US3502548A - Method of electroplating gold on chromium - Google Patents

Description

I March 1970 E. H. LYONS, JR 3,502,548

METHOD OF ELEGTROPLATING GOLD 0N CHROMIUM Filed Oct. 24, 1966 AOSORBEO MOISTURE Path of electrons released from lVi proceeding to cathodic region Corrosion current Precipitated Ni (0H): keeping fi low (EfE ,sinoe there is a metallic path between anode and cathode) FIG. IDEAL/ZED REPRESENTATION OF OORROSION IN PORE IN OHROMIUM-PLATEO NICKEL 0r ADSORBED MOISTURE Electrons do INA 6T! I/E not flow GHROMIUM sun/anon wouLo potentials at anode REQUIRE E. 0.370 T0 REACT) HIGH RESISTANCE AND lVl++ accumulates IMPEDED ION/0 DIFFUSION with little 'ANODE lVi- Ni**+2e" precipitation lvi** z MQ z"036W of NI (0H (too positive to I drive cot/lode reac tron) L A (At equilibrium, E =E J F I 6.2. IDEAL/ZED REPRESENTA TION OF CORROSION IN PORE IN GOLD-OHROMIUM-PLATEO NIOKEL nv VElV TOR ERNEST l-l. LYONS ATTORNEYS United States Patent US. Cl. 204-32 5 Claims ABSTRACT OF THE DISCLOSURE A plating process for adhering gold to chromium. The surface of the chromium is depassivated in anacid bath and is then placed in aqueous acid gold plating bath having a pH between 2.5 and 4.5. It is then plated with a thin layer of gold from the acid bath while preventing the formation of loose displacement deposits on the surface. One way of preventing the formation of loose displacement deposits is by providing anacid plating bath containing from 0.1 gram per liter to grams per liter of a base alloying metal such as iron, nickel, or cobalt in the form of a soluble salt. Another Way is to employ a bath containing chelating agents.

This application is acontinuation-in-part of application Ser. No. 493,560, filed Oct. 6, 1965, now abandoned, which was a continuation-in-part of application, Ser. No. 181,478, filed Mar. 21, 1962, now abandoned.

This invention relates to improvements in gold plating. More particularly, it relates to a novel self-passivating corrosion-resistant duplex metal coating and a method for producing such a coating. i

The color of gold has been highly prized for many centuries, and articles plated with thin coatings of gold possess its appearance, but sooner or later such articles have always become tarnished. As gold itself does not tarnish, what has actually occurred has been the tarnishing or corrosion of the substrate metaL.

T hinelectroplated coatings are invariably penetrated by numerous microscopic andsubmicroscopic holes, pores, cracks, and other discontinuities. The substrate is exposed to corrosive atmospheres through these openings. Corrosion products derived from the substrate spread until the entire surface may appear to be tarnished. The gold coating is often loosened and lost because of corrosive undercutting of the substrate beneath it.'Since the substrate is always anodic relative to the gold coating, corrosion reactions are accelerated by the anodic potential; hence, tarnishing proceeds more rapidly in these exposures than it would in the absence of the gold.

I have found that a thin duplex coating of chromium with an overcoating of .gold has remarkable and unique passivation and de-passivation of the chromium. It is well known that certain metals and alloys may be caused to assume a so-called passive state, in which ordinary corrosion or tarnishing processes are practically extinguished,- andthe passivated metal surface retains its brightness for an indefinite period. With some metals, the passivity is quickly lost, unless it is maintained by suitable chemical resistance to tarnish. In salt spray and other accelerated corrosion tests, it is superior to ordinary coatings, such as copper-nickel-chromium, even though the latter may be as much as ten times thicker. The singular protective power of these gold coatings is exemplified by duplex coatings consisting of 0.00002 inch of chromium and 0.000005 inch. of gold, some applied to a copper and others to a steel substrate, either with or without a nickel undercoating such coatings havewiths tood 72 hours in the common salt spray test, whereas about 0.0005 inch of nickel and 0.00002 inch of chromium are required to give the same protection using the common coating system.

or electrochemical treatment; this is true with ordinary carbon steels. With some other metals, of which chromium is one, passivity, once established, is maintained under a variety of conditions, and a bright finish is preserved.

Heretofore, a standard corrosion-resistant coating has been a duplex electroplated coating of chromium on'nickel. The chromium surface attains and retains passivity without becoming tarnished, although the corrosion of nickel exposed through the inevitable cracks in the bright chromium electroplate is accelerated by the presence of the chromium. Nickel-chromium coatings, if thick enough, have maintained a bright surface finish on metal for a number of years so long as depassivating agents were substantially absent from the environment. Early corrosion of automobile parts and other products plated with nickel and chromium has usually been attributed to chlorides, which are present not only at the sea-coasts, but are spread on highways and streets during the winter in order to melt accumulations of ice and snow. Sulfur dioxide, in part oxidized to sulfuric acid, is commonly found in the air in industrial locations, and it, too, leads to accelerated corrosion on surfaces protected with nickel and chromium plating.

