|Ask the Experts|
April 9, 2020
Tacky Substance Under Zero Clearance Parts
A CM is producing power supplies for us using no-clean leaded solder paste. Some large components (0.5" x 0.5") sit flush on the board. We are having failures after only a few days of operation that look to be the result of an electrical short. When we mechanically remove these large parts we find a tacky substance the consistency of rubber cement underneath.
Our initial thought is that this is uncured no clean flux from the solder paste that cannot escape from under the components during reflow. The no-clean paste manufacturer says that it's impossible for the flux to not be cured after reflow, however in its uncured state it would be tacky and would breakdown under voltage (these parts do see 300VDC).
Is this uncured residue common when no-clean leaded solder paste is used with large no clearance parts? What causes this? Can we bake completed assemblies to ensure the material is fully cured? Does the part need to be changed to one with a clearance gap?
|Expert Panel Responses|
Solder paste flux is made of about 50% flux by volume with about 25% of the flux being various solvents which enable the rheology of the solder paste so it can be printed and have some tack to hold components.
Normally solvents will mostly evaporate during the reflow process. But if solder paste is trapped under a part the solvents may not completely vaporize and the flux residue may not harden completely. Residual solvents will make the remaining active ingredients in the flux mobile; this at times can be a problem depending on current density and terminal spacing.
It is important to use solder paste with an ROL0 classification; this gives the highest reliability. Baking out of all the solvents can be difficult if clearance is very tight.
The only way to eliminate the issue of trapped solvents is clearance but a higher preheat (soak) and longer TAL may be tried to see if this helps in drying the flux residues.
Hardened flux residues of ROL0 type solder pastes are normally quite reliable but complete cure is important.
Senior Market Development Engineer
I am sorry to tell you that you're right and your flux vendor is wrong!!! That is exactly what is happening and the large components are trapping the flux and causing a conductive soup under the part with very good stray voltage pathways.
We see this all the time and the flux vendors do not understand that there no clean flux when not completely heat activated is a corrosive residue. We can test the localized area under the large component and determine how much flux residue is left behind under the component and show you how to complex (heat activate) the flux and teach the vendor about their own flux.
President/Senior Technical Consultant
A few thoughts come to mind right off the bat. Assuming that the substance is flux residue and not originating from the part in any way, I would ask your CM to look at the reflow profile for the assembly and make certain that it is correct. It sounds as if no-clean solder is not being properly reflowed and the flux chemicals are not being fully processed and therefore leaving a tacky substance which should not be there.
Reflow profiling can be validated through the use of thermal probes which are typically connected tolarge (read: high mass) parts such as the one that you described in your question specifically to make certain that it sees the correct reflow profile times, speeds and temperatures such that the paste is properly reflowed for that part in accordance with manufacturer specifications.
Secondly, you should have the CM look at the solder paste foil design (specifically thickness and apertures) to make certain that they are not putting too much paste on the pads which can result in excess residues,especially for large parts. They might want to experiment with different thicknesses and a home-plate or reduced full-aperture design might be required.
It is easy to think that large parts need more solder than they actually require under the assumption that more is better when this is not true. Perhaps they are using a foil that is too thick or feel that they need to unnecessarily step the foil for that part. It also sounds possible that they could be deleting the apertures for this part and hand soldering them, which would definitely leave a significant amount of no-clean residue under and around the parts.
Finally (and not necessarily so obvious), make certain that your CM is using solder paste that is within the manufacturer's shelf life or expiration date, that it is stored correctly, brought to room temperature before use and is not contaminated in any way. Solder paste that is not stored correctly can show separation of the flux from the alloy (especially in tube or syringe format).
Expired paste can show signs of degraded performance in many ways and paste that is not brought to ambient room temperature before use may not perform as expected. Make certain that they are not mixing pastes (old with new) or even between manufacturers and chemistries (water-soluble with no-clean), especially if they are attempting to reclaim paste (which only continues to get more expensive) at the end of a shift.
These are certainly extreme situations but nonetheless valid as you delve down the rabbit-hole to resolve this issue. Bottom line:look at the overall process here as the root cause and not necessarily focus on the paste on its own as the issue.
Sales & Marketing Manager
Technical Manufacturing Corp.
It is very normal for flux residue to remain under these low-clearance components and to potentially bridge between contacts, in fact it is very difficult to remove such residue, even with aggressive cleaning. If the residue is still very soft it is likely that ionic transport can still occur,and at the voltages mentioned it would be far more likely. The end result will be dendritic growth and eventually a short. Evidence of the growth should be present after mechanical removal, provided the residue remains with the PWB and is not too badly disturbed. Optical microscopy should reveal the path. Of the short.
