---

MONTHLY REVIEW

Published by the
American Electroplaters Society
Publication and Editorial Office
3040 Diversy Ave., Chicago

VOL. XVII    MARCH, 1930    No. 3

---

EDITORIAL

Apparently the future of Electroplating is on trial and is being challenged by the rumors of the use of Stainless or Rustless Steels, or of other names of similar composition, replacing Chrome Finishes in the Automotive Industry.

True, these all have more or less value as a protection against corrosion, erosion, etc., and which we must admit is the “Bug Evil” against the present Chromium Plated finishes.

The Automobile Manufacturers and others seem to take an unnecessary frightened attitude toward the continued use of Chromium Plating for their product, because they have found it is necessary to do a better job than heretofore in order to obtain satisfactory results, and is practically an admission of their own mistakes in not thoroughly testing these finishes before marketing, or in cheapening the process in mass production, or speed or insufficient knowledge in the production of Chromium Finishes.

We heard a great deal a few years ago about Japan or Black Enamels replacing the nickel decorations on Automobiles and several New Models came along and these same makers were the first ones to again use Nickel decorations, only more so, new body lines requiring suitable contrasts to relieve the Hearse-like monotony of color effects.

Again the “Bug Evil” presented itself. Slipshod methods of turning out a real job caused several makers to supplant Nickel Silver and Monell metal for some of the Steel Nickeled parts on the cars and was beginning to be spoken of as a cure for the Nickel Evil. Then came the cry of too expensive, color not so good and requires too much hand labor to keep in condition. Then like a bolt out of the sky comes this new wonderful Metal, Chromium, of which we knew little or nothing, with its rich enhancing color and lustre and with all the characteristics of wear and service claimed for it that no other electro deposited finish ever had before. “Marvelous” and the plater over night had to face the problem of mass production before he had time to catch his breath and before the Manufacturer had sufficiently tested his wares to prove to the public that it would do what had been claimed for it, it was placed on the market.

Erroneous statements made by exploiting sellers of processes, etc., misled the manufacturer, the electroplater and the public- to expect and demand even the impossible and with the hallucination that this was the cure for all bug evils. Chromium finishes have taken the public as no other finish ever has. All concerned had to be educated as to its characteristics when not produced under proper standards of operation, and that a flash of chrome, plated on base metals, perhaps porous and subsequent preparatory surfaces, being applied with no consideration for tests against the elements which might otherwise create a breakdown of the plate has caused the present “Hullabaloo.”

Millions of dollars have been expended in equipment for its production, volumes of literature have been published and contain information which was not available a short time ago, so that we have now learned the fundamental principles of proper procedure.

It is rumored that alloy steels which are neither decorative or enhancing in beauty or color will replace chrome and some well known manufacturers have already adopted it, however, it is well known that in their plants they never did a real job of nickel plating. We doubt if the customer would not rather have a beautiful chrome job on his car than the other, providing, he has some assurance that it would stand up for a reasonable time of the life of his car. We believe the customer would, and it is up to the manufacturer and the plating industry to furnish the public with this by using some definite standards of application in the various operations of production. The plater will do his part, if you let him, so that the failures we have encountered have been lessons and have taught us to become more proficient in the art than heretofore. With electroplaters classes learning chemical control in electro deposition of metals and the research committee giving serious thought to this problem for our associates at the Bureau of Standards to continue their investigations on chrome plating. Chromium plating is here to stay, as soon as we get rid of the “bug evil” that our mistakes or failures in the past will be our success in the future.

---

BRASS PLATING

George Hogaboom
Read at Detroit Convention

The subject of brass plating is one that has received as little attention, if not less attention, than almost any other subject in electro-plating. The data available for any study is of questionable value because those who have made a study of brass plating have not had the practical experience in commercial operations. Most of the formulae that have been used up until recent years have been those which have been published in such books as Roseleur’s, which was published in 1852. We have, since that date, improved greatly in the quality of the materials used in electro-plating. Cyanides, metal cyanides and different chemicals that have been used, have been improved so that it is not reasonable to assume that the formula that was acceptable in those days is acceptable today. The requirements of electro-plating or an electro-plated finish, is far different from that of even a few years ago. If you will examine the literature of recent years, you will find that in the transactions of the American Electro-Chemical Society there are only two papers published upon brass plating. One of them is a resume of all the formulae of solutions that had been used up to that date, that had just merely been collected from the then existing literature. That paper was published, I think, in about 1913 or 1914. Later, some work was done by Sturtevant and Ferguson at the University of Michigan. Unfortunately, those investigators did not comprehend the value of using a real brass solution, or brass anodes. It is well known to you who are platers that a solution of a composition that would be acceptable for good copper plating would not be good for zinc plating, and one that is good for zinc plating, cyanide zinc plating, would not be good for cyanide copper plating. And these authors used anodes of brass and zinc, separate anodes, separate metals— not an alloy. And they had slimes formed on their anodes, and they brushed the slimes off and took the difference in weight of the anode as anode efficiency. Inasmuch as brass or copper do not corrode well in a solution of the composition which they gave, naturally there were excessive slimes, and that gave results which are not commercially applicable.

Recently, a paper has been published on the control of brass solutions. That paper is strictly rule of thumb. It is recommended that stock solutions be maintained and they be added from time to time, and the additions made once or twice a day. It is not believed that that is good practice, and we will try to show the reason why it is not believed so. It is also recommended that anodes of 66% copper and 34% zinc can be used. Neither do we think that is good practice, and we will try to show the reason why. Neither do we believe that it is good to add such things as sodium hydroxide or things that will tend to bring up the zinc in the solution, especially if a solution may have a lot of zinc present.

In brass plating, the prime object is the color of the deposit. You may require one color, while the fellow working in the shop next to you requires another color. If that is true, then the conditions that exist in your plant and the control that you do in your plant should be different from the other fellow’s. A good way to ascertain how one should proceed would be to have an analysis made of the deposit of the required color. If you analyze the deposit and it has a color similar to fabricated brass, you will find, if that is a fabricated brass of 65% copper and 35% zinc, commonly called high brass, that it will have a range of from 76% to 84% copper; taking a mean of those two percentages, it is reasonable to assume that the deposit would have approximately 80 copper. And that is true from an experience of a number of years. It has been found that a deposit, electro-deposited brass that has a composition of approximately 80% copper and 20% zinc has, upon finishing, the same color as a fabricated brass of 65% copper and 35% zinc. That is due to what the metallurgists would call “orientation”—reflection of light. The fabricated brass has a close, dense, grain structure. It has been rolled. The electro-deposited brass is crystal. It has long been a question as to whether the deposit of brass is an alloy or a mixture of copper and zinc crystals.

