No commercial polyether monomer handles clay by its head group: HPEG, TPEG and vinyl-ether grades such as EPEG all carry a long polyethylene glycol side chain, and it is that chain, not the methallyl, isoprenyl or vinyl-ether anchor, that slides into montmorillonite and takes the PCE out of the mix. Clay tolerance is built on top of the monomer, through shorter side chains, the right side-chain density, a capped chain end and functional comonomers.
Changing the polymer is not the only fix. Pretreating the sand with a sacrificial agent such as polyethylene glycol or lignosulfonate is cheaper and simpler, though the sacrificial dose has to be optimised against everything else in the mix, and washing the sand is the other alternative. No study we found sets a rule for choosing between these routes and a clay-tolerant PCE.
Whichever route you take, confirm it on your own aggregate. The methylene blue value flags clay in sand, but in an ACI study of mortars it predicted flow loss for clays added to standard sand and did not hold for real job-site sands, so a mortar test on the actual sand is the check. Studies also state clay on different bases, by weight of cement in some and as a share of sand mass in others; compare figures only on the same basis.
Why Clay in Aggregates Takes PCE Away From the Cement
Clay minerals in the sand adsorb PCE far more readily than cement grains do, so the polymer is used up on the fines before it can disperse the cement.
The clay takes up the polymer in two ways. Once cement hydration enriches the pore solution with Ca²⁺, the calcium adsorbs onto the negatively charged clay faces and gives the anionic PCE backbone a cationic layer to attach to, the same kind of surface adsorption by which PCE disperses cement. The PEO side chains also hydrogen-bond, through water, to oxygen on the montmorillonite layers and intercalate between them. Raising the dose can restore flow, but it is a costly fix, because much of the extra polymer ends up on the clay.
Which Clay Minerals Do the Damage
Montmorillonite does the most damage, followed by kaolinite and then illite, the order in which they adsorb PCE. Kaolinite, a 1:1 mineral, and illite and muscovite, non-expanding 2:1 silicates, do not intercalate the side chains, but they still cost flow at low PCE dose.
How Each Polyether Monomer Family Handles Clay
On the measured evidence, the families cannot be ranked by head group: we found no study comparing plain HPEG-, TPEG- and vinyl-ether-based PCEs at the same side-chain length and acid-to-ether ratio in cement with clay-bearing aggregate. The variables with a measured effect act on the PEG side chain all of them share: its length, its density along the backbone and its end group.
Changes outside the polyether add tolerance as well: sulfonate comonomers such as AMPS, phosphate groups, quaternary-ammonium or amine cationic groups and cross-linked structures each reduced clay interference in published work.
HPEG
HPEG-based PCE is clay-sensitive: in montmorillonite studies it is the unmodified control that clay-tolerant designs are measured against, and it intercalates. In mortar containing 1% clay, HPEG-PCE needed about 20% more dosage than a phosphate-modified TPEG-PCE to reach a similar initial flow.
HPEG’s reputation for handling clay comes from calcined clay used as a binder, where highly anionic HPEG designs perform well; raw montmorillonite in the sand is a different system. Our HPEG 2400 is methallyl alcohol polyoxyethylene ether at 2400 ± 100 g/mol, a long PEG chain of the kind these studies show intercalating.
TPEG
TPEG-based PCE cannot be ranked above or below HPEG on clay: we found no peer-reviewed measurement of plain TPEG-PCE’s clay adsorption against plain HPEG-PCE under the same conditions.
The TPEG clay results that exist are for modified polymers: the phosphate-modified TPEG-PCE compared with HPEG-PCE in the HPEG section, and a cross-linked TPEG-PCE that kept more paste flow with montmorillonite than a conventional PCE. In both, the modification accounts for the gain.
TPEG is isoprenyl polyoxyethylene ether. Some literature calls it IPEG, while other papers use IPEG for isobutenyl, a methallyl-type monomer, so check the chemical name before carrying a result across. VPEG is the vinyl-ether family; isoprenyl TPEG belongs to a different one. Our TPEG 2400 is isoprenyl alcohol polyoxyethylene ether at 2400 ± 100 g/mol.
Vinyl-Ether Monomers (EPEG and VPEG)
Vinyl-ether PCEs have no established clay advantage: a 2025 peer-reviewed review notes that studies of EPEG-PCE clay sensitivity remain too limited to define it, and we found no measured clay result for a PCE built on a commercial long-chain EPEG or GPEG macromonomer. VPEG began with 4-hydroxybutyl poly(ethylene glycol) vinyl ether; EPEG and GPEG are newer members of the family.
The only clay-tolerant vinyl-ether PCE in the literature we checked is a research terpolymer of maleic anhydride, a monoalkyl maleate and 4-hydroxybutyl vinyl ether, with short side groups and no long PEO chain. Its tolerance came from dropping the long chain, so it says nothing about vinyl-ether PCEs that keep one.
Our EPEG 3000 is diethylene glycol monovinyl ether polyoxyethylene ether at 3000 ± 100 g/mol, a vinyl-ether macromonomer with a long PEG chain. On current evidence it needs the same side-chain design work for clay as HPEG or TPEG.
Short-Chain APEG and MPEG
Short side chains give the clearest clay-tolerance result in the literature. Allyl-ether APEG-PCEs with 7 EO units sorbed 110–170 mg/g at saturation on sodium bentonite in synthetic pore solution, against 241.5 mg/g for a conventional MPEG-PCE with 45 EO units. With 7-EO APEG, acrylic acid as comonomer gave much better clay tolerance than maleic anhydride.
These are laboratory polymers: 7-EO APEG PCEs are produced only in laboratories, and APEG macromonomer supply is limited. For a buyer, short-chain APEG is a direction for lab trials.
Esterified Polyether Monomers
Capping the side-chain end with a carboxyl group stops intercalation. In a study on montmorillonite in cement, isobutenyl polyoxyethylene ether, a methallyl-type monomer, was esterified at its terminal hydroxyl with succinic anhydride and substituted in part for the plain monomer. With 3.0 wt% clay by weight of cement, the conventional PCE lost 52.0% of its paste spread flow, while the best of the esterified versions lost 32.0%.
Only partial substitution works: the authors recommend keeping the esterified share within 10–20% of the polyether so the PCE keeps its dispersing power.
Side-Chain Length and Density for Clay-Bearing Aggregates
Shorter side chains lower clay sensitivity, and lower side-chain density usually does, though the density evidence is mixed. Both make the PCE disperse cement less well, so it needs a higher dose, and both shift slump retention.
Side-Chain Length
A shorter PEO side chain lowers clay sensitivity. In an MPEG series tested with clay, the 45-EO PCE needed 0.71% bwoc to reach a 26 cm paste spread, against 0.35% bwoc for the 23-EO version.
The side-chain length results are on MPEG and APEG PCEs; we found no study measuring side-chain length against clay on HPEG, TPEG or EPEG PCEs. Supplier claims that longer chains reduce clay adsorption contradict all the paste and mortar data we found.
With quarry sands, one study found a long-side-chain, low-graft acrylic PCE gave the best initial slump flow but lost it quickly in the clay-contaminated sands; the authors suggest using it as a sacrificial PCE alongside another superplasticizer.
Side-Chain Density
No single rule holds. In an APEG study on bentonite, PCEs with lower side-chain density, meaning a higher acid-to-ether ratio, intercalated less and generally resisted clay better, but when the polymers were compared at the same dosed amount, medium density performed best.
A TPEG study on natural montmorillonite reported clay tolerance rising with greater PEO grafting density, the opposite direction, though we could read only its abstract. With TPEG in particular, trial density in both directions on your own sand, and compare at both equal dose and equal flow.