IPC defines a flux as ‘a chemically and physically active formula which promotes wetting of a metal surface by molten solder, by removing the oxide or other surface films from the base metals and the solder. The flux also protects the surfaces from re-oxidation during soldering and alters the surface tension of the molten solder and the base metal.’
Fluxes are used to:
Before reading further, you should consider why assisting in the transfer of heat between parts being soldered, and thus helping to equalise their temperatures, might be particularly important for reflow soldering applications.
A flux may be solid, pasty or liquid, depending on how it is to be used. Its principal constituents are:
When used for wave soldering, fluxes are generally ‘low solids’ with perhaps only a couple of percent of flux, and the balance solvents. The resulting materials are relatively thin and mobile. However, when formulated as part of a paste for reflow soldering, the flux will also contain additives to improve the paste ‘rheology’ (flow characteristics) and ‘tackiness’, and to help slow the sedimentation of solder particles. In many cases, paste components will have more than one function.
For most fluxes:
An additional, essential requirement for electronic soldering is that any residues should be solid, relatively inert, and have good insulating properties.
Traditionally there were three main ways of categorising a flux:
In terms of active ingredients, there are two main categories of flux base material for which details are given in the sections which follow:
While inorganic fluxes are used for metal-working purposes, they are not generally used for applications in advanced printed circuit assembly because of their potential for corrosion.
Most fluxes which are soluble in organic liquids are resins dissolved in a combination of aliphatic alcohols. Most of the resin used is based on colophony (usually called rosin), which is a natural product obtained from the sap tapped from various species of pine tree, steam-distilled to produce liquid turpentine and solid colophony. Each source of rosin has its own chemical make-up and characteristics, and careful selection is needed to ensure even moderate uniformity and repeatability of the resulting flux.
Rosin has been in common use as an electronics flux because it has a combination of favourable properties:
With its solid state and glassy structure, rosin is not an effective flux until it melts, when the material has sufficient mobility to enter into reactions with the surfaces to be joined. It starts to soften at 50–70ºC and is fully fluid at about 120ºC, at which temperature plain rosin will clean lightly oxidised copper. However, joint materials are frequently more severely oxidised and, in order to enhance the flux action and give the right results in an acceptable time, rosin is normally accompanied by a chemical system, called an activator.
Despite the widely recognised term ‘activator’, these added chemicals do not activate the rosin, but simply act directly in addition to it. Two salts were commonly used as activators:
These salts are readily soluble in alcohol or water (though not in colophony itself), and decompose at slightly below the soldering temperature, yielding hydrochloric acid. The ‘activity’ of this type of flux is expressed as a percentage of chloride ions (Cl−) relative to the rosin content.
Both activity and corrosivity depend on the percentage of activator. Figure 1 illustrates the effect on the wetting time for clean copper at 235ºC of increasing the chloride ion content in rosin flux.
Organics containing active functional groups such as carboxyl (–COOH) and amine (–NH2, –NHR, –NR2) are also good fluxing agents, and are now in more common use because of their reduced corrosivity. However, to achieve the same effectiveness as chloride-containing fluxes, a much larger addition of activator is needed, and hence these ‘halide-free’ materials are generally used only for mildly activated fluxes.
In the majority of applications pure rosin (what used to be called ‘Type R’ flux) is insufficiently active to remove surface oxidation and produce good wetting, so activators are added to improve the wetting capability. In many previous specifications, such as the long-established US military QQ-S series of specifications, the level of activation was indicated by the following designations:
Colophony is a natural product and hence variable in quality, depending on source and processing. Synthetic resin substitutes have sought to provide a more consistent and purer product. One commercially available substitute is based on the ester pentaerythritol tetrabenzoate: this fumes less and spatters less than rosin fluxes during the soldering operation, and its residues are even less corrosive than those of colophony, and so need not be removed after soldering, unless further activators have been added.
A wide range of other organic materials may be used as fluxes, and specifications often distinguish between fluxes which are primarily resin compositions, or are substantially free of both colophony and resin materials, in which case they are classified as ‘organic’ types.