In this investigation about which we are to talk, examinations were made of electro-deposited brass with a microscope from 100 to 2500 diameters. The results were disappointing. We could not tell whether it was an alloy or a mixture of the copper zinc crystals. X-ray examination, however, proved conclusively that electro-deposited brass is a true alloy.

Through the courtesy of the Engineering Department of the Bell Telephone Laboratories, several examinations were made, and they predicted within one-half of one per cent the exact amount of copper and zinc present. One alloy was given to them that they predicted had 90% copper and 10% zinc. They found 90.5% of copper and 9.5% of zinc, and chemical analysis showed it to be 90/10. The same was true of a higher zinc alloy, and they predicted again within one-half of one per cent from an x-ray analysis, and stated without any reservations that copper and zinc, electro-deposited, formed an alloy.

In this investigation, which was carried on for a number of months, 9 liter solutions were used, and the anodes and the cathodes were four inches square, giving one-ninth of a square foot on either side. The cathodes were steel. These were prepared by cleaning in the regular manner and then dipping into a soap solution, then into cold water, and drying in sawdust. This gave a soap film on the steel which prevented the deposit from adhering, so that it was possible to break the deposit and strip it. This (showing piece of brass) is a part of a strip deposit. You see there is a piece cut out there for examination. So you could strip it right off, and then examine it either from a cross section, or you could cut a piece out and make an analysis of it. So that gave us a good way of examining the brass, rather than just from the iron or else dissolving some of the iron with the brass when analyzing and then having trouble removing the iron.

The anodes were of three alloys, 65% copper, 35% zinc; 75% copper, 25% zinc; and 80% copper and 20% zinc. The pictures we will show you will be only of the 65/35 and the 80/20 mixture. The anodes were rolled and cast. All anodes were used as rolled or cast, and were annealed at a temperature of 350° C., 500° C., and 700° C., and for a sufficient length of time so that the soaking period was long enough to obtain a free corrosion of the crystal structure It was our wish to determine as much as possible the effect of grain size on the corrosion of the anode, which materially affected anode efficiency. The current employed was that at which we could get a flexible deposit, and that was 2.7 amperes per square foot which was identical with the current density decided upon by Sturtevant and Ferguson. It was not possible, as has been stated, in our opinion, to use 5 amperes per square foot upon a brass and get a deposit that would be flexible in commercial practice. You may do it in a big or a small job for once, but for day in and day out practice it has not been found advisable. 2.7 amperes per square foot is about what can be used.

One of the difficulties we ran up against in this work was the estimation of free cyanide. We got results that were quite at variance for one solution. We would analyze it for free cyanide, say now, and a couple of hours later analyze the same sample and get a different result. We couldn’t get any consistent results from the analysis of the free cyanide content of a brass solution with N/10 Silver Nitrate. Clennell, with his work on cyanides in metals or cyanides, makes a statement-in his volume published in 1900 that zinc interferes with the estimation of free cyanide in a solution which contains zinc and copper. This was found to be incorrect. We found that carbonates was the interfering agent, and that according to the amount of carbonates present in the solution, so the error varied. We then set upon a method which gave consistent results all through the run. These runs were made every forty-eight hours, extending over nine months, and 6 analyses were made every forty-eight hours, so that our cyanide determinations were quite numerous. We took 10 c. c. samples of the solution, put them in beakers, diluted them with distilled water. Then we estimated the carbonate content using barium nitrate. The nitrate was used because we were subsequently intending to use nitrate of silver and there would be no interfering salts added. We found the amount of carbonates in the solution, and found the amount of barium nitrate that was necessary to precipitate those carbonates, then we added that quantity of barium nitrate to the other sample, filtered and washed with distilled water, and then titrated it. We could duplicate results, and would get within an experimental error of the third decimal point. We used that all through, and we found that that was a reliable method and that zinc did not interfere as Clennell had stated. We tried copper sulphate and tried copper nitrate. Neither of those gave us any results.

In any investigation, it is necessary to have as many fixed conditions as possible; the variables should be as few as possible. We did not attempt to find out what was the best solution. We accepted a solution that had been giving good results in commercial practice. Therefore, we set the current density, kept that constant’ we kept the free cyanide constant, kept the temperature constant’ running both cold and warm solutions, cold solutions at room temperature, which was approximately 75 degrees F., and warm solutions running at 100 degrees F., this being controlled with a thermostat. The distance between the electrodes and the size of the electrodes were kept constant. The metal concentration of the solutions to commence was 80% copper and 20% zinc. As a study of the anode was being made, and not directly of the composition-of the cathode, no metal was added to the solution during the entire run, which was for about 846 hours. The reason for this was to find what composition of anode would give the best results so as to have the least change in a solution in commercial practice. Had we added metal so as to keep the metal concentration constant and keep it in the same ratio of 80/20, then we would not have had a true value of the effect of different alloys of copper and zinc used for anodes. The variables were the composition of the anodes. As I told you, they were 65/35 and 80/20, which will be shown to you. The carbonate concentration was changed. In running a large solution, it was found that carbonates built up in the solution quite rapidly. In a solution run of 23 gallons, it was found that in nine months the carbonate content went from two ounces per gallon to thirty-two ounces per gallon, just from decomposition of cyanide, and that was on a full automatic, where the changes were less than what it would be in a still tank. So additions of carbonates- were made at the rate of two ounces per gallon—from two ounces to thirty-two ounces per gallon. The data recorded was the anode efficiency, and in determining the anode efficiency, we took the anode and scrubbed it before removing it from the solution, in the solution, so as to have whatever products formed at the anode go back in the solution as much as possible. That is not a correct way of obtaining anode efficiency, but it was commercial practice, because you leave an anode in the solution during its life and the products that are formed generally go into the solution in the long run, which we found that it did, except in the case of very high zinc anodes.

We then took cathode efficiency. We had to find a method of obtaining cathode efficiency because we had an alloy of zinc and copper. We assumed, for no reason whatever, that it would be well to analyze the alloys in the cathode and then take the electrochemical equivalent of the copper and of the zinc, in the same ratio as that found in the analysis of the cathode, and call that cathode efficiency. The metal ratio in the solution was obtained and recorded, and the cathode composition. The results of these were put in glass, and I would like now to show you a few pictures of the results of the work.

Slide No. 1. This first slide is anode efficiency. The composition of the anode was 65% copper and 35% zinc. And you will notice the curves “as cast” in a weak solution. I probably forgot to mention we used two strengths of solutions; one, what we designed as weak, and the other as strong. We took these arbitrarily, for no reason whatever, except that we wanted some definite basis of metal concentration. We took 12 ounces of metal, total metal, 80/20 ratio, called it weak solution; and we took 3 ounces of metal and called it strong solution. Room temperature was cold and 100 degrees Fahrenheit was warm.

