Process for Treating Aluminum and Aluminum Alloy Surfaces

It is, of course, known that metals such as aluminum and alloys therewith contain an oxide surface thereon, resulting from the metal oxidizing in normal atmospheric environment. This oxide has a very low adhesion itself for organic coatings, and also effectively prevents the buildup of a boehmite or hydroxyoxide coating on the metal surface. For this reason, such barrier oxides must be removed from the metallic surface so as to allow the proper boehmite formation.

Chemical reagents which are suitable for removal of the oxide coating include those compounds which in essence are capable of dissolving aluminum oxides. Solutions of strong bases such as sodium hydroxide or potassium hydroxide are particularly effective because of their capability to rapidly dissolve the naturally occurring oxide film on the metal surface. Solutions of other alkaline reagents, such as amines (e.g., ammonia, ethylamine, ethanolamine, ethylenediamine, etc.) or metal salts (e.g., sodium carbonate, calcium carbonate, magnesium oxide) may also be utilized. In addition, acids such as hydrofluoric acid may be used. Preferably, aqueous solutions of the reagents are utilized, since they are economical and present minimum solvent handling problems.

This chemical reagent dissolves the oxide film on the metallic surface in accordance with the following reactions:

in the case of an alkaline reagent, and

in the case of acidic reagents. Such materials may also mildly etch the underlying aluminum or alloy metal per the following reactions:

with an alkaline reagent, and

with an acidic reagent.

If etching does occur, a small amount of the alloying constituents, such as magnesium in high magnesium-containing alloys will remain on the treated surface, such remainder commonly being termed "smut". In the process of our invention, analysis shows that this smut is incorporated into the hydroxyoxide layer subsequently formed, and does not substantially interfere with the adhesion-promoting characteristics of the hydroxyoxide layer. Therefore, smut removal is not normally necessary in our process.

With the barrier oxide layer removed, the bare aluminum metal is now susceptible to hydrolysis by the action of steam, thereby forming a hydroxyoxide film on the aluminum surface in accordance with the following reaction:

this reaction may terminate as the hydroxyoxide film becomes sufficiently thick to form a new barrier layer.

One critical and unique feature of the process of the invention is that the aluminum surface must remain covered with a thin layer of the chemical reagent when exposure to the boehmite-forming steam begins. If the chemical reagent utilized to dissolve the barrier oxide layer is removed from the metal surface, such as by rinsing, prior to exposure of the surface to steam, it has been ascertained that an insufficient film of the hydroxyoxide will be formed in an allotted reaction time, i.e., about 10 seconds of exposure to steam, which will therefore result in inadequate adhesion of the organic coating thereto, and high speed processing is therefore not available.

One of the distinct advantages of the inventive process, in addition to the reduced time consumption, is that the chemical reagents utilized and reaction products formed are essentially non-polluting. In accordance with the above reactions, waste products that might be discharged from the process would be H.sub.2 AlO.sub.3.sup.- or Al.sup.3+. Neither of these ions are known to have toxic effects. For discarding the chemical reagents, the solutions can be neutralized to a salt if necessary.

The reagent treatment solution temperatures, concentrations, and the time that the solution is to be in contact with the metal surface prior to application of steam thereto depend on the degree to which the reagent dissolves the aluminum oxide barrier layer. With extremely aggressive reagents such as sodium hydroxide, rapid formation of the hydroxyoxide layer will occur if the solution is applied at room temperature and the surface exposed to steam immediately. Less aggressive reagents may require higher temperatures and longer contact times prior to steaming.

Spray application of the reagents may require temperatures, concentration and contact times different from those for dip procedures, and, at least when utilizing sodium hydroxide, the time for processing individual panels can be reduced considerably if the treatment solution is applied by spraying as opposed to dipping.

To minimize the hydroxyoxide-forming treatment time, the concentration level of the reagent should be optimized. If the reagent concentration is too low, the barrier oxide layer may not be completely removed, and rapid formation of the boehmite layer will be prevented, although elevated temperatures, i.e., above about 50.degree. C., will assist the reagent action. If the reagent concentration is too high, etching of the metal surface may continue during exposure thereof to steam, again preventing effective hydrolysis. Reagent concentrations of from about 0.3 to about 5.0 percent by weight have been found acceptable as a general range, depending on the particular reagent chosen and the treatment temperature utilized. In practice, because of the diversity of aluminum oxide solvents available, the surface is exposed to an aluminum oxide solvent-containing solution for a time, at a concentration, and at a temperature sufficient to allow for substantially complete removal of the barrier oxide layer.

For the optional initial metal cleaning step, solvent degreasing or conventional spray or soak cleaners, such as mixtures of sodium pyrophosphate, sodium borate and surfactants may be utilized. Generally, the metal is considered clean if water will wet the surface without beading thereon. However, rigorous precleaning is not necessary if the surface is immersed in the aforementioned reagent solutions long enough to remove contaminants. In fact, precleaning in some instances can be entirely omitted.

