How to stabilise water-in-oil emulsions when adding polar components

A polar additive destabilises a water-in-oil emulsion by competing for the interface and by shifting the required HLB of the oil phase. The fix is to recalculate the target for the modified phase and rebuild the pair to meet it — typically a low-HLB anchor such as Span 80 at HLB 4.3 or Span 60 at HLB 4.7 with a partner such as PEG-400 dioleate at HLB 7–8.

What a polar additive does to a water-in-oil interface

Water-in-oil emulsions are held together by a lipophilic emulsifier film at the outside of each water droplet, with the hydrophilic head group anchored into the droplet and the hydrocarbon tail extending into the continuous oil phase. That film is thinner and less robust than the equivalent in an oil-in-water system, because there is no electrostatic repulsion between droplets to help — the continuous phase is nonconductive — so stability rests almost entirely on steric hindrance from the tails and on the mechanical strength of the packed interfacial layer. Anything that disturbs the packing of that layer has a disproportionate effect.

A polar component is exactly such a disturbance. Glycols, glycol ethers, polar esters, amines, corrosion inhibitors and water-miscible solvents all partition between the oil phase, the water droplets and the interface itself, and the fraction that accumulates at the interface sits between emulsifier molecules without providing the same anchoring. The film becomes more permeable and more easily ruptured, which shows up as coalescence of water droplets and eventual gross separation. There is a second, slower effect: the additive changes the average polarity of the continuous phase, and since the required HLB of an oil phase is a function of its polarity, the number the existing emulsifier package was built to meet is no longer the number the system needs. The emulsion does not fail because the emulsifier stopped working; it fails because the target moved. Both effects act together, which is why a formulation often tolerates a small addition of a polar component and then fails abruptly above a threshold level.

Recalculating the required HLB, and the pairs that reach it

The required HLB of a mixed oil phase is calculated as the weighted average of the required HLB values of its components, weighted by their proportion in the phase, and adding a polar component should trigger that recalculation rather than a guess. Where a published required-HLB value for the additive is unavailable, the practical alternative is to bracket it: prepare a short ladder of blend HLB values spanning the plausible range, roughly one HLB unit apart, using identical addition order, shear and temperature, and take the finest and most stable emulsion as the working target. Blend HLB itself is the weight-average of the component values, so once the target is known the ratio follows directly — the weight fraction of the higher-HLB component equals the target minus the low HLB, divided by the difference between the two.

Grade	Chemical name	CAS	HLB	Form	Acid value (mgKOH/g)	Saponification value (mgKOH/g)	Role in a W/O pair
Span 80 (VS-201A)	Sorbitan monooleate	1338-43-8	4.3	Amber viscous liquid	< 4.0	140–160	Primary low-HLB anchor, liquid handling
Span 60	Sorbitan monostearate	1338-41-6	4.7	Off-white waxy solid	≤ 8	240–270	Low-HLB anchor with a firmer film; must be melted in
PEG400DO	Polyethylene glycol 400 dioleate	9005-07-6	7–8	Pale yellow liquid	≤ 10	100–130	Closest higher-HLB partner for W/O targets
PEG600DO	Polyethylene glycol 600 dioleate	9005-07-6	10–11	Pale yellow liquid	≤ 10	85–105	Partner where the recalculated target rises
PEG400DL	Polyethylene glycol 400 dilaurate	9005-02-1	10.5	Colourless to pale yellow liquid	≤ 10	130–155	Shorter-chain diester alternative
PEG400MO	Polyethylene glycol 400 monooleate	9004-96-0	11–12	Pale yellow liquid	≤ 5	75–95	Use sparingly; high fractions favour inversion to O/W

Values per manufacturer TDS. All grades listed are nonionic.

Worked through, a target of HLB 5.5 from Span 80 at 4.3 and PEG-400 dioleate taken at 7.5 requires a weight fraction of the dioleate of (5.5 − 4.3) ÷ (7.5 − 4.3), or roughly 38 %, with 62 % Span 80. Because PEG-400 dioleate is specified as a range of 7–8, repeating the calculation at both ends gives 38 % at 7.5, about 44 % at 7.0 and about 32 % at 8.0 — a band worth knowing before the ladder is prepared rather than after the results disagree.

Process controls that decide whether the calculation holds

A correct HLB calculation can still yield an unstable emulsion if the process undoes it, and water-in-oil systems are unusually sensitive to how they are made. The emulsifier pair must be fully dissolved in the oil phase before any water is introduced; an undissolved fraction of Span 60, which is a waxy solid at ambient temperature and needs the oil warmed to melt it in, will never reach the interface and effectively lowers the emulsifier level below what was charged. The aqueous phase should then be added slowly into the oil under constant shear, not the reverse, and the rate of that addition is itself a variable: a fast addition produces coarser droplets and will make a good emulsifier package look worse than a poor one added slowly.

Where the polar component goes in the sequence deserves a deliberate decision. Adding it to the oil phase before emulsification lets the interfacial film form in its presence, which generally gives a more representative result than adding it afterwards to an already-formed emulsion, where it must displace an existing film. If the plant process adds it late for practical reasons, the laboratory work must add it late too, or the trial will not predict the batch. Hold temperature constant across variants as well: viscosity, emulsifier solubility and the partition of the polar additive all change with temperature, and a comparison run at two different temperatures is not a comparison at all.

Reading stability, and knowing when the system has inverted

Water-in-oil failures present differently from oil-in-water ones, so the observations to record differ. The primary reading is the volume of free water separated at the bottom of the column, expressed as a percentage of the water charged, taken at 24 hours, 7 days and 30 days at both ambient and the highest expected storage temperature. Alongside it, note the appearance of the continuous phase — a shift from translucent to opaque often signals droplet coarsening before any water is visible — and any grittiness or crystalline deposit, which in a Span 60 system usually means incomplete melting or recrystallisation on cooling rather than an emulsion failure.

The failure worth catching explicitly is phase inversion, in which the system flips from water-in-oil to oil-in-water. It typically follows too high a fraction of the higher-HLB partner, too high a water content, or a polar additive that has pulled the effective HLB of the system upwards. A simple conductivity check distinguishes it in seconds: a water-in-oil emulsion has a nonconductive continuous phase and reads near zero, while an inverted system conducts. A drop test confirms it — a water-in-oil emulsion disperses into oil and not into water. Run that check on every variant at the top of the HLB ladder, because an inverted emulsion can look perfectly uniform and stable in the jar while being the wrong product entirely. Once a variant passes both the separation readings and the inversion check, repeat it at one drum on plant equipment before committing, since plant shear rarely matches laboratory shear and a water-in-oil system feels that difference more than most.