Supercritical CO2 foaming vs chemical foaming for PS trays

Supercritical CO2 foaming gives PS packaging finer and more uniform cells and leaves no blowing-agent residue behind. That combination wins on food contact compliance and on the resin weight per part. Chemical foaming is cheaper to install and drops into existing extrusion lines. What decides the switch is order mix and export exposure, not the price of the machine.

Why PS packaging plants are re-running this comparison

Nobody buys polystyrene foam for the polymer. They buy the air inside it, held in a shape that survives a loaded pallet, and food packaging is where that trick earns the most money. Trays, clamshells, cup sleeves and box liners all sit at a density well under solid resin, and the margin lives in one question: how little material can stay in the wall before the tray starts to fail.

Two processes make that foam. A solid chemical can be cooked until it splits and the gas it gives off inflates the melt. Or carbon dioxide can be forced into the melt under pressure and allowed to escape at the die, with no chemical in the picture.

The choice used to be simple. Chemical foaming was cheaper and nobody argued. Three things changed it. Resin took over the cost sheet, so gram weight per part overtook the additive bill in importance. Food contact enforcement tightened at home and abroad, which turned a residue-free process from a talking point into a document a customer can ask you for. And recycled content targets arrived, which the two routes do not meet at the same rate.

What supercritical CO2 actually is

Above 31.1 degrees Celsius and 7.38 megapascals, carbon dioxide stops behaving like an ordinary gas or an ordinary liquid. It is dense enough to dissolve into a polymer melt and fluid enough to move through it quickly, and without both properties the process would not run on a continuous line.

On a real line, liquid CO2 is pumped up to pressure, metered into the barrel and mixed until it forms a single phase with the melt. Nothing reacts and nothing is consumed. Pressure alone holds the gas in solution. Let the melt through the die and the pressure falls, so the CO2 drops out of solution everywhere across the section at the same moment. Cells start all over the sheet instead of at scattered particles of a decomposing solid.

The gas also softens the polystyrene. Dissolved CO2 lowers melt viscosity, which is the reason a foaming die runs colder than a plain PS die. It is also why the window is narrow. Too hot and the cells coarsen, too cold and the melt sets before it has expanded.

Chemical foaming and supercritical CO2 foaming side by side

Copy the table below into a supplier review. It is short on purpose. The two routes disagree about a handful of things that matter and agree about nearly everything else.

DimensionChemical foamingSupercritical CO2 foaming
Where the gas comes fromThermal decomposition of a solid agent inside the barrelDissolved CO2 released by the pressure drop at the die
Blowing agentAzodicarbonamide, or a bicarbonate and citric acid blend, dosed at 0.5 to 2 percent by weightCO2, dosed at 2 to 6 percent by weight, no solid additive
Onset of gas evolutionA temperature threshold, around 195 to 215 degrees Celsius for an azo gradeNo threshold, controlled by pressure and temperature instead
Typical cell diameter100 to 500 micrometers20 to 100 micrometers
Cell size spreadWide, because it tracks barrel temperature variationNarrow, because every point sees the same pressure history
Residue left in the partNitrogen-rich solid residue and traces of breakdown productsNone from the blowing agent
Position under EU food contact lawAzodicarbonamide suspended as a blowing agent since 2004No blowing agent to declare
Line change requiredDosing unit and a mixing section on an existing barrelGas storage, metering, high pressure port, dedicated screw, new die
Regrind ceilingLow, because residual agent reacts again in the barrel20 to 30 percent
Demand on the operatorModerate, the process tolerates driftHigh, the process punishes drift
Capital costLowHigh
Best fitThick industrial sheet, short runs, domestic non-food workThin food trays, export orders, recycled content targets

Chemical foaming: how the gas is made and what it leaves behind

A chemical blowing agent is a solid powder that splits above a set temperature and gives off gas. The two common families do not behave alike. Exothermic azo grades, azodicarbonamide being the classic example, break down sharply inside a narrow band and release nitrogen, carbon monoxide, carbon dioxide and ammonia along with a nitrogen-rich solid residue. Endothermic blends of bicarbonate and citric acid give off mostly carbon dioxide and water across a wider band and leave an ash that is easier to live with.

The process consequence follows from that. Gas generation depends on temperature and residence time, not on a flow meter. Nothing in the barrel reports how much gas was just made. Density ends up as a function of the melt’s thermal history, so a shift in barrel profile or output rate moves foam density, and the cell population moves with it.