You will notice that the anode efficiency is quite regular. This was due to the polarization of the anode, to the zinc forming on the anode and you will notice that at places it was continuously decreasing on account of the formation of zinc, and there was a large amount of sludge.

Slide No. 2. Here is the anode efficiency of a rolled anode. The first one was cast. Up until we get to ten ounces of sodium carbonate (you see this runs from two to three ounces of carbonates per gallon)—as we get up to 10 ounces, our efficiencies are very erratic. Then later, we get them more uniform and they are higher than with the cast anode.

Slide No. 3. Here is cathode efficiency, with a 65/35 alloy, and annealed at 500 and 700 degrees, in a strong and weak solution. You will notice that the strong solution, with 500 degrees, seemingly after we passed a certain carbonate concentration, gave the best results, but even there, due to the zinc content, the cathode efficiencies were very irregular.

Slide No. 4. Here we come to cathode efficiency from the rolled anodes, and you will notice they are more regular, but they have an effect on the higher carbonates. Again, the 500 degree annealed, strong solution, warm solution, gives in this case the best results

Slide No. 5. Here is metal ratio percentage in solution. When you wish to obtain an 80/20 cathode, it is advisable to use an 80/20 anode and maintain the metal ratio in solution as dose as possible to 80/20. If you wish a different alloy, if you wish a different color, if you wish a color that would approximate 75/25, then, of course, you must change your ratio. And this talk is only from 80/20, which we found necessary in this builders’ hardware, to match fabricated brass. You will notice the metal ratio in solution (and this is a rolled anode)—that it kept quite constant, but it got down here (pointing) so it was quite low on the zinc side. It went very high in the beginning.

Slide No. 6. Here is cathode composition from a 65/35 anode, and this was rolled. The cathode composition started up high, and then fell, and we got as high as around between 35 and 45% zinc in the final run. That means this: It is not possible to control the carbonate content of the solution; and as your solution grows old and the carbonate content increases, so the composition of your cathode will be changed.

Slide No. 7. Here is the anode efficiency of an 80/20 anode. What I showed you before was 65/35. Here is one of 80/20. You will notice, if you can recall the other, that that is much more regular than with the 65/35.

Slide No. 8. The anode efficiency in the warm solution was better than in the cold solution, and just as soon as it received about 8 ounces of carbonates per gallon, then the efficiency was quite regular.

Slide No. 9., Here is the anode efficiency in an 80/20, as rolled, and you will notice there that the rolling, where it was hard rolled, and not annealed, that it had a very poor effect, but where it was annealed at 500 degrees, in a strong solution, you will notice how uniform it was, and therefore annealing must improve it. Where you have “as rolled” in a cold weak solution and in a strong solution, look how it came out (very zig-zag curve). And then, with the 5()0 degree, in the strong solution, why your anode efficiency was more regular. Here (Slide No. 10) is the anode efficiency of an 80/20, rolled, in a warm solution, and you see how much different that was from the cold solution. Again we have the 500° C. that is a straighter line.

Slide No. 11. Here we have the cathode efficiency of a cast anode, 80/20, in a warm solution. There is a marked difference between the farm solutions and the cold solutions, and the weak solutions and the strong solutions. There is a direct ratio between them and you can see that the warm strong solutions give you higher efficiencies and more regular efficiency than the cold solutions.

Slide No. 12. Here is cathode efficiency of rolled, used as rolled, and annealed at 500 and 700 degrees. And again you will see that the annealed at 500 in a strong solution gave seemingly the best results, while the others were affected by the carbonate content of the solution.

Slide No. 13. Here is metal ratio in solution. We have the strong solutions again and the weak solutions. That is the metal ratio in the solutions kept entirely by the anode. You will recall I stated in the beginning that no additions of metal were made during the entire run. And each of these runs was from two to thirty-two ounces of carbonates per gallon as added. No ammonia or any other salt was added to the solution, as we did not wish to have any interfering salts or anything that would bring the copper and zinc closer, because we were studying the anodes, and not the cathodes.

Slide No. 14. Here is the cathode composition of a warm solution, 80/20, and you will notice again that the 500 degree annealed, in the strong solution, stands out quite prominently.

Slide No. 15. I wish to compare the two principal points of a brass solution, or, in fact, of any solution where there is an alloy, and that is comparison of the cathode composition and the solution composition. Rolled anodes, 65/35; annealed at 500 degrees; in a strong solution. It started out with 80/20 solution and 65/35 anode. It wasn’t long before the solution had the same composition as the anode. But note the cathode composition. It constantly increased in zinc as the carbonates increased, and that is true of commercial operations. If you use an anode that has high zinc, as your solution gets older it is going to be more difficult to obtain the required color. If we wish to bring this back to a composition so that we would get the required color, it would be necessary to add copper. If you add copper, then you must add it at a ratio of 4 to 1, because you have an 80/20 ratio and therefore must add four times as much copper as zinc. If that had been done anywhere along the line here, then the metal concentration of the solution would have increased to such a point that it would almost have been unmanageable and you probably would have to go beyond a point where the solution would be usable at all. Or else you would have to remove a part of the solution, then add copper and cyanide to bring it back to the normal point. Now let us see what will happen with an 80/20 anode.

Slide No. 16. Here we have a cathode composition that will run up to a trifle above 90%. The solution composition will run down to about 75/25. I told you that no addition agent was added. It has been found in brass plating that it is advisable from time to time to add small quantities of ammonia, not as a regular addition. It seems to bring the copper and zinc into a closer relation, and then you will deposit at that color for a long time. In operating a 2300 gallon solution, sometimes for twenty-four hours a day for five days a week or six days a week, we found it was only necessary to add ammonia once a week, probably Monday morning. Then as the solution was kept working it was not necessary to make any other addition. When we added sodium hydroxide, we brought our zinc up to such a relation that we depleted our solution very rapidly of zinc, and we got an off color. We could not maintain a uniformity of color. Our anodes polarized because we were releasing out the zinc more readily than the copper, and when our anodes polarized, then the needle of our ammeter was constantly going down. By maintaining nothing else but cyanide in the solution, we were able to keep clean anodes, not take the anodes out for their entire life for cleaning—just to clean the hooks and the bus bars. If we had added ammonia to this solution, we would have brought more zinc into the cathode deposit. That would have brought that zinc up, and would have brought that copper down, and we would have approximated then the 80/20 cathode which we desired to obtain; It is much more easy to add zinc to a solution than it is to take it out.

