Separation Techniques for Mixtures

Every mixture hides a way to be taken apart. Sand in water, salt in water, ink on paper, air in a balloon — each combination stays together only because nothing forces the components apart, and separation science is simply the art of applying exactly that force. The trick is always the same: find a physical property where the components disagree, then build a process around that disagreement. Particle size, boiling point, solubility, and adsorption affinity are the four disagreements exploited by filtration, distillation, chromatography, and evaporation/crystallization respectively. None of these techniques changes the chemical identity of anything present — they only sort matter that was mixed, which is why they qualify as physical changes rather than chemical reactions. This article treats each technique as a quantitative tool, not just a lab ritual. Recovery percentages, boiling-point separations, and retention factors are all calculable numbers, and a first-year chemist should be able to predict, not just perform, a separation.

Key takeaways

Separation techniques exploit differences in a physical property (particle size, boiling point, solubility, polarity) between components of a mixture.
Filtration separates by particle size using a porous barrier; it works on heterogeneous mixtures (suspensions), not true solutions.
Distillation separates by boiling point; simple distillation suits widely separated boiling points, fractional distillation handles close ones.
Chromatography separates by differential adsorption/solubility between a stationary and a mobile phase, quantified by the retention factor $R_f$ .
Evaporation and crystallization separate a dissolved solid from its solvent by removing the solvent, leaving the solute behind.
Percent recovery and mass balance let you check that a separation was quantitatively successful, not just visually complete.
Key idea: choose the property that differs most between components, then design the apparatus so only that property controls where each substance ends up.

The core insight

A separation technique is a selective filter for one physical property at a time — mass, size, volatility, or affinity — and the quality of a separation is measured by how cleanly that one property, and only that property, determined the outcome.

The physical picture

Picture a muddy suspension of sand grains in salt water. Three "substances" are tangled together: sand, dissolved salt, and water. No single technique untangles all three, because sand differs from the liquid in particle size, while salt differs from water in boiling point and solubility behavior — two entirely different properties. This is the central lesson of separation science: one technique exploits one property, so a complex mixture usually needs a sequence of techniques, not one heroic step. Filtration removes the sand: the solid particles are too large to pass through the pores of a filter paper (typical pore size 5–25 μm), while water and dissolved ions pass freely. The filtrate — now a homogeneous salt solution — still contains two substances at the molecular level, invisible to any sieve. To split those apart, distillation is needed: heat the solution until water (boiling point 100 °C at 1 atm) vaporizes and leaves, while dissolved NaCl (decomposition/melting far above 800 °C) stays behind as a solid residue. Chromatography would instead be the right choice if the mixture were dyes or pigments dissolved together, since they share a boiling point but differ in how strongly they cling to a stationary surface versus how readily they travel with a moving solvent.

Figure

Figure 1 — Simple distillation apparatus with a temperature gradient

Distillation works because the vapor phase is always enriched in the more volatile component relative to the liquid it came from. As boiling proceeds, the flask temperature — read at the side-arm thermometer — should hold nearly steady at the boiling point of the volatile substance until that substance is exhausted, then climb. That plateau is the experimental signature of a clean separation, and a temperature that drifts upward from the start is the signature of a mixture whose components are too close in volatility for simple distillation to resolve, calling instead for fractional distillation with its longer, packed column offering many small vaporization–condensation cycles.

Figure

Figure 2 — Paper chromatogram with retention factors

The core relations

\%\ \text{recovery} = \frac{\text{mass of substance recovered}}{\text{mass of substance originally present}} \times 100\%

where: "recovered" mass is measured after the separation and any drying step; the denominator is the true initial mass, obtainable from the original mixture's composition or a control measurement.

R_f = \frac{\text{distance traveled by spot}}{\text{distance traveled by solvent front}}

where: both distances are measured from the same origin line, along the same direction, on the same chromatogram; $R_f$ is dimensionless and lies between 0 and 1.

\%\ \text{mass composition of component } i = \frac{m_i}{m_{\text{total mixture}}} \times 100\%

where: $m_i$ is the mass of one component isolated by a separation, and $m_{\text{total mixture}}$ is the mass of the original mixture before separation.

Derivation APercent recovery and mass balance in a filtration–evaporation sequence

Goal: show how the percent recovery of a solid, obtained by dissolving away an impurity and then filtering and drying, follows directly from a mass balance, and confirm the formula behaves correctly in a limiting case.

