Ionic Contamination Testing: ROSE Method and Extraction Control

Ionic contamination is the residue that carries current when it becomes wet, and it is measured by extracting the board in a solvent and watching the conductivity of the liquid. The test is simple, cheap and often misunderstood, because it reports one number for the whole board and says nothing about where the residue sits.

What Ionic Contamination Is

The ions come from flux activators, from plating chemistry, from handling and from the water used in cleaning, and they remain on the surface as salts once the solvent has evaporated. A clean looking board can carry enough of them to corrode a fine pitch circuit in a humid environment over months.

The effect is electrochemical rather than immediate, which is why ionic contamination is a long term reliability issue rather than a yield issue. It shows up as dendrite growth between conductors, as corrosion on a termination and as leakage that increases with humidity.

The ROSE Test and What It Reports

In the ROSE test the board is placed in a tank of a solvent mixture, usually 75 percent isopropanol and 25 percent water, and the conductivity of the liquid is monitored while the board is extracted. The rise in conductivity is converted to an equivalent mass of sodium chloride per unit area, and the result is quoted in micrograms per square centimetre.

A common acceptance limit is 1.56 micrograms of sodium chloride equivalent per square centimetre, and the figure comes from a standard rather than from the product. That makes it a screen rather than a specification, and the number has to be interpreted with the size and the density of the board in mind.

Extraction Method and Its Variables

The extraction depends on time, temperature, agitation and the ratio of solvent to board area, and changing any of them changes the result. A board that is extracted for five minutes instead of ten reports less contamination, not because there is less on the board but because less has dissolved.

The solvent also has a finite capacity, so a large panel in a small tank saturates and stops extracting. Most systems therefore quote a maximum board area for a given tank volume, and the limit is checked before a large panel is tested rather than after a surprising result.

Setting Limits That Mean Something

A single figure for a whole board hides the fact that the residue is not distributed evenly, and the areas that matter are the ones where conductors are close together. Where a product has fine pitch devices, the limit is often tightened or the test is supplemented with a local measurement.

The limit also depends on the environment the product will see, because a coated board in a dry enclosure is a different case from an uncoated board in a humid one. The cleanliness requirement therefore belongs to the product specification, and the ROSE result is evidence rather than the requirement itself.

What the ROSE Test Does Not See

The test measures ionic material that dissolves in the solvent within the extraction time, so a residue trapped under a component or inside a connector body may not be reached. It also cannot see non-ionic residue, such as oils and some rosins, which do not affect conductivity but do affect adhesion and coating.

That blind spot is why a passing result on a highly populated board is treated with caution, and why the process is controlled by the cleaning parameters rather than by the test alone. The test detects a change in the process, which is what it is good at, rather than proving a board is clean. The same residue also decides what a coating will bond to, and our conformal coating notes describe the adhesion side of that measurement.

Local Extraction and Other Methods

Where the residue is expected under a specific component, a local extraction is used: a small volume of solvent is applied to one area, collected and analysed. Ion chromatography goes further and identifies which ions are present, which separates a flux residue from a plating residue.

A surface insulation resistance test measures the electrical consequence rather than the chemistry, and it is the method that comes closest to the failure mode. The three are complementary, and the choice depends on whether the question is what, where or how much.

Process Variables That Drive the Result

The variables are the flux chemistry and the amount applied, the cleaning agent and its concentration, the rinse water quality, the bath life and the drying step. A cleaning bath that is at the end of its life leaves more residue than one that is fresh, and the change is gradual until it is not.

Rinse water is often the overlooked item, because water that carries ions from a previous stage deposits them on the board as it dries. The water is monitored by its own conductivity, and the limit for the rinse is set from the cleanliness requirement.

Sampling and Monitoring

The test is run on a sample rather than on every board, and the sample is taken after the last process that can contaminate the board, which is usually the final clean. Testing a board before a subsequent handling step measures the wrong thing, however carefully it is done.

Sampling frequency follows the stability of the process, with a higher rate after any change to the flux, the cleaning chemistry or the equipment. A trend on a control chart is more useful than any single result, because it shows a bath drifting before the limit is reached, and a step change on the chart usually follows a bath change that was made late or a rinse stage that was bypassed. Our board failure notes describe how an electrochemical failure is recognised when it appears months later in the field.

Records and Corrective Action

The record carries the solvent, the extraction time and temperature, the tank volume, the board area and the result, because a figure without its method cannot be compared with a supplier’s. The date of the last bath change belongs in the same record.

When a result exceeds the limit, the action starts with the bath and the rinse rather than with the board, because those are the variables the process owns. Our assembly cleanliness notes describe the same sequence from the assembly side, where the residue affects coating adhesion as well as corrosion.

FAQ

Is a passing ROSE result proof that a board is clean? It is proof that little ionic material dissolved in the solvent under those conditions, which is a narrower statement. Residue trapped under a component or non-ionic material is outside what the method can see.

Why is the limit expressed in micrograms of sodium chloride? Because the measurement is a conductivity change, and an equivalent mass of salt is a convenient way to express it. Any ionic species would produce a similar change, which is why the test is not specific.

Can the test be run on a board with a conformal coating? Only on a board that has been coated and damaged, or on an uncoated sample from the same build, since the coating prevents the solvent from reaching the residue. That is one reason the cleanliness check is made before the coating is applied.