One object of the present invention is to provide a protective coating that can be plated on metal surfaces and which assumes and retains passivity even in the presence of considerable concentrations of depassivating agents commonly encountered in present day atmospheres, so that tarnishing and dulling does not occur, and corrosion is inhibited. The duplex coating of the present invention shows resistance to corrosion greatly superior to chromium-nickel coating in atmospheres containing these corrosive agents.

Another object of the invention is to provide a non-tarnishing, self-passivating plated coating which affords relatively permanent protection to a metal substrate even though employing substantially thinner coatings than are required by common copper-nickel-chromium coating systems.

Another object is to provide a non-tarnishing, self passivating plated coating which does not accelerate the corrosion of the substrate metal through breaks and openings in the protective coating.

A further object is to provide a non-tarnishing, selfpassivating plated coating which has the desirable color of gold Without requiring substantial quantities of that precious metal.

Another important object of this invention is to provide methods for applying these new duplex coatings in a way that results'in true adhesion of the gold to the chromium.

Chromium can be applied to the substrates by electroplating from the conventional chromic acid bath operated so as to obtain a bright, hard coating; or by vapor deposition or chromizing processes. Electrodeposits of chromium made from high temperature baths do not have the de- Furthermore, my novel duplex coating can be applied to low-cost substrates, and to impart and maintain the highly desired gold color.

The phenomena of this invention are concerned with long been established in the art, namely, that the coating is adhered so strongly to the-substrate that it cannot be dislodged or separated by mechanical forces, such as might be exerted by vigorous rubbing, by application of a chisel, by large variations in temperature, or by bending. of the plated article. According to Electroplating Engineering Handbook (A. K. Graham, editor, Reinhold Publishing Co., New York, 1955, page 65), Thereneeds to be an atomic lattice continuity at the interface between the coating and the basis-metal in order to obtain true adhesion. True adhesion is equivalent to what I mean hereinby adhesion. and adherence.. I am speaking of adhesion in the sense in which H. B. Linford used the term in Plating 41, 1954, page 283, when he said the adhesion should be of the same order of magnitude as the tensile strength of the metals themselves. Another investigator, C. L. Faust, Monthly Rev. Am. Electroplaters Soc., March 1946, page 266, says, Adhesion should be considered as solely a matter of workmanship in plating, and like that, is either good or bad. Adhesion of a degree should not be accepted. If adhesion is not perfect, the preparation and plating process are either not under proper control or are improperly selected for the job.

If a coating can be rubbed off by repeated rubbing with a soft cloth to which pressure is applied, or if it can be lifted off by atfixing adhesive tape and then removing the tape, its adhesion is unsatisfactory; it is not an adherent coating, as I use the term herein and as the term is characteristically used in the electroplating art. In other words, physical adsorption by Van der Waal forces is not adhesion, in the present sense.

An important fact is that heretofore no method has been published or made known to me which shows how to obtain an adherent coating of gold on bright hard chromium (which is the normal chromium plate obtained from a chromic acid bath at conventional temperatures and current densities). In this connection, chromium must be differentiated from such alloys as stainless steel. The behavior of chromium toward gold plating baths is quitediflerent from that of stainless steel. Hence, such patents as Lukens Re. 20,306 and 2,039,326 are not pertinent, as will now be briefly explained.

A notable distinction between depassivated stainless steel and depassivated chromium as a surface on which to deposit gold is that the standard potential of active chromium is -0.744 v., whereas that of stainless steel, although not measurable or definable with precision, appears to lie somewhere between 0.3 and -0.4 v. Since chromium is 0.34 to 0.44 v. more negative than stainless steel, chromium offers serious problems in making the deposit, for chromium tends to produce loose immersion deposits of gold which do not adhere, as electroplaters use that term, for such coatings can be wiped off, even with a soft cloth. Baths which operate well with stainless steel may not be satisfactory with chromium, because of this essential difference between the substrates, and the effect on the protective action which is obtained.

1 have prepared a bath according to Lukens 2,039,326, page 2, column 1, lines 36-41. The pH of this bath was 7.2 as measured with a glass electrode. Panels of stainless steel were depassivated according to page 1, col. 2, lines 46-50 of that same patent and rinsed in accordance with lines 51-55. Brilliant gold coatings were deposited from the bath, but they were readily removed by one light wipe with .a cloth. Although operation at various current densities and temperatures ranging from 25; to 90 C. was tried, satisfactory adherent depositswere not obtained. Lukens states that he used litmus paper to determine acidity, so that it was deemed appropriate to. vary the pH rather widely inan. attempt to obtain the desiredadherence. Crystals of sodium cyanide were therefore added, in accordance with Lukens instruction on page 1, col 2, line 31, and plating was attempted at pH 7.6 and 8.2. However, the coatings could stillbe wiped off, in other at ure from 35 C. to 90 C., without obtaining satisfac-j words, they were not adherent. Then,- additionalhydrochloric acid was added cautiously, and when the pH dropped to 4.2, the adhesion of the coatings was markedly improved; satisfactory coatings were obtained at 60 C. with pH 3.4-3.7. At pH 2.8, a precipitate, doubtless AuCN, formed, but brilliant, adherent plates were still obtained. 3 v I Attempts were then made to deposit gold"on chromium, using, themodified Lukens bath,'.but the deposits, though brilliant, did not adhere (i.e.',.were lacking in adhesion), although the bath operated satisfactorily on stainless steel. The pH was then variedfrom 2.2'to; 8.1, and the tempertory adhesion. These results, confirmed by many other experiments, indicate that Lukens bath is quite inoperative on depassivated chromium, as distinguished from stainless steel. I

I have found that gold can be adherently electroplated on bright hard chromium by: i

(1) Using an acid gold bath, inwhich between 2.5 and 4.5,and I (2) Adding an alloying metaLpreferabIy iron, nickel or cobalt, in amounts about'0;1 gram to about 10 grams per liter of plating bath.

