Azeotropes

Why distillation stops working

Distillation works because the vapor rising off a boiling mixture is usually richer in the more volatile component than the liquid it came from. Boil a 10% ethanol wash and the first vapor is maybe 50% ethanol. Condense that, boil it again, and you climb toward purer and purer ethanol. Each "plate" of a distillation column is one of these little re-boilings stacked on top of the last. That is the entire trick.

An azeotrope is the place where the trick stops paying off. At the azeotropic composition the vapor coming off has the exact same composition as the liquid underneath it. Condense it, re-boil it, do it ten thousand times — you get back the same mixture every time. The column has nothing left to enrich. You have hit a thermodynamic wall, and no number of plates, no taller column, no slower reflux will get you through it.

Normal mixture          Azeotrope
liquid:  x = 0.30        liquid:  x = 0.894
vapor:   y = 0.55  ↑     vapor:   y = 0.894  =  (no enrichment!)
         enrichment               vapor identical to liquid

The name says it all: from Greek a (without) + zeo (boil) + tropos (change) — "boiling without change." The composition does not change as the mixture boils.

The vapor–liquid equilibrium condition

For an ideal mixture, each component follows Raoult's law: its partial pressure is its mole fraction times the pure-component vapor pressure.

p_A = x_A · P°_A           (ideal, Raoult)
p_B = x_B · P°_B
P_total = p_A + p_B

Real mixtures deviate. We patch Raoult's law with an activity coefficient γ that captures how much molecule A "likes" being surrounded by molecule B:

p_A = γ_A · x_A · P°_A      (real, modified Raoult)
p_B = γ_B · x_B · P°_B

When γ > 1 (positive deviation), the components dislike each other — A and B push each other into the vapor, so the total vapor pressure is higher than ideal and the boiling point is pushed down. When γ < 1 (negative deviation), they attract strongly, vapor pressure is suppressed, and boiling point goes up.

The vapor composition is set by the ratio of partial pressures:

y_A = p_A / P_total = (γ_A · x_A · P°_A) / P_total

The azeotrope is the composition where the deviation has bent the total-pressure curve enough to create a maximum (positive) or minimum (negative), and at that extremum the math forces y_A = x_A for every component simultaneously. That single equality — vapor composition equals liquid composition — is the formal definition of an azeotrope.

Minimum-boiling vs maximum-boiling azeotropes

There are exactly two flavors, and they come from the sign of the deviation:

Positive (minimum-boiling) azeotrope. Positive deviation from Raoult's law (γ > 1) gives a vapor-pressure maximum, which is a boiling-point minimum. The azeotrope boils below both pure components. This is the common case — about 90% of known azeotropes are minimum-boiling.

Temperature (at 1 atm)
  │
100°C ●─ pure water
  │    ╲                              ╱
  │     ╲  bubble (liquid) curve    ╱
  │      ╲___                    __╱
  │          ╲__              __╱  ● 78.4°C pure ethanol
  │             ╲__        __╱
  │                ╲______╱  ← 78.2°C azeotrope
  │              (curves KISS here, y = x)
  └──────────────────────────────────────→
  0% EtOH          89.4 mol%         100% EtOH

Negative (maximum-boiling) azeotrope. Negative deviation (γ < 1) gives a vapor-pressure minimum and a boiling-point maximum. The azeotrope boils above both pure components. Hydrochloric acid is the textbook case: HCl–water boils at 108.6 °C, hotter than pure water and far hotter than pure HCl (a gas at room temperature). Boil dilute or concentrated HCl long enough and it both drifts to the same 20.2% HCl — which is why "constant-boiling HCl" is a usable analytical standard.

Common azeotropes and their fixed points

System	Type	Azeotrope b.p. (1 atm)	Azeotropic composition	Pure b.p.'s
Ethanol + water	Minimum	78.2 °C	95.6 wt% (89.4 mol%) EtOH	78.4 / 100 °C
2-Propanol + water	Minimum	80.4 °C	87.7 wt% IPA	82.3 / 100 °C
Benzene + water	Minimum (heteroazeotrope)	69.2 °C	91 wt% benzene	80.1 / 100 °C
Ethanol + benzene	Minimum	68.2 °C	32.4 wt% EtOH	78.4 / 80.1 °C
Water + formic acid	Maximum	107.1 °C	77.5 wt% formic acid	100 / 100.8 °C
Water + hydrogen chloride	Maximum	108.6 °C	20.2 wt% HCl	100 / −85 °C
Water + nitric acid	Maximum	120.2 °C	68 wt% HNO₃	100 / 83 °C
Acetone + chloroform	Maximum	64.4 °C	20 wt% acetone	56.1 / 61.2 °C
Benzene + toluene	none	—	—	80.1 / 110.6 °C

Notice the last row: benzene and toluene are chemically similar (γ ≈ 1, nearly ideal) and their boiling points are 30 °C apart, so no azeotrope forms and they distill cleanly. That is exactly why this pair is the standard teaching example of a well-behaved distillation.

Worked example: how far can a still take you?

Suppose you ferment a sugar wash to 12 wt% ethanol and run it through a perfect column. Where do you end up, and why not pure?