One of the members of the Society told me yesterday that in operating a brass solution, last month they had deposited brass upon two and one-quarter million pounds of skid chains. During that time they wanted to maintain an 80/20 ratio in their deposit. They added 500 pounds of metal, cyanide and metal, mostly zinc. It was the opinion that maybe he should change his composition. It would be detrimental to him to change his composition because it is so much more easy to add small quantities of zinc than it is to get an excess of zinc and then have to add a ratio of 4 to 1 of copper, because zinc you only have to add a ratio of 1 to 4 and it is so much easier to add zinc than it is to take it out.

So that you see there, with a strong solution, run warm, having an anode of 80/20, and that anode has been annealed at 500° C. so as to get a uniform crystal structure, and get uniform corrosion, that you obtain a better and more uniform deposit.

Slide No. 17. I have here a very interesting chart, and that is the operation of a 2300 gallon solution over a period of one year. An analysis was made each week, and here you see the time in weeks, from 1 to 52 weeks. At the bottom here, it gives you the solution content in ounces per gallon. The straight line here is the copper content. See how closely we kept it between 12 and 2 ounces. The free cyanide content is this broken line, which followed very closely the copper content. The bottom line is the zinc content, and look how constant that was kept. That was without the additions; the analysis was made before any additions were made. When the analysis was made, the chemist then calculated the amounts of copper cyanide, zinc cyanide and sodium cyanide that should be added to the solution. This was a 65/35 anode, and we were forced to maintain the required color. When the laboratory made their additions, they were faithfully followed, but we also found that it was necessary to make other additions so as to maintain our color and to maintain the required number of amperes that were flowing, because we watched our ammeter very closely, and we wanted a certain number of amperes flowing through our solution constantly so as to have as near as possible the same weight of metal.

Here (at top side) you will see the additions of sodium cyanide in pounds, and this line here will be the laboratory report, and up here will be the actual additions. In all probability the difference between them was due to the then known method of analysis of free cyanide. Had we at that time had available the method I outlined to you a little while ago by precipitating with barium nitrate and titrating with D/10 silver nitrate solution, then probably our curves would have been closer. As I stated, in all probability the difference is due in a large measure to the trouble experienced in titration of free cyanide.

In the addition of zinc you will notice the laboratory and the actual additions followed very closely, and the same thing with the additions of copper cyanide, that the laboratory and the actual operations followed very closely. We also kept a record of the carbonate construction. Here (pointing to line) you see where the tank was repaired, the solution removed, and we had then a trifle over ten ounces of carbonate per gallon. And this, as you will see, was at a time when it wasn’t winter; it was summer, and we couldn’t remove the carbonates. We know of no method to remove the carbonates economically except by lowering the temperature of the solution and freezing out the carbonates. The carbonates constantly built up until you will see here that it attained 32 pounds per gallon. Then we reduced the solution to 26° F., took it out and put it in a tank that was available for that purpose, and we went down to about 13 ounces per gallon. Had re then taken that solution and added water to it, and then recrystallized it, we would have lowered it more. One crystallization is not enough because there is a certain amount of water that is necessary to combine with the sodium carbonate because you crystallize it out as sal soda crystals, with the ten molecules of water. So that you would not have enough free water to combine with your crystals or with your carbonate of soda, so as to crystallize it. But if you will then remove those crystals, and then add water and re-crystallize, then you can bring it down, and we lave done it, brought it down as low as three ounces per gallon.

Now that shows you a 65/35 anode. Identically right alongside of this was another 2300 gallon tank in which we operated an 80/20 anode, and with the same work for the same period.

Slide No. 18. Here you see again the zinc content, the copper content and the free cyanide (at bottom of slide). There (at top) you will see the sodium cyanide additions, the laboratory report of additions; and here, the zinc cyanide, and here the copper cyanide, and you will notice how closely the copper cyanide actual additions and the laboratory reports agreed. So it is an indication that laboratory control, and not by the additions of a stock solution, is the only way in which to control a brass solution. Had we added a stock solution, we would have overbalanced our solution with zinc, and we would have had trouble, and I will show that in the next slide.

Slide No. 19. We kept a record for six months of the amount that was added to these 23(0 gallon solutions. No. 1 was 65/35 and the other was 80/20. We added copper cyanide, 1,000 pounds to No. 1, 375 to No. 2, making a minus in favor of the 80/20 of 625 pounds. Since, we added 328 pounds to No. 1, 365 pounds to No.2, in favor of the higher zinc of 37 pounds. That would be expected in a lower zinc solution. We got considerable sludge through from the 65/35 which we do not get from the 80/20. The sodium cyanide, the free cyanide content, was 1995 against 1545, a minus in favor of the 80/20 of 450.

Assuming just arbitrarily—not chemically, but arbitrarily—that it takes one pound of sodium cyanide to put one pound of zinc cyanide or copper cyanide in the solution, then we could calculate from that the amount of combined cyanide or the total cyanide used in the entire time. Here we have, for metal salts of cyanide, a total of 3323 pounds, that is, for the salt and the free cyanide for the No. 1. For No. 2, 2285 pounds, making a difference of 1038 pounds of cyanide less that was used in the 80/20 solution to maintain it at a constant, than used in the solution having 65/35. Both solutions run identically the same on the same class of work, full automatics, so that there was no danger of the human element entering into it at all, and the same class of work at the same current density was run. Now you can see from here, had we added a stock solution to this (No. 2) we would have gotten into trouble on the 80/20. Had we added it here (No. 1) the (65/35) we would have also gotten into trouble. So a stock solution is not advisable.

It also indicates clearly that it is much more economical, along with the advantages that I showed you in cathode composition, it is more economical to use an 80/20 anode because there is a saving of 1038 pounds of cyanide in a six months’ run in such a solution. It also indicates that if a solution is run by chemical analysis, the amount of salts that is added can be very small; that if you will keep your anodes clean and you get good anode efficiencies that it is better to run a solution in that way than it is by the addition of salts. Here, in six months’ time, 1,000 pounds of copper cyanide and 328 pounds of zinc cyanide in the 65/35; and here, in the 80/20, 375 pounds of copper cyanide and 365 pounds of zinc cyanide was run. The class of work was builders’ hardware. We were running through door-knobs and escutcheons. We ran through 2500 door-knobs every seventeen minutes; 2500 escutcheons every seventeen minutes. And that had to have enough deposit on it for brush finish. That was 20 pieces to a rack. If we were doing window fasteners that had 60 pieces to a rack, then we were running 7500 pieces every seventeen minutes.