The paste manufacturer's statement leaves out some key information. Although the paste should be completely reacted after reflow, the unique situation where you have such a large component with a very low standoff will effectively exclude oxygen from the environment under the component. Most pastes rely to some degree on oxidation processes to harden the residues, in fact some will have extremely soft residues immediately after reflow if reflowed in nitrogen, and much firmer residues if reflowed in air.Over time, most no-clean residues will harden further.
It may be possible to accelerate the process of the oxidation reactions by baking. The presence of the large part will still inhibit the process, and it will be necessary to perform testing to determine when and if the residue achieves acceptable electrical properties. Recognize that it may never do so, and it may be necessary to investigate a change in paste and/or a change in the part standoff.
The term "cure" really doesn't apply to solder paste fluxes -- curing generally applies to polymers and cross-linking behavior with heat or other means of cross-linking (UV, etc.). When solder paste fluxes are heated,they don't really cross-link in the same way that polymeric materials do.
The solder paste system contains high-boiling solvents which act as the carrier for the various solid flux contents. Those solvents will generally evaporate completely (or nearly completely) with sufficient heating. However, with very low standoff components, the solder paste solvents are entrapped underneath the component, which prevents the solder paste residue from everfully drying out.
So despite an appropriate amount of heat applied to the system, the area isn't ventilated enough to allow the flux solvents to evaporate, leaving the semi-solid, tacky material under the component that you have described.
So based on your description of the situation, I would assume that you have no-clean residues that have not been properly dried out.
Given this, the next question is "why would no-clean residues give me this problem? After all, isn't reliability without cleaning the main benefit of no-cleanfluxes?" And the answer to this is "not really." No-clean solder paste fluxes really need to be mostly dried out to a semi-solid state in order to preserve the reliability of the system.
If no-clean residues are overly fluid (like jelly or worse), this allows for mobility of the other flux ingredients upon power-up. (Normally, those flux ingredients would be "trapped" in a solid residue if the solvent was evaporated off, and no mobility equals lower risk of electromigration.) The flux ingredients will have some ionic behavior, and when left on the board in a liquid form, they can easily migrate and give the kinds of failures that you describe.
Once these failures are formed, it will require full rework of those components.
Prevention is the bigger (and trickier) concern. This will happen again without a change somewhere in the process. Either the component needs to be redesigned with some clearance to allow for proper ventilation, or the system somehow needs to be cleaned. Cleaning will be very difficult given the low standoff, so my recommendation would be to design in some small standoff into this component.
General Manager - Electronic Assembly Americas
I have heard of instances of this effect occurring in narrow gap situations. It is not a cure issue since fluxes do not cure, in that there is no deliberate chemical polymerisation or cross linking of the flux components during the reflow process. What I suspect is happening here is the the flux as it heats up is flowing under the component, at this stage the flux contains a high boiling solvent once under the component the evaporation of the solvent is inhibited.
As a result not all the solvent is driven off during the reflow. The remaining solvent will act as a plasticiser to the resin that was drawn under with the solvent (hence the gummy characteristic of the residue). By plasticising the resin the ionic material present in the flux medium are now mobile under a sufficiently strong electric field, causing your failure.
Senior Applications Chemist
In order to answer this question, we first need to determine the cause of the failure and understand the nature of this residue. We recommend conducting analytical tests such Ion Chromatography / Ionic contamination tests to see whether the residue is ionic in nature or not. If not ionic, further investigation could be done using a FTIR equipment to analyze the non-ionic residue. The client could also send a sample of this residue to the flux manufacturer or a third party lab , to compare the flux formulation to that residue to see if they would match.
Since in the field, the atmospheric humidity and temperature levels cannot be controlled (unless it is climate controlled environment), failures such as electrochemical migration and leakage current would most likely to occur unless the boards go through a defluxing / wash process. Has the client considered washing the boards to see how cleaning could improve the reliability?
Application Technology Manager
If the contaminant is causing you a reliability problem, then it needs to be eliminated or controlled. It would be a good idea to use a small spacer underneath the components if that is possible. If not possible, these parts can be cleaned in a vapor defluxing (vapor degreasing) process. The hot vapor has an infinite surface tension and can penetrate the smallest crevices in order to remove "no clean residues" or uncured paste or other contamination.
However, even with hot vapor cleaning, if the solute(contamination) is substantial, then the condensed hot vapor will not be able to remove the contamination because the solvent / solute mixture will be too thick to slide out from under the component. A 2 to 3 mil spacing under the component would make a hot vapor cleaning process the best choice. With a 2 to 3 mil spacing other types of low surface tension solvents could also be used.
There are some issues to consider, one of which is the height of the component after reflow and the ability of the cleaning solvent to get within that space. More than likely the solvent cannot penetrate that space and if it could then it would have to be flushed out to clean out the contaminated material and this is very difficult due to the surface tension of the cleaning solvent and the size of the product.
