Creation of the DSM NeoResins Adhesive Challenge blog!
We have developed a new adhesive for laminated wood, but there’s one issue that we have not solved so far. Can you help us? There’s a € 10.000 reward for the implemented solution!!
Feel free to ask, discuss or comment on this challenge, we are looking forward to hearing from you!
Advertisement
I have my doubts about the water insensitivity of the adhesive itself. Have you tried putting a fresh laminate in the deepfreezer?
This reminds me of simple wood glue, by the way. Looks like it’s just not waterproof – should it be?
Interesting thought Martijn.. we will definitely take this into consideration
Below you find summary of the various discussion points that have been send so far:
Several questions on the character of the adhesive.
Without going into too much detail (because of IP and not to make our competitors too wise, guess you understand), the E-850 is an adhesive dispersion. Think in the group of acrylics, urethanes, alkyds.
a) From the failure-mode displayed in the presentation I conclude that you get primarily cohesive failure when exposing the laminate to water. By absorption of water the wood swells substantially thereby exerting stress on the adhesive. The lack in cohesive strength of the adhesive causes the described failure mode.
This is an idea that we got ourselves also. Instead of fighting the problem;(making the adhesive stronger, could we not fight the cause; prevent adsorption of water in the wood to minimize swelling and thus stress? Why would this not happen with current adhesives, are they stronger or do they prevent water absorption?
We found that when we dry the laminate in air at room temperature, delamination seems to be less severe/absent. Drying at elevated temperatures, thus faster drying, seems to increase the delamination. This could show that warping of the outer wood layers because of evaporation of moisture during drying is causing stresses that cause the adhesive to fail.
In fact, we already had started experimenting with adding additives / resins to our adhesive that reduce the water-uptake of the wood. We see some promising results when following this route.
b)The problem is the layer at the edge
The wet film layer at the edge tends to be less than the internal film layer due to the singular region tensile, so it is important to verify the wet film rheology.
I would try to use an anti-sagging agent ( it could be a microgel polymer) or increase the insolubles during the E-850 polymerization step, so this effect ( higher quantity of insolubles) would also correct the formulation with a lower cost.
The adhesive does not dissolve in water, but what happens to the tensile properties
Try making a free film of the adhesive and testing the tensile strength before and after soaking in water.
This is indeed what we did not do yet, but it is now at the top of our action list.
If a low wet tensile is the problem, then cross-linking might help, or try adding something to make the film more hydrophobic like an organosilane.
c) Since there is no adhesive failure, I assume the failure is 100% cohesive. This would mean internally, the adhesive is not able to transfer the peel force from the laminates, which have the tendency to curl up, as shown in your presentation.
It is well known that water is actually a swelling agent for polymers. Therefore, with the ingress of water the Tg is lowered, which makes the (tough) polymer much softer, almost like it hasn’t cured at all, since the solvent is water. This could be checked fairly easily by a DSC measurement where the Tg of a piece of a adhesive as well as one exposed to water is measured.
Like testing the mechanical properties of a wet film, we will also do this analysis.
d) One of the ways of solving this would be to either introduce a hydrophobic chain segment to the polymer, which would in principle keep the water out once the adhesive is cured. Another option would be to create more cross-links between the polymeric chains, which give less space for water to enter the polymer when the adhesive is cured
e) Just several comments
This problem could occur because of wood quality and it is not the matter of the adhesive. If you still have wood samples, try a standard adhesive to exclude the wood issue.
We did that indeed, and the wood is (unfortunately) good and not responsible for failure
f) It is very important to investigate tensile properties of the free film after water treatment. It looks like we don’t have anything outstanding in the formula, and if wood is Ok, the reason of the problem is a water sensitiveness of the adhesive.
Like said above, we will do that soon
g) I don’t know how the coating/adhesive is applied. I imagine that they sandwich the coating between two layers of wood as the coating is still wet. In this case, drying occurs in a mechanically confined geometry between porous walls. Would the surface tension still play a role then? At which scale? (interface, porosity?)
The parts are indeed made in a press. Several layers of veneer are coated with the adhesive, stacked and put in a closed press. Pressing occurs at approx 100 – 120 °C for about 10-15 minutes.
h) I agree about the residual stress, no matter how it’s induced (surface tension gradient, adhesive shrinking upon drying, wood shrinking upon drying after being swollen by the resin water, …)
i) My first impression is that it is a Marangoni effect (i.e. surface tension gradient) which is observed for waterborne paints and which is called edge de-covering.
