<<

Electrodeposition of - Alloys from Arnrnoniacal Electrolytes By B. Sturzenegger Department of Chemical , Swiss Federal Institute of , Zurich and J. C1. Puippe Werner Fluhmann A.G., Diibendorf, Switzerland

A viable technique for the electrodeposition of palladium-silver alloys could find application during the manufacture of electrical contacts. Previously such a process has not been available but some alloys with useful physical properties can now be deposited from an ammoniacal system, while a better understanding of the mechanisms involved has been gained during the work reported here.

Applications of palladium and palladium Deposits from -based electrolytes alloys in the electrical contact field have showed a broader compositional range, up to 70 recently been reviewed by Antler (I). Among per cent palladium (2, 7-12), the best coatings these materials, palladium-silver alloys, from this type of bath containing 2 to 10 per particularly the 60 per cent palladium-40 per cent palladium. A bath formulation based on cent silver composition, are well established in palladium and silver ammino-hydroxy salts the wrought form, as claddings, weldments or gave semi-bright deposits with palladium con- inlays, with advantages over the pure in tents from 15 to 85 per cent and good cathode terms of cost and durability. More recently, efficiency of 85 to 95 per cent (13-16),while interest has extended in the direction of bright coatings have been reported from an electrodeposited coatings of palladium and electrolyte based on amino-acid complexes alloys as economic substitutes for , (17, 18). of palladate and argentate and in this context it is clearly attractive to con- salts have been stated to give semi-bright coat- sider the possibilities of producing palladium- ings with palladium content between I o and 60 silver alloys by electrodeposition, a technique per cent, with thicknesses up to 15,um being which would permit, for example, the applica- obtainable (I 9). tion of coatings to pre-formed contact fingers Work has also been carried out using con- and other components in cases where the inlay centrated halide baths with chloride to technique is not practicable. While, however, a permit increased solubility of silver (20-22). considerable amount of research work has been This type of is very aggressive to carried out to this end, notably from Russian due to vigorous displacement reactions; sources, it would appear that no viable process however, good deposits were reported, with has yet been developed. palladium contents ranging from 30 to 60 per Early attempts used a solution based on cent. Attempts to deposit alloys from complexes of the two metals (2,3), but ammoniacal solutions of nitrito-complexes were the alloys produced, with a palladium content unsuccessful (2), but when complexes of 20 to 22 per cent, were of poor appearance are employed it has been reported that the full and the cathode efficiency was very low. range of compositions can be obtained Improvements were later obtained in terms of simply by adjustment of the current increased cathode efficiency by optimising (23). Later reports claim the production of process parameters and bath formulation (4-6). bright, pore-free deposits from this type of

117 PVC jacket-- Cathode __ Anode __

ENLARGED VIEW ON A-A Outlet Teflon sleeve DIMENSONS IN rnm

Fig. I The elrctrolytes were studied in the cell shown here. Deposits were made onto a rotating cylindrical cathode, the anode being a con- centric platinised cylinder

solution, with thicknesses up to 20 pm (24). was prepared for plating by an initial polish The object of the present work was to study with I pm alumina paste, followed by cathodic the mechanism in palladium-silver alloy deposi- degreasing at I o mA/cm’ in an alkaline cyanide tion in order to gain a better understanding of solution, and a final activating dip in con- the influence of process and compositional centrated sulphuric acid, with intermediate parameters on alloy composition and properties. water rinses. Immersion in the electrolyte was For this purpose the ammoniacal solution with made under an applied voltage. palladium and silver was selected in Deposition was carried out at room view of the possibility of obtaining a wide range temperature (22OC) for times selected to of alloy compositions. produce coating thicknesses in the order of 2pm. The applied current density (j&I was Experimental Conditions varied between 5 and 40mA/cm2, and the The electrolyte compositions studied are cathode rotational speed between o and 10 given in the opposite. revolutions per second. Deposition was made on a copper cylinder (10mm diameter, 10 mm high) attached to a Deposition Mechanism Studies Teflon shaft, the rotational speed (a)of which Polarisation curves for (a) a solution contain- could be continuously adjusted from o to 10 ing 20 g/l of palladium only, and*(b) a solution revolutions per second. The anode was a con- containing 20g/l of palladium and 2gfl of centric cylinder of platinised titanium, which silver are shown in Figure 2. From these it is was itself provided with a further jacket of PVC seen that silver begins to deposit at potentials for temperature stabilisation. The electrolytic less negative than for palladium, and that cell is shown as Figure I. The cathode surface palladium deposition occurs when silver is