Consistent results seem to be favored by using potassium salts rather than sodium salts, e.g., potassium gold cyanide. I

Being somewhat more specific, the invention as applied to electroplating calls for the steps of keeping it depassivated until it is gold plated, and plating it in a gold bath at pH 2.5 to 4.5, with the gold bath containing about 0.1 g. to 10 g. per liter of bath of an alloying base metal.

A pure gold adherent coating on bright hard chromium can also be obtained by first depassivating the chromium and keeping it depassivated while coatingit in a displacement-and-chemical-reduction gold bath having a pH lying in the 2.5-4.5 range, even though the manufacturers instruction say to useno pH below 5.5.

Other objects, advantages and features of the invention will appear from the more detailed description that follows.

In the drawings:

FIG. 1 is an idealized representation of corrosion in a pore in chromium-plated nickel.

FIG. 2 is an idealized representation of corrosion in a pore in gold-chromium plated nickel, illustrating some of the advantages of the present invention.

Compared to the common copper-nickel-chromium platlng, equivalent protection is afforded to steel or copper, brass, or other copper alloys by a duplex chromium-gold coating only 10-25% as thick.

This singular result of the invention apparently can be explained as follows, although the invention is not limited to the explanation:

Like other plated coatings, the thin gold top-coat is undoubtedly penetrated by submicroscopic pores, fissures, and other openings, exposing the underlying chromium to corrosive attack. However, chromium becomes passive spontaneously, and this effect is further reinforced by the fact that active chromiumwould be anodic to the gold, and thus would immediately be passivated. In thepassive state, it is not tarnished or corroded appreciably. Since the gold is intrinsically-proof againsttarnish and corrosion, and the chromium where exposed is likewise tarnish-free, the duplex coating is not noticeably corroded.

However, microscopic examination indicates that many discontinuities in thegold fall immediately over the cracks which forma network in theunderlying chromium. Accordingly, the substrate of base metal is exposed through these openings. Whyis corrosion of'the base metal'exposed through these-openingslstrjongly suppressed? It might be thought that the accumulation of corrosion the pH is held products in these pores stifles' 'the corrosion reaction, as

was suggested by Lukens inconnectiojrr gold'coat-' ings on stainless steel; However, in studies of: nickel- 5. chromium coatings, it has been found that the nickel undercoat is corroded steadily (though slowly) through the pores in the chromium. See, for example, W. E. Lovell, E. H. Shotwell, and J. Boyd, Tech. Proc. Amer. Electroplaters Soc. 47, 215 (1960); E. I. Seyb, ibid. 47, 209; W. H. Safranek and R. W; Hardy, Plating 47, 1027 (1960). The corrosion products do not stifle the corrosion reaction in these coatings, and it seems unreasonable to suppose that they would do so simply because 0.000005" of gold is applied over the'chromium.

It is generally accepted that nearly all corrosion reactions arise from electrolytic cells comprising anodic and cathodic'regions which are contiguous, or nearly so, in the corroding metal. At the anode, the metal dissolves and produces ions. The cathode reaction is usually the discharge of hydrogen from water. The electrolyte is a film of moisture adsorbed on the metal surface, ordinarily too thin to be visible. Corrosion seldom occurs in the absence of moisture.

In the submicroscopic pores under consideration, the dimensions of the exposed substrate are assumed to be too small for contiguous anodic and cathodic regions to coexist within the substrate metal surface. Consequently, the chromium apparently functions as cathode, when no precious metal coating is applied. Considering the ordinary nickel-chromium coatings, the anode reaction is:

Ni Ni+++2e and the cathode reaction is: 2H O+2e-- H +2OH.

The question then arises whether the potentials prevailing at these electrodes are sufficient to account for the observed corrosion. According to the Nernst equation, and the accepted standard potential of nickel, the anode po tential is given by E,,=0.250+0.03 log c The cathode potential is E =0.000.06 pH+0.06 log p Assuming an effective partial pressure of atm. for hydrogen in the adsorbed electrolyte, and a pH of about 7, E =0.00 v. These assumptions can be varied over a wide range without affecting the conclusions significantly.

' To these static potentials, the effective overpotentials must be added. The corrosion data suggest a corrosion current of the order of a microampere per square centimeter; the nickel overpotential is therefore negligible. The hydrogen overpotential can be calculated from data given by G. Milazzo (Electro-chemistry, Elsevier Publishing Co., New York, 1963, p. 230) for chromium, using the Tafel law, which is believedto hold for this electrode. At 1 uamp./cm., it is 0.37 v. t

For the corrosion reaction to proceed, the anode potential must be more negative than the cathode potential; that is, it must be more negative than 000-037 V. Setting E,,=-0.37 v., it is found that the equilibrium value of C is 0.0001 M. Therefore, corrosion apparently proceeds as long as the concentration of nickel ions in the film of electrolyte adsorbed on the nickel substrate does not exceed about 0.0001 M. The passive chromium will function as cathode. This state is shown schematically in FIG. 1.