Essentially the surface tension of a paint is a function of concentration. For a waterborne material, this is normally a decreasing function of solids concentration. During lamination, the surface area at the edges is covered just a little bit less, which means that in time the surface tension there will be lower. Hence there is a trend for the water to move away from the edges: a wellknown phenomena for waterborne paints.
It could be that the viscosity of the laminate is so high that there is no edge-decovering during, but that the residual stresses built up by the surface gradient causes delamination in a subsequent phase. Anyhow, I think surface tension gradients are the culprit
Interesting thought. One way we will check this is to cut out a piece of laminate from the center of a product to see if delamination also occurs here
j) adhesive water sensitive: you claim not because it’s not soluble once dry, but how much is it swollen by water? And how much soluble part is in it? If your resin is a latex, it probably contains surfactants that are usually water-soluble. And these may also create water channels in the tie layer that might affect the bulk properties.
Can you compare the swelling of your dried resins with the swelling of the wood you use?
The issue of surfactants / dispersants was suggested a number of times. We will check the properties of a wet adhesive film.
k) Resin material: is a water-soluble material that react during drying, or is a latex? In the first case it should somehow swell when dried. In the latter case, it should contain soft and hard materials for film formation and final properties. What is the structure of these soft/hard phases?
The resin is a latex that is not water soluble after film formation. I can not reveal too much regarding he nature and structure of the polymer, but we will analyze a dry film to see if water has a negative influence on its properties.
l) Did you try to anneal the laminate before applying water on it? Any improvement?
Yes, does not have an effect
m) The classical question for the end: how does it fail “exactly”?
Any pictures of the resulting interfaces? You write that adhesive remains on both side, but if you have pictures of it – microscopy – that helps to understand the mechanisms of rupture, thus what might have happened).
We did IR analysis of the ruptured surfaces which shows there’s still adhesive present. We do not have any (SEM)pictures
n) What can happen that induce delamination?
Water swells the wood more than the E-850 tie layer: the wood extension induces shearing of the tie layer that fails in the bulk (mechanically confined system). In this case, you could think of either increasing the swelling kinetic/level of your dried resin to reduce the shear, or decreasing the stiffness of the resin to reduce the effect of the shear.
Water swells the tie layer more than the wood. Well, it’s then the reverse that above, with the interesting difference that a more swollen adhesive layer may have lost its strength and fails more easily.
The role of the stress induced during initial drying of the laminate: when you dry your laminate first with wet resin layers in it, water has to go out, and the volume of your resin layer decreases (shrinks). This usually induces residual stress applied on the dry laminate. It depends how much the wood swells (when it dries later, it will shrink and induces stress on the tie layer), when does your wet resin start to “solidify” (i.e. forms a solid network, the earlier the higher the final stress due to more shrinkage) and the elasticity of your forming solid (the lower the stiffness, the lower the stress induced). Depending which of these point is the most critical, you may try to delay the solid formation of your wet resin, by delaying the particle coalescence of your latex or adding surfactant in it), or increase the amount of soft phase to decrease the stiffness of the coating.
Alternatively, did you try to let the resin dry longer on the wood before putting the next layer of wood? This can not be done in practice, but we could try for testing purposes
Interesting comments. This somewhere touches upon an earlier discussion that water penetrates into the wood, upon forced drying the wood warps and creates stresses that break the adhesive joint. Reducing absorption of water could reduce swelling differences between wood and adhesive and thus failure. A thought: the adhesive is not that brittle and should probably be able to deform a bit. In the fatigue test, it outperformed an alternative (more brittle) resin.
o) The water-based resin giving away the water while drying. Ideally the moisture would leave the coating towards the environment whereas in the present case the majority is picked up by the wood and results in the serious swell/delamination that you observe. A reason why this can happen is that the drying process of the applied coating is inhomogeneous, read the external surface of the coating will quickly form a (cured) skin which obviously is impermeable to water and/or moisture. The remainder of the applied coating will continue its drying process but the only way the moisture can leave the layer is through the laminated wood. There it will or swell the wood fibers or interact with the gluing component (potentially water soluble) resulting in the observed delamination. In this case, the deterioration of the wood will be a function of the applied coating thickness, the water content of the coating and maybe the drying process.
What about the coating being specific in a way that it allows moisture migration but hinders water migration. Could one expect in this case that the wood will absorb water to an equilibrium state while will be hindered to realease the adsorbed moisture?
