Historical Articles

November, 1953 issue of Plating

 

---

Thickness of Electrodeposited Coatings by the Anodic Solution Method

---

Presented at the 40th Annual Convention of the American Electroplaters’ Society, June 17, 1953.

C.F. Waite, King-Seeley Corporation, Ann Arbor, Michigan

 

INTRODUCTION

With the advent of federal restrictions curtailing the use of nickel for decorative plating came the problem of substitute plating and the attendant thickness measurement problem. As decorative plating at the author’s company was entirely interior hardware, plating chromium directly over copper sufficed to solve the first problem. The thickness measurement problem came when a customer complained of insufficient chromium on plated parts. After experimentation, it was decided that the hydrochloric acid drop test was not dependable for the measurement of chromium plating thickness upon copper. This left an urgent need of an accurate method of measuring chromium thickness. In investigating possible methods, the modification of the anodic solution procedure of measuring plating thicknesses developed by Dr. H. F. Francis of the Armour Institute was discovered1. After demonstrating through repeated tests, the accuracy of this method with an experimental test model, it was found that one equipment supplier was manufacturing a production instrument (Electronic thickness tester, Kocour Company, Chicago, Ill.) which offered advantages in speed and versatility over the experimental model mentioned above. That instrument was purchased and has been used with excellent success in the routine measurement of the thickness of chromium plated upon copper. Its use has been extended to other plated coatings as well.

Difficulties with the conventional tests for chromium thickness will be described first in this report. Then details will be given on the application of the Francis modification of the anodic solution method for measurement of chromium over copper, and the later use of the commercial instrument for this purpose. Finally, there will be discussed the use of that instrument for other combinations of plated and base metals.

DISCUSSION

Hydrochloric Acid Drop Test
When the customer’s complaint of inadequate chromium thickness on some of the plated parts occurred shortly after the change to a chromium plate directly upon copper, determinations of the thickness of the chromium plate were being made by conventional methods. For the measurement of the thickness of chromium plated upon nickel, the author’s company had always used the hydrochloric acid test described in the literature2. When this method was applied to chromium plated over copper very erratic results were obtained. Various methods of cleaning the sample were tried, but still the results were not reproducible. In an effort to prove or disprove the validity of the method, a number of 3 x 6 inch panels were plated. Half of them were copper plated and then chromium plated, and the other half were copper plated, nickel plated and then chromium plated. They were all chromium plated at the same time with the same current density. To avoid edge effect, a 2 x 3 inch portion was machined from the center of each panel. The chromium thickness on several of these 2 x 3 inch panels was measured by weighing the panels before and after stripping the chromium with hydrochloric acid. A quantitative analysis for chromium of the strippings of each panel proved that the weight method was reliable and that only chromium was stripped by the hydrochloric acid. This method of weighing a sample before and after stripping the chromium from a measured area with hydrochloric acid has been used from that time as an umpire method of measuring the thickness of plated chromium. Trials were made using the hydrochloric acid drop test to measure the chromium thickness on the remaining 2 x 3 inch panels. The panels with a nickel layer gave readings that were reproducible and were within ten per cent of the results obtained by the loss in weight method. The panels that had chromium plated directly on copper did not give reproducible readings. The tests indicated chromium thicknesses from 8 to 35 millionths of an inch on immediately adjacent areas of the panels. Thus it was decided to abandon the hydrochloric acid test as a means of determining the thickness of chromium plated directly on copper and to search for a more accurate test.

Anodic Solution Method
An immediate inquiry was launched to discover a satisfactory test for measuring chromium thickness. Superficial investigation was enough to eliminate most standard test procedures. The magnetic method was not applicable and the thinness of the chromium prevented the use of the microscopic method or the chord method. An article in the May, 1948 issue of the ”Journal of the Electrochemical Society”, however, described a method which seemed to have possibilities. This article, entitled ”An Electrolytic Thickness Tester for Plated Coatings”, was written by Howard F. Francis of the Armour Research Foundation.

The method discussed in this article was a development of the coulometric method. Faraday’s Law states that, ”the quantity of any element (or radical) liberated at either anode or cathode during electrolysis is proportional to the quantity of electricity that passes through the solution”. If there is 100 per cent current efficiency, it is therefore possible to calculate the weight of any metal which will be plated or deplated by a particular amount of electricity. The weight of metal is a function of its density, thickness, and area, and the amount of electricity is a function of the current and time during which it flows. Since the density is known, therefore, the time required to deplate a known area with a known constant current is a direct measure of the plating thickness. This ”anodic solution method”, therefore, provides an absolute measurement which does not depend upon an empirical calibration.