From the solubility product of nickel hydroxide, which is 2 10' according to F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, Interscience Publishers, New York, 1962, p. 733, it is calculated that any hydroxide ion concentration greater than about 10* M is sufficient to cause nickel ions to precipitate as Ni(OH) As the cathodic half of the corrosion reaction produces hydroxide ions, as shown in the equation given above, the hydroxide ion concentration will generally exceed this limitingvalue; nickel ions will be removedfrom the electrolyte in the precipitate, the nickel ion concentration will be less than 0.0001 M, and consequently the anode potential will be sufficiently negative to cause the corrosion reaction to proceed. This conclusion is in agreement with observations.

However, if a coating of gold overlies the chromium, the situation is altered, asshown in FIG. 2. The hydrogen overvoltage observed on the gold isfar lower than that on chromium. Consequently, the cathodic reaction is shifted from chromium to the gold. The nickel ions produced at the anode, and the hydroxide ions produced at the cathode, are now separated by an intervening face of passive chromium, which apparently takes no part in the cor- I'OSlOIl process.

In order to precipitate Ni(OH) the ions must diifuse' across the chromium face. As the passive film on this face is hydrophobic, the film of moisture thereonwill tend to be discontinuous. If moisture were completely repelled by the passive chromium, the corrosion cell would lack electrolytic contact, and corrosion would stop; this is quite conceivable. But more generally, there will be a thin, irregular film through which ionic diffusion is very slow. Because of this impediment to diffusion, nickel ions will tend to accumulate at the anode. The concentration may exceed 0.0001 M, especially in view of the very minute quantity of electrolyte involved. Consequently, the potential of the nickel anode would become too positive to cause the corrosion reaction to proceed. The high ohmic resistance to be expected in the film of electrolyte lying on the passive chromium would also be an important factor. My current thinking is that the corrosion reaction is strongly inhibited in this manner, as observations discussed in Example 17 apparently confirm. I

In summary, the peculiar inhibition of corrosion observed when gold is deposited on chromium is apparently due to the fact that the cathodic and anodic regions are separated by a passive chromium face which impedes both ionic diffusion and electrolytic conduction, shifting the corrosion potentials to levels which are insuflicient for corrosion to proceed. The hydrophobicity of passive chromium tends to interrupt couple action between sub strate and the gold coating by causing the electrolyte film to break away. Note that in gold coatings on stainless steel, no such eifects could be observed, since there is no base metal to be exposed and function in the manner described.

CORROSION TESTS The following corrosion tests indicate the unexpected corrosion resistance conferred by chromium-gold duplex coatings on substrates which possess little intrinsic resistance to corrosive atmospheres:

EXAMPLE 1 Panels were prepared as follows: 3" x 6" panels were cut from bright finished copper sheets thick. They were cleaned by first swabbing with carbon tetrachloride, then boiling in a cleaning bath containing 45 g./l. sodium orthosilicate and 1 g./l. sodium lauryl sulfate, and finally scrubbing with pumice and a soft brush under flowing Water. After cleaning, the panels did not exhibit water break. They were dipped momentarily in 1:1 hydrochloric acid rinsed in flowing water, and immediately plated.

Three of these panels, designated A, B, and C, were then plated directly with chromium in a solution containing 400 g./l. CrO 4 g./l. H 50 at 50 amp./ sq. ft., 35 C., for 25 minutes, to give 0.00002" chromium. The plating cell (which was similar to that for nickel plating described below) was equipped with a lead anode about A smaller in length and width than the cathode.

Six of these panels, designated D, E, F, G, H, and I were nickel plated instead of being'chromium plated at this time. v

' The plating bath consisted of Udylite proprietary bright nickel, the pH being held between 2.8 and 3.4, current density 50 amp./ sq. ft., at 40 C. Time and current were measured with a calibrated ammeter and stop-watch, to give 0.0003" nickel calculated at 100% efficiency. The nickel anodes measured 2 x /8", placed in a' glass cell 3%" x 7 /2" x 6" (inner dimensions). The anode was I placed at one end of the cell, and the cathode rested against the glass wall at the opposite end. Although pressed against the tank wall, the obverse faces of the panels were nevertheless covered with nickel of undetermined thickness.

Upon removal from the nickel bath, the panels were rinsed in flowing water, dipped momentarily in 1:1 hydrochloric acid, rinsed, and panels D, E, and F were then plated with chromium in the same manner as that described for panels A, B, and C.

The six panels A, B, C, G, H, and I, but not the three panels D, E, and F, were then gold plated. Immediately after plating, each of the six panels was rinsed, and treated cathodically in hydrochloric acid at 40 amp./ sq. ft. for 12 seconds. After this depassivation, the panel was rinsed in 0.5% hydrochloric acid, and plated in a gold bath containing 6 g./l. commercial grade KAu(CN) 1.0 g./l. potassium ferrocyanide, g./l. KCl, and enough HCl to bring the pH to 2.8. Temperature was held between 68 and 80 C., the current density was 3 amp/sq. ft., and the time 74.5 sec., to give 0.00001" of gold.