Following this precleaning, a thorough rinsing of the metal is recommended. However, trace contamination by phosphate or borate ions from the cleaning solution will not inhibit the boehmite formation as is true in other processes, since the chemical reagent in effect will remove surface contaminants still present following a precleaning step.

The process is applicable to high purity aluminum, e.g., Alloy 1100, as well as to aluminum containing alloying components, e.g., Alloys 3003, 5182 and 5352. It is probable that differing alloys may require different treatment conditions for optimum formation of the hydroxyoxide on the metal surface, because high alloy metals probably generate greater smut buildup than the high purity aluminum surfaces.

The invention will now be more specifically defined by the following non-limiting examples, wherein all parts are by weight unless otherwise specified.

A 14 cm by 15 cm panel of 0.024 cm thick Alloy 5352 (containing approximately 2.5 percent by weight magnesium as the major alloying additive) was cleaned by dipping for one minute in a solution consisting of 95 grams sodium borate, 95 grams sodium pyrophosphate, 6.0 grams Maprofix NEU, tradename for sodium lauryl sulfate surfactant commercially available from Onyx Chemical, 2.0 grams Ultrawet DS, tradename for sodium linear alkylate sulfonate, a surfactant commercially available form Arco Chemical, 2.0 grams Igepal CO 730, tradename for nonylphenoxypoly(ethyleneoxy) ethanol, a surfactant commercially available from the GAF Corp., and 4000 ml deionized water at 70.degree. C. After cleaning, the panel was rinsed by spraying with deionized water at room temperature for 10 seconds and was then dipped for 5 seconds in a 2.5% solution of NaOH in water at 50.degree. C. The excess liquid was drained from the aluminum surface for 5 seconds and the panel immersed in a chamber containing steam at 92.degree. C. After 10 seconds of exposure to steam the panel was withdrawn, rinsed with deionized water and dried with a jet of warm air.

The effectiveness of this treatment for improving the adhesion of organic coatings was tested by the following procedure, which is useful for non-brittle coatings. The primed surface was coated with a vinyl-epoxy lacquer, DAV 210, commercially available from the Mobil Chemical Co. using a No. 30 Meyer bar. After drying in air at room temperature, the coated panel was cured for 105 seconds at 210.degree. C., cooled and cut into 5 cm by 15 cm strips. The thickness of the cured lacquer layer was about 1.8.times.10.sup.-3 cm. To test the adhesion of the coating, each strip was bent to a sharp crease, and 3M Brand No. 610 high tack tape was applied over the bend area and peeled back at a 180.degree. angle. The width of the zone from which the lacquer was removed by the tape was then measured starting from the bend area. The adhesion of the lacquer has an acceptable level if the failure zone is less than 0.08 cm wide. In this case, the zone was 0.04 cm wide, i.e., the lacquer adhesion was good.

Panels of 0.024 cm thick Alloy 5352 aluminum alloy were cleaned and rinsed as described in Example 1, then dipped in various aqueous reagent solutions for 5 seconds, drained for 5 seconds and exposed to steam at 92.degree. C. for 10 seconds. After the steam treatment the panels were rinsed with deionized water, dried, and coated with the lacquer and tested as described in Example 1. The results are shown in Table 1.

The adhesion of the lacquer to all samples was therefore acceptable.

In a series of experiments Alloy 5352 panels were cleaned and rinsed using the procedure described in Example 1, then dipped for 5 sec. in a solution of NaOH in water at 50.degree. C. The concentration of the NaOH solution was varied over a wide range, i.e., from 0.16% to 10%. After draining the excess solution from the surface for 5 sec. the panels were exposed to steam at 92.degree. C. for 10 sec., rinsed with deionized water, dried, coated with lacquer and tested as described in Example 1. The results are illustrated in Table 2.

The above results show that NaOH solutions in the concentration range between 0.32% and 5.0% provide acceptable lacquer adhesion.

A 7.5 cm by 15 cm panel of 0.024 cm thick Alloy 5352 aluminum foil was cleaned by dipping in a 0.7 percent by weight solution of NaOH at 82.degree. C. for 30 seconds, rinsed for 15 seconds with deionized water, dipped in 20 percent by weight nitric acid for 30 seconds, and rinsed for 18 seconds with deionized water. The panel was then dipped for 2 seconds in a 2 percent by weight solution of NaOH at 50.degree. C., drained for 10 seconds and immersed in a chamber containing steam at 92.degree. C. After 10 seconds of exposure to steam the panel was withdrawn, rinsed with deionized water and dried with a jet of warm air. Lacquer adhesion was tested as described in Example 1. The lacquer failure zone was 0.04 cm wide, i.e., the adhesion was acceptable.