Mr. Oberender ran this tank, or one of the tanks, because there was only one installed at this time there, and he can confirm my statements.

Slide No. 20. One of the things that is very interesting, and I thought it might interest this group of men, was stratification of solutions. We found that a large percentage of trouble in nickel solutions is due to stratification, that you take a sample at the top of the solution, or you have a solution that becomes stratified, and you get a heavy deposit at the bottom, and you get pitting at the bottom, and making an analysis of the solution and taking the pH from the top of the solution and then attribute it to the whole solution. Look what happens. Here was a nickel solution. Here was the top, 1050, and the bottom, when that solution stood still, idle, for 48 hours (and this was again about 2300 gallons) the specific gravity changed to 1140. The nickel content was 1.55 to 6.91. The solution had been previously running, but stood idle 48 hours. The atnmonium chloride was 1.12 to 3.20. The pH colorimetric was 6.8 to 4.4. The solution was worked five hours after stirring, and the results were 4.72 to 4.44; 2.13 to 2.13; 5. to 5.7.

So you see if you work a solution, keep working it and drawing work out, then it will keep a better, even concentration, but if the solution stands, or if you are doing shallow work and not stirring the solution well by drawing out your work, then you will have stratification. In brass solutions, such a thing does not occur, and therefore it is not necessary to pay any attention to it.

Slide No. 21. Here you have a brass solution, 2.13 to 2.22; .62 to .57 for zinc cyanide. 1.38 to 1.5 for a still tank, 2.13 to 2.18; .60 to .65; 1.69 to 1.84. So you have no trouble whatever with stratification of brass solutions, but you do with nickel solutions, and that should be taken into consideration in determining difficulties being experienced due to poor deposits or poor anode corrosion.

Slide No. 22. I thought probably it would be interesting to show you a slide of the operation of a cyanide copper solution of 2300 gallons over a year’s time, the same as I did with the brass. Here, you have the free cyanide content, here you have the copper content, here you have the free cyanide additions, and here you have the copper additions (pointing to different portions of slide). Here you have the building up of the carbonates. Here, the solution was frozen and then allowed to build up again, and then it was frozen again later on, and so we brought it down to about eleven ounces per gallon.

I thank you. (Applause. )

I failed to mention that one of the effects of carbonates in the solution is upon the character of the deposit. And also, the difference between cast and rolled anodes on the character of the deposit. In all cases, when using a cast anode, your trouble with rough deposits and with the rolled anode, you obtained smooth deposits.

We found also, with the low carbonate, you could get a deposit that was very flexible, even quite heavy, and it would be very flexible. It was a little rough but a thinner deposit was absolutely smooth, and very flexible. But when you had high carbonates in the solution, run identically under the same conditions it was like this (demonstrates deposit was very brittle). So you get your same ratio,- you get your same cathode efficiency, but the character of your deposit is materially changed by the carbonate concentration, and this is more so true when you have sodium hydroxide in the solution; it materially decreases the time in which you will obtain flexible deposits, and is apt to make the deposit very brittle.

Here (showing piece of brass) is a piece from a barrel which nay interest some of you. It formed on a rod that extended across the barrel and it was in there a whole year, and it is just merely rings of brass, and you see how smooth it is. This was done by our good friend Bill Stratton or Sheldon in a plant where he does a great deal of brass plating in barrels and you will see that and see the color. (Applause.)

CHAIRMAN SMITH: Gentlemen, you have heard the reading of Mr. Hogaboom’s paper. Are there any questions you would like to ask him at this time?

QUESTION: May I ask Mr. Hogaboom how much ammonia he adds?

MR. HOGABOOM: To 2300 gallons, we added one gallon; never over two.

QUESTION: I would like to know if a copper anode would be just as efficient or easy to control?

MR. HOGABOOM: Can a copper anode be used in a brass solution? No, because copper anodes do not come down in a brass solution equally as good as brass anodes. And it is very difficult to maintain a color by the constant addition of zinc salts. It is better to get your zinc salts or your zinc from your anodes, rather than keep adding it all the time.

MR. ZADOWSKI: I would like to ask Mr. Hogaboom if he freezes out his carbonates.

MR. HOGABOOM: Yes.

MR. ZADOWSKI: Do you take your copper out at the same time.7—the crystals ?

MR. HOGABOOM: No, in freezing out the carbonates, we found that we took out a total of less than one tenth of one per cent of the metal content of the solution. The carbonates were taken out of a 2300 gallon solution—we would take out, say in the Christmas shut-down, we would take out about two ton of carbonates from a 2300 gallon solution. We would drain them, just drain off the solution, and we would use them in the rolling room for cleaning work, and the firm never had to buy any soda ash for rolling. We always used from the three solutions. We got approximately ten tons of sodium carbonates from our solution in a year’s time.

MR. OSCAR SERVIS: Would it be possible to eliminate your carbonates or hold the carbonates down by an addition of barium nitrate to the solution or would electrolysis interfere?

MR. HOGABOOM: By the addition of barium nitrate and adding to the carbonates, it would require filtration and then you would be building up sodium nitrate in your solution, which would be bad news, because Dr. Watts has shown very clearly that nitrates are not beneficial additions to cyanide solutions.

The amount of removal, when it can be removed, would depend entirely upon the locality in which you live. If you live in Los Angeles, you will probably have trouble with freezing out the carbonates; but it is my opinion that eventually it will be found economical to install some refrigerating process.

---

PITTING OF NICKEL SOLUTIONS

F. P. Romanoff
Read at 18th Annual Meeting of Chicago Branch

The deposition of nickel for decorative and art purposes is often accompanied by that defect known as “pitted nickel.” This troublesome phenomenon has been bothering electroplaters since nickel was first deposited. But until a few years ago, the pitted surface was not as common nor as serious as at the present. With the inauguration of high pressure production methods, involving the use of high current densities and heavy deposits, from high metal content solutions, the real battle against the pitted surface was started.

Today, the more common causes of pitting are easily controlled, but the pits are still with us, spasmodically, if not chronically. The effect of the surface of the metal to be plated, we know, may cause pitting, especially if irregular in composition. Sheet iron of certain grades has been the cause of much grief due to the fact that although to the average eye no flaws are evident, still, the microscope, or even very close examination discloses irregularities which are subsequently much magnified when plated. Often a sheet of steel will contain pits which are not detected in the polished state. These pits and other irregularities, will serve as nuclei for reducing the over voltage of hydrogen. The result is the formation of bubbles of gas at these points, which cling sufficiently long to cause pitted surfaces. It has been found that copper plating before nickel plating, can exert an important influence on the quality of the nickel, with regard to porosity. The defect is usually decreased somewhat, although complete elimination of porosity may not be attained. This action of reducing pitting may be ascribed to the over voltage of hydrogen which is less at copper surfaces than at steel. In other words, the hydrogen discharge takes place more easily with copper than iron and steel, resulting in a more uniform nickel coating.