This category is based on a solution of strong activators which are designed to produce post-soldering residues which dissolve in water. The activators are generally more aggressive than those found in rosin fluxes, so are better able to deal with poorly solderable surfaces.
However, both the activators themselves and the metal salts produced during the soldering process are potentially more damaging to the circuit and thorough washing is necessary. This is not always easy to achieve with high-density circuits: the difficulty of cleaning under a component increases as the fourth power of the inverse of the stand-off – a 0.1mm gap is 16 times more difficult to clean than a 0.2mm gap!
Colophony (or rosin) is a common constituent of electronic fluxes. What characteristics does it have which make it useful for this purpose? What is normally added to a rosin flux to improve its effectiveness?
There has always been a confusing mixture of national and international specifications for fluxes, dominated by US military specifications. However, over recent years:
The descriptions below cover the most commonly used tests, most of which are described both in ANSI/J-STD-004 or ANSI/J-STD-005, and in the IPC-TM-650 Test Methods Manual. The sections of this manual may be freely downloaded at http://www.ipc.org: go to ‘Online resources’ and then ‘IPC documents available for download’. Similar tests are to be found in British Standard BS EN ISO 9455.
IPC-TM-650 Method 2.3.34 is a measure of the material left behind when the flux is heated for one-hour periods at 85°C until a constant weight is obtained.
There are two tests used: the silver chromate test (IPC-TM-650 Method 2.3.33) detects halide concentrations above a pre-set level by looking for a colour change in a chemical absorbed on a paper when a drop of flux is placed on it; specific ion titration (IPC-TM-650 Method 2.3.35) gives a direct measure of halide content, and can detect halide levels as low as 20ppm.
In IPC-TM-650 Method 2.6.15, a fixed weight of the non-volatile
matter from the flux is added to a fixed weight of 60:40 tin-lead
solder in a depression in a sheet of copper. This sheet is floated
on a solder bath to reflow the alloy, and then stored at 40°C
and 93% RH (relative humidity) for ten days, after which it is inspected
for ‘signs of corrosion’ in or around the flux residues.
This test is most severe on low-solids fluxes (which have to be concentrated for the test) and difficult to interpret.
A measure of the chemical corrosivity of a flux, the copper mirror test is referenced in a number of specifications (such as IPC-TM-650 Method 2.3.32), and is in common use as an indicator of the level of soluble ionic and aggressive activators in a flux. The test assesses the corrosion by cold flux (23°C) in a humid environment (50% RH) of a very thin film of copper on a glass slide: if there is significant corrosion, the film becomes transparent.
For liquid fluxes, the test involves placing two drops of a 35% solution of the flux on a ‘copper mirror’. This is a glass microscope slide on which has been deposited approximately 30–50nm of copper, defined as a Cu thickness which allows the transmission 10±5% of normally incident 500nm wavelength light. Typically, a flux made of pure water-white gum rosin and isopropanol acts as a control, and a drop of it is also placed on the mirror. The mirror is then placed in a controlled environment of 23±2ºC and 50±5% relative humidity for 24 hours, rinsed in isopropanol and examined for any spots or areas where the copper has been removed by the flux.
The test has several drawbacks which have led critics to question its relevance.
However, the test is quick and easy to perform, and its results correlate well with those of the more lengthy and involved surface insulation resistance test. It is widely accepted and is especially useful as a pre-screening test when testing a new formulation or as a routine incoming materials check.
In some versions of this test, it is applied to solder paste by first extracting any ionic species by boiling a sample of paste in solvent (propan-2-ol) and then concentrating the extract to approximately 35% solids by weight. It has been suggested, correctly, that this procedure is unfairly weighted against pastes with a low-solids flux content, so IPC-TM-650 Method 2.3.32 applies solder paste direct to the mirror.