In this example, the aluminum panel was passed through a series of spray chambers in a semi-automatic treatment apparatus. A 20 cm by 25 cm panel of 0.024 cm thick Alloy 5352 aluminum foil was cleaned by spraying for 10 seconds with a solution consisting of 120 grams sodium pyrophosphate, 12 grams sodium gluconate, 3 grams Pluronic L-61, tadename for ethylene oxide condensate polyol surfactant commercially available from BASF Wyandotte and 4000 ml deionized water at 68.degree. C. The panel was rinsed for 10 seconds by spraying with deionized water at room temperature, sprayed for 4 seconds with a fine mist of a 2.5 percent by weight solution of NaOH in deionized water and immediately passed into a chamber filled with steam at 90.degree. C. After 5 seconds of exposure to steam the panel was passed into a spray rinse chamber for 10 seconds and was dried with a jet of warm air. The panel was coated with lacquer and tested for adhesion as described in Example 1. The width of the lacquer failure zone was acceptable at 0.04 cm.

A 7.5 cm by 15 cm panel of 0.024 cm thick Alloy 5352 aluminum foil was cleaned by wiping with ispropyl alcohol. The panel was then dipped for 5 seconds in a 2 percent by weight solution of NaOH at 50.degree. C., drained for 10 seconds and immersed in a chamber containing steam at 90.degree. C. After 10 seconds of exposure to steam the panel was withdrawn, rinsed with deionized water and dried with a jet of warm air. Lacquer adhesion was tested as described in Example 1. The width of the lacquer failure zone was acceptable at 0.04 cm.

This example illustrates the insensitivity of our process to contamination by phosphate and borate ions.

A panel of Alloy 5352 aluminum foil was cleaned as described in Example 6, then dipped for 5 sec. in a 50.degree. C. solution, consisting of 980 ml deionized water, 20 grams NaOH and 5 ml of an aqueous solution containing 2.26 percent sodium borate, 2.26 percent sodium pyrophosphate, 4 percent Maprofix NEU, 0.048 percent Ultrawet DS and 0.048 percent Igepal CO 730 by weight. After withdrawing and draining for 10 seconds, the panel was rinsed, dried and coated with lacquer and tested for adhesion as described in Example 1. There was no failure of the lacquer coating, indicating excellent adhesion.

A 14 cm by 15 cm panel of 0.024 cm thick Alloy 5352 aluminum foil as cleaned and rinsed as in Example 1. The panel was then dipped for 5 seconds in a 2.5 percent solution of NaOH in water at 50.degree. C. The excess liquid was drained from the surface for 5 seconds and then rinsed off by spraying with deionized water at room temperature for 10 seconds. While still wet with the water, the panel was immersed in a chamber containing steam at 92.degree. C. for 10 seconds, withdrawn, rinsed with deionized water and dried with a jet of warm air. Lacquer adhesion was tested as described in Example 1. The width of the failure was a 0.32 cm indicating poor, unacceptable lacquer adhesion.

This illustrates that even though a film of water wetted the metal surface, the lack of chemical reagent on the surface resulted in unacceptable adhesion.

A panel of 0.024 cm thick Alloy 5352 aluminum with mill finish was immersed in an aqueous 2 percent by weight solution of NaOH at 22.degree. C. until uniform gas evolution occurred over the surface of the sample, taking approximately 20 seconds. The panel was then withdrawn from the solution, excess liquid drained off and the sample exposed to steam for 7 seconds. The panel was then rinsed, coated with lacquer and tested for adhesion as described in Example 1. There was no failure of the lacquer coating, indicating excellent adhesion.

This example illustrates that precleaning of the panel was not necessary.

Panels of aluminum Alloys, 1100, 3003, 5182, and 5352 were cleaned, treated with 2.5 percent NaOH solution, exposed to steam, coated with lacquer and tested for adhesion as described in Example 1. The adhesion test results are summarized in Table 3.

These results illustrate that acceptable lacquer adhesion can be achieved to very pure aluminum as well as to high magnesium content alloys by subjecting the metal surface to our treatment process.

Four panels of Alloy 1100 aluminum were cleaned and rinsed as described in Example 1. Two of the panels were treated with NaOH and exposed to steam as provided in Example 1.

Two relatively brittle coatings were then applied to one untreated and one treated aluminum surface. The first coating was Kel F-827, tradename for a copolymer of chlorotrifluoroethylene and vinylidene fluoride, which was applied usng a No. 30 Meyer bar from a 5 percent by weight solution in methyl ethyl ketone, air dried for 15 minutes and cured for 20 minutes at 60.degree. C. The second coating was ethyl cellulose, which was applied with a No. 30 Meyer bar from a 10 percent by weight xylene/ethanol solution (9:1 by weight ratio), air dried for 60 minutes and cured at 60.degree. C. for 15 minutes.

In this instance, because of the relative brittleness of the coatings, adhesion was tested by scratching the coating in a straight line and pulling 3M Brand No. 610 tape across the scratch area. The degree of adhesion is summarized in Table 4.

These results clearly illustrate the improved adhesion resulting from the process of our invention.