A Novel Palladium Stripping Solution - Its Performance & Maintenance

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

A Novel Palladium Stripping Solution—

Its Performance & Maintenance


By H.K. Straschil, J.J. Maisano, I.V. Kadija, F. Humiec and J.A. Abys

Palladium finishes have become widely established in Although many procedures can be used, a chemical salt-
electronic and decorative applications. A reliable ing-out of the complexed precious metals has been found
stripping solution for rework and metal recovery has preferable for safe and efficient precious metal recovery from
not been commercially available, however. A three- the spent solution. The cyanide-containing supernatant liq-
component system has been developed that provides uid can then be treated by conventional means (e.g., reaction
satisfactory and economical processing of palladium, with sodium hypochlorite solution to form carbon dioxide
gold-flash palladium or palladium-nickel finishes. The and ammonia).
system is exceptionally selective for palladium and
palladium-nickel vs. copper, and can be applied in Prior Developments
rack, barrel or reel-to-reel operational modes. Formu- Palladium Stripping and Related Information
lation data, etch rates, performance, maintenance, Both electrochemical and chemical stripping methods have
metal recovery, and analytical procedures associated been proposed for palladium removal. E.B. Saubestre pro-
with the process are presented. Ways of recovering posed anodic stripping in a solution of NaCl and NaNO2 at pH
precious metal from the spent stripper solution are 4-5, 2-4 V;13 similar methods are also given by N. Hall et al.14
discussed, both for in-house recovery and for cost- It appears, however, that most electroplaters prefer chemical
effective shipment to outside refining. methods in which an oxidizing chemical is used to dissolve
the metal. T.A. Shankoff proposed the use of dichromate ion
In the fabrication of electronic devices and in the decorative as the oxidizing agent and chloride ion as the complexing
industry, palladium and its alloys are often used to replace agent.15 Concentrated sulfuric acid has been suggested as a
gold.1-6 They are sometimes finished with a thin overlayer of solvent for chemical stripping with nitric acid added as an
gold. Palladium and its alloys, notably with nickel or cobalt, oxidant, as well as an electrolyte for anodic stripping.
afford corrosion protection and superior wear, combined Another group of oxidants in use for precious metal strip-
with excellent electrical properties.7-11 Palladium is also use- ping includes “nitro-substituted aromatic compounds,” which
ful for providing a barrier against diffusion (e.g., between have been suggested for “dissolving metals of the class
gold and nickel). It is economically favorable because it costs consisting of nickel, copper, cadmium, silver, zinc, and gold
significantly less than gold. But even though the metal cost is by chemical action alone, etc.” in 1949 by R. Springer and
reduced, a reliable palladium and palladium alloy stripping W.R. Meyer.16 This class of oxidants is used in combination
process is highly desirable for two reasons: It will permit with ammonium and alkali cyanides, which, as the inventors
recovery of precious metal from discarded or worn-out parts, point out, are rendered unstable or even explosive by inor-
and rework of expensive parts after faulty plating. In the latter ganic oxidizing agents of high oxidizing potentials. J.J.
case, especially, it is important that the substrate be fully Grunwald17 proposed the use of complex tetracyanide ions,
preserved. Therefore, a stripping method is desirable that such as [Zn(CN)4]-2, in combination with nitro-substituted
functions with a minimum of substrate corrosion, that is, with aromatics, for gold stripping with the aim of diminishing
pronounced selectivity for the precious metal. corrosion of the substrate, especially copper. L.R. Fountain18
While the stripping of gold from base metal substrates is in used a solution of alkali cyanide and a nitro-substituted
common use, the stripping of palladium and its alloys is much aromatic compound for removing a gold alloy braze having
less well known. One reason may be that the available palladium or niobium in combination therewith. H.M. Solidum
processes lack selectivity. The lack of a feasible stripping and R.C. Zobbi19 recommended use of a dry powder contain-
method presents an obstacle for the otherwise commercially ing cyanides and nitrobenzene derivatives with an auxiliary
attractive conversion from gold to palladium. agent for making up and replenishing stripping solutions for
The stripping solution12 to be described is selective for gold and silver (palladium is not mentioned). Nitrobenzoate
palladium with regard to nickel, copper, copper bronzes, and derivatives as oxidant and alkali cyanide as ligand are also
steel substrates. The proposed process easily dissolves any used in a palladium stripping solution proposed by A. Fletcher
pure or hard gold finish on top of palladium or a palladium and W.L. Moriarty.20
alloy. If palladium is plated over nickel, both layers can be
dissolved in the same solution, but nickel is removed more Corrosion Problems with Copper Substrates
slowly than palladium by a factor of 10 or more. If the nickel Inhibitors are commonly in use for selective stripping of
layer is thick, it may be preferable to strip the nickel sepa- nickel or gold from copper substrates. M. Weisberg and F.P.
rately. On the other hand, the large difference in stripping Butler21 claimed a long list of compounds of sulfur in the -2
rates permits selective stripping of palladium from nickel. oxidation state in combination with a nitro-substituted or-
The corrosion of copper or copper bronzes is so inhibited that ganic compound and an amine for removing metals, espe-
a clean removal of palladium or palladium alloy can be achieved cially nickel, from bases, especially copper and its alloys.
virtually without loss of substrate material. This leaves expen- E.H. Too and D.R. Marx22 proposed addition of an organo-
sive work pieces reusable after palladium stripping. mercapto compound for selectively stripping gold from a

