters already report.