Platinum Metals Rev., 1984, 28, (3) 118 being deposited under limiting current condi- tions. The large plateau of curve (b) is associated with the limiting current of silver deposition, as will be quantitatively evidenced later. The fact that palladium, a more metal than silver, is deposited only at more negative potentials than the latter, is explained by the much greater stability of the palladium- ammino complex, the relevant stability con- stants being as follows (25):

-200 -400 -600 -80C ELECTRICAL POTENTIAL, Ev,,*p,A,c,,rnV Fig. 2 Polarisation curves for palladium The influence of current density and and palladium-silver electrolytes. rotational speed of the cathode on the composi- Voltage scanning speed: I mV/s (a) Palladium: 20 g/l tion of the alloys was studied for electrolytes (b) Palladium: 20 g/I; silver: 2 g/l containing I and 2g/l of silver. The results, shown in Figures 3 and 4, are consistent with 40 the hypothesis of a mass transport-controlled r codeposition of silver, in that the silver content increases with agitation rate at constant current density, and decreases with increase in current density at constant agitation rate. As shown by comparison of Figure 3 with Figure 4, the increase in the silver content of the deposit is proportional to the increase in silver concentra- tion in the electrolyte, which is again consistent with the mass transport model for silver 2 4 6 81012 ROTATIONAL SPEED.n,rev/s incorporation. Fig. 3 Silver content of deposits a5 a Figures 3 and 4 also show the experimental function of rotational speed of cathode conditions under which bright, semi-bright, and current density for electrolyte 1. and matt deposits are obtained. The thickness 0 bright deposits; 0 semi-brighl deposits; matt deposits of the deposits was 2pm. In general, deposits with a silver content below 25 atomic per cent are bright, while higher silver contents are mode for palladium, while at higher rates of associated with semi-bright or matt coatings. silver incorporation the morphology of the This may be explained on the basis that at low deposit is progressively affected by the silver incorporation rates the deposit structure crystallisation of silver. Since silver deposition is essentially governed by the crystallisation takes place under mass transport controlled

Electrolyte Compositions

1 2 3

Palladium, as Pd(NH,l,(NO,), 20 911 20 911 20 911 Silver, as Ag(NH,),(NO,) 1 911 2 911 3 gA pH (adjusted by gaseous NH,) 11.5 11.5 11.5

Platinum Metals Rev., 1984, 28, (3) 119 conditions, there is a tendency for the forma- that the current is under full mass transport tion of dendrites and powder, leading to control, providing a quantitative demonstration rougher deposits. The morphology of matt, of the fact that similar conditions apply to the semi-bright and bright deposits with various incorporation of silver in the palladium-silver silver contents is illustrated in the scanning alloy. This being so, a model may be derived for electron micrographs of Figure 5. the calculation of alloy composition as a func- The variation of current efficiency with tion of deposition parameters, as given below. silver content is shown in Figure 6. The During the electrodeposition of palladium- decrease in efficiency with increasing silver silver alloy, the total current density is due to content can be explained by the increasing contributions from the deposition of palladium, roughness of the deposits, leading to enhance- silver and . ment of hydrogen evolution at asperities and it01 = iPd + jAg + iH [ivl hence to a decrease in current efficiency for The molar-ratio of silver in the deposit is: metal deposition. The current efficiency was calculated on the basis of Faraday's law from xAg = jAg/bAg + 112 jl'dd) [VI the metal contents of deposits, as determined by Assuming a I 00 per cent current efficiency and atomic on solutions prepared by combining Equations [iv] and [v], dissolution in , as follows: xAg = 2 iAg/(iAg + jd [Vil