I would suggest getting aspectrographic analysis of the material beneath the component evaluated by a lab and to compare that to the original flux which would provide an answer to whether or not it is the same material.As for this particular material reaching soldering temperatures, this is a gamble. Its physical size will definitely impact the temperature it sees as compared to the smaller components.
This part should be individually profiled to prove it reaches the soldering temperature. I know you are saying it reflowed that is good enough;well I don't believe that as I think we need a higher temperature and a longer dwell time to complete the transformation of those fluxes to a benign condition at this particular location. Keep in mind there is no exit for the liquid volatiles to escape from under a component .5 x .5 inches, so although the soldering temperature for the solder to reflow was reached more than likely it was the minimum temperature and this is where the problems begin.
Your question as to whether or not you can bake to make the material cured, means you would have to bake the boards at over 250C for a given period of time and this may impact the reliability of all the other components on the boards, so from my perspective this is not a solution to this issue.
I agree with the possibility of using a component with higher clearance, which would give you a better chance of cleaning beneath the component.
If more is needed please contact me at your convenience and we can discuss it further.
Vice President, Technical Director
Highly dense bottom termination components decrease conductor pitch, spacing, and standoff height from the board surface to the bottom of the component. The complexity arises from the variability of components and their function. For example, standoff height isn't an issue with a BGA pitch, while for other components, pitch may not be the issue but stand-off is.
Current spacing trends for some component types result in flux residues bridging conductors. The combination of flux bridging a gap and incomplete volatilization under components may expose a reliability risk. Additionally,the low stand-off can result in incomplete cleaning, which is also a significant factor.
Tighter spacing between conductors results in a higher electric field. Past researchers reported reliability risks due to voltage swings, high frequencies, leakage currents, and high impedance.
Is this uncured residue common when no-clean leaded solder paste is used with large no clearance parts? Yes, when the component gap decreases, flux residues flow under the component during reflow. This effect is similar to how underfills work. During reflow the liquid flux in combination with surface tension effects is drawn under the component. The tighter the gap, the greater the probability of underfilling the gap with flux residues. What causes this? Surface tension effects and the low surface areas next to the interconnects are driving causes of flux underfilling the bottom side of the component.
Can we bake completed assemblies to ensure the material is fully cured? Unlikely, due to level of residue building up under the component, and as such, incomplete volatilization occurs.No-clean flux residue that has not been fully activated resides next to the conductors (I/Os). As the part is powered-up, low levels of moisture can create a corrosive electrolytic solution that creates a path for electro chemical migration. This is most likely the root cause of the part failure.
Does the part need to be changed to one with a clearance gap? One option is to remove the soldermask from under the component. Non solder mask defined pads increase the gap height. If cleaning is performed, higher gap heights open the process window.If the board is a no-clean assembly, high gap heights increased the probability to activating and rendering the residue non-conductive.
The zero/low standoff is preventing complete solvent volatalization that is in turn preventing the rosin residue to form an interprotectiver layer over mobile ions. This allows for the ions to form a conductive path in the presence of the 300 V and causing failures.
The 300V DC may be the key factor here as bottom-termination components such as QFNs & LGAs are built with no-clean Pb-free pastes all the time with no issues, where the same concern of potential incomplete solvent volatalization applies.
In your case, you may want to consider solder solutions where the flux quantity is exponentially reduced.
(a) Flux-coated solder prefroms where the flux quantity is 1% versus a standard no-clean paste where the flux quantity is 50% by volume. A preform can be made as 0.5 x 0.5 inches, secured in place with 4 small paste dots and reflowed
(b) ultra-low residue fluxes where the flux residue is 1-2% versus a standard no-clean paste where the post reflow flux residue is 35-45%. These low residue pastes are very low on solvents as well.
Technical Manager - Europe
One of the methods is how the board is placed the board in Vapour Degreaser. The board has to be placed vertical so the condensed solvent can drip down dragging the flux residue and contamination. In case the boards are placed flat, or at an angle, the flux residue may not drain out completely. Definitely stand-off will be best solution to give sufficient clearance for cleaning.
Astra Microwave Products, Hyderabad, AP India
Even with components essentially sitting flush with the board, the solder paste deposit can provide enough height to allow outgassing. Modify the stencil design to a cross hatch pattern that will provide channels between the solder paste that to allow the solvent gases to escape. Have as long of a soak phase as tolerable to give time for the solvent gases to escape. If you can outgas the solvents, the remaining flux should harden and eliminate your shorting issue.Bill Meyers, Argo EMS
|Submit A Comment|
Free Newsletter Subscription
Circuitnet is built for professionals who bear the responsibility of looking ahead, imagining the future, and preparing for it.
Insert Your Email Address