Although a light copper plate of even five minutes helps; an hour plate will not do away with pore formation entirely.
We can eliminate, however, to a great extent the effect of the surface and its irregularities, by applying a heavy deposit of soft copper first, and then buffing- this deposit sufficiently to close or cover most of the irregularities. Of course the copper must be non-porous, or the difficulty may be further exaggerated.

Eliminating the effect of the surface of the material we often find that pitting persists apparently due to either manipulation of the nickel solution, its composition, or some external cause. In searching for the particular reason, we may check up on our current density. It is known that excessive current densities will cause pitted as well as burnt deposits from nickel baths. The current density may be correct, and a check is then made on the acidity on pH value and nickel content of the solution.

Now, in listing the known causes of pitting of a nickel solution, we might tabulate them as follows:

1. High current density.
2. High acid concentration.
3. Low acid concentration.
4. Low nickel concentration.
5. Low temperature.
6. Suspended particles. (Organic, inorganic or gases.)
7. Dissolved impurities.

High current densities are more or less readily determined, and corrected. The influence of the acid concentration is rather marked. It is found, however, that while more hydrogen is generated, and less malleable metal is made in an-acid bath, the condition is seldom the primary cause of pits, but will multiply them if the primary cause is present. Small suspended particles cause pits, but the types most troublesome are not so caused. Excess air will cause pitting, but is not a primary cause, unless it exists in a certain condition which will be explained further on in paper.

In the case of suspended particles, the usual procedure is the filtering of the solution. Most of the visible particles in the bath may thus be eliminated. In the case of dissolved impurities we would require the aid of a chemist, or the experienced plater can often tell from past experience and apparent intuition just what impurities are present, and take the proper steps to eliminate them.

Today, it has been found that although the proper current density, acid concentration or pH value, nickel content, and temperature are maintained, pitting will often persist in apparently clean solutions, on excellent surfaces. To remedy this, additions of various ingredients have been made which have more or less successfully eliminated the condition at least for a period of time.

The additions recommended by platers include the following:

Common salt 5 to 15 pounds per 100 gallons
Sal ammoniac 5 to 20 pounds per 100 gallons
Nickel chloride 2 to 10 pounds per 100 gallons
Sodium perborate 5 to 20 ounces per 100 gallons
Potassium permanganate 2 to 8 grams per 100 gallons
Hydrogen peroxide 2 to 12 ounces per 100 gallons

It will be noted that all of these but the last are electrolytes. Now each of these materials has been advocated, and each has its strong adherents. The question arises as to why the addition of these electrolytes causes pitting to disappear.

Two theories are available which will explain the specific action of the various additions or treatment. The first takes into account that the addition of electrolytes and boiling will reduce the solubility of gases in a solution.

The second theory takes into account that the addition of electrolytes and high temperatures or boiling tends to precipitate colloids.

In the case of the dissolved gases there is no particular reason for the gas traveling and settling on the cathode, unless the gas molecules are electrically charged. Under certain conditions this is just what happens. When the dissolved gas molecules are electrically charged, they behave the same way as do colloids. If we assume our pitting due to colloids or to gases electrically charged, a number of phenomena are explained.

The study of colloids dates back to the middle of the 19th century when attention was called to the diffusive properties of those substances which can be readily crystallized from water, such as inorganic salts in general, and those substances which cannot be obtained in the crystalline form or only with great difficulty. Of this latter class, gums and gelatines are typical examples and are referred to as colloids. These substances also have the property of passing through the ordinary filters. The application of the word colloid does not infer a definite class of materials, as used originally, but infers a particular physical condition, in which form it is now possible to obtain many substances. A colloidal solution is now understood to indicate, not necessarily a solution of glue or gelatine, but any solution which possesses specific properties comparable with those of a gelatine solution. Colloidal solutions of iron, sulphur, metals, and metallic salts are more or less readily prepared.

Colloidal solutions may be considered, for our purpose, to be really suspensions in a liquid of extremely minute particles, which are retained in suspension by the tenacity of the medium. Unless a specially equipped high powered microscope is utilized, the presence of colloids often cannot be suspected unless the color of the colloid is distinguished in the solution, as is the case of colloidal metallic gold.

If a current is passed through a colloidal solution by means of two electrodes, a migration of the colloid occurs, which may be similar to that of ions in a true solution. In this case, however, the mechanical movement is in one direction only, either to the anode or to the cathode, depending upon the colloid, and the solvent. Apparently, the colloids carry electric charges, those which migrate to the anode, carrying a negative charge, such as rubber latex, and those which migrate to the cathode, carrying a positive charge. Iron hydroxide is one of the latter type.
Colloids are usually divided into two classes:

1. “Suspension” colloids. These give non-viscous solutions and are precipitated by small quantities of electrolytes, such as salts and acids. They can be restored to the colloidal form only by indirect or roundabout means. These are known as “Irreversible Colloids.”
2. “Emulsoid” colloids. These give viscous solutions and are not so readily precipitated by electrolytes. They can be brought into colloidal solution again by simple contact with water. These are known as “Reversible Colloids.”

The latter class may give the plater his greatest trouble, if present, but fortunately, it is rare that impurities of this nature find their way into the nickel bath. The only simple remedy suggesting itself might be the operation of the bath at high current densities over a long period. In copper refineries, colloids, such as glue or similar substances are added to the acid copper sulphate vats to give a more desirable crystal structure. Analysis of the copper deposit or cathode, shows that the glue is also deposited in very small quantities between the crystal layers. Infusorial earths and similar materials, are often used to filter solutions containing colloids, as they have the property of absorbing colloidal substances. The first type or “suspension” colloid, is of common occurrence in nickel plating solutions. If low current densities are used, such as were common a few years ago, when double salt solutions were entirely employed, troubles due to the presence of these impurities are rare. Today, with the use of high current densities, this difficulty is rather common, although not often recognized. It has often been found that the addition of a small quantity of nickel salts has corrected the difficulty, and pitting was then attributed to a deficiency of metal in the solution. However a test would show that the nickel content was satisfactory, and often the amount of single or double salts added was insufficient to raise the metal content even 1/10th ounce per gallon.

