Why does salt disrupt ice formation?
When water freezes, its molecules must line up into a rigid, repeating crystal lattice. Dissolved salt ions get in the way: they are surrounded by their own shell of water molecules and disrupt the hydrogen-bonding network that would otherwise organize into ice [3]. This disruption means the water must be cooled further before the remaining water molecules can arrange themselves into a solid. The traditional explanation is that the ions increase entropy—a measure of disorder—in the liquid, making the solid state less favorable until the temperature drops enough to overcome that disorder [3].
At low salt concentrations, the amount of freezing-point lowering depends mainly on how many dissolved particles are present, not on what those particles are. This is why the effect is called a colligative property: it is a property of the number of particles, not their chemical identity [4]. For example, one molecule of sodium chloride (NaCl) splits into two ions in water, so it has roughly twice the effect of a single-particle solute like sugar at the same molar concentration.
Does it matter which salt you use?
At higher salt concentrations, the simple colligative picture breaks down and the specific identity of the ions starts to matter. A 2025 study of five different lithium salts found that the freezing-point trends did not follow the familiar Hofmeister series—a ranking of ions by how strongly they affect water structure [3]. Instead, the researchers found that the mobility of water molecules within the ions' hydration shells—the layers of water clinging to each ion—was a primary determinant of freezing behavior [3]. This suggests that how tightly water is held by the ions, and how easily it can move, matters more than the ion's intrinsic ability to disrupt hydrogen bonds.
The effect also depends on what else is in the water. In mixtures containing a polymer like PEG along with salt, some salts (NaF, NaCl, NaI) produced freezing-point depressions larger than the sum of their individual effects, consistent with strong ion hydration pulling water away from the polymer [5]. In contrast, NaSCN gave a nearly additive response, reflecting weaker hydration and direct association with the polymer chains [5]. So the answer to 'how much does salt lower the freezing point' depends on the whole solution, not just the salt.
What happens in real materials like concrete?
In porous materials such as concrete, salt can become trapped in thin liquid films between growing ice and charged surfaces. A 2022 theoretical study proposed that this nanofluidic salt trapping can explain why concrete pavements damaged by de-icing salts experience pressures far greater than simple water expansion would predict [1]. The trapped salt lowers the freezing point locally, but as ice grows in charged pores, the rejected ions create enormous disjoining pressures—forces that push surfaces apart—until the ions are removed by precipitation or surface adsorption at lower temperatures, releasing the pressure and allowing complete freezing [1]. This mechanism could also help explain how some organisms and wet materials survive temperatures far below the normal freezing point of bulk water [1].
How do scientists model and measure this?
Researchers use a combination of experiments and computer simulations to understand freezing-point depression. A 2022 study tested a computer model called Madrid-2019 against experimental measurements for several salts, including NaCl, KCl, LiCl, MgCl2, and Li2SO4, and found that it reproduced the measured freezing-point depression fairly well across all compositions up to the eutectic point—the concentration at which the solution can no longer lower the freezing point further [4]. This kind of model helps predict how salt will behave in real-world conditions, from seawater to battery electrolytes.
Earlier work from 1974 developed equations for freezing-point depression in porous materials and presented experimental results using sodium chloride, linking changes in salt concentration during freezing to the amount of unfrozen water remaining [2]. Together, these studies show that the basic physics of salt lowering the freezing point is well understood at low concentrations, while the details at high concentrations and in complex materials remain an active area of research.
Sources used in this answer
Freezing point depression and freeze-thaw damage by nano-fuidic salt trapping
Trapped salt in thin liquid films between ice and charged surfaces can lower the freezing point and create large disjoining pressures, potentially explaining freeze-thaw damage in concrete and cryo-tolerance in porous materials.
Effects of salt concentration changes during freezing on the unfrozen water content of porous materials
Equations for freezing-point depression and experiments with sodium chloride show how changes in salt concentration during freezing affect the amount of unfrozen water in porous materials.
Molecular Insights into Anion-Specific Freezing Point Depression in Lithium Salt Solutions
In lithium salt solutions, freezing-point depression arises from a complex interplay of ion-water interactions, and the mobility of water in ions' hydration shells is a primary determinant of freezing behavior, with trends deviating from the Hofmeister series.
Freezing point depression of salt aqueous solutions using the Madrid-2019 model
The Madrid-2019 computer model reproduces experimentally measured freezing-point depression for several salts, including NaCl, across a range of concentrations up to the eutectic point.
Quantifying Hofmeister Effects on Polymer–Water Interactions through Freezing Point Depression
In salt-polymer-water mixtures, strongly hydrating salts produce synergistic freezing-point depressions beyond additive effects, while weakly hydrating salts give nearly additive responses, linking ion hydration strength to macroscopic freezing behavior.