The relative volatility α tells you the enrichment per plate. For ethanol–water at low ethanol content, α ≈ 8–10 (ethanol escapes far more easily than water), so a few plates take you quickly from 12% toward 80%+. But α is not constant — it falls as you approach the azeotrope:

α = (y_EtOH / x_EtOH) / (y_water / x_water)

x_EtOH = 0.10  →  α ≈ 8     (huge enrichment, easy)
x_EtOH = 0.50  →  α ≈ 2.2   (still working)
x_EtOH = 0.85  →  α ≈ 1.2   (barely enriching)
x_EtOH = 0.894 →  α = 1.00  (azeotrope — DEAD STOP)

When α reaches exactly 1.00, the vapor is no richer than the liquid and the column delivers no further enrichment. That is the azeotrope at 89.4 mol% (95.6 wt%). The energy cost makes it worse: the number of plates needed to climb the last few percent toward the azeotrope explodes, because each plate buys almost nothing. The practical ceiling for ordinary distillation of fermented spirits sits at the azeotrope — 95.6 wt%, which is about 97.2% ABV (≈194 proof) — for exactly this reason, and "190 proof" (95% ABV) Everclear sits just shy of that azeotropic limit.

Breaking the azeotrope: getting to 100%

Anhydrous ("absolute") ethanol is needed for biofuel blending, pharmaceutical synthesis, and as a solvent. Since you cannot distill past 95.6%, industry uses one of four tricks to get the last 4.4% of water out:

• Azeotropic distillation (entrainer). Add a third liquid — historically benzene, now cyclohexane or toluene — that forms a new ternary azeotrope boiling even lower (the cyclohexane–ethanol–water ternary boils at ~62.1 °C). That ternary carries the water out the top, leaving anhydrous ethanol in the pot. This was the dominant industrial route for most of the 20th century.
• Extractive distillation (solvent). Add a high-boiling solvent such as ethylene glycol or glycerol that hydrogen-bonds to water, suppressing water's volatility and lifting the ethanol–water azeotrope out of the operating range. The solvent leaves with the bottoms and is recovered separately.
• Pressure-swing distillation. The azeotropic composition shifts with pressure. Ethanol–water's azeotrope vanishes entirely below about 70 mmHg (≈9.3 kPa), so a vacuum column can in principle reach pure ethanol. Two columns at different pressures can also ping-pong the composition across the moving azeotrope.
• Molecular sieves (the modern winner). 3 Å zeolite beads have pores that admit water (kinetic diameter 2.6 Å) but exclude ethanol (4.4 Å). Pass the 95.6% azeotrope through a packed bed and the sieve physically adsorbs the water, delivering >99.5% ethanol with no entrainer and far less energy. Today this is how most fuel-grade and lab-grade absolute ethanol is made.

Where azeotropes show up

• Spirits and fuel ethanol. Every vodka, whisky, and bottle of Everclear runs into the 95.6% ceiling. Fuel-grade E85 and E100 need molecular sieves or entrainers because flex-fuel engines and gasoline blending demand anhydrous ethanol — water causes phase separation in the tank.
• Constant-boiling acids as standards. 20.2% HCl and 68% HNO₃ are azeotropes, so they boil at a fixed temperature and composition. Analytical chemists historically used constant-boiling HCl as a primary standard precisely because the azeotrope guarantees a reproducible concentration.
• Solvent recovery and recycling. Industrial plants that recover spent solvents constantly fight azeotropes — water/THF, water/acetonitrile, and chlorinated-solvent mixtures all azeotrope and require pressure-swing or extractive columns to purify for reuse.
• Refrigerant blends. "Near-azeotropic" refrigerant mixtures like R-410A are engineered to evaporate and condense at almost constant composition so the blend doesn't fractionate in the loop. R-507A and the older R-502 are true azeotropes used for exactly this reason — they behave like a single fluid in the cycle.
• Dean–Stark traps. Organic chemists drive esterifications and acetal formations to completion by boiling off the water as a benzene or toluene heteroazeotrope, condensing it in a Dean–Stark trap where the water layer separates and the dry solvent returns to the flask.

Common misconceptions and pitfalls

• "An azeotrope is a compound." No — it is a mixture held by intermolecular forces, not bonds. The dead giveaway is that its composition changes with pressure; a real compound has fixed stoichiometry. Drop the pressure on ethanol–water and the azeotrope moves and eventually disappears.
• "A taller column will eventually get past it." No. The azeotrope is where relative volatility α = 1, so the enrichment per plate is zero. Adding plates multiplies zero. You need to change the chemistry (entrainer, solvent) or the conditions (pressure), not the hardware.
• "All azeotropes boil low." Only positive ones. Maximum-boiling (negative) azeotropes like HCl–water and HNO₃–water boil higher than either pure component. In a max-boiling azeotrope the still removes the volatile excess and drives the pot composition toward the azeotrope.
• "95% and 100% ethanol are basically the same." For many uses, yes — but that residual ~5% water ruins moisture-sensitive Grignard reactions, dehydrating syntheses, and fuel blending. The whole point of breaking the azeotrope is that last few percent matters.
• "Heteroazeotropes are the same as homogeneous ones." A heteroazeotrope (like benzene–water) is two immiscible liquid layers that together boil at constant composition; the condensed distillate splits into two phases. That phase split is what makes Dean–Stark water removal possible and is exploited in entrainer-based dehydration.
• "Azeotropic composition is a universal constant." It is only fixed at a given pressure. Tabulated values (95.6 wt% EtOH, 78.2 °C) are always quoted at 1 atm. Change the pressure and both the composition and the boiling point move.