After gold plating, the samples A, B, C, G, H, and I were rinsed in water and dried with a soft paper towel. They were then placed in a salt spray cabinet operated according to Proc. A.S.T.M. 53, 267 (1953), using brine. The following results were obtained:

Hours to first Hours to estimated Specimen visible corrosion 10% corrosion Cr-Au A 48 84 B 48 64 C 64 96 Ni-Cr D 8 24 E 12 16 F 12 24 Nl-Au G 16 24 H 12 24 I 24 32 The above results show that the chromium-gold plated panels outperformed the nickel-chromium panels by better than four fold, although the total thickness of the plated coating was only 0.00003" compared to 0.00032" for nickel-chromium. Chromium alone does not give good protection, even in much greater thicknesses (H. Silman, Metal Ind. (London) 78 (17), 327 (1951). The Cr-Au panels far outperformed the conventional Ni-Au coatings. The unique effectiveness of the chromium-gold duplex coating is striking.

EXAMPLE 2 A set of six samples, duplicates of the panels A, B, C, D, E, and F just described, were prepared and submitted to test in a humidity chamber containing a 2% solution of NaHSO to which a little hydrochloric acid was added; the amount of S0 in the cabinet was not determined but was estimated at 0.5 to 3%. After 72 hours, circular areas of black corrosion products surrounding deep pits were observed on the Ni-Cr plated panels; the three Cr-Au panels were unaffected after 200 hours, although the coating was less than one-tenth as thick.

EXAMPLE 3 Steel panels of the same dimensions as those described above were plated in the laboratories of the Udylite Corporation with 0.0002" of copper, 0.0003" of bright nickel, and 0.00002" of chromium. A duplicate set was plated with 0.00002" chromium directly on the steel. After being received in atfiants laboratory, the panels were de passivated and gold plated, following the procedure outlined in Example 1. These panels were submitted to the CASS test (A.S.T.M. Standard B638-61T). Three Cu-Ni- Cr panels showed marked corrosion after hours. Three Cr-Au panels showed'corrosion after 48 hours. They outlasted the conventional coating by two-'to-one, although the platingwas less than one-tenth as thick.

EXAMPLE 4 Two sheet silver panels of the same dimensions were plated with 0.00005" gold. Two others were plated with 0.00004" of chromium, followed by depassivation and 0.00005" of gold, as outlined above; Neither set of panels was affected by 48 hours exposure in the humidity chamber containing sulfur dioxide;'accordingly, ammonium sulfide was introduced. The panels which had not'received chromium were badly stained and pitted within 12 hours after the sulfide was added. The panels with the chromium undercoating did not show corrosion for 72 hours.

EXAMPLE 5 in each set.

EXAMPLE 6 Two brass panels were plated with 0.00002", of chromium, depassivated, and platedwith 0.00005 gold. Two other panels, were plated with 0.0002 of Watts nickel and 0.00005" gold. These sets were attached to the front and rear bumpers of a car driven in a semi-urban area during the winter months. After seven weeks, gold was no longer visible on the nickel, and only small areas were golden after a light polish. The panels which had chromium underlying the gold were restored to their original appearance by a light scrubbing with a soft brush, and water.

EXAMPLE 7 Specimens which were coated with 0.0005 inch of zinc and 0.00002 inch of gold were badly corroded, with gold peeling off, after one hour in the salt spray. Specimens with 0.0002 inch of manganese and 0.00002 inch of gold suffered complete loss of gold after two hours in the salt spray. Furthermore, the zinc-gold coatings were substantially destroyed in six hours in the humidity cabinet. These observations indicate the properties of chromium as an undercoating. Neither zinc nor manganese are readily passivated under these conditions, and coatings based on them have virtually no value.

As nickel is readily passivated, it is surprising that the nickel-gold duplex coatings did not perform better in these tests, as well as in actual service. Perhaps the potential required to passivate them is so high that it is not always provided by the cathodic reaction on the gold, so that passivation is-not always achieved. It is well known that chromium is much more readily passivated than is nickel. Passivation of nickel would not be expected in the presence of chlorides, as in the salt spray tests. In the tests, there is a tendency for nickel to be corroded out from under the gold; this is not observed with the chromium-gold duplex coating.

EXAMPLE 8 Brass panels were plated first with 0.001 inch of bright nickel and then 0.00004 inch of chromium. Three groups were then gold plated, some with 0.00005 inch, some with 0.000025 inch, and some with 0.000005 inch of gold. They were submitted to the copper accelerated salt spray (CASS) test; after 96 hours, corrosion was notably more widespread on the panels with the thingold coatings. However, the improvement with thicker coatings was not at all in proportion 'to theincrease in thickness, evidencing that the protective action is primarily due to the self-passivation effect. It it-were due to the closing of pores as the thicknessof gold is increased, a more pronounced improvement would be expected, for the gold thickness is increased by a factor of ten. 7

9 EXAMPLE 9 Two sets of brass panels were plated. One received 0.00002 inch of chromium, the other 0.00004 inch of chromium. Each then received 0.000005 inch of gold. They were submitted to the CASS test. Initial corrosion appeared for both sets at 84 hours. This indicates that the thickness of the chromium undercoating is not a significant factor in the performance of the duplex coating. It also indicates that the presence of microcracks in the chromium does not have a major influence on the performance of the duplex coating, since it is well known that these cracks develop extensively when the thickness is increased beyond 0.00002 inch.

