Where should a solvent selection exercise start?

With a written statement of the job, not with a shortlist of candidates. Say what has to dissolve or be removed, in what matrix, at what stage, and what state the article has to be in afterwards. Add the constraints that are already fixed: the equipment, the containment, the drying or recovery capacity, and any customer or sector restriction on what may be present in the finished article. Most disappointing trials can be traced back to a requirement that nobody wrote down, so the candidates were screened against an incomplete target.

Is a single solvent always better than a blend?

Not always, although it is simpler. A blend can place solvency and evaporation behaviour where no single component sits, which is why coatings and cleaning formulations routinely use them. The costs are real, though. Blends can change composition as they evaporate or as they are recovered, so performance drifts with use, and distillation recovery becomes a separation problem instead of a straight recycle. Choose a blend when a single component genuinely cannot meet the profile, and then specify how its composition will be monitored in service.

How much weight should occupational and environmental rules carry against technical performance?

They frequently decide the outcome before the technical comparison finishes, and it is cheaper to discover that early. Classification, exposure control expectations, emission limits, and restrictions on particular substance families vary by territory and change over time. A candidate that performs marginally better but forces a containment upgrade, a permit variation or an abatement system is not the cheaper option. Establish the regulatory position in the country of use at the screening stage, and confirm it with the authority or a competent adviser rather than inferring it from a supplier document.

What does total cost of ownership include beyond the purchase price?

Consumption per unit of production, which depends on carry over and evaporation losses as much as on the formulation; the energy used to apply, heat or dry it; the cost of recovery if the stream is distilled, and the yield of that recovery; the disposal cost of what cannot be recovered, which is driven by waste classification; the cost of the containment, detection and ventilation the material requires; and the insurance and storage implications of its fire classification. A cheap solvent with a poor recovery yield and an expensive waste route is rarely cheap.

What should be requested from a supplier before a trial is authorised?

A safety data sheet for the destination market, a technical data sheet describing the grade rather than the substance in general, a typical analysis showing which impurities are controlled and at what stage, a statement of the packaging and the storage assumptions behind the stated shelf life, and a representative sample from a production lot with its own certificate. If the material will contact a regulated article, also ask what compliance statements the supplier can issue and, just as important, which ones remain the user's responsibility.

Choosing an industrial solvent: a selection method

Technical article · Eapearl Chemical · 19 May 2026

Solvent choice is usually made once and then lived with for years, which is why it repays being made deliberately. This article sets out a sequence a process engineer or a technical buyer can follow, from defining the task through to the documents that should exist before a trial is authorised.

Define the task before naming any candidate

A solvent is not selected for a product; it is selected for a step. Degreasing a machined part, dissolving a resin for a coating, extracting an active from a plant matrix and stripping a photoresist are four different jobs whose ideal materials share almost nothing. Write the requirement down in operational terms: what must go into solution, what must remain untouched, what temperature and residence time the step allows, what happens to the solvent afterwards, and what residue the finished article may legally and commercially carry.

Two constraints belong in that statement from the start, because they eliminate more candidates than any laboratory result. The first is the equipment you already own, including its materials of construction, its containment and its ventilation. The second is any restriction imposed on you by your own customers or by the sector you supply, which is often stricter than the law and is rarely negotiable at short notice.

Screen on function: solvency, volatility, and what happens in between

Functional screening answers whether the material can do the job at all. Solvency is best approached through a polarity and hydrogen bonding framework rather than by intuition: solubility parameter approaches let a formulator predict whether a resin will dissolve and whether a blend will remain stable, and they turn a trial and error exercise into a short list. Miscibility with water matters in both directions, deciding whether an aqueous rinse is possible and whether a spill can be diluted or must be contained.

Volatility governs how the step behaves in practice. A fast evaporating material leaves a part dry quickly but is lost to the atmosphere and to the operator’s breathing zone, while a slow one gives levelling and open time in a coating but has to be driven off with energy. Relative evaporation rates, quoted against a common reference, are the practical comparison, and they are more useful during screening than any single physical constant. Surface tension, viscosity and the ability to displace water decide whether a liquid actually reaches a blind hole or a capillary gap, which is why parts cleaning results so often fail to match beaker tests.

Purity is a functional property, not a virtue

Grades differ in what is controlled, not only in how much of the main component is present. Water, acidity, peroxides in ethers, aldehyde content in alcohols, non volatile residue and trace metals each matter to some processes and to no others. A high assay figure with an uncontrolled residue is worse for a precision cleaning step than a lower assay with residue tightly held. Ask what the grade controls; do not read the number alone.

Check the fit with the plant before the fit with the product

Compatibility work is unglamorous and it prevents the expensive failures. Confirm every wetted material against the specific candidate: gaskets, hose liners, pump seals, sight glasses, filter media, coatings inside vessels and the labels on the containers themselves. Generic resistance charts are a screening tool, not a decision. Where the solvent will be recovered, the recovery train is part of the compatibility question and so is whatever azeotrope may form with water or with a co solvent.

Fire and static behaviour set the installation requirements. A liquid handled above its flash point, or sprayed, or transferred at a rate that generates static in a low conductivity fluid, needs an engineering answer rather than a procedural one. Chlorinated candidates such as methylene chloride avoid the fire question but bring toxicological and stabiliser considerations of their own, and high boiling aprotic materials such as N-methyl-2-pyrrolidone or dimethyl sulfoxide trade easy removal for exceptional dissolving power. There is no free choice, only an informed one.

Establish the compliance position early

Three questions decide whether a candidate survives, and all three are territorial. Is the substance registered and permitted for this use in the country where the process runs? Does its hazard classification force controls the site cannot currently provide? Does the finished article carry restrictions on residual solvent that the step must meet? Answers differ between jurisdictions and they move, so treat published guidance as a starting point and confirm the current position with the competent authority or a qualified adviser. A supplier can tell you what a substance is and what it has been registered as; the lawful use of it in your process remains your assessment to make and to document.

Count the cost over the life of the step

Price per drum is the least informative number in the comparison. Consumption is driven by losses, and losses by volatility, drag out and housekeeping. Energy appears in heating, in drying and in condensing. Recovery may turn most of the purchase into an asset or, with an unfavourable separation, into an operating burden. Disposal cost follows the waste classification of the used stream, which depends on what the solvent picked up as much as on what it was. Add the containment, ventilation and monitoring the classification demands, and the ranking often reverses against the purchase price order.

Run the trial like an experiment

Obtain production lot material with its own certificate, never a laboratory reagent, and record the lot.
Fix and record the variables the plant can actually hold: contact time, agitation, temperature setpoint, bath loading.
Measure the outcome against the written requirement, including the acceptance test the customer will use.
Run a soak test on every wetted material for a period representative of service, not of a shift.
Test the used bath, not only the fresh liquid, since ageing and contamination change both performance and waste classification.
Confirm the recovery or disposal route with the intended contractor before the first full batch, not after it.

Frequent mistakes

Selecting on a single physical property, usually boiling behaviour, while ignoring solvency and residue. Carrying a laboratory result to plant scale without checking mass transfer or geometry. Substituting a material mid contract because the price moved, without repeating compatibility work. Treating a hazard reduction as automatic when a substance is replaced, when in fact the risk depends on exposure and containment. And qualifying a supplier on a single sample, rather than on the consistency of several lots, which is what a process actually consumes.

A grade recommendation is only as good as the description of the step it serves. Sending the task definition, the plant constraints and the acceptance test alongside the enquiry, through contact, usually shortens selection from months to weeks.