Now, in removing colloids from solution, precipitation may be effected by low electric charges on very small quantities of electrolytes. Boiling also tends to coagulate colloids and allows their removal by filtration. In order to consider the precipitation of colloids further, we must take note that although the concentration of electrolyte required to precipitate a colloid may be small, it has a definite value below which precipitation is not effected. When we add a certain quantity of sodium or ammonium chloride which will have some result in reducing pitting, it is found by trial that the amount required is more than that of nickel chloride. More of the latter is required, however, than sodium perborate, and more of this than potassium permanganate.

In considering the type of colloid which would cause pitting in a nickel bath, it is obvious that the colloid which travels to and is deposited on the anode will not affect the work, since it does not come in contact with it. The colloid which is attracted to the positive pole is commonly referred to as “electro-negative.” The colloid which is attracted to the cathode is likewise called “electropositive.” This latter type of colloid is that which results in pitting. It migrates to the work and is deposited thereon. Iron in nickel solutions under certain conditions causes pitting due to the formation of colloidal iron hydroxide, which was mentioned before as electro-positive.

An electro-positive colloid of this type will be precipitated by the electro-negative ion, or acid radical of a salt in solution. We have noted before that the concentration of the electrolyte necessary to precipitate a colloid is very small, but it has a minimum value below which precipitation is not effected. This minimum value is of the same order for all electrolytes in which the precipitating radicals have the same valencey. Thus, the minimum concentration of monovalent acid radicals required to precipitate electro-positive colloids, will be identical irrespective of whether the radical is chloride, nitrate or bromide. Moreover, the higher the valence of the precipitating ion, the lower its minimum concentration. Therefore, the requisite quantity of a divalent ion to overcome pitting is less than that of a monovalent ion, but greater than that of a tri-valent ion.

This outline of some colloidal properties will suffice as a preface to the use of electrolytes to eliminate pitting in nickel solutions.

In adding monovalent ions, platers have practically limited themselves to the addition of chlorides, since this is usually one of the ingredients of- the bath. Sulphates such as magnesium and sodium sulphates have usually comprised the divalent additions outside of nickel salts. These additions are not made to improve conductance, for in the cases we are considering the conductance is satisfactory, and pitting exists.

When perborates or permanganates are added, their oxidizing reactions on organic matter, iron in the ferrous form, or other substances results in the producing other negative ions, due to the introduction of acid with the perborate, or bivalent manganate ions. The negative ions then serve to neutralize the charge on the colloids present or formed by the reactions.

In the case of the addition of chlorides, however, the most to be expected is the neutralization and precipitation of colloids by the negative chloride ions.

To explain the formation of colloids we will use iron in the solution as an example. Iron probably enters the solution as a ferrous salt. Some of the iron is always being oxidized to the ferric form due to the oxygen in the air. At the usual pH of nickel solutions, ferric hydroxide is formed. The formation is influenced by conditions at the time and may form colloidal hydroxide. As a colloid in the solution it will remain suspended, and if the ordinary filtering machine is used, will pass through back into the tank. Its presence cannot be detected by the ordinary visual means. The passage of current through the solution causes the iron hydroxide colloid to travel toward the cathode or work, and be precipitated thereon, causing the phenomenon of pitting. The pH of the solution, metal content, temperature, and current density may be correct. The addition of a chloride at this time precipitates the existing colloids, but affords only temporary relief. The reason for this seems to be that the chloride may have precipitated all the-colloid present. New colloidal formation, however, may continue to proceed, due to more ferrous iron being oxidized to the ferric form. This, therefore, persists in causing the formation of pits according to the mechanism outlined.

The addition of hydrogen peroxide results in depolarization of hydrogen at the cathode as well as oxidation of all the existing ferrous iron to the ferric state, which forms ferric hydroxide. We will not consider cathodic-polarization here, however. The formation-may or may not permit iron hydroxide to precipitate as a colloid. In the presence of the salts present in the nickel bath the tendency would probably be to precipitate any colloids that are formed. However, we know that peroxide of hydrogen is not always as satisfactory as might be in eliminating pitting of the solution. Under certain conditions, when polarization may not exist, the iron might be oxidized and form colloidal hydroxide and not precipitate.

The addition of sodium perborate presents another situation. If polarization by the hydrogen at the cathode were the sole cause of pitting, hydrogen peroxide alone should overcome our pitting. Considering colloids, however, we have the advantage of adding an oxidizing agent as well as an electrolyte. The perborate releases free hydrogen peroxide, which serves to oxidize iron to the ferric state; then the electrolytic action of the negative sulphate or chloride ion, depending upon the acid used for neutralization, serves to depolarize any colloidal iron which has been formed, and causes its precipitation. The addition of acid to correct the pH of the perborate furnishes the negative ion in the precipitation of colloids. The affect of adding perborate is usually more lasting and apparently more successful than that obtained by adding hydrogen peroxide, chlorides, or sulphates. We must understand that the pH of the solution is kept constant at all times.

The addition of potassium permanganate also serves to oxidize iron and precipitate any colloids formed as well as depolarize the gases formed at the cathode surface. The introduction of the manganese ion, however may in time lead to complications. It is found that upon addition-of permanganate, very good results are effected by extremely small quantities, of the order of five grams per hundred gallons of solution. The permanganate is reduced to manganese dioxide. This is an excellent depolarizer, and probably would overcome any hydrogen polarization at the cathode. However, the formation of a gel suggests that when formed, it came down in a colloidal state. It is also probable that colloidal manganese dioxide has a negative charge, causing neutralization of the iron colloid, and precipitating both the latter and itself.

Colloid particles are much larger than ions, and it is found that their electrical charge is also much larger. It is also known that colloids of opposite electrical charge exert a mutual precipitating action. Therefore, if a manganese dioxide colloid is formed, its charge which is probably negative will be more effective in precipitating electro-positive colloids than are electrolytes ordinarily.

The oxidizing agents also serve to remove other substances, such as organic matter and foreign bodies which may cause pitting. They also function as depolarizers or neutralizers of hydrogen gas at the cathode. This latter action may be most prevalent, but does not quite explain the successful use of perborates where hydrogen peroxide fails.

Iron has been used as an example because it is one of the most common impurities in nickel solutions. It is introduced with the salts, anodes, work, and water. Other metals or impurities may function likewise. Nickel, copper, and zinc hydroxides, as well as other metallic compounds, under certain conditions also form colloids, as well as do organic substances.

Experiments to determine just how the addition of various electrolytes affect nickel solutions, with respect to precipitating action, showed the following results with a filtered solution. A blank was run along with all experiments.

