Measurement of Thickness of Fired-on Gold Coatings K. H. BALLARD, E. I. du Pont de Nemours & Company, Perth Amboy, N. J.
A
crystals into a 1-liter glass-stoppered bottle, add 100 ml. of distilled water to the iodide-iodine mixture, and shake until the solids are dissolved. Pour 900 ml. of distilled water into a large graduate and add 8 grams of Gardinol (du Pont, WA flakes). Stir until the Gardinol is dissolved and then add the Gardinol solution to the iodide-iodine solution. Add 31 ml. of hydrochloric acid (concentrated c. P., 1.19 sp. gr.), shake, and then filter the reagent. Store the reagent in a glass-stoppered bottle in a darkened cabinet. Preparation of Reagent B. This solution is recommended for tests on the heavier type gold decorations. It is prepared like Reagent A, using the amounts shown in Table I.
STRONG need for a practical method of measuring the local thickness of fired gold coatings on ceramic base materials has been felt for some time both in the laboratory and in commercial operations. The jet-test method here described was developed primarily for this purpose, but with only minor changes it should be applicable to electroplated and other types of gold coatings. The oldest and best-known test for determining the local thickness of base metal coatings is the Preece test (5),which consists in the gradual removal of the coating by repeated immersion in a suitable strip ing solution. Later investigators (1) showed that the rate of sorution of zinc and cadmium coatings was more constant if the solution was applied to the coating in successive drops instead of all at once as it is in immersion. Hull and Strausser (4) used a fast dropping rate and found that the rate of solution of electroplated coatings then becomes nearly independent of the rate of dropping, making it possible t o express thickness merely in terms of the ti.me of dissolution of coating. In 1937, Clarke (2) described the jet test, consisting of a continuous stream of reagent from a small orifice im inging on the surface instead of successive drops. A stripping factor was calculated for each particular solution and the thickness of zinc and other base metal films was computed directly from the time required for penetration of coating and the respective stripping factors.
TABLE I. REAGENT B Iodine (resublimed). erama
11.1 27.8
1000 8.5
37
Standardization of Reagents The ammonium iodide-iodine-hydrochloric acid reagents used in this work were standardized by two independent methods.
T h e major part of this investigation consisted of developing a suitable reagent for the jet test on gold coatings. Since the customary gold decoration is composed of gold metal plus a flux, it was necessary that the reagent independently attack both components of the coating. An acidified ammonium iodide-iodine solution, containing Gardinol as a wetting agent, met this requirement.
1. A jet test was made on a gold decoration of known area, the remaining old coating was dissolved, and the weight of gold was determine1 by a standard colorimetric test ( 3 ) . The strippin factor of a particular reagent was calculated from the time of t%ejet test, the weight of gold, and the area of coating. 2. A large porcelain crucible was coated on the outside with gold. The stripping factor was again calculated from the time of the jet test, the known area covered, and the weight of gold as determined by weighing the crucible on an analytical balance. STRIPPING FACTORS. The thickness of gold decoration may be calculated directly by taking the product of the time of jet test and the following factors:
Test Apparatus and Reagents APPARATUS. The jet-test apparatus consists essentially of a 250-ml. separatory funnel, the stem fitted with two glass stopcocks instead of the usual single cock. The lower end of the stem is cut to produce a plane face making an angle of 30” with the long axis of the funnel. A ringstand is used for supporting the separatory funnel and specimen and a large dish serves as a receiver for the dropping reagent. Figure 1 shows the assembled apparatus nith specimen in correct test position. REAGENTS.There are two common types of gold decorations used on ceramics-the so-called “liquid bright” type, which is usually below 0.00025 mm. (0.00001 inch) in thickness, and the ‘‘burnish’’ or “coin” gold type, which is usually several times thicker than the liquid bright gold decoration. In order to have the time of dissolution of coating in both cases fall within a practical test period, two solutions designated as Reagent A and Reagent B were used. The latter reagent is much stronger than the former and is therefore recommended for determining the thickness of burnish gold dxorations. The hydriodic acid contained in the reagents is relatively unstable, and for most accurate results the reagent should not be used after standing longer t h a n 2 t o 3 weeks. tleagent A . This solution is recommended for thickness measurements of liquid bright gold decorations. Weigh FIGURE 1. 2.2 grams of iodine crystals (resublimed) JET TEST and 5.5 grams of ammonium iodide APPARATUS
Reagent A. One minute equivalent to 4.1 X 10-6cm. Reagent B. One minute equivalent to 24 X 10-6 em. The weight of gold metal per square centimeter may be calculated from the time of dissolution and the following factors: Reagent A. One minute equivalent to 7.9 X 10-6 gram per sq. em. Reagent B. One minute equivalent to 5 X lo-’ gram per sq. em.