Another hypothesis relies in the micro-nature of the applied coating. I am aware of some technologies, where a coating while drying develops a monomer gradient resulting in an inhomogeneous composition (in the direction surface -> wood) of the coating. Although somewhat farther fetched, this gradient in coating composition could further enhance/block moisture transport through the layer and result in a high
moisture level in the wood, thus swell.
A general approach would be to increase moisture permeability of the coating to allow surface evaporation. I fear that this might not be acceptable to the original requirements of the coating, but realize, moisture permeability is different from water permeability.
I wonder if the drying technique can be adjusted to avoid skin formation on the outer surface (moist conditioning). What about addition of permeable fillers that will keep surface “open”. What about soaps to reduce surface tension and retard the skin formation (some strange thoughts).
An alternative approach, maybe technically difficult in the final application would be a double impregnation with the resin, first a very thin layer, that will harden at decent speed and like assumed above not be moisture permeable (this time towards the wood). The second applied layer will than have to dry “outwards” and substantially reduce the moisture migrating into the wood.
Another option might be the modification of the wood to increase hydrophobicity, addition of a special ingredient to the coating to immediately migrate to the wood surface and reduce moisture absorption, envision other laminates (e.g. different glues, wood types),…
Although we are talking about an adhesive and not a coating, this is an interesting thought also. It might me that our adhesive forms a rather water-impermeable film between the veneer layers that influences drying speed of the various layers, giving rise to internal stresses indeed. A solution to this might be difficult though…
We found that when we dry the laminate in air at room temperature, delamination seems to be less severe/absent. Drying at elevated temperatures, thus faster drying, seems to increase the delamination. This could show that warping of the outer wood layers because of evaporation of moisture during drying is causing stresses that cause the adhesive to fail.
In fact, we had started experimenting with adding additives / resins to our adhesive that reduce the water-uptake of the wood. We see some promising results when following this route.
p) With timber adhesive customers traditional problems are the clamping force applied to the sheets of veneer the type and the moisture content prior to pressing. The application methods used to be spraying ( controlled) or squeegee (uncontrolled, this is what our customer does) all of which can be important.
What is the solids and viscosity of your emulsion? What type of chemistry is this adhesive : pvac, acrylic, polyurethane…..? approx 30 – 40% without fillers, but we use fillers to increase solid content. A too high water content creates problems during hot pressing the (closed) mould.
PVAc adhesives, which are also emulsions, always penetrated the wood (or actually absorbed into the wood pores) and always gave wood failure. If they did not show wood failure, usually there was something else wrong with the parts or the process.
Knowing UF, it dries quite hard. I noticed the presentation mentioned, you had less fatigue, I took this as better creep resistance correct, so I thought that this was a tough adhesive. Perhaps some improvement to the toughness (less elastic) of the adhesive could improve the results? Maybe the adhesive softens in the drying process enough for the stress of the wood to pull apart,,, but you also said you saw the same issue with air-dry, so perhaps this is not the issue…..
We will check the Tg and mechanical properties of the adhesive when wet and when warm
q) Even with the best PVAc’s, there were applications (especially multiple lay-ups like the picture I saw in your presentation) when the surface preparation was too poor for the adhesive to overcome alone. Indeed UF worked better to “bridge” the gaps in the wood preparation. Here you can only increase pressure, but at a certain point either you squeeze too much glue out of the joint, or the surface preparation is so bad that the adhesive must perform as a gap filler, as well as the normal mechanical bond.
Testing a piece cut in the centre might confirm if pressure is an issue, but if pressure is weak across the entire part, or the surfaces are rough (you can see saw marks under a hand-magnifying glass or you colour the surface with a crayon and you see saw marks) then this might be too much for the E-850 to overcome as it is today.
Interesting comment, we will check this. From what we know now, the veneer layers are quite smooth/flat
r) Are you able to make any samples (even just two pieces of wood) where you get wood failure? Just to understand if under controlled circumstances the adhesive is capable to result in wood failure. Most wood workers expect this as a qualifier for an adhesive.
Yes, we have been able to create wood failure with this adhesive and wood
s) If you allow me, I ask some of the standard questions:
How are the parts made, in a hot press, cold press, RF dying? Hot press (120 °C, 10 minutes), air drying
For how long are parts pressed? Do you know the pressure (?) at the glue line?
Are they allowed to further cure before processing? Yes, there is quite some time between pressing and lacquering
What is the moisture content of the wood before pressing? What is the moisture content of the panel after adhesive is cured?