EXAMPLE 10 Copper panels were plated with 0.00002 inch of chro- I mium and 0.000005 inch of gold. They were tested by salt spray and the results were substantially identical with those cited above.

EXAMPLE 1 1 Brass panels were plated with 0.001 inch of silver. One set was then coated with 0.0001 inch of chromium, another with 0.00004 inch of chromium, and a third set received no chromium.

All of these panels were then plated with 0.00005 inch of gold. They were submitted to 48 hours exposure to the humidity chamber, with the addition of sulfur dioxide. Following this, they were exposed to the fumes from ammonium sulfide. The panels which had no chromium A steel sample was plated directly with 0.00002 inch of chromium and 0.00005 inch of gold, and submitted to saltspray testing. The first evidence of rust appeared after 52 hours; it spread to cover about 15% 0f the panel after 64 hours.

EXAMPLE 13 A suitable coating schedule for gold-plated steel objects is:

' Inches Copper 0.0001 Nickel 0.0002-0.0005 Chromium 0.000020.00008 Gold 0.0000010.000005 The heavier the nickel and chromium undercoats, the more pronounced is the ultimate corrosion resistance. For

- a brass object, the copper may be omitted. For white metal, it is desirable to use at least 0.0002 inch copper. The minimum practical thickness of chromium is about 0.000005", because thinner coatings are so full of discontinuities that the full protective effect is not developed. The maximum practical thickness for most uses is about 0.0002", because thicker coatings develop stresses and cracks which do not provide a satisfactory surface for depositing precious metal; still thicker coatings have considerable corrosion protection in their own right, but are much more costly than thin duplex coatings.

The thickness of gold applied depends on the intended use. For ornamental coatings for indoor exposures, about 0.000002 inch is adequate. But if the finished object is likely to be subjected to wear, as are doorknobs or tableware, or is to be used outside where an occasional burnishing will be applied to remove accumulated soils, substantially thicker gold coatings are advisable, e.g., up to 0.005 inch or so. Heavier gold coatings may be applied over the preliminary gold coating from the conventional alkaline cyanide or chloride type baths. Instead of pure gold, one or another of the well-known gold alloys, such as Hamilton gold, may be used as coatings to provide a harder, better wearing surface, provided that the alloy is not subject to tarnishing. Plating with gold alloys gives a thicker coating than when the same amount of gold in its pure form is used, and where a coating of a specific thickness is desired, less gold will be needed if an alloy is used. Plating a gold alloy over a thin coating of pure gold produces various tones and Shades of gold color depending upon the alloy and its thickness and provides a more durable product.

Instead of electroplating the gold, chromium-surfaced articles may be dipped into a displacement-and-chemicalreduction gold bath such as the proprietary Atomex bath of Engelhard Industries, Inc. to give a good adherent and corrosion-resistant coating. It is important, however, to use a pH between 2.5 and 4.5, which is contrary to the instructions of the manufacturer, who recommends a pH not below 5.5. The manufacturer does not indicate that the bath can be used to plate gold on chromium.

It is evident from these examples that the duplex coatings use such small amounts of precious metals that they are not excessively costly, but may be considered for many applications. On the other hand, because the duplex coating is extremely thin, it is not serviceable where the surface is subject to considerable abrasive wear, or where it is apt to be nicked or otherwise damaged mechanically. Depassivated chromium is so extraordinarily active that it offers special difficulties in plating precious metals thereon. In alkaline baths, or baths containing an active oxidizing agent, repassivation occurs before the deposit is obtained; the coating then has little adhesion to the substrate. Many soluble compounds of the noble metals are active oxidizing agents; it is necessary to select com pounds which do not exhibit much oxidizing power. Furthermore, on active chromium, noble metals tend to be deposited by displacement; this, too, interferes with good adhesion, and occasionally with the integrity of the coating.

The chromium itself is deposited directly on the substrate, or on an undercoating of copper or copper-nickel or of nickel, made in the conventional manner. Any suitable chromium plating process may be used. The current density, temperature, and time of plating are preferably chosen to give a coating thickness of 0.00002 inch.

After the chromium-plated surface is thoroughly rinsed, it is depassivated by dipping it for 2-150 seconds in dilute hydrochloric, sulfuric, or acetic acid. Typical times are 5 to 30 seconds in 5 to 10% hydrochloric acid at room temperature. It may be helpful to pass current through the depassivating bath, making the chromiumplated surface cathodic, typically at current densities of 10 to amperes per square foot. The electric current makes the chromium surface cathodic, hastens depassivation, and helps to prevent excessive dissolution of the chromium by the acid, for as soon as the chromium is depassivated, it is actively attacked by the acid.

The depassivation treatment should be regulated so that the thickness of the chromium coating is not greatly diminished. A treatment of 5 seconds in 15 by volume hydrochloric acid at room temperature, with 30-40 amperes per square foot cathodic current, is usually satisfactory.