Then there is the residue. Some of what the agent leaves behind stays in the sheet, and whatever can migrate has to be tested before the article touches food. In Europe the argument is mostly settled already. Azodicarbonamide was suspended as a blowing agent in food contact materials back in 2004, after semicarbazide showed up as a breakdown product of concern. That suspension, more than any environmental pitch, is what made the residue-free route interesting to exporters.

Supercritical CO2 foaming: what the line actually has to control

A CO2 line is not a chemical line with a different additive in the hopper. Four loops run at once, and each one moves the others.

Control loopTypical settingWhat happens when it drifts
CO2 purity99.9 percent or betterNucleation weakens, pinholes appear
CO2 dose2 to 6 percent by weight of meltLow dose starves the cells, high dose collapses walls
Injection pressure10 to 25 megapascalsTwo-phase flow, density surges and sags
Melt seal pressureAbove the solubility pressure for the doseGas escapes upstream and the dose reading becomes a lie
Die temperature110 to 140 degrees CelsiusWarm gives coarse cells, cold freezes the melt off
Pressure drop rate at the dieAbove 10 to the power of 9 pascals per secondToo few nuclei, wide cell size distribution
Screw designMixing section plus a sealing elementUndissolved gas, streaks in the sheet

Two of those rows deserve a sentence of their own. The melt seal is the loop plants learn last. Drop the pressure in the metering zone below what keeps the dose dissolved and gas leaves through the feed, while the mass flow meter keeps reporting a number the sheet never receives. The pressure drop rate is not a knob on a standard die either. Die geometry fixes it, which means nucleation is decided at the design stage, long before anyone runs the line.

Where the two routes separate on cell structure

Substitution stops at cell size.

A chemical agent nucleates wherever its particles happen to sit at the moment they decompose, and decomposition tracks temperature, so the number of nuclei drifts with the barrel. The result is workable foam with a wide spread, typically 100 to 500 micrometers and a tail at each end. A dissolved gas has no particles to nucleate on. Nucleation comes from the pressure drop at the die, a machine parameter rather than a dispersion quality, so the number of nuclei is designed in. Supercritical CO2 foaming typically lands between 20 and 100 micrometers with a narrow spread.

The gram weight claim follows from that. Foam modulus scales with relative density, so a foam of uniform thin walls carries the same load at a lower density than one with a mix of thick walls and holes. Fine cells also get rid of the oversized voids that concentrate stress, which is why a tray of equal density resists denting better. Below about 100 micrometers a second effect joins in. Gas trapped in a small cell conducts less heat than bulk gas, because the mean free path of the molecules is no longer small next to the cell. Fine foam insulates better at equal density, and for a hot food container that is a second saving on the same purchase.

Food contact and export compliance: the documents differ

Three regimes decide most of these projects.

GB 4806.7-2023 has applied since 6 September 2024. It folded the old resin standard and the old article standard into one text, sets limits on overall migration and potassium permanganate consumption, adds a migration limit for aromatic primary amines, and requires the maximum use temperature to be printed on the product. On a foam tray that label is not decoration. It is the claim.

EU Regulation 10/2011 runs off a positive list. Only authorized substances are allowed, overall migration has to stay under 10 milligrams per square decimeter, the article needs a declaration of conformity, and non-intentionally added substances must be assessed. Since azodicarbonamide is suspended for food contact use, no converter running it can reach that declaration by pointing at an authorized additive.

The United States works from 21 CFR 177.1640, which covers polystyrene and rubber-modified polystyrene as food contact basic polymers, and the paperwork there follows a different logic.

One caveat belongs under all three. A CO2 line takes the blowing agent out of the residue question. It does not take migration out of the question. Antioxidants, slip agents, mineral oil and styrene oligomers are in the resin long before any gas arrives, and they migrate on the same test. What changes is the list of substances you have to defend, not whether you defend any.

The defect map: three failures that look like operator error

Most CO2 problems land as one of three sheet defects, and all three get blamed on the operator when the cause is usually a relationship between two settings.

