Experimental

Samples of a separated lap joint with minor corrosion inside the joint were tested for surface energy and surface polarity (fraction of the overall surface energy which is made up of polar component). This was done according to the Fowkes method. Fowkes theory is described in detail in technical note TN306e on our website (information database), if you need details. Diiodomethane and water where used as probe liquids and were found to have average contact angles of 56.5° and 67.8° respectively on exposed surfaces from the partially corroded aluminum lap joint. These angles are as opposed to average angles of 63.9° for diiodomethane and 80.5° for water on the customer's freshly formed and pre-treated aluminum stock. All reported angles were measured by the sessile drop method on a KRÜSS Drop Shape Analysis System DSA100.

Applying Fowkes theory to these angles gives us the following data - which tells us that we are certainly not dealing with the same surface for partially corroded aluminum versus fresh aluminum.

Just due to the corrosion that has taken place (and these samples did, by no means, represent what most people would call a heavily corroded joint - rather they were just slightly discolored and had signs of only modest change) the surface energy of the aluminum increased more than 5 mJ/m2 and the surface polarity more than doubled.

To see the problem this poses for CPC treatment one only needs to study the surface properties of the CPC candidates which our customer provided us for analysis. We'll call them samples A through E.

We measured the surface tensions of these samples by the Wilhelmy plate method using a KRÜSS K100 Tensiometer, and also measured their contact angles on poly(tetrafluoroethylene) PTFE on the same KRÜSS DSA100 used to characterize the aluminum samples. The following average surface tension and contact angle values were obtained.

Of course, the reason to do this type of analysis is so that the PTFE contact angle data can be used along with Fowkes theory to separate the overall surface tension of each CPC into polar and dispersive components and % surface polarity. Technical note #TN306e, referenced earlier, also describes this part of Fowkes theory as applied to liquids. Using the above data, and the Fowkes theory, the following surface tension components were calculated for the various CPC's.

You will note that the surface tension values in the data set only vary over a range of about 3 mN/m. However, the fractional surface polarity values vary by more than 11%. All of the CPC's show much lower surface tensions than the surface energy of the corroded aluminum from the joint - so one would think the wetting would be very favorable if the joint was dry and exposed. Also, some of the higher surface polarity CPC's have surface polarities approaching that of the corroded aluminum.
So, initially at least, it would appear we could easily pick the most compatible CPC for the corroded aluminum surface from this list. It would be CPC E, since CPC E has the highest surface polarity and, thereby, the most compatible surface polarity to the corroded aluminum).
However, when we consider that the goal is to find the best CPC for potentially moist lap joints, the choice proves not to be so easy. When the surface is wet you are trying to replace a water/aluminum interface with two interfaces within the joint: a CPC/aluminum interface and a CPC/water interface. The free energy for such a transition is:

Free Energy for CPC to Displace Water = γ CPC/Aluminum + γ CPC/Water - γ Water/Aluminum where γ = interfacial tension.

Interfacial tensions between the corroded aluminum and the CPC's, as well as for corroded aluminum and water, must be calculated based on the known properties of water (46.4 mN/m polar component and 26.4 mN/m dispersive component) and the properties of the aluminums and CPC's as given above. There is no direct way to measure the interfacial tension at a liquid/solid interface. The calculations can be made using Good's equation - also described in technical note #TN306e, referenced above.

CPC/Water interfacial tensions can be measured directly, since both are liquids. And, they were measured for this work using the pendant drop method on the same KRÜSS Drop Shape Analysis System DSA100 used to characterize the CPC's themselves and the aluminum surfaces. The following results were obtained.

From these data the following free energies for water displacement by the CPC's can be calculated:

Here we give data for both the corroded aluminum and a fresh aluminum for comparison. Given that negative free energies for water displacement are favorable, you will quickly note that water displacement by each of the CPC's is favorable if the surface is fresh aluminum. However, with the corroded aluminum, water displacement is unfavorable for two of CPC's (A and B). For those CPC's which are predicted to support water displacement (show negative free energies of displacement) the prediction is that C will displace water better than E, and that displacement favorability for D is likely to be intermediate.

As to the cause for these trends - for the most part the interfacial tensions for the CPC's on the aluminums are rather low. This is to be expected, because it is well known that the CPC's are very effective in treating the surface of even corroded aluminum in the absence of moisture. The interfacial tensions for the CPC's are what is high, and what drives some of the free energies to be higher than is desirable.

There is, however, some trade-off to be had with CPC's regarding interfacial tension. The interfacial tension with water cannot be made too low or else it will easy for water to displace the CPC once it is in place. In other words, the CPC will wash off, and/or, the CPC film will emulsify water - thus bringing water closer to the surface. On the other hand, as we see here, if the interfacial tension is too high then the free energy for water displacement is high. We don't necessary understand those limits currently and the degradation of a CPC coating over time after treatment may be the subject of a future paper.

For now, we continue to focus on the ability of CPC's to displace water and penetrate lap joints, To that end, we designed one more experiment which attempts to directly confirm the impressions which the free energy of water displacement data above give us.

The experiment was conducted as follows. We took corroded aluminum pieces from the customers separated lap joints (having dimensions of 2 cm x 10 cm), wet them with water by dipping, and combined them in pairs corroded-face to corroded-face, thus creating controllably moistened lap joints. The pieces were held together with binder clamps (four used to hold together each two piece sample). They were then brought down so that one of the 2 cm edges of the sample just touched the surface of a beaker of CPC, as shown below. Each sample (model lap joint) was allowed to sit touching the CPC surface for a period of 6 hours. It was then removed, separated into two individual aluminum pieces again, allowed to dry, and recombined backwards, with the two corroded surfaces exposed. Adhesive was used to firmly combine the samples this time - creating a dual sided sample having two main surfaces which were previously the "lap joint" interiors.