Its accuracy depends upon careful measurement of the variables noted above. An empirical calibration may be determined from standards of known plate thickness to avoid error introduction by a faulty measurement of the areas. The time necessary to deplate at constant current the known area is the variable in measuring thickness, and so it must be checked carefully. A change in voltage is used as an indication that the plated metal is entirely deplated from the base metal. An electrolyte must be selected such that the plated metal will dissolve anodically more readily than the base metal or, in other words, for such an electrolyte the plated metal is higher in electrochemical activity than the base metal.

Dr. Francis used a small current to insure 100 per cent current efficiency and this was held constant. He made use of a stainless steel cell and a rubber washer to limit the area to a known value. An ingenious system was used to record the tine necessary to deplate the metal in question from the known area. An electric timer was started when a test began. A vacuum tube relay was actuated by the voltage change at the anode when the plated metal was completely deplated from the known area. This relay stopped the electric timer and automatically recorded the time necessary to deplate the plated metal.

Experimental Cell Design
In order to investigate this method an experimental cell was designed and made. This cell is pictured in Fig. 1. The stainless steel cylinder is pressed into the Lucite cylinder. The rubber washer fits snugly without deformation into the recessed portion of the Lucite. The rubber washer insulates the cathode, the stainless steel cathode cylinder from the metal to be checked which is the anode; and also defines the area to be stripped. It is therefore important that the area of the hole be measured as accurately as possible and this is the only critical dimension of the cell. The washer used was placed on a comparator and blown up to fifty diameters to enable a more accurate measurement of the area. Current was furnished by a 90 volt dry cell; with a large variable resistance in series with the deplating cell. A high voltage dry cell was used so that variations of the current caused by variations of the voltage at the deplating cell, would be kept as small as possible. The emf of the deplating cell and the time were recorded on a recording potentiometer (Brown Instrument Co., Division, Minneapolis-Honeywell Regulator Co.). A 10 per cent solution of sodium hydroxide was used as the electrolyte.

Test Procedure
To determine the accuracy of the experimental instrument chromium thickness standards were prepared. Several 3 x 6 inch panels were plated as described previously and 2 x 3 inch panels were machined from them. The loss in weight method was used on a number of these to determine the thickness of the chromium. The panels selected for this were carefully weighed and then painted on the back and edges with acid proof paint. Then they were immersed in hydrochloric acid until the chromium was stripped. After rinsing and drying the acid proof paint was removed with thinner and after careful drying the panels were reweighed. The thickness of the chromium was assumed to be the same over the entire six square inches. The chromium thicknesses thus determined were assumed to be the same as those of the other panels plated with the same current density and at the same time.

Thickness determinations with the deplating cell the were made on these corresponding panels using a current of 2 milliamperes in the cell. The potentiometer trace on its record automatically plotted the emf of the deplating cell against time. Time, as a variable, was read directly from the potentiometer record. The start of the deplating was recorded by the beginning of the trace as the deplating current was turned on. While deplating, the emf of the cell was nearly constant, but at the moment the deplating was complete, the emf rose sharply. Thus the time elapsed while deplating was in progress was read directly from the potentiometer record. Each calibration panel was checked several times. From the thickness determined by the loss of weight test, it was then possible to calculate the thickness of plate removed per unit of time. The calibration of the deplating unit with one particular rubber washer was 2.05 millionths of an inch of chromium for each minute of deplating time with a current of two milliamperes. This value agreed within 1 per cent of the value calculated from the equivalent weight, density, area, current, and time. The reproducibility of readings in the range of 8 to 20 millionths of chromium was within 5 per cent. The variation between these readings and the thickness by the loss of weight method for those panels was less than 5 per cent.

To determine if a smaller current was being used than was necessary, experimental runs were made using a current of 5 milliamperes and 2 milliamperes. A current of 5 milliamperes gave results which were not always reproducible, and usually were slightly high. This was attributed to the fact that the current efficiency was not 100 per cent at 5 milliamperes, It was found experimentally that a current of 22 milliamperes could be used. However, it was decided to use 2 milliamperes in production to be assured of 100 per cent current efficiency.

The anodic solution or coulometric method was adopted immediately by the inspection section for the routine checking of chromium plating thicknesses on production parts. It was found to be very accurate and satisfactory for this purpose. Shortly after the adoption of this anodic solution method to measure the thickness of chromium electrodeposits, announcements of a commercial instrument were noted. Thickness comparison tests were made using the two instruments and the results were found to be comparable.