DefectMost likely causeFirst parameter to move
Coarse cells with a wide spreadNucleation rate too low for the growth rateRaise the pressure drop rate, then check the dose
Merged cells and collapsed wallsMelt strength too low for the expansion rateLower the die temperature, trim the dose
Pinholes on the sheet surfaceGas breaking through a skin that has not setCool the die lip, adjust line speed
Density drift along the rollDose or melt pressure not steadyCheck the mass flow meter and the melt seal
Rippled or wavy surfaceCooling asymmetry across the sheet widthBalance the cooling rolls

Merged cells deserve the warning. They look like a dose problem and are almost always a strength problem. CO2 lowers viscosity and makes expansion easier, so a plant chasing lower density pushes the dose up and gets collapse instead of finer foam. The correction runs backward: colder die, less gas, and if that still fails, the resin grade does not carry the melt strength the density target needs.

The counterintuitive part: the blowing agent is not the cost story

Most comparisons of these two routes open with the price of the blowing agent. That is the wrong number to open with.

A chemical agent runs between a fraction of a percent and roughly two percent by weight. Even at several times the price of resin, that stays in the low single digits as a share of material cost. Carbon dioxide costs less, but there is not much of it on either route. The additive saving does not pay for a new line.

Three other levers do, and each moves by tens of percent rather than by single digits.

Gram weight per part is the first. It is the cell structure argument carried onto a purchase order: a tray that holds the same load with eight percent less resin repeats that saving on every tonne shipped.

Scrap is the second. A route that accepts 20 to 30 percent regrind turns a disposal cost into raw material, and sheet lines throw away a lot of edge trim.

Market access is the third. A tray that cannot produce a clean declaration does not ship to the customer who asks for one, and a lost order outweighs any additive bill.

This also explains something that looks irrational from a distance: a costly new CO2 line running beside an old chemical line, both of them profitable. The chemical line serves thick industrial sheet, where none of the three levers bite. The CO2 line serves thin food trays, where all three do.

Scrap, regrind and the 20 to 30 percent ceiling

Regrind is the cheapest row to leave out of a cost sheet and one of the most expensive to get wrong, and it is where the routes diverge furthest.

Chemical scrap still holds unreacted agent, plus agent that has already made gas once. Put it back through the barrel and it decomposes again. Density turns unpredictable and the sheet can run away from whoever is on the line. Plants answer by capping the regrind ratio well under what the material would allow, and by writing trim off as waste, which is a cost that never goes away.

Carbon dioxide leaves no residual agent, so regrind goes back in and behaves. That is what lets the route quote 20 to 30 percent recycled content without the foam fighting back.

Two limits survive, and neither is chemistry. Heat history comes first, since the scrap has already been melted once, so its molecular weight is down and its melt strength with it, which caps the load the foam can carry. Contamination comes second, because a food contact article made with recycled content still has to hold the same declaration as a virgin one. Drying stays mandatory on both routes.

What the switch costs and when it pays back

The capital list runs well past the machine price. Add gas storage and a pump, mass flow metering, a high pressure injection port, a screw with mixing and sealing sections, a die built for a controlled pressure drop, tighter temperature control, local extraction, and a gas safety study, since carbon dioxide quietly displaces oxygen in a pit.

Payback reads simply even when the arithmetic does not. Take the capital gap and divide it by the yearly sum of material saving, scrap saving and whatever price the compliance story earns. Put real numbers in and the answer usually falls out in one pass.

Two variables decide most cases. Thin gauges at high volume help, because material saving scales with output and thin sheet leaves the most room for cell structure to remove resin. Export exposure helps, because a compliance advantage only becomes money when a customer is asking for the document.

Selection rules for five common plant situations

  • Domestic food trays on an existing line with tight cash: stay on chemical foaming, but fund a migration test program now rather than at the first customer audit, and book the trim as waste honestly.
  • Food contact packaging bound for the European Union: the CO2 route is effectively the only one that does not ask you to defend a suspended blowing agent, so treat the capital gap as a license to sell.
  • A plant with heavy scrap and a recycled content commitment: the regrind ceiling often covers the payback on its own, and it is the easiest lever to model before committing.
  • Thin sheet under about one millimeter: cell structure has the most room to work, chemical foaming struggles to hold a narrow distribution at that gauge, and the density argument is at its strongest.
  • Thick industrial sheet for domestic non-food use: stay chemical, because none of the three cost levers move and the gap never closes.

Limitations

This comparison comes from process logic and from the standard engineering ranges for both routes. It does not come from a measured line trial. The 20 to 30 percent regrind figure is the process window quoted for a well run physical foaming line, not an audited plant average, and a plant with wet regrind or dirty trim will land lower. Cell diameter bands are order of magnitude guides that move with resin grade, die design and output rate.

