Ice Melt Products and What They Do to Concrete

Most de-icing on commercial property is done by habit. One product, one rate, one bag on the truck all winter — spread the same way at 28°F and at 2°F.

That's how you end up with an untreated lot on the coldest night of the year and a spalled slab in April. Neither is a snow problem. Both are a product-selection problem.

Here's what's actually in the bag, what temperature it works at, and what it does to your concrete.

The temperature on the bag is not the temperature it works at

Every de-icer has a eutectic temperature — the lowest temperature at which its brine will theoretically stay liquid. It's a lab number, and it's the number the marketing prints.

It is not a working spec. The Minnesota Pollution Control Agency, whose Smart Salting manual is the most useful reference in this business, says it flatly: at the eutectic temperature "it would take a very long time to melt ice." What you want is the lowest practical melting temperature — the point below which the product stops doing useful work at a real application rate.

The spread between the two is enormous, and it's why "calcium chloride works to −60°F" is one of the most repeated wrong things in the trade.

Product	Practical melting temp	Eutectic temp
Sodium chloride (rock salt)	~15°F	−6°F
Magnesium chloride	~−10°F	−28°F
Calcium chloride	~−20°F	−60°F
CMA (calcium magnesium acetate)	~20°F	−18°F
Potassium acetate	~−25°F	−76°F
Sodium acetate / sodium formate	~0°F	−7°F / −16°F
Sand	Never melts — traction only	—

Practical and eutectic figures per the Minnesota Pollution Control Agency's Smart Salting for Parking Lots and Sidewalks manual (2022).

Two things fall out of that table immediately.

Rock salt is a mild-weather product. Below about 15°F, Minnesota's guidance moves dry salt to "not recommended" and sends operators to abrasives instead. The instinct to throw more salt at a lot when it gets brutally cold is exactly backwards: that's when salt melts the least and washes off the most.

Sand doesn't melt anything. It's traction. On a genuinely cold Front Range morning, sand on a walk is the difference between grip and no grip — and it's the right answer, not a cop-out.

Chlorides vs. acetates

Chlorides — sodium, calcium, magnesium — are the workhorses. Cheap, effective, and everywhere. They also carry every one of the downsides below.

Acetates and formates — CMA, potassium acetate, sodium acetate, sodium formate — are chloride-free. This is why they're the standard on airfields: the FAA prohibits chloride de-icers on aircraft operational areas outright, because chlorides corrode aircraft. Acetates and formates are what's left.

They cost a great deal more — Iowa State Extension pegs CMA at roughly 20 to 30 times the price of rock salt — and they carry their own environmental catch. Being organic carbon, they biodegrade, and the FAA notes that this process "often cause[s] a large drop in the dissolved oxygen levels of receiving waters." Chloride-free is not consequence-free.

And here's a trap: "chloride-free" does not mean "concrete-safe." More on that in a moment.

What de-icers actually do to concrete

The folk version is that salt eats concrete. That's mostly wrong, and the real explanation matters, because it changes what you'd do about it.

The main mechanism is physical

The National Ready Mixed Concrete Association's guidance on scaling is unambiguous. Concrete that is saturated with water and cycled through freezing and thawing is what scales. Water expands as it freezes, and that generates stress inside the slab. De-icing chemicals make it worse by increasing surface saturation and by increasing the number of times the slab crosses freezing.

So the de-icer is an aggravator of a freeze-thaw problem, not primarily a solvent chewing on the cement.

The evidence for this is better than most people realize. The Portland Cement Association ran a 36-year outdoor exposure study, applying rock salt and calcium chloride to slabs every winter. The finding: both showed minimal scaling on concrete that was properly formulated, finished, and cured with enough time to mature before its first winter.

Which points the finger at the slab and the spreader — not the chemical. Four of the six causes of scaling that NRMCA lists are workmanship or mix-design defects: too little entrained air, low strength, bad mixture proportions, improper finishing, and insufficient curing. Only one is the de-icer, and it's specifically excessive amounts of it.

But magnesium and calcium chloride do have a real chemical attack

This part is genuine, and it's the part the trade press gets wrong in both directions.