Recommended Procedure Select a gold band over 4 mm. in width, which has been given the proper firing cycle, and determine local thickness of gold band by the use of the jet test in the following manner: Fill the special separatory funnel with the proper stripping reagent at a temperature of about 25” C. Open both stopcocks and adjust lower stopcock so that solution discharges a t the rate of 7 ml. ( * 0.2 ml.) per minute. Clamp specimen fimly in position, with the surface of test spot parallel to face of delivery tip and with a slight clearance between the two surfaces. The surface of test area should then make an angle of 60’ with the horizontal. In order to produce streamline flow, the reagent should be allowed to flow a distance of about 1 em. over surface of specimen before reaching the particular spot whose thickness is to be determined. Degrease the surface of test area by use of organic solvents or by rubbing lightly with a piece of cotton containing a dilute solution of Gardinol. Open the upper stopcock and note the number of minutes required for removal of decoration in the flow area-i. e., appearnnre of the base glaze or glass surface. Calculate the gold thickness from the stripping factors given above and the experimentally determined times of dissolution of coating. 868
ANALYTICA L EDITION
November 15, 1942
Practical Application The method has'been applied for practical control work with entirely satisfactory results. About 3 minutes is required for each thickness test. It has also been used b y certain potteries for estimating the costs of different type of gold decorations. Some tests have been made on the substitution of Reagent B for hydrofluoric acid in removing undesired gold spots from decorated ware.
width, at a temperature of 25" C., and for gold coatings containing no interfering substances-e, g., platinum and silver. The method would be applicable t o thickness measurements of ceramic coatings of gold alloys, b u t a stripping factor for each particular alloy would have t o be ascertained. The method is rapid and with reasonable care should be accurate to *6 per cent.
Literature Cited
Summary A jet-test method for measuring the thickness of liquid bright and burnish gold coatings on ceramics and two independent standardization methods for determining the stripping factors Of the ammonium acid-Gardinol reagents have been developed. The stripping factors given apply to tests made on gold bands over 4 mm. in
869
(1)
Blum, W., Strausser, P. W. C., and Brenner, A , , J . Research
Natl. Bur. Standards, 16, 185 (1936). (2) Clarke, S. G., J . Electrodepositors' Tech. soc., 12, 1 (1937). (3) Fink, C. G., and Putnam, G. L., IKD.ENG.CHEM.,ANAL.ED., 1 4 , 4 6 8 (1942). (4) Hull, R. O . , and Strausser, P. W. C., Monthly Rev. Am. Electroplaters Soc., 22, 9 (1935). (5) Pettenkofer, M.,-\'at. Tech. Comm. .%fdnchen, 1, 159 (1857).
Spectrophotometric Determination of Iron With o-Phenanthroline and with Nitro-o-phenanthroline J. P. MEHLIG AND H. R. HULETT', Oregon State College, Corvallis, Ore.