Scope stays inside extrusion foaming of PS sheet. Expandable polystyrene bead molding is a different process, using a pentane blowing agent and a steam chest rather than an extrusion line, and it still holds many molded packaging shapes. Cost figures are left out on purpose, because resin, energy and capital prices vary enough between regions that one number would mislead. Regulations move, so check the current text of GB 4806.7-2023 and Regulation (EU) 10/2011 before building a compliance plan on either.

Bottom line

Price is the wrong variable to decide on. Thin food trays headed for a market that enforces food contact rules belong on supercritical CO2 foaming, and thick industrial sheet sold at home still belongs on a chemical line. Everything in between is a spreadsheet, and the regrind ceiling is the row most plants leave empty.

FAQ

What is supercritical CO2 foaming?

Supercritical CO2 foaming dissolves carbon dioxide into a polymer melt under pressure and releases it as gas at the die. The gas comes out of solution all at once, so cells nucleate throughout the melt rather than at particles of a chemical agent. CO2 above 31.1 degrees Celsius and 7.38 megapascals is supercritical, which is why the process needs no solvent and no chemical blowing agent.

Chemical foaming or supercritical CO2 foaming: which costs less to run?

Chemical foaming costs less per kilogram only while resin is cheap and scrap is low. Add compliance testing, a low regrind ceiling and off-spec sheet, and the gap narrows or reverses on thin food trays. The blowing agent itself is a small share of material cost, so the price of the agent is a poor deciding number.

Does the CO2 route leave residue in a food contact tray?

No. Carbon dioxide leaves the sheet as gas and leaves nothing behind, so the article carries no blowing agent residue to declare. The migration risk does not disappear, because antioxidants, slip agents and styrene oligomers in the resin still migrate on the same test. A CO2 line changes which substances you defend, not whether you defend any.

What cell size does a PS foam tray need for cushioning?

Food trays usually sit between 50 and 300 micrometers, and finer cells carry load more evenly at the same density. Below about 100 micrometers the gas inside the cells also conducts less heat, which helps a hot food container. Very fine cells cost more pressure drop to make, so every line has a practical floor.

Why do pinholes appear on CO2 foamed sheet?

Pinholes come from gas escaping through a skin that has not set yet, usually at the die lip or on the first cooling roll. The fix is almost always thermal: cool the lip, change the line speed or trim the dose. Raising the dose to chase a lower density makes pinholes worse, which is why the cause gets misread.

Can an existing chemical foaming line be converted?

Sometimes, but the job is a rebuild rather than a retrofit. You need gas storage and metering, a high pressure injection port, a screw with a mixing and sealing section and a die with tighter temperature control. The take-off and winder usually survive, while the control system and the gas safety work rarely do.

How much regrind can each route take?

The physical route can take 20 to 30 percent regrind without runaway foaming, which is the window quoted for a well run line. The chemical route sits much lower because residual agent decomposes a second time in the barrel. Contamination and heat history usually cap the real number below the chemistry limit.

Which route clears EU food contact rules more easily?

The CO2 route clears them more easily because azodicarbonamide has been suspended for food contact use in the EU since 2004. A converter using it for EU bound trays has to show that no restricted breakdown product migrates, which is hard on a foam with a large surface area. CO2 removes that argument and moves the work to the resin declaration.

How much CO2 does the process consume?

The dose is usually a few percent by weight of the polymer melt, so a tonne of sheet takes tens of kilograms of carbon dioxide. Most of it leaves the sheet within seconds of the die and is captured by local extraction. It is a process gas rather than an ingredient, and its cost per tonne of finished sheet stays small.

What operator skill does a CO2 line demand?

The line fails on parameter drift rather than on chemistry, so the operator has to hold four loops steady at once. Dose, melt pressure, die temperature and pressure drop rate interact, and a small move in one needs a compensating move in another. Plants that treat it as a chemical line with a different additive spend months chasing coarse cells.

When does supercritical CO2 foaming pay back the extra capital?

Payback comes fastest on thin food trays sold in high volume, where gram weight per part and compliance both move in your favor. Add an export customer or a recycled content target and the case usually closes within a few years. Thick industrial sheet sold domestically rarely pays back at all, because none of the three levers move.

What does the CO2 route not fix?

It does not fix melt strength, so a resin grade with a low molecular weight still collapses at the die no matter how the line is tuned. It does not remove the need for migration testing, and it does not make regrind free of heat history. It also does not help a plant that cannot hold pressure and temperature steady across shifts.

Tags: