When pH and Salt Decide How Much Water One Kilo Takes

Two drums of the same grade, from the same lot, dosed at the same rate, produced gels that looked nothing alike. One set into a firm mass that passed the tilt test. The other stayed soupy for twenty minutes and had to be re-dosed twice. The powder had not changed. The water had: one drum was filled from the mains, the other from a borehole drawing water at 780 mg/L hardness. Nobody had measured either.

The polymer is a pH-dependent salt

Sodium polyacrylate works because its backbone carries carboxylate groups, neutralised to the sodium form at 65-75% during manufacture. Those fixed negative charges repel each other, and that repulsion is what forces the network to expand. Anything that suppresses the charge suppresses the expansion.

Below pH 5, hydrogen ions associate with the carboxylate groups and the charge largely disappears. Absorption falls steeply: a powder that returns 300 g/g at pH 6.5 typically returns 120-150 g/g at pH 4 and under 80 g/g at pH 3. In practice, absorption collapses somewhere between pH 4 and 5, and if your process liquid is an acidic cleaner, a pickling bath or fruit process water, assume you need two to three times the dose before you even test.

Above pH 10 the picture is slower but still unfavourable. The gel holds up initially, then loses strength over hours as the polymer backbone hydrolyses and the network relaxes. For caustic streams, dose for a shorter holding time and expect the gel to soften if it sits overnight.

The optimum is narrow and boring

Between pH 5.5 and 8, performance is close to flat and near maximum. That is the range where a datasheet number means something. Outside it, the datasheet describes a product you are not actually using.

Ionic strength does more damage than pH

Salt attacks the same mechanism from the other side. Mobile ions in the water screen the fixed charges on the backbone, shrinking the effective repulsion, and multivalent ions go further by physically crosslinking chains together. The effect is large and fast.

Rough figures for one commercial grade, 60 minutes, room temperature:

– Deionised water: 320-380 g/g
– Mains water at 250 µS/cm: 220-260 g/g
– 0.9% sodium chloride: 55-65 g/g
– Seawater, about 3.5% salts: 40-50 g/g
– 2% calcium chloride: 20-30 g/g
– Landfill leachate at 12,000 µS/cm: 25-45 g/g

The drop from deionised to physiological saline is roughly a factor of five, and it happens within seconds. This is why agricultural and hygiene applications quote different numbers for the same product: urine is not water.

Calcium and magnesium are the expensive part of the story because they crosslink rather than merely screen. At 100 mg/L as CaCO3 the loss is modest, perhaps 10%. At 300 mg/L, expect 35-45% less uptake than the same water softened. Above 500 mg/L the gel may form a skin at the granule surface that slows further absorption badly, so the batch looks under-dosed even after twenty minutes.

Decision rule: measure conductivity and hardness before you set a dose. If conductivity exceeds 1,000 µS/cm or hardness exceeds 200 mg/L as CaCO3, apply a correction factor of 1.8-2.5 on the datasheet dose and confirm with a jar test in the real liquid. Softening the process water is sometimes cheaper than doubling polymer consumption, and worth costing both ways.

Honest limits: adding alkali to raise pH will not rescue a high-hardness stream, because the calcium is still there and precipitation may make things worse. And pH adjustment adds dissolved solids, which costs you in the ionic strength column. There is no additive that makes a polyacrylate behave in brine the way it behaves in tap water – only grade selection and dosage will.

For streams with known chemistry, the SAP product grades include higher-crosslink and surface-treated options that tolerate salinity better than standard powder.