Ethanol cannot be distilled past the 95.6% azeotrope, so to reach above 99% you must remove the last water chemically: quicklime binds 3.1 g of lime for every gram of water, and a dried molecular sieve takes water out while leaving the ethanol. Neat 90-95% ethanol is enough for engines built for it, but a blend with gasoline needs the dry product, because wet ethanol separates from fuel.
Part of Ethanol Fuel Production
Dehydration removes the last 4-5% of water from distilled ethanol that distillation alone cannot take out, producing anhydrous fuel-grade alcohol.
Why Water Stays After The Last Distillation
At 95.6% ethanol by mass the vapour and liquid have the same composition (Achieving 90%+ Purity). More boiling changes nothing, so the remaining water has to be taken up by something that holds water more tightly than ethanol does.
Do you need it? The parent says 90-95% ethanol works adequately in many engines, especially in warm weather, and dehydration is optional unless you run a flex-fuel system with exact calibration. Blending is the exception: hydrous ethanol and gasoline do not stay mixed, so an E10-E20 blend needs the dry product (Ethanol in Gasoline Engines).
Method A: Quicklime
Quicklime (CaO) reacts with water: CaO + H₂O → Ca(OH)₂. Molar masses 56.08 and 18.02 give 56.08 ÷ 18.02 = 3.11, about 3.1 g of lime per gram of water. Fuel-grade (anhydrous) ethanol means above 99%.
Step 1: Get fresh quicklime. Lime that has sat in damp air has already turned to Ca(OH)₂ or CaCO₃ and takes up no water (Limestone Burning describes making it).
Step 2: Add it in portions, to ethanol in an open vessel with a loose cover, away from any flame. Do not seal a bottle.
Step 3: Stand for at least 24 hours, shaking or stirring from time to time.
Step 4: Separate. The parent says filter after 24 hours. Filtering alone leaves the ethanol with fine lime and with whatever water the lime did not take; the standard laboratory practice, recalled here and not sourced in the corpus, is to distil the ethanol off the lime instead (gentle heat, condenser, same vapour precautions as any still).
Step 5: Test by the gasoline test below.
Worked Example: Quicklime For A Litre Of Azeotrope
A litre of 95.6% ethanol weighs about 804 g (recalled), of which 4.4% is water: 0.044 × 804 = 35 g. The lime that exactly binds it is 35 × 3.11 = 110 g. Use more than that, because the lime is never pure and the reaction slows as the water runs out; a working figure of about twice, 220 g, is an assumption to adjust from your own results.
The reaction also gives off about 65 kJ per mole of water (Slaking). 35 g of water is 35 ÷ 18.02 = 1.96 mol, so 1.96 × 65 = 127 kJ. A litre of ethanol at about 2.4 kJ/kg·K and 0.8 kg needs 127 ÷ (0.8 × 2.4) = 66 K to absorb it. From 20°C that reaches 86°C, above ethanol’s 78.4°C boiling point, if all the heat went into the liquid at once. The lime and vessel absorb part of it and the reaction is gradual, but the arithmetic is why the lime goes in by portions and the vessel is vented.
Method B: Molecular Sieve
A zeolite with pores of about 3 angstroms (type 3A) admits water and excludes ethanol, so a column of dried beads takes water from 95% ethanol passed through it. Not every desiccant is a sieve: silica gel and larger-pore zeolites also take up ethanol and are not suitable. The parent suggests salvaged desiccant packs; check the type, and test the result. Dry the beads before use by heating, and reheat to regenerate them after they take up water. The corpus has no sourced capacity figure; weigh the beads before and after to learn how much water a bed holds.
The parent also lists anhydrous calcium chloride. Calcium chloride forms a compound with ethanol and is generally avoided for drying it (recalled, not sourced in the corpus), so prefer lime or a sieve.
Test The Result
Mix a small amount with an equal amount of gasoline in a clean clear jar, away from flame, and shake. A clear single liquid suggests the ethanol is dry enough to blend; cloudiness or two layers means too much water. This is a field check inferred from how ethanol, water and gasoline separate, not a published specification, so confirm it with a hydrometer reading of the dried ethanol as well.
Common Mistakes
- Using spent lime. It has no water capacity left.
- Sealing the vessel. The reaction heat can boil the ethanol and burst a closed bottle.
- Skipping the test, then finding fuel that will not stay mixed.
Quicklime burns, and the vapour burns
Quicklime is caustic to skin, eyes and lungs and reacts violently with water; wear eye protection and gloves and keep it out of the air (Protective Gear, Spill Response). The warm ethanol gives off vapour that ignites at 3.3-19% in air: work outdoors, with no flame.
What’s Next
- Learn next: Engine Use, to see what strength each engine needs
- Combine with: Fuel Line Material Compatibility, since ethanol attacks rubbers and some plastics