64 PLATING & SURFACE FINISHING


Fig. 1—Etch rates vs. Immersion time for individual metals. Fig. 2—Etch rate vs. shelf-life as indication of stability of the stripping
solution.

metal substrate, especially copper, in a solution also contain- oxidant solution and a small amount of inhibitor. Ten mL of
ing a ring-substituted nitrobenzoic acid compound, a bath- inhibitor solution per liter has been found optimal for most
soluble cyanide, and a lead or bismuth compound. C.H. decorative and electronic components. For components with
Huang23 described the function of benzotriazole as a corro- discrete geometries and/or uneven deposit thickness, how-
sion inhibitor for copper in a sodium chloride solution. None ever, such as electronic packaging lead-frames, high inhibi-
of these corrosion inhibitors, however, has yet been applied tor concentrations (as high as 80 mL/L) have been found
for stripping palladium from copper. We tested solutions as advantageous to avoid attack on even the thinnest of leads.
proposed by Too and Marx22 for selective gold stripping, for The solution may become turbid when inhibitor is added, but
removal of palladium from copper substrates. The attack will clear on heating. Just before use, the solution is heated to
rates were 0.18 to 0.30 µm/min for palladium, and 4.5 to 4.8 40 to 50 °C; temperatures above 55 °C should be avoided,
µm/min for copper, at 40-43 °C. These initial findings sug- because solution decomposition, with loss of stripping activ-
gest that conventional copper inhibitors may be ineffective in ity, may set in. Before immersion, the parts are cleaned
palladium stripping. thoroughly and rinsed. Good agitation during stripping is
essential. After stripping, the parts are rinsed repeatedly;
Process Description water from the first rinse can be used to prepare more fresh
Palladium Stripping with Copper Passivation stripper, or added to the spent stripping solution for eventual
While corrosion inhibitors for copper do not appear to be metal recovery. Cathodic cleaning of the parts is recom-
useful by themselves, it was found that a suitable combina- mended before re-plating.
tion of additives with synergistic action can provide a good
palladium stripping rate, as well as very satisfactory copper Performance Tests
protection. The resulting stripper meets all practical demands Individual Metal Stripping Rates
with regard to speed of operation, selectivity, safety, and Selectivity for palladium in the proposed stripping solution
metal recovery. It is freshly prepared prior to each use from was evaluated by testing etch rates for individual metals, as
two liquid concentrates and a salt mixture containing cya- well as for multilayer finishes. For the individual metal tests,
nide. electroplated palladium, nickel, cobalt-hardened gold, and
rolled copper foil and unplated Alloy 725 were used. The
Preparation and Operation of the Stripping Solution workpieces were immersed in the stripper for five subsequent
The stripping solution is prepared by dissolving 150 g of a salt periods of 30 sec each, at 42 °C, with moderate agitation.
mixture in water, followed by the addition of 250 mL of After each interval, the pieces were rinsed, dried, weighed,

Fig. 3—Etch rate vs. Pd conc. as illustration of “aging” of stripping Fig. 4—Effect of salt replenishment on etch rate.
solution with use.