Step 1. Suppose a solid mixture contains a desired product P together with a soluble impurity X. Choose a solvent that dissolves X completely but leaves P as an insoluble solid. Let the total starting mass be $m_0 = m_P + m_X$ , where $m_P$ and $m_X$ are the true masses of pure P and pure X locked inside the mixture.

Step 2. After adding solvent and filtering, the solid retained on the filter paper is (ideally) pure P, and the filtrate carries all of X plus the solvent. Denote the mass of dried solid actually collected as $m_{\text{collected}}$ . In a perfect separation with no losses, $m_{\text{collected}} = m_P$ .

Step 3. In practice a small mass $\delta$ of P is lost — some dissolves slightly, some sticks to the filter paper — so

m_{\text{collected}} = m_P - \delta

Step 4. The percent recovery is then

\%\ \text{recovery} = \frac{m_{\text{collected}}}{m_P} \times 100\% = \frac{m_P - \delta}{m_P} \times 100\%

Step 5. This expression makes the limiting behavior transparent. As the loss $\delta \to 0$ (an ideal, lossless filtration), the recovery ratio $\to 1$ , so percent recovery $\to 100\%$ exactly, matching the physical expectation that a perfect separation returns all of the original product. Conversely, as $\delta \to m_P$ (total loss of product, perhaps because the "insoluble" assumption failed and P dissolved along with X), the recovery correctly collapses to $0\%$ . Any experimental value between these extremes is diagnostic: recoveries noticeably below 100% signal either a genuinely soluble product, careless technique, or an underestimate of $m_P$ in the original mixture, and recoveries above 100% signal that the collected solid still contains residual X or solvent that failed to fully evaporate.

Worked examples

Example 1 — Filtration and evaporation with a non-round mass balance. A 12.40 g sample is a mixture of sand and sodium chloride. The sample is stirred with 60.0 mL of water, which dissolves all the NaCl but none of the sand. The suspension is filtered; the sand is dried and found to weigh 7.85 g. The filtrate (salt water) is then evaporated to dryness in a separate dish, leaving solid NaCl behind. Step 1 — mass of sand. The dried filter residue is pure sand:

m_{\text{sand}} = 7.85\ \text{g}

Step 2 — mass of NaCl by difference. Since the only two components are sand and NaCl,

m_{\text{NaCl}} = m_{\text{total}} - m_{\text{sand}} = 12.40\ \text{g} - 7.85\ \text{g} = 4.55\ \text{g}

Step 3 — check by direct evaporation. Suppose the evaporating dish, after all water boils off, is found to hold 4.48 g of white solid. This is the experimentally recovered NaCl, slightly below the mass-balance prediction — a realistic outcome, since some fine NaCl crystals are lost as spray or remain dissolved in residual moisture on the sand. Step 4 — percent recovery of NaCl.

\%\ \text{recovery} = \frac{4.48\ \text{g}}{4.55\ \text{g}} \times 100\% = 98.5\%

Step 5 — percent composition of the original mixture.

\%\ \text{sand} = \frac{7.85}{12.40} \times 100\% = 63.3\%,\qquad \%\ \text{NaCl} = \frac{4.55}{12.40} \times 100\% = 36.7\%

These two percentages sum to $100.0\%$ , confirming the mass balance is internally consistent even though the recovered NaCl mass (4.48 g) fell slightly short of the true NaCl mass (4.55 g) because of the small experimental loss.

Example 2 — Chromatography: computing and comparing retention factors. A green ink is spotted 1.0 cm above the bottom edge of a chromatography strip and developed in a solvent chamber. After development, the solvent front has traveled 15.6 cm from the origin (the spot's starting line, not the paper's edge). The ink resolves into two spots: a blue spot that traveled 9.36 cm from the origin, and a yellow spot that traveled 13.26 cm from the origin. Step 1 — retention factor of the blue component.

R_f(\text{blue}) = \frac{9.36\ \text{cm}}{15.6\ \text{cm}} = 0.600

Step 2 — retention factor of the yellow component.

R_f(\text{yellow}) = \frac{13.26\ \text{cm}}{15.6\ \text{cm}} = 0.850

Step 3 — interpretation. The yellow dye has the larger $R_f$ , meaning it spent proportionally more time associated with the mobile solvent phase and less time adsorbed to the stationary paper fibers; the blue dye is comparatively more strongly retained by the stationary phase (more polar interactions with the paper, in a typical polar stationary phase / less-polar mobile phase system). Step 4 — limiting-case sanity check. If a hypothetical third dye moved exactly with the solvent front, its distance traveled would equal 15.6 cm, giving $R_f = 15.6/15.6 = 1.00$ — the maximum possible value, corresponding to a component with essentially no affinity for the stationary phase. If a dye never left the origin, its distance traveled would be 0 cm, giving $R_f = 0/15.6 = 0.00$ , the minimum possible value, corresponding to total retention. Both bounds confirm $0 \le R_f \le 1$ for any real spot, and the computed values 0.600 and 0.850 sit properly inside that range.

