Ask three suppliers which de-icing salt is best and you get three answers, because “best” is being measured on a different axis each time. The two axes that decide most winter maintenance contracts – how much ice a salt can melt and how much damage it does to the pavement underneath – do not rank the three mainstream salts in the same order. In fact they rank them in exactly opposite orders.
This page sets out both rankings, explains why the corrosion ranking looks counter-intuitive, and shows how the two rankings translate into the split most exporters already work to: the dihydrate for Japan and Korea, the near-anhydrous grade for North America and Russia.
The two rankings
| Salt | CAS | Melt performance (best to worst) | Corrosion to the pavement surface (best to worst) |
|---|---|---|---|
| Calcium chloride | 10043-52-4 (anhydrous) / 10035-04-8 (dihydrate) | 1 – strongest | 1 – least damaging |
| Magnesium chloride | 7791-18-6 (hexahydrate) | 2 | 2 |
| Sodium chloride | 7647-14-5 | 3 – weakest | 3 – most damaging |
Read that table again, because it is the whole commercial argument in three lines: the cheapest salt is also the weakest and the most destructive, and the strongest is also the gentlest on the surface. If melt performance and pavement damage were the only criteria, sodium chloride would never win. It wins on one criterion only – delivered cost per tonne – and the entire de-icing market is a negotiation between that criterion and the other two.
That is not a marketing claim. It is the reason calcium chloride commands a price multiple over rock salt and still owns the premium end of the market: airports, bridges, ramps, multi-storey car parks and any surface where resurfacing costs more than the salt.
Why the corrosion ranking runs that way
The mechanism is not chemical attack in the first instance – it is physical.
Salt damages pavement through salt scaling: brine enters the surface pores, and when the water freezes again the salt crystallises inside the pore structure and the hydraulic and crystallisation pressure spalls the surface layer. Repeated freeze-thaw cycles with a salt that dries out and re-crystallises do the most damage. Sodium chloride is the classic case: it has limited moisture retention, so the surface cycles between brine and dry salt crystals, and each cycle prises a little more paste off the surface.
Calcium chloride behaves differently because it is strongly hygroscopic. It holds water, keeps the surface in a liquid or damp state, and therefore spends far less time in the crystallising phase that produces scaling pressure. The surface stays wet rather than cycling through crystallisation, and wet-and-drained is measurably kinder to a concrete or asphalt surface than wet-freeze-crystallise.
Magnesium chloride sits between the two, which is exactly what its hygroscopicity would predict.
The important caveat: this is a pavement ranking, not a steel ranking
All three salts are chlorides. On reinforcing steel, galvanised steel, aluminium and zinc coatings the ranking is a different conversation, and it is not the one above. Chloride ions depassivate steel and initiate pitting regardless of which cation delivered them, and calcium chloride is particularly aggressive to galvanising in direct contact. A specifier who reads “calcium chloride is the least corrosive” and applies it to a bridge deck with galvanised guardrail and post-tensioned tendons has misread the sentence.
The correct reading is narrow and precise: on the pavement surface itself, where scaling and freeze-thaw dominate, sodium chloride does the most damage and calcium chloride the least. Everything to do with embedded metal is governed by the chloride content of the mix and by the drainage, not by this ranking.
Melt performance in numbers
Melt performance is governed by how far a salt can depress the freezing point of water before the brine itself freezes. Approximate eutectic temperatures for the three saturated brines:
| Salt | Approximate eutectic temperature | Behaviour | Dissolution heat |
|---|---|---|---|
| Sodium chloride | About -21 deg C | Stops working in a hard cold snap; the workhorse for ordinary winters | Slightly endothermic – takes heat from the surface |
| Magnesium chloride | About -33 deg C | Extends the working range and holds moisture on the surface | Exothermic |
| Calcium chloride | About -50 deg C | Keeps melting in conditions where the others have stopped | Strongly exothermic – releases heat as it dissolves |
Two practical consequences follow. First, the exothermic salts melt faster at the start, because the heat of dissolution is doing part of the work – which is why a calcium chloride application shows a result in minutes where rock salt takes considerably longer. Second, the working range matters more than the average winter: a network that only ever sees -5 deg C can use rock salt indefinitely, and a network that sees -25 deg C cannot.
Eutectic figures are concentration-dependent and are quoted here as approximate values for the saturated brine. For a tender, ask the supplier for the freezing-point curve of the specific grade rather than relying on a single number.
Why Japan and Korea buy the dihydrate, and North America and Russia buy the near-anhydrous grade
This is the second decision, and it is a logistics decision rather than a chemical one. Once calcium chloride has been chosen, the question becomes which hydrate to ship.
| Market | Grade normally specified | CAS | Why |
|---|---|---|---|
| Japan, Korea | Dihydrate, 74% flakes | 10035-04-8 | Short sea freight, so the water of crystallisation costs little to move; flake form suits the spreading equipment in use; humid maritime climate means the deliquescent 94% grade is harder to store |
| USA, Canada, Russia | Near-anhydrous, 94% | 10043-52-4 | Long inland freight, so every tonne of water is paid for twice – once to ship it and once to evaporate it; severe continental winters need high-concentration brine; long winter seasons justify the better packaging discipline |
The underlying rule is simple arithmetic: the dihydrate is 74% calcium chloride and 26% water, and the near-anhydrous grade is 94% calcium chloride. Shipping the dihydrate inland means paying freight on water. Shipping it across a short sea lane means paying almost nothing extra for water that is harmless in the application. That is why the same product decision reverses between two export markets, and why a supplier who quotes one grade to every destination is not doing the freight calculation.