XAg = 2/(I + J) [vii]

where J =& = 'to' [viii] LA^ 85.5 nn'ss9.CAg With Equation [vii] one can thus predict the silver content in the deposit for different total = 100 [ii] current , different silver 110, in the electrolyte and different hydrodynamical where IAg and I,, are partial currents for silver conditions. and palladium, Ilo, is total current, mAgand mPd For not negligible drops in current efficiency, are weights of silver and palladium deposited, Equation [vii] has to be rewritten as follows: AAg and A,, are the atomic weights of the XA~=2/( I + J . CE) [vii'] metals, F is the Faraday constant and t is the time. From the silver content of deposits, and on the assumption that silver is deposited under mass transport control, the limiting current density for silver deposition can be calculated from the results of Figures 3 and 4, and is plotted as a function of rotational speed of the cathode in Figure 7. The limiting current density values so obtained fit the following relationship: 4 jhg= 85.5 f2n.sr9. CA~ [iii]

That is to say, the limiting current density is 2 4 6 81012 directly proportional to the silver ROTATIONAL SPEED, n, rev/s CAgin the electrolyte and is a function of 12°.ss9. Fig. 4 Silver content of deposits as a An exactly similar relationship between limiting function of rotational speed of cathode and current density for electrolyte 2. current density and rotational speed was found 0 bright deposits; 0 semi-bright in separate experiments with a ferri- deposits; matt deposits ferrocyanide system, where it is well known

Platinum Metals Rev., 1984, 28, (3) 120 Fig. .5 Scanning electron micrographs of palladium- silver alloy deposits from rlectrnlytc 3 (palladium 20 g/l; silver 3 g/l).

(a) R=0 rev/s; ,i =5 mA/cm2; Ag = 52.5 at.% (matt)

(b) R=0 rev/s; ,i = 15 mA/cm2; Ag = 29.3 at.% (srmi-bright )

(c) 0 = 0 rev/s; .i = 30 mA/cm*; Ag = 20.8 at.% (bright)

(d) 0 = 3 rcv/s; ,i = 30 mA/cm2; Ag = 33.0 at.% ( semi-bright)