When testing the parts to confirm “adhesion to wood is good enough” and “outperform current adhesive”, do you realize wood failure yes(see broken bits of wood or only glue-line failure)? Typically, adhesive should form a mechanical bond with the wood, so the wood should break before the adhesive.
If you take a glued panel, then cut it in half and finish this piece (so you are finishing the middle of the piece instead of the edges) do you still get delaminating? This is what we will try next time indeed
If you dry the parts in the same process without finishing the part with lacquer, do you still see delaminating? no
The issue is in the glue-line. It is undercured or too thick. While the adhesive might not be water sensitive, wood is sensitive to water content. Since the delaminating is occurring during the drying of the lacquer or later, this is stress in the wood going from a high moisture content and being dried to a lower moisture content. The issue remains that the mechanical bond of the glue-line is not enough to overcome the stress of the wood as it is drying. See the earlier replies regarding this hypothesis
Adhesive water sensitivity. The adhesive may well not be generally water soluble but can still be moisture sensitive. I don’t know what testing has been done to determine water solubility of the cured adhesive but is there any way you can test the adhesive (HPLC?) under sensible conditions to determine if there is a hydrolysis pathway that produces short chain segments or products? That would be indicative of a breakdown of the adhesive? If you want to get really sophisticated then you may want to try using deuterated (D20) water to see where it ends up. If hydrolysis of the adhesive is really the problem then the deuterium will end up in the hydrolysis products. This of course will not solve your problem but it will help you home in on it. A change in the chemistry is almost certainly going to be needed. We will test the properties of a wet film (Tg and mechanical properties)
Has the adhesion been tested on different substrates other than wood? Did this phenomenon show up in these tests?
It appears that the cohesive strength of the “cured/dry”polymer is exceeded by the enormous tensions at the wood warping during moisture absorption.
Other substrates with less/mo water sensitivity could shed some light on the nature of the failure.
Dear Carlos,
Thank you for you comment.
Indeed, we also have observed/proposed that the cohesive strength of the polymer film is exceeded by the stress that the warping wood is causing.
Instead of increasing the strength of the polymer film itself, we have recently started looking at routes to minimize the warping of the wood itself by trying to minimize water penetration into the wood. As a result of our lab trials, we now have scheduled another commercial application test in the coming three weeks. I will keep “the world” up to date through the various channels.
I strongly feel you have wrong chemistry for the wrong substrate, wrong process, lack of understanding on key variables.
Dear Zubar,
Thank you for your response in this blog, to our website and through Linked-In. Amazing how social media work and what coverage they have. In fact, using social media for this case was part of our learning objectives.
Regarding your reply, of course I hope you’re not right. In some sense, you are right in your comment that we have a lack of understanding of wood adhesives. We are a resin producer to start with, but we have obtained quite some understanding of the use of our resins as adhesive resin. However, this does not compare to the knowledge and experience an adhesive formulator/producer has. this is the reason that, through open innovation, we create partnerships with key players where we both bring in our expertise.
Referring to the challenge itself. The resin we use is a modification of a resin that is used, amongst others, for wood coatings. Here, there are no issues with adhesion, water sensitivity etc. This strengthens us in our view that we are on the right track.
For your information, based on the numerous reactions we were allowed to get on this challenge, and based on our own ongoing research, we have made good progress. This progress has led us to decide to do another commercial application trial within three weeks from now.
I will keep you and “the world” updated of the results of these trials. I hope we will be succcessful. As said, I hope we can prove you were not right in your remarks, but let’s wait and see.
With kind regards,
Erik
Perhaps that has already tried, but what about using an aliphatic isocyanate at the application to increase film stiffness?
Dear Carlos,
This has been tried indeed, and it works..
Without revealing too much details, the resin we used also contains PU-type chemistry
Dear all,
Through this note we want to thank you for your contribution to the “E-850” challenge that we shared with the world through various channels in October.
Why did we do it?
First because we were seeking input for a technical problem and second, because we wanted to experiment and innovate in the way we interact with the world. The E-850 challenge was part of this program.
We want to thank you for your very useful contributions. Your numerous questions and suggestions helped us in getting closer to a solution, and based on them we will do a an industrial trial in December.
You also gave us very positive and helpful feedback on our approach which helps us to improve the way we use social media to serve our customers.
We remain open to any comments, suggestions or remarks and look forward to interacting with you in the near future.
With kind regards, on behalf of the DSM NeoResins team.
Erik Pras