Harmful residues from the flux (or indeed other process chemicals) may provide a conductive path across a circuit. The surface insulation resistance (SIR) test of IPC-TM-650 Method 184.108.40.206 detects the effect of flux residues on the electrical characteristics of conductor patterns. The board is lightly brushed with flux and then floated on the surface of a bath of solder, flux side down. The resistance of the soldered circuit is then checked to be a minimum of 100MW both at the outset and after a period of storage in warm, damp conditions.
SIR is a lengthy but worthwhile flux evaluation test. Of all the test methods available, this one best simulates the real board assembly application. This is because:
Sample preparation simulates the assembly process as closely as possible, including any delays in transfer between processes if this is significant. For wave soldering, preparation includes fluxing the test board and soldering it; for solder paste, this becomes printing and reflow. Cleaning would also be carried out if the process normally includes this. The parts are then subjected to a defined combination of voltage, temperature and humidity determined by the classification of the product and re-tested over an extended period (for example, measuring after one, four and twenty one days).
To maximise the sensitivity of the test, leakage current is measured between long parallel lines at different electrical potentials. Instead of making very long and narrow circuits, electrically equivalent inter-digitated ‘combs’ are typically used, as indicated in Figure 2. Traditionally, measurements are reported as insulation resistance (in MW) rather than as leakage current . The diagrams in Figure 2 show typical circuitry used to apply a bias voltage during test and to measure the resistance with a reverse polarity test voltage applied.
This test is not only used for evaluating fluxes, but is often found as a measure of surface contamination or the purity of surface coatings or encapsulations.
The classification is based on the materials used and the response of the flux to each of the tests above, judged against the requirements of Table 1.
|Flux type||Halide level||Copper mirror||Silver chromate||Corrosion||Conditions for passing 100 MW SIR|
|L0||0.0%||No evidence of mirror breakthrough||Pass||No evidence of corrosion||Both cleaned and uncleaned|
|M0||0.0%||Breakthrough in less than 50% of test area||Pass||Minor corrosion acceptable||Cleaned or uncleaned|
|M1||0.5 – 2%||Fail|
|H0||0.0%||Breakthrough in more than 50% of test area||Pass||Major corrosion acceptable||Cleaned|
The first pair of code letters denotes flux composition, and is chosen from RO (rosin), RE (resin), OR (organic) or IN (inorganic).
The third code letter shows flux/flux residue activity, and is chosen from L (low), M (moderate) or H (high).
The final code character indicates whether the flux contains halide (1) or is halide-free (0).
Examples of these codes are given in Table 21.
1 Note that there are substantial differences between the ANSI/J-STD-004 classification and that used previously in IPC-SF-818, which did not always specify the material type but included a usage classification.
|ROL0||Rosin-based flux with no halide content, and low flux or flux residue activity. Passes the SIR requirements even when not cleaned.|
|REM1||Rosin-based flux with 0.5–2.0% halide, and moderate flux or flux residue activity. Depending on the specification, may or may not need to be cleaned in order to pass the SIR requirements.|
|ORH1||Organic-based flux with >2.0% halide, and high flux or flux residue activity. Must be cleaned in order to pass the SIR requirements.|
ANSI/J-STD-004 offers a guideline table (Table 3) which illustrates ‘an appropriate, though not necessarily always accurate, analogy between L, M and H type fluxes with the traditional classes of rosin-based fluxes’.
|L0 type fluxes||All R, some RMA, some low solids no-clean|
|L1 type fluxes||Most RMA, some RA|
|M0 type fluxes||Some RA, some solids no-clean|
|M1 type fluxes||Most RA, some RSA|
|H0 type fluxes||Some water-soluble|
|H1 type fluxes||Some RSA, most water-soluble and synthetic activated|
It must be stressed, however, that there is no direct correlation between the ANSI/J-STD-004 designation of a flux, based on results, and any designation based purely on materials. Some fluxes which were formerly considered to be potentially corrosive can now be used; others believed ‘safe’ have had doubt cast on their long-term reliability.
Many common fluxes have in the past borne the designation ‘RMA’. What does this mean? What alternative way of classifying fluxes is now being adopted? Is this new way better?