Platinum Metals Rev., 1984, 28, (3) 121 treatment at 280mV in a solution of 200g/l copper sulphate, g/l sulphuric acid, to leave 100. 50 + an isolated foil for measurements. A series W U of foils was produced in this way under the K 90. W following conditions: > 0 U .O 0 Electrolyte I: jlOl= 5, to, IS,ZO dcm’ -$80- .. . . u m. Electrolyte 2: j,,,= 10, 15, 20 mNcm2 LL . ,“ 70. . Alloy Composition Alloy composition as a function of current density under the above conditions is shown in 10 20 30 40 ATOMIC PER CENT SILVER Figure 9, together with analogous results Fig. 6 Current efficiency for metal pertaining to coatings produced on the copper deposition as a function of silver content of deposits. electrodes used in the earlier work, with zero 0 bright deposits; 0 semi-bright rotational speed. From the results it is clear that deposits; matt deposits the change from the one type of cathode to the other has no significant influence on the Equation [vii] and the experimental results for composition-current density relationship. R > 2/s are plotted in Figure 8. Agreement between calculated and measured values was very good. Only small deviations occur for experi- The Knoop hardness of coatings was deter- mental values corresponding to matt deposits. mined as the mean of five measurements at various locations on the surface of 12pm Physical Property Measurements deposits at a load of 15p, with indentation For measurement of properties such as of 30 seconds. It has been experimentally hardness and specific resistance it was necessary determined that the alloy thickness of I 2 pm to produce alloy coatings of thickness at least was sufficient to avoid any influence of the I o pm and of good surface finish, both on flat substrate on the Knoop-hardness values. Figure substrates (for hardness) and as isolated foils 10 shows the hardness of deposits from the two (for specific resistance). Since, at thicknesses in electrolytes as a function of composition. Since excess of about 2pm, coatings tended to both curves are more or less parallel, it appears develop cracks due to internal stress associated that the hardness is independent of deposition with hydrogen occlusion, and also to deteriorate conditions and increases with increasing silver in surface finish, a multi-layer technique was content. adopted for this purpose, in which coatings were built up in successive increments of 2 pm, Specific Resistance with intermediate of the cathode with Results of measurements of specific I pm alumina. Thus it was possible to produce resistance by the four-point probe method bright, crack-free coatings up to I 2 pm thick. (ASTM F390-78) are shown in Figure I I. As Deposition was made on both sides of brass for hardness, this property increased with plate cathodes (20 x 70 mm), which were increasing silver content, being practically masked to expose a total area of 16cm’ and independent of electrolyte composition and suspended between platinised titanium anodes bath conditions. in an open vessel, with stirring by natural con- vection. To produce unsupported foils as required, the coating was removed from one Contact resistance measurements were made side of the plate by grinding, and the support- on 2pm coatings by the method according to ing substrate was then dissolved by anodic the ASTM B667 specification, the test surface

Platinum Metals Rev., 1984, 28, (3) 122 >

I 1 2 3 4 5 678910 ROTATIONAL SPEED,IL. mv/s 4 8 12 16 20 24 CURRENT DENSITY, J, dimensionless

Fig. 7 Limiting current density for silver Fig. 8 Silvrr content in the deposit as a as a function of the rotational speed of funrtion of the dimensionless current the cathode. density J. ( I ) Electrolyte I ; (2)Electrolyte 2 - according to equation [ vii] ; 0 experimental values

4 50

VI 9 x I400 v; W z E 2 350

10 20 30 40 CURRENT DENSITY, 1. mA/cm2 300 5 10 15 20 25 Fig. 9 Silver content of deposits as a ATOMIC PER CENT SILVER function of current density for cylindrical (0) and vertical plate Fig. I0linoop hardness of deposits as a ( x ) electrodes by natural convection. function of silver content. ( I ) Eleclrolyte I ; (2) Electrolyte 2 ( I ) Elertrolylc- I ; (2) Electrolyte 2

4

C 5401a? E w Y3 I- 2 / W a2 I- u 2 / 81 10 15 20 25 ATOMIC PER CENT SILVER 10 15 20 25 ATOMIC PER CENT SILVER Fig. I I Electrical resistivity of electrodeposited palladium-silver foils Fig. 12 Conlact resistance of palladium- as a function of the silver content. silver alloy deposits as a function of ( I ) Elertrolyte I ; (2) Electrolyte 2 silver content, for different loads

Platinum Metals Rev., 1984, 28, (3) 123 being flat, and the contact partner being a gold the mass transport model, but this technique hemisphere of diameter 3.2 mm. The results, permitted no increase in the silver content for for loads between 10 and IOOg, are shown in bright coatings as compared to d.c. conditions. Figure 12, indicating that, in contrast to From the applicational viewpoint, alloy hardness and resistivity, contact resistance is deposits so far obtained show useful physical relatively little affected by alloy composition, properties, despite the limitation of the silver but does depend strongly on the test load. content. From the process viewpoint, the very high content of the bath may con- Concluding Remarks stitute a potential disadvantage, certainly The present investigation has resulted in a necessitating effective ventilation. However, it better understanding of the mechanism govern- appears to have no adverse effect, as might have ing electrodeposition of palladium-silver alloys been expected, on the adhesion of deposits to from an ammoniacal system, in particular in copper and copper alloys, and, in total, provides the clear demonstration that silver incorpora- a system of good stability, with very simple and tion occurs under mass transport controlled easily monitored bath chemistry. conditions. On this basis it is possible to predict the silver content of deposits as a function of Acknowledgements the experimental conditions. The palladium- The work described in the present was silver system from ammoniacal baths provides, started under the direction of the late Professor Dr. Norbert Ibl, and was later supervised by Dr. 0. in fact, a very good example of the “regular” Dossenbach. Thanks are due to Mr. F. H. Reid for type of alloy deposition, according to the helpful advice at various stages of the work, and for classification of Brenner (26). assistance in the preparation of this paper, and to the firm of Werner Fliihmann AG, Dubendorf, for the Results of pulsed current plating studies, to provision of financial support and facilities for be reported elsewhere, are also consistent with analytical work.