It is also possible to use 5-10% sulfuric acid, 1525% phosphoric acid, or 1030% acetic acid. In fact, these last acids are preferred to hydrochloric, since they do not attack the chromium quite as rapidly.

After depassivation, the surface is kept wet with a rinse solution of 0.1-0.5 acid, to prevent repassivation by air when the article is transferred to the plating bath.

If the depassivated chromium surface enters an alkaline bath, it is instantaneously repassivated, and subsequent electroplates do not adhere. Hence, it is necessary to use acidified baths. However, .a conventional bath, such as an to add them as simple salts. Iron (II), cobalt, or nickel sulfates may be added in amounts ranging from 0.1 to 10 grams of metal per liter, and other soluble salts of these metals work equally well. Furthermore, salts .of tin, cadmium. and Zinc are occasionally useful.

The addition of alloying metals to gold cyanide baths is not new in itself, but it has heretofore been practiced in order to modify the physical or mechanical characteristics of the gold deposit, as to harden it, or to modify its color. In this application, however, the added metal apparently exerts a beneficial effect in preventing the formation of immersion deposits on the depassivated chromium surface, which is, of course, one of the most active metals. This result is entirely unanticipated, and the way in which it works is not understood. Since the added metal may co-deposit with the gold, it is to be understood that the term gold coating may include such deposits, containing up to about 2% of an alloying metal.

The following examples are given to illustrate the process. Many other variations can be made, in accordance with the discussion above.-

tgemperature and time are adjusted to suit the concentration of the act B. Rinsing: Hydrochloric acid, percent 0.3 0. 01-1.

0. Gold plating:

KA11(CN):, g./l 6 2. 0-50. 0 NiSO4-7H2,0 g./l 4 0. -45 HCl to pH 3.0 2. 5-4. 0 Current density (amp/sq. it.) 3 1-25 Temperature F.) 130 90-160 EXAMPLE 15 Permissible Preferred range A. Depassivation:

Sulfuric acid, percent 5 3-20 Temperature C F.) 90 70-100 Time (sec.) 6 2-20 Current density (ampJsq. it.) 30 -100 B. Rinsing:

Sulfuric acid, percent 0. 5 0. 01-1 0 C. Gold plating:

KAu(CN)2 (g./l.) 10 2. 0-50 H P04 to pH 3.0 2. 5-4. 0 COSO4.7H2O (gJl) 4 0. 5-45 Current density (ampJsq. ft.) 3. 5 1-25 Temperature C F.) 145 90-160 (Nickel or ferrous sulfate can be substituted in equal parts for tge) cobalt sulfate, as can other Salts of the metals mentione EXAMPLE 16 So far as I am aware, the so-called chemical or displacement gold plating baths. when operated in the ordinary way, on depassivated chromium surfaces, produce non-adherent gold deposits. These baths are specially formulated to have the capacity of forming reguline metallic gold deposits instead of the dark, spongy immersion deposits which are formed in ordinary salt solutions of the metal to be plated, and which are especially objectionable when it is attempted to deposit gold on chromium in the conventional manner from the ordinary electroplating bath. With some of these immersion plating baths, however, modification of the pH in accordance with my invention, enables the deposition of gold on chromium with adequate adhesion, even though no alloying metal is present.

A proprietary immersion gold plating bath, Engelhard Atomex, believed to be one specified by H. W. Robinson in US. Patent No. 3,230,098, was heated to 90 C. When depassivated chromium-plated specimens were immersed, smooth, bright gold deposits were formed; but when the specified pH range (55-14), was used, the deposits could be lifted from the chromium by the fingers. However on adding sufficient hydrochloric acid to lower the pH of the bath to 4.5 which is in my range of 2.54.5, the gold deposits became adherent, and were apparently fully equivalent to those produced by electrodeposition. This bath is believed to have had approximately the following composition: KAu(CN) 5 g./l. ammonium citrate 20 g./l. and chelating agent, such as ethylene diamine, 25 g./l.

EXAMPLE 17 To indicate whether nickel ions actually accumulate where nickel is exposed through discontinuities in the chromium-gold duplex coatings in accordance with the suggested mechanism, whereas they are largely precipitated in the absence of the intermediate chromium layer, two copper panels were plated as described in Example 5. One panel received nickel, chromium, and gold coatings, in that order. The second received only nickel and gold coatings. After plating, the panels were immersed for an hour in a 2% solution of NaCl. They were then rinsed and allowed to dry. The presence of nickel corrosion products was detected by applying strips of filter paper wet with a solution containing 0.1 g. of dimethylglyoxime in 20 ml. alcohol and ml. water. After five minutes, the papers were removed and examined.

The paper removed from the nickel-gold duplex deposit showed a few very faint pink spots, indicating that only traces of soluble nickel salts existed on the metal surface. The test paper removed from the nickel-chromiumgold triplex deposit showed extensive but irregular stains, which though but weakly pink, were nevertheless much deeper in shade than the colors from the duplex deposit. Evidently the accumulations of soluble nickel salts were much more extensive on the triplex coating than on the duplex.

