Document dDneEgoDOV1LOky3RgN9wx9gR

r e c e iv e d May 24, 1955 10:00 AM r MAY 3 liibL THE SHERWIN-WILLIAMS LAB. G. G. Schurr .... J 13 CHROMATE PIGMENT EVALUATION Hie Sherwin-Williams Co. P. V. and L. Annual Symposium May, 1955 Cleveland, Ohio Abstract: A cooperative government industry committee has evaluated nine chrcsnate pigments in three different vehicles in primers for steel, aluninum and magnesium. An Important aspect of this in vestigation was the control of CPVC by adjusting the proportions of a micronlzed and a fibrous magnesium silicate. This allowed the establishment of a constant PVC/CPVC ratio at a constant PVC. Primers formulated in this way to give equivalent per meabilities showed little If any differences in performance due to variation in the type of chromate pigment. CONFIDENTIAL r~ For Sherwin-Williams Technical Personnel Chly. 0007-SWP-043954 There are a number of metallic chromate pigments now being used or considered for use in corrosion lnhibttlve metal primers. In 1950 the Engineer Corps instituted a project to determine which of these pigments was the most effective in primers for the protection of steel, aluminum or magnesium. The Engineer Corps, as represented by H. W. Hill of the Protective Coatings Section of the Engineer Re search and Development laboratories at Port Belvoir, Virginia, en listed the aid of an industrial committee to guide them in this work. The Paint Research Department has participated actively in this com mittee over the past five years. This Is a report on the work done by and through this committee. Hie conmittee agreed in the original meeting to a basic plan that was followed precisely throughout the project. Nine chromate pig ments were selected for evaluation. These vere; Regular zinc chromate Zinc tetroxy chromate Normal lead chromate Basic lead chromate Barium-potassium chromate Strontium chromate Calcium chromate Barium chromate Salt free zinc chromate These nine chrcmate pigments were to be evaluated in three vehicles, oleoresinoua, alkyd, and phenolic; over three metals, steel, alumi num, and magnesium; under three different conditions, water immersion, salt fog, and exterior exposures. It also was decided to use a pigmentation consisting of 40$ of the chromate pigment, 15$ iron oxide and 45$ inert, all by weight, in each of the three vehicles. However, beyond this general formu lation there was considerable disagreement on the details. It is In the working out of these details that the more pertinent part of this report lies. In particular, the interest is in the method of substituting one chrcmate pigment for another while attempting to keep all other factors of the formulations constant. Quite often new pigments are evaluated by merely substitut ing them at a constant velght in an established formula. Such a method may be satisfactory when comparing two pigments of the same composition, differing only in their method of preparation. Even here there is a danger of radical differences in particle size and consequent radical difference! in performance. Obviously, these are 0007--SWP--043955 0007-SWP-000127305 n 2- - not due to difference# in the inherent inhibiting qualities of the pigment Itself. In the case where the nine different chromates are compared there is an even more obvious objection. For example, in equal weights of lead chromate and calcium chromate there is much more chromate content in the calcium chromate type. This is evident if one considers the higher percentage of lead in lead chromate com pared with the percentage of calcium in calcium chromate. Two other possible approaches would be through a substitu tion on an equal chromate or an equal available chromate basis. Again, any comparisons between lead chromate and calcium chromate are con founded by the disparities in weight and solubilities of lead and calcium. Simple substitutions on an equal volume basis, although bet ter than a weight Dr chromate content basis, also may lead to errone ous results through effects on the critical pigment volume concentra tion. Substitution on an equal cost basis may be of some practical interest but has no place in a research project. Consideration of the foregoing leads to the conclusion that the following factors should all be controlled in such a project: 1. Tie chromate pigment content, preferably on a volume basis. 2. !he total pigment content again on a volume basis. 3. The composition of the pigoentation other than the chromate. It. the ratio of the pignent volume to the critical pigment volume. (Control of this factor should also control film permeability.) The difficulty lies in attempting to control the fourth factor, the ratio of the PVC/CPVC, which in effect amounts to maintaining a con stant CPVC. This was done by manipulation of the inerts. The CPVC of a paint la determined by the vehicle binder ab sorption of the pigsentation and the way in which that pigment packs In a dried film. The packing of the pigmentation in turn depends upon the site end shape of pigment particles or agglomerates. Mag nesium silicates are available with considerably different particle size and shape, such as the fibrous talc, Fibrene C-400, and the mlcronized talc, #549 Microne Talc. It was considered possible that the use of each of these inerts separately with each chrcmate pigment might lead to formulations that would all bracket a eonmon CPVC. Ihen, by blending the two Inerts in the proper proportion in a formulation for each chromate pigment, it should be possible to arrive at a comnon CPVC for all of the formulations within one vehicle type. 0007-SWP-043956 0007-SWP-000127306 -3- It was determined that such a result could he obtained. Thus, nine formulations varying only In type of magnesium silicate vere developed for the oleorealnous vehicle with a CPVC of 375^ Similar series of paints were developed for the alkyd and phenolic vehicles at CPVC's of 36.5$. If the fact that fibrous and mlcronlzed talcs will show only minor, if any, differences in performance In metal primers of equal permeabilities is accepted, the original goal of a series of nine paints differing only in type of chromate pigment has been accomplished. A sidelight of this investigation may be of interest at this point. It appears that the shape of pigment particles is more influ ential in affecting the CPVC than Is