Which material is meant by propylene glycol ethyl ether?

It is the P-series glycol ether 1-ethoxy-2-propanol, CAS 1569-02-4, molecular formula C5H12O2 and molar mass 104.15 g/mol. Commercial material is an isomer mixture in which the secondary alcohol isomer strongly predominates, with a minor proportion of the primary alcohol isomer. The abbreviations used for it are not standardised across suppliers, so the identifier belongs on the specification and the abbreviation belongs only in conversation. The isomer ratio is a specification line worth asking about, because it is the one difference between nominally identical grades that formulators most often notice later.

Can a solvent damage the substrate rather than the coating?

Yes, and it is one of the least examined failure modes in coating work. On thermoplastics a solvent can soften the surface, swell it, or promote crazing where the part is under moulded-in or applied stress. On wood it can raise grain or mobilise extractives. On a previously coated surface it can soften, lift or wrinkle the old film. None of these are properties of the solvent alone. They are properties of the pairing of that solvent blend with that specific substrate in that specific state, which is why the question can only be answered by test.

Why does stress in a plastic part matter so much?

Because environmental stress cracking needs both a chemical and a mechanical contribution, and a test that supplies only the chemical one will pass a part that fails in service. A moulded part carries residual stress from moulding, and an assembled part carries stress from fasteners, snap fits and thermal movement. A flat plaque of the same polymer, unstressed, can tolerate a solvent that will craze the real part within minutes. Test on the real geometry, or on a deliberately stressed specimen, and test on parts from the same moulding conditions as production.

How do glycol ethers compare with stronger solvents for plastics?

As a class, glycol ethers are moderate in their attack on many thermoplastics compared with ketones, esters and chlorinated solvents, which is a large part of why they appear in coatings destined for plastic parts. Moderate is not the same as safe, and the comparison is polymer-specific: a solvent that is unremarkable on one engineering plastic can be the worst choice for another. Use the comparison to decide what to test first, never to decide what not to test.

What should be recorded from a substrate compatibility trial?

Record the substrate identity down to the grade and supplier, the moulding or preparation history, whether the specimen was stressed and how, the full solvent blend rather than the single solvent of interest, the application method, the film thickness regime in qualitative terms, the exposure time before the film set, and the inspection interval afterwards. Most of the value of such a trial is lost because only the result was written down. Without the conditions, the next reformulation has to start again from nothing.

Coating solvents and the substrate they are painted onto

Technical article · Eapearl Chemical · 25 March 2026

Solvent selection is usually argued out between the resin and the drying schedule. The surface being painted rarely gets a vote, and it is the one that fails in the field.

The part of the specification that gets skipped

The solvent portion of a coating is normally chosen against the binder: what dissolves it, what keeps it open long enough to flow out, what leaves at the right moment. The substrate enters the conversation only as a source of adhesion problems, and usually after the first field failure. That order is backwards for any coating applied to plastics, to wood, or over an existing film, because in those cases the solvent blend touches the substrate directly and has time to act on it before the film sets.

The material under discussion here is the P-series glycol ether propylene glycol ethyl ether, that is 1-ethoxy-2-propanol, CAS 1569-02-4, molecular formula C₅H₁₂O₂, molar mass 104.15 g/mol. It carries both an ether oxygen and a hydroxyl group, which is what gives it a broad solvency range and full water miscibility, and it evaporates in the middle of the range rather than at either end. Those properties are why it appears in so many coating formulations. They are also why it stays in contact with the substrate long enough to matter.

Thermoplastics: softening, swelling and crazing

Three distinct things can happen when a solvent blend meets a plastic part, and they have different remedies. The surface can soften, which can help adhesion or can destroy the finish depending on degree. The polymer can swell and then shrink back as the solvent leaves, which distorts thin sections and leaves a stressed surface. Or the part can craze, developing a network of fine cracks that may not be visible at handover and becomes visible under load, under light, or after months in service.