Commercial Thickness Tester
The commercial thickness tester has a deplating cell similar to the experimental unit. It is made of stainless steel and uses a rubber washer to limit the deplating area. An electronic relay is used in conjunction with an automatic counter to measure the time of deplating. When the deplating current is turned on, the counter is automatically started. When the plated metal is deplated, the rate of change of voltage actuates the relay which turns off the instrument including the counter. The counter is calibrated directly in thickness units (millionths of an inch for chromium and hundred thousandths of an inch for other metals). The electrolyte is agitated during deplating by the ingenious use of pulsating air which is the secret of its greater speed, as will be explained presently. The thickness of several plated metals can be measured by adjusting a dial on the instrument’s front panel to the appropriate reading and then using the appropriate electrolyte supplied.

The commercial unit was found to have the advantage of speed—about thirty times faster—over the experimental one. In production, the author’s company tries to hold the thickness of decorative chromium plate between 10 and 20 millionths of an inch. With the experimental unit it took approximately eight minutes plus set-up time to measure 16 millionths of an inch of chromium. The commercial tester was found to deplate about one millionth of an inch of chromium or about one hundred-thousandth of an inch of another metal per second. This is because the air agitation allows 100 per cent current efficiency at a higher current density. The author feels the commercial instrument is much more versatile than the experimental unit since it is designed to measure, by the use of appropriately furnished electrolytes, the thickness of chromium, silver, tin, cadmium, zinc, brass, copper, and nickel on steel; the thickness of chromium, silver, tin, cadmium, zinc, and nickel on brass or copper, and the thickness of chromium on nickel. Composite coatings, such as copper, nickel and chromium on steel, may be measured successively by changing the electrolyte solution without disengaging the cell from the sample.

Production Experience
It has been used successfully for about a year to measure the thickness of nickel and chromium on copper and chromium on nickel. For these measurements the anodic solution method is preferred over all others. The instrument has been used also to determine the thickness of cadmium, zinc and copper upon steel. It is believed to be more accurate than a magnetic type instrument* for these determinations; however, the latter is preferred because it does not damage the parts. The thickness of copper upon zinc die cast metal can not be determined by the anodic solution method because as far as is known, no electrolyte has been discovered in which copper will dissolve anodically more readily than zinc.

The thickness of nickel deposits over steel has been found by many to be difficult to evaluate by the anodic solution method. As this was a problem the author’s company might shortly be facing, the thickness of nickel plated upon steel was checked by this method. Results were found to be erratic. The rate of change of voltage after the nickel was deplated from the steel was not great enough to actuate the relay and turn off the instrument even though the electrolyte was such that the nickel dissolved anodically more readily than the steel upon which it is plated. The results indicated that the nickel thickness could be measured if it were plated upon copper, but not if it were plated upon steel. To prove or disprove this point, another set of experimental panels were plated. These panels were cold rolled steel and were 4 x 5 inches. They were numbered in opposite corners as shown in Fig. 2. Twelve panels were copper plated and twelve were unplated steel. All 24 panels were placed in the nickel tank and plated at 50 asf. At the end of one minute, 3 of the copper plated and 3 of the unplated steel panels were removed. At the end of 4, 20, and 60 minutes, this procedure was repeated. Thus, there were produced 3 copper plated steel panels; 3 steel panels plated with 1 minute of nickel; 3 of each with 4 minutes of nickel; 3 of each with 20 minutes of nickel; and 3 of each with 60 minutes of nickel. Two samples of each type were selected and cut in half as is shown by the dotted line in Fig. 2. A piece about 1 x 2 inch was cut from one half of each sample as illustrated in Fig. 2. These small pieces were mounted in bakelite, polished, and the nickel thickness was measured microscopically. The nickel thicknesses did not vary with the base metal and are tabulated in Table 1. These values were considered to be the same as the nickel thickness of the adjacent edges of the corresponding samples. The nickel thicknesses in these areas were then measured with the commercial anodic solution type measuring instrument. The thicknesses of nickel plated upon copper are shown in the above table. The amount of nickel on the one minute panel is not shown because it was below the minimum limit of accuracy of the instrument, which is 50 millionths of an inch for all metals except chromium and 5 millionths of an inch for chromium. The percentage of error of these values compared to those determined by the microscopic method is also tabulated. The range of values for each panel was about equal to the error. The thicknesses of nickel plated upon steel could not be determined on the commercial anodic solution tester with the electrolyte available during the test period noted above.