the size. It was found in a spe cific case that substituting fibrous talc for mlcronlzed talc raised the CPVC from 34.2 to 37-0, an increase of 8.4$. On the otter hand, with a similar formulation, the substitution of ASP 400 clay for ASP 100 clay raised the CPVC from 45*6 to 47.8, an increase of 4.9. It Is reported that the average particle size of fibrous talc is 8 microns compared with 4 microns for mlcronlzed talc. The average particle size of ASP 400 is reported to be 4.2 microns and the ASP 100 to be 0.55 microns. Thus a 2:1 ratio in particle size has had a greater effect on CPVC than an 8:1 ratio. The only apparent explana tion is in the particle shape -- the talcs differ significantly be tween themselves while the clays are similar to each other in particle shape. The only problem remaining was the establishment of the PVC/ CPVC ratio. This was done hy setting the PVC at a point 5$ of the CPVC below the CPVC or a PVC/CPVC ratio of O.95. This was done somewhat artlbrarlly. The attempt was to establish a ratio as close to one as possible without there being a danger that the PVC exceed the CPVC. It Is estimated that CPVC can be determined within 2$. The experi mental error in preparing laboratory batches of paint can be held to .10 pounds in weight par gallon, which in tteBe paints Is less than 0.5^1 in PVC or less than 1*5^ of the CPVC. Even If these errors are additive there la a margin of safety of 1.5$. Admittedly this Is rather thin ground since the CPVC actually does not mark an abrupt change. However, the results Indicate that it was a sufficient margin of safety. When it came to designing the exposure series It was decided for a variety of reasons that a number (actually about 70) of control or check paints should he Included. One group of 27 paints utilized the alkyd vehicle, three levels of chromate pigment (15, 35 and 65^ at the expense of the Iron oxide), with all substitutions on an equal weight basis. The CPVC of each paint was determined and the PVC ad justed to 5$ of the CPVC below the CPVC. There also were several specification primers, both chromate pigment primers and standard primers based on other types of pigmentation, such as red lead or zinc dust-zinc oxide. In addition, each committee member was allowed to submit three proprietary systems if he wished. 0007-S'WP-043957 0007-SWP-000127307 -4- These paints were then applied by an experienced spray nan at a film thickness of 1.5 0.15 mils to degreased cold rolled steel, phosphoric acid cleaned aluminum, and Dow #7 treated FS-1 magnesium. A topcoat of the specification paint 3-173 was applied over half of the surface of those panels destined for exterior ex posure. The panels vere edged and exposed in triplicate. The water immersion tests vere run in the Cincinnati water works in running water at uncontrolled temperatures (generally between 40 and 5 I1) The salt spray tests vere run In Parker Rust Proof's 20$ salt fog room. The exterior tests vere run at National Lead's Long Island site. Vie will concentrate on the results of the exterior tests on steel substrates which are of the most interest. They vere graded last after 30 months exposure. At this time one each of the tripli cate panels was stripped and examined for underfilm corrosion. The remarks here are based on this observation. A cursory examination of the entire set reveals immediately that the "Evaluation" primers show the most consistently good perform ance. This is the group in which all factors but the type of pigment was controlled. As a group and individually they are better than the specification primers and In many cases they are better than the pro prietary primers. The "Check" paints as a set vere more erratic and generally poorer. This is surprising in that the PVC/CPVC ratio was controlled although the CPVC was not. It is believed that a pert of this may be due to deficiencies in technique, both in the formula tion and in the preparation of these primers. The interesting point here is that excellent primers have been formulated strictly on a theoretical basis in three different vehicles. The main basis for this theoretical work was the CPVC concept. It be hooves us to give strong consideration to this concept not only in formulating new primers but also in evaluating new pigments. The method used here to control the CPVC should be applicable to almost any primer and should be very useful. Actually there is little basis for selecting any one chromate pigment as being outstanding for the Inhibition of the corrosion of steel. It would'seem that this method of comparing pigments tends to equalize their performance. Such a result is not too surprising since It is quite possible that the metal Ion from the pigment has no in fluence on the inhibition given by the chromate ion in that even the most Insoluble chromate is capable of furnishing more than the required amount of chromate to impart Inhibition. This is a valuable conclusion since It allows the selection of chromate pigment on the basis of avail ability and/or cost rather than performance. 0007--SVJP-043958 0007-SWP-000127308 r' -5- A complete report on this work prepared, by the Engineer Corps Is scheduled to be published. A preliminary copy of this report has been reviewed. Hie conclusions arrived at are in direct contra diction to the conclusions stated here. It is felt that the main reason for this difference In opinion Is in the use of a rather pe culiar scale for rating corrosion. This consists of two ratings which are averaged to arrive at the final rating. These are as follows: Distribution Degree None Few Spots Scattered Generally Scattered General Concentrated Continuous 10 9 8 7 6 5 4 None Trace Very Slight Slight Slight to Moderate Moderate Severe (Pitting) 10 9 8 7 6 5 4 It would seem that In this scale too much emphasis (10 down to 6) is paid to differentiating between degrees of corrosion that veil could arise from experimental error. Too little emphasis (6 down to 4) Is paid to degrees of corrosion that probably are significant. Thus In the "Evaluation" set of primers no rating was shown of less than 6 for corrosion, and yet an attempt has been made to rank all of the chromate pigments In their order of performance. An effort has been made to reconcile the divergent views of these results prior to publication. GGS:ls 0007-SWP-043959