The addition of chlorides gave a small flocculent precipitate. The addition of sulphates resulted practically a similar precipitate. The addition of hydrogen peroxide gave an appreciable increase in the precipitate. The addition of perborate, properly neutralized to the pH of the solution gave a still further increase of precipitate. The addition of permanganate resulted in a large flocculent precipitate of manganese dioxide which included iron hydroxide.

It is also known that gases can function as colloidal aggregates as well as solids. Therefore, the result of the addition of electrolytes reduces the solubility of gases by neutralizing the charges on the colloidal aggregates. The hydrogen gas adhering to the cathode has a charge, and therefore also functions as a colloid in the solution. These gases may be oxygen, hydrogen or air. Removing the work from the solution for a period of 5 to 20 seconds is also a known remedy for pitting. The action here probably depends upon allowing the gases to escape from the surface of the work when the restraining tension of the liquid is removed. This method would help reduce pitting especially if caused by the evolution of hydrogen on the surface of the work.

Heating to a high temperature or boiling are also used to overcome pitting. That this procedure is often successful is more or less well known, especially if carried out in lead lined tanks. However, it is also found, at times, that when the solution has been heated or boiled by means of steam in wooden tanks, pitting is worse than ever. The reason for this is due to the leaching out of the resinous oils or other matter from the wood at these temperatures. Some of these products form electro-positive colloids and also may deposit on the work.

From a practical point of view, the most adaptable plan to follow in overcoming pitting in a nickel solution would be, first, a checking up of the following factors:

1. Current density.
2. Temperature.
3. Acid concentration or pH.
4. Metal content.
5. Chloride content.
6. Fair amount of freedom from suspended dirt.

If all of these factors are correct insofar as regular practice is concerned, the addition of an electrolyte would be most logical. The addition of chlorides might be best if the chloride content is low. Addition of more nickel salts may also be made. In adding these salts, care must be taken to prevent further contamination of the solution. If temporary relief is obtained only, and pitting continues, the addition of perborate or permanganate may be well considered.

The addition of these should be carefully made. When perborate is added, the pH of the dissolved perborate should be reduced to that of the solution with acid. The quantity of permanganate added is extremely small, and the-effect of too great additions may result in complications. Excessive quantities of perborates can be made practically harmless by the addition of acids, thereby serving as a fairly safe addition for overcoming pitting from the sources considered.

References:

1. McNaughton. Trans. Faraday Soc., Feb. 1929.
2. Phil. Trans. 1861, V151, 183.
3. Zsigmondy. The Chemistry of Colloids. Translated by E. B. Spear 1917, Jon Wiley & Sons.

---

CONTAMINATION OF CHROMIUM PLATING SOLUTIONS WITH BUFFING COMPOSITIONS

By C. Kocour

The purpose of this investigation was to determine the effects of impurities that contaminated chromium plating solutions. Inasmuch as the subject would be too large to cover in a single investigation it was decided to determine only those effects due to contamination from buffing compositions introduced into the bath on the work to be plated. This contamination results from improperly cleaning the work, or not cleaning it at all, for there are many platers that believe that the chromium solution itself is a good cleaner.

It was decided to determine the chemical effects of this contamination rather than the effect on the working of the solution. As work has been published on the relations existing between the working of a solution and its chemical components it was not necessary to repeat it.

The following experiments were all performed with samples of a chromium solution having 393.6 g/l (52.4 oz/gal) of chromic acid. The effect of individual components of buffing compositions was noted, it being assumed that a buffing composition would cause the same- contamination as the sum of its components would.

Experiment 1—A 25 ml sample of the chromium solution was placed in a flask and .1 gram of stearic acid added. (This is equal to a concentration of 4 grams per liter (.533oz/gal) of stearic acid.) The solution was now boiled for three hours, a reflux condenser returning the condensate to the flask, thus keeping the volume constant. At the end of this period it was found that all the stearic acid was not decomposed. The solution was now analyzed for hexavalent chromium and found to contain 372.2 g/l (49.6 oz/gal) a decrease of 21.4 g/l (1.8 oz/gal). It is logical, therefore, to suppose that this amount of hexavalent chromium was reduced to the trivalent form.

Experiment 2—This was an exact duplicate of Experiment 1, except that .1 gram of tallow was used in place of stearic acid. Again three hours boiling in this strong oxidizing solution did not completely decompose the tallow. Analysis of the solution showed 371.6 g/l (49.5 oz/gal) of hexavalent chromium, a decrease of 22 g/l (1.9 oz/gal). The tallow left in the flask looked very much like the original sample that was introduced. No attempt was made to determine what percentage of the tallow had actually been decomposed.

Experiment 3—This was also a duplicate of experiment 1, except that .1 gram of paraffin wax was added. Analysis showed that 391.2 g/l (52.1 oz/gal) of hexavalent chromium were left, a decrease of 2.4 g/l (3 oz/gal)

Experiment 4—In this case petrolatum was used ad the experiment run in the usual way. Analysis showed that the hexavalent chromium was reduced 4.6 g/l (.6 oz/gal).

Experiment 5—The same experiment was performed using .1 gram of Vienna Lime. This reduced-the hexavalent chromium but 1.2 g/l (.08 oz/gal). This decrease could not be explained, except by experimental error for the precipitate that first formed completely dissolved giving a clear solution.

Experiment 6—This last experiment was performed, using .1 gram of flour tripoli and the hexavalent chromium was found to be reduced but .6 g/l (.04 oz/gal). This loss might also be traced to experimental error.

Summary

1. Saponifiable parts of buffing compositions, such as stearic acid and tallow, reduce the chromic acid to the trivalent form. The belief that chromic acid is a good chemical cleaner for buffing compound dirt is shown to be fallacious as two of the experiments showed that three hours of boiling failed to decompose as little as .1 gram of stearic acid or tallow, two of the widely used bases of buffing compositions.
2. Unsaponifiable parts of buffing compositions, such as paraffin and petrolatum, have but little effect on chromium solutions and are in turn little effected by the chromium solution. Reduction of the chromic acid is slow, even under the best conditions for chemical action.
3. If a chromium solution successfully cleans saponifiable or unsaponifiable matter from the work to be plated, it does so mechanically by the production of hydrogen or in some other manner. It certainly is not chemical action to more than a small extent.
4. Vienna Lime (basis of all white finish compounds) has no effect on the hexavalent chromium. The effect of the dissolved calcium and magnesium dichromates has not been studied.
5. Tripoli flour has no effect on the chromium solution, it no doubt remaining in suspension in an agitated solution. What effect this has on the working of a solution is not known.
6. Chromium solutions are contaminated in three ways from buffing compositions, first by reduction of some of the hexavalent chromium to the trivalent form and second, by some of the impurities going into solution, and third, by - some remaining in suspension.