References I M. Antler, Platinum Metals Rev., 1982, 26, (3), 14 N. T. Kudryavtsev, 1. F. Kushevich and N. A. I 06 Zhandarova, Russian Patent 291, 988; 1971 2 G. Grube and D. Beischer, Z. Elektrochem., I 933, 15 I. F. Kushevich and N. T. Kudryavtsev, 39, (31, 13 1 Elektrokhim. Osazhdenie Primen. Pokrytii 3 J. Fischer and H. Barth, German Parent 688, 398; Dragorsennymi Redk. Metal., I 972 I940 16 I. F. Kushevich and N. T. Kudryavtsev, Zashch. 4 N. P. Fedotev, P. M. Vyacheslavov, B. Sh. Met., 1974,m 78 Kramer and V. V. Ivanova, Zh. Prikl. Khim. 17 J. Culjkovic and R. Ludwig, German Patenr

(Leningrad), 1967, 40, (7) I474 29 445,538; I974 5 Nippon Co. Ltd., Japanese Parent 81, 18 H. J. Schuster and K. D. Heppner, German 156,790; 198I Patent 2,657,925; 1976 6 Z. G. Stepanova, Russian Parent 425,98 I ; I 974 19 N. T. Kudryavtsev, I. F. Kushevich, L. P. Vladimirova and L. V. Galkina, Russian Parent 7 N. P. Fedotev, P. M. Vyacheslavov, B. Sh. Kramer and G. K. Burkat, Russian Patent 22 I, 379Y 676; 1973 452; 1968 20 Extended Abstracts No. 309, Electrochem. SOC. Fall meeting, Denver, Oct. I 98 I, I I 8 B. M. Kramer, Zh. Prikl. Khim. (Leningrad), 21 A. K. Graham, S. Heiman and H. L. Pinkerton, I 973,469 (61, 1242 Plating, I 948,35, (I 2), I 2 I 7 9 B. M. Kramer, Zh. Prikl. Khim. (Leningrad), 22 Bell Telephone Laboratories Inc., US. Parent 1973,469 (10h 2326 432699671; 1981 10 B. M. Kramer, Zh. Prikl. Khim. (Leningrad), 23 V. E. Medina, US.Parent 3,053,741; 1961 1975~48,(I), 226 24 A. Ya. Pronskaya, S. I. Krichmar and S. D. I I B. Sh. Kramer, Obmen Opytom Radioprom., 197 I, Okhrimets, Dokl. Akad. Nauk SSSR, (I 1) Elekrrokhim., 1971,7,(6), 778 I 2 Kumamoto Prefecture, Japanese Patent 82, 076, 25 R. M. Smith and A. E. Martell, “Critical Stability 196; 1982 Constants”, Volume 4, Plenum, New York, I 98 I 13 N. T. Kudryavtsev, I. F. Kushevich and N. A. 26 A. Brenner, “Electrodeposition of Alloys”, Zhandarova, Zashch. Met., 1971,7,206 Volume I, Academic Press, New York, 1963,76

.Platinum Metals Rev., 1984, 28, (3) 124