Limiting cases & common traps

Scenario	What actually happens	Common trap
Two liquids with nearly identical boiling points (e.g., within a few °C)	Simple distillation gives poor separation; vapor is only slightly enriched in the more volatile liquid	Assuming any distillation "purifies completely" — fractional distillation or another method is needed
Filtering a true solution (e.g., salt fully dissolved in water)	Everything passes through the filter; nothing is retained	Expecting filtration to remove dissolved solutes — filtration only separates by particle size, not by solubility
Two dyes with identical $R_f$ in a given solvent	They co-migrate and appear as one spot; chromatography fails to resolve them	Concluding a mixture is "one pure substance" just because it shows one spot in a single solvent system
Evaporating a solution that contains a volatile solute (e.g., alcohol dissolved in water)	The solute leaves along with the solvent; nothing selectively remains behind	Assuming evaporation always leaves the solute — it only works cleanly when the solute itself is nonvolatile

The chemist behind it

Modern chromatography traces to the Russian-Italian botanist Mikhail Tsvet, who in 1901–1906 passed plant pigment extracts through columns of powdered calcium carbonate and watched them separate into colored bands — he coined the very word "chromatography" ("color writing") for the striped pattern. Tsvet's insight, that different molecules cling to a solid surface with different strengths while a solvent carries them along at different rates, is precisely the physical principle behind the retention factor used today, whether on paper, in a thin layer of silica, or inside a modern liquid-chromatography column. His technique was largely ignored for decades — pigment chemists doubted that "colorless" separations of similar-looking plant extracts meant anything chemically real — until the 1930s and 1940s, when chemists rediscovered and generalized it into the essential purification and analysis tool it remains.

Origins · 1 minMikhail Tsvet and how this was discovered

Mikhail Tsvet

1872–1919 · early-20th-century plant biochemistry, adsorption chromatography

Ask the tutor: If two liquids have boiling points only 2 °C apart, why does fractional distillation succeed where simple distillation fails?

Can filtration ever separate two dissolved substances from each other?

No. Filtration only sorts by particle size, and once a substance is truly dissolved, its individual particles (ions or molecules) are far smaller than any filter paper's pores. Two substances dissolved together in the same solvent will both pass straight through; separating them requires a technique sensitive to some other property, such as differing boiling points (distillation) or differing polarity/adsorption behavior (chromatography).

Why does the thermometer in a distillation setup sit at the side-arm, not down in the liquid?

The side-arm thermometer measures the temperature of the vapor about to leave and condense, which is what tells you which substance is currently distilling over. A thermometer bulb submerged in the liquid would instead report the boiling liquid's temperature, which can be misleadingly high or fluctuate with the liquid's changing composition as distillation proceeds, rather than cleanly plateauing at the boiling point of the component currently escaping as vapor.

Is a higher

R_f

always "better" in chromatography?

Not inherently — a higher $R_f$ just means a component interacts more weakly with the stationary phase and travels closer to the solvent front. Good chromatographic separation depends on components having sufficiently different $R_f$ values from each other, not on any single component having a high or low value in isolation. An $R_f$ near 1.0 for every spot in a mixture is actually a bad sign, since it means nothing is being resolved from the solvent front.

Does evaporation destroy the dissolved solute?

No — evaporation is a physical change, exactly like the other separation techniques discussed here. The solvent's molecules are converted from liquid to vapor and driven off, but the solute's chemical identity is untouched; it simply loses the solvent molecules that were surrounding it and is left behind as a solid (or, if it is itself volatile, it leaves along with the vapor, which is why evaporation only cleanly isolates nonvolatile solutes).

How is percent recovery different from percent yield in a chemical reaction?

Percent recovery, as used here, compares the mass of a substance retrieved by a physical separation to the mass that was already present in the original mixture — no new substance is created. Percent yield instead compares the mass of a product actually obtained from a chemical reaction to the theoretical mass predicted by stoichiometry from the balanced equation; the two percentages are calculated with an analogous ratio-times-100% structure, but they describe fundamentally different processes.