Fresh wet test paper was then applied to the same specimens, which were then covered with a wet cloth and allowed to stand in this condition overnight. Subsequently it was found that the test paper removed from the nickel-gold duplex deposit was quite generally stained with relatively strong pink, whereas the test paper removed from the nickel-chromium-gold triplex deposit was scarcely more colored than that obtained in the first test, which had lasted for only 5 minutes. Evidently the duplex deposit had furnished substantially more nickel by corrosion than had the triplex deposit, which had a chromium intermediate layer.

These observations appear to support the supposition that in the presence of an intermediate chromium layer between coatings of nickel and of gold, the corrosion products of nickel tend to remain in soluble form and suppress further corrosion, at least to some extent; whereas in the absence of chromium, they tend to be precipitated as insoluble matter, presumably the hydroxide, and thus allow corrosion by local cell action to proceed uninhibited by accumulating nickel ions.

It is remarkable that those conditions which enable nickel corrosion products to appear earlier with the nickelchromium-gold triplex coatings than with the nickel-gold duplex coatings actually seem to lead to later suppression of the corrosion reaction, whereas those in which initial detection of corrosion products is delayed (that is, in the duplex coatings) lead eventually to far more extensive corrosion.

This is apparently the consequence of precipitation of the corrosion products in early stages of corrosion of the nickel-gold duplex coating, though resistance to local cell action resulting from the hydrophobicity of the intermedi' ate chromium layer in the triplex coating, as postulated above, is probably also a factor. Determination of the prevailing pH in the film of moisture on the surface of the metal coatings would be revealing, but no methods for indicating pH in such submicroscopic quantities of moisture are available.

The order of protection conferred on the base substrate by the chromium-gold duplex system, whether it is applied over an undercoating of nickel to form a triplex coating, or is deposited directly on the substrate, is so far beyond that afforded by other types of coating systems (at least, those which are cathodic to the substrate metal), that it is believed to represent an entirely new protective coating system, of a type hitherto unknown. It is 'for this reason that a new protective mechanism, here termed self-passivation, is proposed, and has been experimentally confirmed to the extent detailed under this example.

To those skilled in theart to which this invention relates, many changes in construction and widely difiering embodiments and applications of the inventionwill suggest themselves without departing from the spirit and scope of the invention. The disclosures and the description herein are purely illustrative and are not intended to be in any sense limiting.

I claim:

1. A method of plating a base metal article with gold, comprising the steps of (a) providing a chromium plate 0.00002 to 0.0001 inch thick on said base metal article,

(b) depassivating the surface of said chromium plate by subjecting it to a cathodizing current in an acid bath,

(c) placing the depassivated chromium surface in an aqueous acid gold plating bath having a pH between 2.5 and 4.5 and containing from 0.1 g./l. to g./l. of a base alloying metal in the form of a soluble salt, and

(d) electroplating said depassivated surface with a layer of 0.000002 to 0.00005 inch thickness of gold, from said gold plating bath.

2. The method of claim 1 wherein said base alloying metal is chosen from the group consisting of iron, nickel, and cobalt.

3. A plating process for gold-plating a metal article having a copper-containing surface comprising electroplating thereon from about 0.0002 inch to about 0.0005 inch nickel, electro-plating from about 0.00002 to about 0.00008 inch chromium on the nickel, depassivating said chromium coated article, by subjecting it to a cathodizing current in an acid bath, placing the depassivated chr0- mium coated article in an aqueous acid gold plating bath having a pH between 2.5 and 4.5 and containing. from 0.1 =g./l. to 10 g./l. of a base alloying metal in theform of a soluble salt, and then electro-plating gold on the chromium, from said gold plating bath.

4. The plating process of claim 3 wherein thefmetal article comprisesa non-copper base on which a coating of at least about 0.0001 inch copper has been electroplated.

5. A process for plating gold on a chromium surface, comprising the steps of (a) depassivating said chromium surface in a 3% to 20% hydrochloric or sulfuric acid solution at a temperature of to F. for two to twenty seconds at thirty to forty amperes per square foot cathodic current,

(b) rinsing said depassivated surface in a 0.01 to 1% solution of an acid chosen from the group consisting of hydrochloric and sulfuric acids, and

(c) electroplating said depassivated surface in an aqueous acid gold bath containing 2 to 50 grams per liter of potassium gold cyanide and 0.1 to 10 grams per liter of metal in the form of 'a soluble salt of iron, nickel, or cobalt, at a pH of 2.5 to 4.5 at a current density of 1 to 25 amperes per square foot at the cathode, and at a temperature of 90 to F.

References Cited UNITED STATES PATENTS Re. 20,306 3/1937 Lukens 20446 XR 1,892,051 12/1932 Gray 20446 XR 2,039,328 5/1936 Lukens 20446 XR 2,133,995 10/1938 Lukens 20446XR 2,133,996 10/1938 Underwood 20446 XR 2,491,126 12/1949 McGill 20434 2,886,499 5/1959 Schaer et a1. 20441 2,967,135 1/1961 Ostrow 204 -43 3,104,212 9/1963 'Rinker et al. 20446 3,214,292 10/1965 Edson, 106--1XR OTHER REFERENCES 0 JOHN H. MACK, Primary Examiner G. L. KAPLAN, Assistant Examiner U.S. Cl. X.R.

Images