The Hydrophobic Effect

What the hydrophobic effect actually is

The hydrophobic effect is not a real attractive force between nonpolar groups. Two oil molecules in a vacuum barely feel each other beyond weak London dispersion. What makes them associate in water is the behavior of the solvent. A nonpolar solute cannot donate or accept hydrogen bonds, so the water molecules touching it cannot form their usual roughly tetrahedral network in every direction. To avoid losing hydrogen bonds, the surface waters reorient and become more ordered, wrapping the solute in a semi-structured shell.

Because this shell is more ordered than bulk water, it carries an entropy cost. When two nonpolar solutes come together, their combined surface area is smaller than the sum of the two separate surfaces, so fewer water molecules are forced into the ordered shell. Releasing those waters back into disordered bulk increases the entropy of the system. The apparent 'attraction' between the solutes is really the solvent's drive to maximize its own entropy. This is why chemists emphasize that the hydrophobic effect is a property of water, not of oil.

The thermodynamics: entropy, not enthalpy

Consider transferring a small nonpolar molecule such as methane or butane from an oily phase into water. Kauzmann and later Frank & Evans found the striking result: the process has ΔH near zero (sometimes even slightly negative — the ordered shell makes good hydrogen bonds) yet ΔG is strongly positive. The only way ΔG = ΔH − TΔS can be positive with ΔH ≈ 0 is for ΔS to be large and negative. Dissolving the nonpolar solute orders the water.

• ΔS < 0 — the ordered hydration shell reduces water's entropy; this is the classical hydrophobic penalty.
• ΔH ≈ 0 — enthalpy is nearly neutral near room temperature, which is why the effect is called 'entropy-driven.'
• ΔC_p > 0 — the true fingerprint. Nonpolar hydration produces a large positive heat capacity, because the ordered shell absorbs heat as it 'melts' with rising temperature. This large ΔC_p makes ΔH and ΔS strongly temperature-dependent.

A consequence of the positive ΔC_p is that the hydrophobic effect strengthens as temperature rises from 0 °C toward roughly 60–70 °C, because TΔS grows. At still higher temperatures the balance flips and the effect becomes enthalpically driven. This temperature dependence is exactly why many proteins can be denatured by cold as well as heat — cold denaturation is a direct signature of the hydrophobic effect.

The 'iceberg' picture — and why it's oversimplified

Frank and Evans in 1945 coined the vivid image of 'icebergs': quasi-crystalline, ice-like cages of water forming around nonpolar solutes. The related clathrate hydrates are real crystalline compounds — for example methane clathrate, in which water forms polyhedral cages that trap methane; these ice-like solids exist on the ocean floor and are a serious hazard in gas pipelines. The iceberg model captures the essential physics: hydration shells are more ordered than bulk water.

Modern spectroscopy and simulation have refined the picture. The shell waters are not frozen and are not truly ice-like; they are only modestly more orientationally ordered and, crucially, more slowed in their reorientation than bulk water. For large nonpolar surfaces (bigger than about 1 nm) the physics changes: water cannot maintain its hydrogen bonds against a large flat surface, so it partially dewets and the effect becomes dominated by the enthalpy of the depleted interface rather than by shell ordering. This crossover between small-solute (entropic) and large-surface (interfacial) hydrophobicity, formalized by Lum, Chandler, and Weeks in 1999, is central to how large biomolecular assemblies behave.

Where it drives chemistry: micelles, membranes, and folding

The hydrophobic effect is the master architect of aqueous self-assembly:

• Micelles and surfactants. Amphiphiles like sodium dodecyl sulfate have a nonpolar tail and a polar head. Above the critical micelle concentration (CMC), the tails aggregate to hide from water while the heads face outward. The CMC is set by the free-energy gain of sequestering the tails — each additional CH₂ group in the tail lowers the CMC by a roughly constant factor, a hallmark of hydrophobic driving.
• Lipid bilayers. Phospholipids with two tails form bilayer membranes spontaneously in water. The hydrophobic effect keeps the tails buried in the membrane core and supplies the cohesion that makes cells possible without any covalent bonds holding the sheet together.
• Protein folding. Kauzmann's 1959 insight: a globular protein buries its hydrophobic residues (Leu, Ile, Val, Phe) in an oily core and exposes polar/charged residues to water. The entropic gain from releasing ordered hydration water on collapse is a principal contributor to the modest net stability of proteins (typically only 20–60 kJ/mol).
• Drug binding and molecular recognition. Ligands often gain affinity by displacing ordered water from a hydrophobic pocket — a key lever in medicinal chemistry and in host–guest systems such as cyclodextrins.

Measuring it: partition coefficients and the hydrophobic scale

The practical proxy for hydrophobicity is the octanol–water partition coefficient, reported as log P — the log of the ratio of a compound's concentration in octanol versus water at equilibrium. A more positive log P means a more hydrophobic molecule. It is one of the most cited numbers in drug design; Lipinski's 'Rule of Five' flags oral drug candidates when log P exceeds about 5.

The free energy of transferring a nonpolar group from water to a nonpolar phase scales, to a good approximation, linearly with buried surface area: roughly 20–25 cal per mol per Ų of nonpolar surface removed from water. This surface-area proportionality lets structural biologists estimate folding and binding contributions directly from atomic coordinates, and underlies implicit-solvent models used in molecular simulation. The additivity of methylene increments — each CH₂ contributing a near-constant increment to log P and to transfer free energy — is itself experimental evidence that the effect is fundamentally about surface exposed to water.

A short history

The thermodynamic anomaly was first laid out clearly by Henry Frank and Marjorie Evans in 1945, who introduced the 'iceberg' metaphor for structured water around nonpolar solutes. In 1959, Walter Kauzmann synthesized the evidence and argued that these same interactions are the dominant force stabilizing folded proteins, coining the framing that has defined the field ever since — 'hydrophobic bonding.' Charles Tanford's 1973 monograph The Hydrophobic Effect made the subject a discipline. The theoretical crossover between small and large length scales was clarified by Lum, Chandler, and Weeks in 1999. Today the hydrophobic effect is understood as an emergent solvent property whose entropic and enthalpic balance shifts with temperature, pressure, solute size, and added salts (the Hofmeister series), rather than a single fixed 'force.'

Thermodynamic signature of dissolving a small nonpolar solute in water at 25 °C

Quantity	Sign / trend	Molecular interpretation
ΔG (transfer to water)	Positive (unfavorable)	Nonpolar solute prefers to leave water
ΔH	≈ 0 or slightly negative	Ordered shell makes good H-bonds
ΔS	Large and negative	Water ordered into a cage around solute
ΔC_p	Large and positive	Ordered shell 'melts' as temperature rises
Effect of warming	Association strengthens	Entropy penalty grows, then reverses at high T