Recrystallization

The seven-step protocol

1. Choose the solvent. The compound dissolves freely in hot solvent and barely in cold. Test small batches first.
2. Dissolve in minimum hot solvent. Add solvent in small portions to the boiling slurry until the solid just dissolves. Adding extra is the most common mistake — it suppresses crystallization on cooling.
3. Decolorize if needed. Add a small spatula tip of activated charcoal, boil briefly, and remove all of it in the hot filter. Charcoal binds colored impurities indiscriminately.
4. Hot filter. Pour through a pre-warmed fluted filter paper into a warmed flask. Speed matters: each minute of cooling on the filter risks crystallization in the funnel.
5. Cool slowly. Let the flask cool at room temperature first (not in a draft), then in an ice bath. Slow cooling produces big, clean crystals.
6. Filter the crystals. Vacuum-filter through a Büchner funnel. Press the cake to remove mother liquor.
7. Wash and dry. Two washes with a small volume of cold pure solvent. Dry on the filter under vacuum or in a desiccator.

The setup

STEP 1: dissolve         STEP 2: hot filter        STEP 3: cool & filter
         + boil chips
   ┌──────────────┐         ┌──────────────┐        ┌──────────────┐
   │ ╔══════════╗ │         │ fluted paper │        │              │
   │ ║░ solid + ║ │         │  ┌────────┐  │        │              │
   │ ║  hot     ║ │  ─────▶ │  │ ▽▽▽▽▽▽ │  │ ─────▶ │  ╔════════╗  │
   │ ║  solvent ║ │         │  │ insol. │  │        │  ║crystals║  │
   │ ╚════╤═════╝ │         │  │ trapped│  │        │  ╠════════╣  │
   │      │ heat  │         │  └───┬────┘  │        │  ║ mother ║  │
   └──────┴───────┘         └──────┼───────┘        │  ║ liquor ║  │
                                   │ filtrate       │  ╚════════╝  │
                                   ▼                └─────┬────────┘
                          warmed receiver flask           │
                                                          ▼
                                                   Büchner + vac
                                                   wash with cold
                                                   pure solvent

Picking the solvent

The ideal solvent dissolves your compound 10–20× more at boiling point than at 0 °C. A two-minute screen on 50 mg of test sample: dissolve in ~1 mL solvent at room temperature — if it dissolves cold, reject (too good). Heat to boiling — if it doesn't dissolve in up to 5 mL, reject (too poor). Otherwise cool and watch: if crystals form on cooling, the solvent is good.

Common single-solvent options

Solvent	Polarity	BP (°C)	Best for	Hazard	Notes
Water	Very polar	100	Salts, sugars, polar amides	None	Slow drying; freezes if cooled too far
Methanol	Polar	65	Phenols, polar acids	Toxic vapor	Hygroscopic; collects ambient water
Ethanol (95%)	Polar	78	General organic	Flammable	Universal default for student labs
Acetone	Polar	56	Hydroxylated, ketones	Flammable	Reacts with primary amines
Ethyl acetate	Medium polar	77	Esters, mid-polarity solids	Flammable	Easy to evaporate; may solvate
Toluene	Low polar	111	Aromatic hydrocarbons, amides	Toxic vapor, flammable	Higher BP gives wider window
Hexane	Non-polar	69	Long-chain alkanes, fatty acids	Flammable, neuropathy	Often used as anti-solvent
Chloroform / DCM	Low polar	61 / 40	Steroids, lipophilic drugs	Toxic, suspected carcinogen	Dense; vapor depresses cooling

Mixed-solvent recrystallization

When no single solvent fits, mix a "good" solvent (compound dissolves freely) with a "poor" solvent (compound is nearly insoluble). The two must be miscible. Famous pairs:

• Ethanol / water — the universal pair for moderately polar organics. Ethanol dissolves; water crashes.
• Acetone / water — for compounds too polar for ethanol alone.
• Ethyl acetate / hexane — for non-polar synthetic intermediates.
• Diethyl ether / petroleum ether — low-temperature recrystallization for thermally fragile solids.
• Methanol / diethyl ether — for ammonium salts and hydroxylated drugs.

