Evaporation and selection effects

Evaporative cooling is tempting to explain as a selection effect. The molecules that leave the liquid are not a random sample. They come disproportionately from the high-energy tail, so the molecules left behind have a lower average kinetic energy. In social science terms, the high-energy molecules select out of the liquid, and the remaining population looks colder.

That is true, but it is not the main energy accounting. It explains which molecules are likely to leave, not what the escaping molecules have to spend their energy on.

A molecule in a liquid is not just bouncing around freely. Its neighbors tug on it. In water, those tugs are especially strong because water molecules are polar: the slightly positive hydrogen side of one molecule is attracted to the slightly negative oxygen side of another. That is hydrogen bonding. Nonpolar liquids have weaker versions of the same basic attraction, mostly from temporary shifts in electron clouds. The details differ, but the result is the same: each molecule is held, loosely, by the molecules around it. To become part of the vapor, it has to pull away from those neighbors. That is more like separating magnetic blocks than like simply removing the fastest blocks from a box. The departing molecule is selected for having enough energy to escape, but as it escapes, much of that energy is spent getting unstuck.

This is why evaporation really is endothermic. Escaping molecules are not just carrying motion away. They are using much of that motion to pull themselves free from the liquid. That energy has to be supplied by the liquid or its surroundings, which is why the liquid cools. For water, the price of getting unstuck is large: the latent heat of vaporization near room temperature is about 44 kJ/mol. The residual “fast molecules carried off extra kinetic energy” effect is on the order of 1 kJ/mol. So for water, the selection effect is only a few percent of the cooling budget; the work of getting molecules unstuck dominates.

The same logic applies to less polar liquids, though the numbers are smaller. Nonpolar hydrocarbons have weaker intermolecular attractions than water, but evaporation still mostly costs energy because molecules have to separate from one another. Selection tells us which molecules get out. Separation tells us why the process absorbs so much heat.

One way to see this is to imagine turning down the attraction knob. At exactly zero intermolecular attraction, there would not really be a liquid, but in the limit there is no potential well to climb out of. The only cooling mechanism left is the selection effect: the particles that leave carry above-average kinetic energy. As intermolecular attraction increases, the well gets deeper, and more of the energy budget goes into pulling molecules apart. Water is far over on that side of the spectrum.

Condensation is the reverse. When vapor molecules return to the liquid, they fall into the attractive potential well created by neighboring molecules. Potential energy becomes kinetic energy as the molecules accelerate towards each other via attraction. That is why condensation is exothermic. It is also why steam burns are so severe: when steam condenses on skin, it releases this energy right at the surface, in addition to being hot water.

This is where the social science instinct can mislead. Selection effects are powerful, and they explain a lot. Here, they explain the filtering process, not the main thermodynamic bill.

Correction, July 4, 2026: The original version of this post overstated the selection-effect interpretation and wrongly implied that evaporation is not genuinely endothermic. That mistake came from leaning too hard on a social-science intuition: when selection is such a dominant force in the problems you usually think about, it is easy to see selection everywhere. In evaporation, selection is real, but the dominant energy cost is separating molecules from intermolecular attractions.