Crazing is the failure mode that catches teams out, because it requires both a chemical contribution and a mechanical one. A moulded part carries residual stress from the moulding process; an assembled part carries stress from fasteners, snap fits, press fits and thermal movement between dissimilar materials. An unstressed flat plaque of the same polymer will tolerate exposures that crack the real part within minutes. Any test programme that uses plaques alone produces a comfortable result and no information.

Glycol ethers sit in the moderate part of the aggression scale for many thermoplastics, well below ketones, esters and chlorinated solvents. That generalisation is useful for ordering a test programme and useless as a substitute for one, because the ranking changes with the polymer. Sensitivity is strongly polymer-specific, so the decision has to be made against the exact grade in the exact part.

Wood: grain, extractives and moisture

Wood reacts to the whole volatile fraction, not only to the water in a waterborne system. A solvent blend that penetrates readily can mobilise resins and colouring extractives and carry them into the film, which shows as staining or as discoloration that develops over days. Hydrophilic solvents can raise grain on species that are prone to it. Very fast blends can be drawn into open grain and leave voids behind, and very slow blends can stay in the wood long enough to delay hardness development in the coating above.

The variable that governs most of this is not the solvent but the moisture content and preparation of the timber at the moment of coating. Any comparison between two solvent blends on wood is worthless unless both were applied to stock from the same batch, conditioned the same way, sanded to the same grit and coated within the same window after sanding.

Metal, pretreatment and what is left on the surface

On metal the substrate interaction is less about attack and more about what the solvent finds when it arrives. Residual drawing oils, fingerprints, salts from handling and the remains of a cleaning stage are all mobilised by the coating solvent and redistributed within the wet film. A blend with strong solvency will dissolve a contaminant that a weaker blend would simply have covered over; whether that is an improvement depends entirely on where the dissolved material ends up.

Conversion coatings and other pretreatments add a second consideration, because the coating solvent contacts a thin inorganic layer that was designed for adhesion rather than for solvent resistance. In waterborne systems the further question is flash rust during the open time, which is a function of how long free water stays on the surface, and therefore a function of the co-solvent choice. Materials such as butyl glycol and propylene glycol methyl ether acetate sit at different points on that scale and are the usual comparators when this question is being settled.

Recoating: the substrate is the previous film

When a coating goes over an existing one, the old film is the substrate, and it is the one substrate whose condition is never specified. An aged film that has continued to crosslink resists solvent differently from one that was overcoated within its recoat window. Lifting, wrinkling and loss of adhesion at the interface are the usual outcomes, and they appear in patches because the old film was never uniform to begin with.

The practical consequence is that any recoat qualification has to state the age and the history of the film being coated over, not just its identity. A panel coated yesterday is not a valid stand-in for a structure painted several years ago, and a trial that used one and claims the other is a trial that will be repeated after the failure.

A test programme that answers the question

Obtain substrates from production, not from a sample cabinet: same grade, same supplier, same processing.
Include stressed specimens for any thermoplastic, using a deliberate applied strain, and include assembled parts where assembly stress is real.
Test the whole blend, not the single solvent under discussion. Substrate attack is a property of the mixture.
Apply by the method production will use, because dwell time on the surface differs between spray, dip and flow application.
Inspect immediately, after the film has set, and again after a defined storage interval. Crazing frequently develops late.
Run adhesion and the relevant mechanical checks on the same specimens, so that one set of parts answers more than one question.
Record the conditions alongside the result, so the trial has value the next time the formulation moves.

What to put on the enquiry

State the substrate as part of the enquiry, not just the resin system. A supplier who knows that the coating is destined for a stressed engineering plastic, for oak, or for recoating an aged industrial finish can point at the relevant comparisons and supply samples of more than one candidate for a side-by-side trial. Ask for the isomer specification of the glycol ether, for the certificate of analysis lines that actually vary between lots, and for a change-notification commitment, because a supplier-side change in isomer ratio or in trace composition will reach you as a substrate problem long before anyone thinks to look at the solvent. Samples and specifications for coating trials are quoted against a named application through our contact page.