One possible reason for this difficulty was that the rate of change of voltage across the cell when deplating was complete was insufficient to actuate the relay of the instrument. To confirm this, a recording potentiometer was placed in the circuit to record any change in voltage and the time at which the change in voltage occurred. This indicated a small change in voltage and also provided a positive measure of the time which did not depend upon the rate of change and satisfactory results were obtained. Passivity was also considered to be a cause of faulty readings in the determination of nickel thickness on any base metal. Any electroplater can testify to the difficulty of preventing nickel plate from becoming passive. If this occurs during deplating, it will give the nickel an apparent thickness greater than it actually is because the current efficiency at the anode is not 100 per cent.
The difficulty experienced with nickel thickness determinations over steel was reported to the supplier who in turn, furnished a sample of a new electrolyte for future trials. Experiments with the nickel plated panels were repeated using the new electrolyte and were wholly successful. With it, it was possible to measure the thickness of nickel plated on either steel or copper within five per cent.

CONCLUSION
In summation, it has been found that the anodic solution method can be used successfully to measure the thickness of electrodeposited coatings of copper, zinc, cadmium and nickel upon steel, the thickness of chromium and nickel upon copper and the thickness of chromium upon nickel. For the determination of nickel upon steel, chromium and nickel upon copper, and chromium upon nickel, the author prefers to use the anodic solution method for accuracy and convenience over the drop test or magnetic methods. For the determination of cadmium, zinc, and copper upon steel, the magnetic method is preferred because it does not damage the parts. At present, the author knows of no electrolyte by which the thickness of copper plated upon zinc die castings can be determined using the anodic solution method. In general, the thickness of any metal plated upon any other metal can be found by this method if an electrolyte is obtainable in which (1) the first metal will dissolve anodically more readily -than the second, or (2) the metals exhibit different potentials.

ACKNOWLEDGMENT
The author wishes to acknowledge the helpful assistance of Dr. E. A. Hodges of the Kocour Company; the co-operation of many of the personnel of that organization and of his own company, the King-Seeley Corporation, for their contributions to this paper.

REFERENCES
1. Howard F. Francis, J. Electrochem Soc., 9.o (May, 1948).
2. Plating and Finislling Guidebook, 20th Editioil, Metal Industry Publishing Co., N. Y. (1959).

---

DISCUSSION

---

MR. H.C. SCHLAUPITZ (R. Wallace & Son, Wallingford, Conn.): In the figures that you presented on the board, you have compared the thickness determined by the Kocour thickness tester with microscopic thickness measurements. Your greatest differences are in the low thickness range. I think Dr. Read has determined in his research work at Penn State, and we have also found it to be the case at, R. Wallace, that the microscopic method is the least reliable in those ranges. Perhaps it might be that the anodic solution method is the more accurate one and that the error given represents that of the microscopic measurement rather than that of the anodic solution measurement.

MR. WAITE: That could be true, though I know of no easy way to prove it. One of the big difficulties in that range is that any errors that may take place, which are more or less constant through a series of measurements, will have much more bearing, percent wise. That is they cause a larger per cent error in the measurement of a thin plate than in the measurement of a thick one.

There is an auxiliary unit which can be used with the Kocour unit which will allow you to measure to 0.000005 inch on coatings other than chromium and, I believe, to about 0.000001 inch on chromium.

MR. A. W. CAGLE (Western Electric Company, Winston-Salem, N. C.): Has erratic performance been experienced with the Kocour testing of tin on copper?

MR. WAITE: I have never checked tin over copper. Supposedly it can be used as such. I don’t know whether Mr. Hodges is in the audience or not—I think he can answer that.

MR. A. E. HODGES (Kocour Company, Chicago, Ill.): There should be no erratic results in testing tin over copper with the anodic solution method, providing the surface oxide film is removed. This can be accomplished by rubbing the surface lightly with a pencil eraser.

MR. WILLIAM H. FORDNEY (Hamilton Watch Company, Lancaster, Pa.): Are the results of microscopically made thickness measurements an average of several observations or of one point on the specimen?

MR. WAITE: BY several, as I showed you by the slide—we cut a piece, 2 inch x 1 inch, out of the center at that edge. The Kocour readings, of course, were taken directly opposite. We mounted the 1 inch portion in Bakelite. We had to bend it in the middle so we took no measurements exactly at the bend, but we took the average of measurements on both sides of that throughout the entire piece.

MR. DAMON C. ANTEL (Westinghouse Electric Corporation, East Pittsburgh, Pa.): Has any work been done, to your knowledge, in checking the thickness of plated coatings on aluminum by this method?, If so, what do you find the accuracy, and what particular coatings have been checked?