Procedure: dissolve in boiling good solvent, then add boiling poor solvent dropwise until persistent cloudiness, then a few drops back of good solvent to reclear. Cool slowly. The blend ratio at the cloud point sets the solubility just above saturation, ideal for slow nucleation.

Worked example — purity convergence

You have 5.00 g of a 90%-pure compound (10% soluble impurity). Solubility in ethanol/water 4:1 is 25 g/100 mL hot, 2.5 g/100 mL cold. Single recrystallization:

1. Dissolve in 22 mL hot solvent (just under saturation).
2. Cool to 5 °C. Solubility = 2.5 × 0.22 = 0.55 g remains dissolved.
3. Crystallize: 5.00 − 0.55 = 4.45 g; ~90% is product → ~4.00 g product.
4. Mother liquor carries 0.55 g (mostly impurity) + ~0.5 g lost product.

Yield      ≈ 4.00 / (5.00 × 0.90) = 89% of available product
Purity now ≈ 4.00 / (4.00 + 0.05) = 98.8%

A second cycle on those 4.00 g takes purity to ~99.8%, yielding ~3.20 g. Roughly 10× impurity reduction per pass. Two cycles get you from synthesis-grade to publication-grade.

Recrystallization vs other purifications

	Recrystallization	Column chromat.	Distillation	Sublimation	Extraction
Sample state	Solid	Solid or liquid	Liquid	Volatile solid	Liquid + solute
Scale	mg to tonne	mg to kg	mL to ML	mg to g	mg to kg
Cost / gram	Low (solvent only)	High (silica)	Low (energy)	Medium	Low
Best purity	>99.9%	~99%	>99.99%	>99%	~95%
Speed	Hours / cycle	0.5–4 hours	Minutes–hours	Hours	Minutes
Industrial use	Pharma final step	Rare at scale	Petroleum	Caffeine, iodine	Most workups

Recrystallization is the dominant final purification for solid pharmaceuticals because it scales cheaply and gives FDA-acceptable purity. Column chromatography is rarely tolerated past pilot scale because silica costs and waste are prohibitive.

Variants

• Slow vapor diffusion — for X-ray-quality single crystals. A small flask of good-solvent solution sits inside a larger flask of poor solvent; vapor migrates over days.
• Layering — gently layer poor solvent on top of good-solvent solution. Single crystals form at the interface.
• Sublimation–recrystallization — for compounds that decompose in solution. Vaporize under vacuum, condense on a cold finger.
• Industrial repeated batch — pharma final-step crystallization is often two or three sequential recrystallizations to control polymorph and purity simultaneously.

Common pitfalls

• Adding too much solvent. The most common mistake. Solvent is added impatiently in big chunks; the solution becomes dilute; on cooling, only a fraction crystallizes. Add solvent in small portions and confirm full dissolution at boiling before adding more.
• Cooling too fast. Quenching in ice traps mother liquor in fast-growing crystals. Let the flask cool to room temperature first.
• Skipping hot filtration. Insoluble dust or undissolved starting material gets trapped in the crystals. Always pass through a fluted filter.
• Charcoal that doesn't get filtered out. Activated carbon stays as black particles unless captured by a fine filter or Celite plug.
• Washing with hot solvent. Dissolves the crystals away. Always pre-cool the wash solvent.
• No crystals form. Solution is too dilute (boil off solvent), too pure (need a seed crystal — scratch the flask interior), or supersaturation hasn't propagated past the kinetic barrier (cool further, leave overnight).
• Wrong polymorph. Cooling rate, solvent, and seeding all influence which lattice forms. If the literature mp doesn't match, you may have a different polymorph.

Knowing when to stop

Each recrystallization halves or better the impurity but costs 10–20% yield. After two or three cycles you're usually at >99% pure with 60–70% material left. A fourth round costs more material than it removes impurity. Switch to HPLC or column for the final 0.1%. For most prep and research, two recrystallizations from a well-chosen solvent are sufficient.