The full comparison of the two forms – assay basis, water in the process, caking, packaging – is in our guide to calcium chloride 74% vs 94%.
Where sodium chloride still wins
Cost. Nothing else. Rock salt is the cheapest de-icer by an order of magnitude, and a highway authority treating hundreds of kilometres of road cannot buy calcium chloride at that scale. The rational position for most networks is a split: rock salt for general roads where resurfacing is budgeted and the drainage is adequate, calcium chloride for the structures where the pavement is expensive and the failure is visible – bridge decks, ramps, elevated sections, car park decks, airport aprons.
The mistake is not using rock salt. The mistake is using it on a structure whose replacement cost is orders of magnitude higher than the salt saving, and then discovering the chloride has reached the reinforcement.
Where the non-chloride option fits
A third route exists for sites where chloride itself is the problem – not the pavement, but the run-off: drinking-water catchments, airport aprons where aircraft skins and galvanised equipment are exposed, and any site with an environmental consent that caps chloride discharge. Sodium formate and potassium acetate are the two products used there, and they are bought for what they leave behind rather than for melt speed.
The trade-offs are real: several times the cost of rock salt, no exothermic kick, and a high oxygen demand in the receiving water. Our note on sodium formate as a de-icing agent sets them out in full.
What to specify in a de-icing tender
- Which salt, on which of the two axes? State the lowest expected surface temperature and the pavement type. Those two answers usually decide the salt.
- Which hydrate form, and at what assay basis? 74% dihydrate and 94% near-anhydrous are not interchangeable on a per-tonne price; convert to cost per tonne of calcium chloride delivered to site.
- Freezing-point curve for the grade offered, not a single eutectic number.
- Corrosion and scaling test data to the method named in the tender – and make sure the method tests the pavement or the metal, whichever the tender is about.
- NaCl content of the calcium chloride. Technical grades permit up to about 5% sodium chloride, which is 50 kg per tonne. In de-icing it is harmless; in desiccation or in a brine specification it is not.
- Particle size and caking behaviour in the storage you actually have, not in the supplier’s warehouse.
- Packaging and moisture protection for the 94% grade, which has far less tolerance for half-used bags and humid silos.
FAQ
Which de-icing salt melts ice fastest?
Calcium chloride, then magnesium chloride, then sodium chloride. Calcium chloride also releases heat as it dissolves, which shortens the time to a visible melt still further.
Which de-icing salt damages the road surface least?
Of the three chlorides, calcium chloride. It is strongly hygroscopic, so it keeps the surface damp rather than cycling through the crystallisation phase that produces salt scaling. Sodium chloride does the most surface damage.
Does that mean calcium chloride is safe for reinforced concrete?
No. The ranking above is about the pavement surface. All three salts are chlorides, and chloride reaches reinforcing steel and initiates corrosion regardless of the cation. Concrete with embedded steel is governed by a chloride limit, which is a separate question – see our note on calcium formate versus calcium chloride as a set accelerator.
Why do Japanese and Korean buyers take the dihydrate?
Short sea freight makes the water of crystallisation cheap to ship, the flake form suits local spreading equipment, and a humid maritime climate makes the deliquescent near-anhydrous grade harder to store.
Why do North American and Russian buyers take the 94% grade?
Long inland freight means paying to move water, severe continental winters need high-concentration brine, and a long season justifies the packaging discipline the 94% grade requires.
Is magnesium chloride still used for de-icing?
Yes, extensively, usually as a liquid or in treated-salt blends. It sits between the other two on both rankings: less melt power than calcium chloride, less pavement damage than sodium chloride, and lower cost than calcium chloride.
Product pages
The two calcium chloride grades are specified on the calcium chloride 74% flakes (dihydrate, CAS 10035-04-8) and calcium chloride 94% near-anhydrous (CAS 10043-52-4) pages, with the specification lines decoded in our guide to reading a calcium chloride specification and the applications side in calcium chloride applications by grade. Magnesium chloride hexahydrate is on the magnesium chloride hexahydrate page, and the chloride-free option on the sodium formate page.
Disclaimer. The rankings on this page describe the two commercial axes that govern de-icing salt selection – melt performance and pavement surface damage – and reflect the experience of producers and exporters of these salts. Eutectic temperatures are approximate values for saturated brines and depend on concentration; pavement damage depends on concrete or asphalt quality, age, drainage and freeze-thaw exposure. Neither ranking is a durability or structural opinion for a specific installation, and neither covers corrosion of embedded steel, which is governed by chloride limits and by design.
For a current COA on either calcium chloride grade, a freezing-point curve, or pricing on bagged and bulk de-icing quantities, contact PolyBlueChem.