MR. WAITE: Mr. Hodges may be able to answer this better than I can. I know of no way of doing such from personal experience. I think it would be extremely difficult to find a solution in which metals ordinarily plated upon aluminum are anodically more active than the aluminum itself.

MR. HODGES: Actually, such tests have been made, including nickel over aluminum and chromium over aluminum. We have not made any accuracy checks on the method, but the end points’ (voltage changes) lead us to suspect the accuracy would be very good.

MR. LIONEL CINAMON (Special Chemicals Corporation, New York, N. Y.): I notice in your column that you have a differential about 0.000010 inch consistently down the line. Does that hold true of all of the various times put in? If so, is there any reconciliation between the instrument and thickness along the 0.000010 inch line? Do you, therefore, in that manner get an instrument calibration?

MR. WAITE: I have not noticed that before, so I cannot answer your question.

MR. CINAMON: If we assume that the instrument is correct and you have 0.000100 inch with a four minute plate, both the 20 minute plate and one hour plate become very consistent as far as accuracy is concerned. This is because if you take 15 times the four minute plate, you have about 0.00150 inch. Do the same thing on the 20 minute plate and you obtain equivalent results. It would make the instrument fantastically accurate.

MR. WAITE: The instrument is capable of extreme accuracy; it is only a question of how closely you can measure the variables. The washer that we used with the Kocour instrument, we blew up to 50 diameters and when our precision gage department drew a circle corresponding to it, we could find no error in the area-’ from being out of round. The current must be held; constant. This is probably a source of error. The density and electrochemical equivalent are exceedingly accurate, so inherently it is possible to make extremely accurate measurements with the instrument.

The reading of four minutes is done microscopically. We feel that one hundred millionths (0.000100)’ is about as low as it is accurate to go with the measurement of any thickness with the microscope. We feel that the error is too great if microscopic measurement of thicknesses under 0.000100 inch is attempted.

MR. FRANCIS T. EDDY (Chase Brass & Copper Co., Waterbury, Conn.): In checking chromium on nickel, what is the reproducibility of results by the electrolytic method as compared with the hydrochloric acid spot test?

MR. WAITE: The results are very reproducible. They consistently have been less than 5 per cent. We found that the hydrochloric acid drop test gave us results closer to 10 per cent accuracy even with an experienced operator because small variations of temperature or error in timing with a stop watch can affect it materially. On the other hand, both the reproducibility and the accuracy are within 5 per cent with the anodic solution method.

MR. EDDY: Ordinarily, how long does it take to run off a test like that?

MR. WAITE: It takes about a second with the Kocour unit.

MR. EDDY: Including preparation time?

MR. WAITE: Practically none at all. It is used routinely in inspection by inexperienced operators who simply run through their operations mechanically.

MR. EDDY: IS it necessary to clamp the sample to the testing apparatus?

MR. WAITE: We do, yes. The Kocour instrument has an arrangement that makes it very easy. In the original we had the cell built right into the clamp. I have a sample of our original cell and some of the samples for you to see.

MR. EDDY: We use a ring of wax around the chromium and’ apply a drop of hydrochloric acid. ‘ We check the temperature very carefully and the time with a stop watch. We reproduce results within 5 per cent readily and rapidly, so that we use quality control methods all the way through on all production.

There is absolutely no time lost as far as setting up the sample is concerned.

MR. WAITE: That is on chromium directly on copper?

MR. EDDY: We use the wax ring method for chromium directly on copper or brass with equally good results. What I am trying to find out is why there would be an advantage for the electrolytic method over a method as simple as the hydrochloric method.

MR. WAITE: One advantage is that you could clamp the part into the instrument, there is no time lost. It is just as fast, probably a little faster than drawing a ring of wax, measuring your temperatures and determining that everything is at the same temperature.

MR. EDDY: But if you have an irregular surface, you cannot get down into a cylindrical piece or anything similar.

MR. WAITE: We measure curves within a half-inch radius in our particular parts and have experienced no trouble. We have no parts with enough variation to cause trouble. Once we have clamped the piece in position we can measure successively first the chromium and then the nickel and copper on steel without removing the piece. It could be measured in a matter of minutes.

MR. FRANK G. BEUCKMAN (Eastman Kodak Company, Rochester, N. Y.): How do you keep the rubber washers from deteriorating with time?

MR. WAITE: We buy extra washers, they are very inexpensive, and we keep them in distilled water.

MR. BEUCKMAN: How do you maintain stability of size when a clamping method is used?

MR. WAITE: It has to be done carefully. We do not use too much pressure—you get the feel of it, with experience. Our operators do it very well.

---