I
N RECENT years various spectrophotometric metho& of analysis have been developed, making use either of a reference curve correlating transmittancy a t a given wave length with concentration or of the extinction coefficient of the color system at a given wave length. The senior author has determined manganese as permangafiate (4) by the former procedure and copper with ammonia (6) and iron with salicylic acid (6) b y the latter. The orange complex formed by the combination of three molecules of o-phenanthroline with one ion of ferrous iron, first reported by Blau (I), has been used for a number Of years as a n internal indicator in oxidimetric titrations. A more recent development has been the use of the cherry-red nitro-o-phenanthroline ferrous complex for the same purpose. Saywell and Cunningham (7) have applied o-phenanthroline to the visual colorimetric determination of small concentrations of iron in various food products and Hummel and Willard (3) have used it for the colorimetric determination of iron in biological materials. Fortune and Mellon (2) have made a critical spectrophotometric study of the SaywellCunningham method with special attention to the effect of diverse ions and hydrogen-ion concentration and to the stability of the color system. The purpose of the work described in this paper was to apply both o-phenanthroline and nitro-o-phenanthroline to the spectrophotometric determination of iron in ores, making use of the fundamental Lambert-Beer equation Y
z
=
lo
x
10-kZc
in which lo represents the intensity of the light of a given wave length entering the system, Z the intensity of the light transmitted by the system, 1 the length in centimeters of the solution through which the light passes, c the concentration in grams per liter of the substance absorbing the light, and k the specific extinction coefficient, a constant which is a measure of the absorption due to a single molecule.
Apparatus and Solutions -411 spectrophotometric measurements were made with a Cenco-Sheard spectrophotelometer. 0-PHESANTHROLINE. A 0.1 per cent solution of tGe monohydrate prepared by dissolving in water warmed to 80 NITRO-O-PHEXANTHROLINE. A 0.1 per cent solution in 95 per cent ethanol.
.
1
Present address, U. S. Army.
HYDROXYLAMISE HYDROCHLORIDE. An aqueous solution con-
T1.0.25
t a ! $ ~ ~ ~ $ a ~ L ~ ~ solution ~ ~ o prepared by dissolving 56.49 grams of the dihydrate in 350 ml. of 6 M hydrochloric acid and diluting. to 1000 ml. with water. STAXDARD IROSSOLU&ON.A standard solution was prepared by dissol.i.ingo.7029 gram of 99.9 per cent pure ferrous ammonium sulfate hexahydrate in water and diluting to 100 ml. Ten milliliters of this solution were made up to 1000 ml. with distilled viater slightly acid with hydrochloric acid. Each milliliter Of this final solution contained 0.01 mg. Of iron. DIVERSE10s SOLUTIOYS.Standard solutions, each milliliter containillg mg, of the ion in question, were prepared from the chloride or nitrate salt of the cations and from the sodium, potassium, or ammonium salts of the anions.
The Color Reaction To produce the color system the amount of the standard solution of iron required t o give the desired concentration of iron was measured into a 100-ml. volumetric flask and 1 ml. of hydroxylamine hydrochloride solution was added to reduce any iron which might have been oxidized by dissolved oxygen. The solution was diluted to approximately 75 ml., 10 ml. of o-phenanthroline or nitro-o-phenanthroline solution were added, and the volume was accurately made up to 100 ml. The solution containing nitro-o-phenanthroline was allowed to stand 2 hours for the full development of the color. All transmission measurements were made with a Cenco-Sheard spectrophotelometer by setting the wave-length scale, adjusting the incident light by means of a diaphragm to provide the desired displacement of the pointer on the photometer scale, and reading the transmission of the standard solution and of the blank solvent. The transmittancy was calculated by dividing the former by the latter. A 1-cm. cell was used throughout. The spectral transmission curves for the two systems are very similar with the peak of the absorption band in each case at about 508 mp. Since they closely resemble the curves obtained for the o-phenanthroline-iron system by Fortune and Mellon ( 2 ) using the self-recording PurdueGeneral Electric spectrophotometer, they are not shown. T h a t Beer's law is followed b y both color systems is proved
by the straight line which resulted when the logarithms of the observed transmittancies at 505 m p for five solutions containing from 0.5 to 4 p. p. m. of iron were plotted against the respective concentrations. This is in accord with the observations of Fortune and Mellon (2) for o-phenanthroline.
Determination of Specific Extinction Coefficients Specific extinction coefficients were determined for both color systems a t 490 and 505 m p by obtaining the respective transmittancies a t those wave lengths of a series of known iron