Calcium oxychloride. Calcium chloride and magnesium chloride react with the calcium hydroxide in cement paste to form calcium oxychloride — an expansive compound. It's documented in FHWA-affiliated research, it shows up as spalling and scaling concentrated near joints, it is not a freeze-thaw process (it forms in non-freezing conditions), and — counterintuitively — the concentration needed to form it drops as it gets colder. Notably, the same research reports this damage with calcium and magnesium chloride but not with sodium chloride.

Magnesium chloride attacks the binder itself. This is the more serious one. Magnesium ions decalcify calcium silicate hydrate — the actual glue holding cement paste together — converting it into magnesium silicate hydrate, which is not cementitious, along with brucite. Iowa State's experimental work found magnesium-bearing de-icer solutions caused severe paste deterioration, and that in their test conditions sodium chloride was the least damaging of the chemicals tested.

NRMCA's verdict is blunt enough to just quote: calcium chloride and rock salt are considered acceptable for concrete; never use magnesium-based salts as a de-icer, because they are chemically aggressive and destroy concrete surfaces.

And the sting: CMA is calcium magnesium acetate. It is magnesium-bearing, and it showed the same severe paste deterioration in the Iowa State work. "Chloride-free" and "concrete-safe" are not the same claim.

Rebar corrosion

Chlorides penetrate the concrete cover and destroy the passive film protecting reinforcing steel, initiating pitting corrosion. The corrosion products take up more volume than the steel did, and that pressure cracks and spalls the slab from the inside.

There's no single clean threshold — research on that varies widely — but the operationally important fact is simpler: chloride is cumulative. It doesn't degrade and it doesn't leave. Every winter's application adds to the load already in the slab.

New concrete is the vulnerable concrete

If you take one thing from this: keep de-icers off concrete in its first winter.

That's not a rule of thumb we made up. It's NRMCA's stated guidance — avoid de-icing chemicals in the first winter, and use them in moderate amounts thereafter. Use clean sand for traction instead.

The reason is that fresh concrete is still saturated with mix water it hasn't lost, and hasn't developed the strength to resist internal ice pressure. You'll see "wait 30 days" repeated all over the internet. We couldn't find that number in any authoritative source, and it badly understates the real guidance, which is measured in seasons, not weeks.

Two more things that matter more than which bag you buy:

Air entrainment. Deliberately entrained microscopic air voids act as pressure-relief chambers for freezing water. It is the single biggest factor in whether a slab survives. FHWA research also notes a hard limit: past roughly 86% saturation, concrete can be damaged in a single freeze-thaw cycle regardless of air content.

Finishing. Over-troweling an exterior slab drives the entrained air out of the surface layer — the exact layer that scales. For most exterior flatwork, a broom finish is what you want. A beautiful steel-troweled exterior slab is a slab that will scale.

Sealing with a breathable silane or siloxane sealer, applied to dry concrete in late summer, helps. A film-forming sealer that traps moisture makes it worse.

Rate discipline is the whole game

Over-application is the actual problem in this industry — and it isn't only an environmental one, though it's that too. It only takes about a teaspoon of salt to permanently pollute five gallons of water, and chloride is not removed by wastewater treatment. It goes in and it stays.

But the property-level argument is simpler: more product does not melt more ice past a point. Excess dry salt bounces off target. Excess liquid can leave a lot more slippery and tracks into your lobby on everyone's shoes. You paid for it, it didn't work, and it's in your landscape beds.

The discipline that actually works: plow to bare pavement first — every application rate in the guidance assumes a cleared surface, because chemical is not a substitute for a plow — then apply to the pavement temperature, at a rate, with a calibrated spreader. And write down what you did.

The short version

• The eutectic temperature on the bag is not the working temperature. Rock salt stops earning its keep around 15°F.
• Below that, it's chloride blends, calcium chloride, or — honestly — sand for traction.
• Most de-icer damage to concrete is physical freeze-thaw scaling, made worse by saturation and extra freeze cycles.
• Magnesium chloride has a real chemical attack on cement paste. So does CMA, which is magnesium-bearing despite being chloride-free.
• New concrete in its first winter is the thing to protect. Sand it; don't salt it.
• Air entrainment, curing, and a broom finish matter more than which product you buy.
• Over-application is expensive, ineffective, and permanent.

If you poured a slab this year, tell your contractor before the season starts so it gets flagged and sanded instead of salted. That's part of how we choose product on every visit — and what we write down afterward. Get your property on a route.