66 PLATING & SURFACE FINISHING


Fig. 5—Scrap Pd stripping trial. Fig. 6—Scrap Pd-Ni stripping trial.

and immediately re-immersed for the next interval. The etch Applying these rate figures to a multilayer finish, where
rates are expressed in µm/min and plotted at the end of each palladium is plated over nickel on a copper substrate, the
interval (Fig. 1). following can be expected: If only palladium stripping is
The plot shows that cobalt-hardened gold and the base intended, with the nickel underplate to be left intact, the low
metals were much more slowly etched than palladium; more- initial nickel etch rate offers clean palladium removal with
over, it shows that the palladium etch rate stayed constant little loss of nickel. On the other hand, if the nickel layer is to
over the whole 2.5-min period, during which a total of 1.65 be removed, the autocatalytic acceleration of nickel etching,
µm was stripped, while the nickel etch rate increased, and the as expressed in Table 1, is favorable; moreover, the nickel
copper etch rate decreased over the same period. Palladium- stripping rate can be increased by adding nickel salt to the
nickel alloy (20% Ni) behaves like palladium. That hard gold stripper. The simultaneous removal of the nickel layer has
could be etched, even though the rate is slow, demonstrates been found especially advantageous in lead-frame applica-
that gold-flashed palladium and palladium nickel finishes are tions. Finally, the slight initial corrosion of the copper or
amenable to treatment. Initial and final (after 2-min immer- copper-based substrate will be further diminished with time.
sion) etch rates were as shown in Table 1.
Stripping of a Multilayer Finished Workpiece
Table 1 The above considerations were demonstrated on a multilayer
Etch Rates finished lead-frame strip with 50 to 100 µin. of nickel and 5
µm/min to 10 µin. of palladium plated over Alloy 725. The results of
this experiment are shown in Table 2.
Metal Initial Final This experiment shows how important a high degree of
Hd. Au 0.125 0.125 selectivity is in the stripping of pieces with coatings of
Pd 0.67 0.67 uneven thickness, because in such cases two layers of metal
Ni 0.075 0.35 may be exposed simultaneously for relatively long periods.
Cu 0.109 0.019
In the above instance, consideration might also have been
Alloy 725 0.094 0.043
given to switching to a substrate-protecting nickel stripper
after all palladium had been removed. No general preferences
Table 2 can be given here, for each application must be judged on its
own merits.
Immersion Weight
Metal Time Loss Comments Stability of the Stripper Solution
Pd 0.5 min 6.8 mg Palladium layer totally removed The made-up stripper solution is stable at room temperature.
Ni 1.5 min 11.5 mg Nickel layer totally removed We also tested the palladium stripping rate in a solution kept
Alloy 725 >2.5 min ≤0.5 mg No visible attack on lead-frame at 42 °C, after 1, 2, 4, and 8 hr. The results are shown in Fig.
pattern 2. Over long periods, a slight drop in the etch rate is expected
as a result of the loss of free cyanide.
Table 3
Aging of the Stripping Solution with Use
Complexing Stripping Because of the variety of stripped materials and stripping
Pd Salt Rate operations, the aging behavior of the stripping solution can-
g/L g/L µm/min not be exactly predicted. In an attempt to obtain a useful
Fresh stripper 0 150 0.75 simulation of actual operating conditions, incremental
Unused stripper 0 132 0.54 amounts of palladium metal powder were added to the
aged 40 days stripping solution and permitted to react completely before
Used stripper, 13.5 80.3 0.36
after 1 day
the etch rate was determined. The results (Fig. 3) show a
Used stripper, 13.5 80.3 0.32 decline in the etch rate with depletion of the reagents, caused
after 13 days by complexing of free cyanide with palladium, and reduction
Same stripper, 13.5 150 0.60 of the oxidant.
replenished A connection between the etch rate of palladium and the
availability of cyanide in the solution can also be seen in Fig.

January 1996 67
3, where the “free” cyanide content has been plotted against 1. Proximity of a precious metals refinery.
the amount of palladium metal dissolved; the curves for the 2. Volume and precious metal concentration of stripping
etch rate and free cyanide run parallel to each other. solution to be refined.
3. Cost of state and local government hazardous waste
Effect of Replenishing and Storage transportation permits.
Etch rates approximating the initial conditions can be at- 4. Availability of chemical or electrochemical process equip-
tained by replenishment of the components, even if the ment, air-borne emission exhaust system and wastewater
stripping solution is used infrequently. A “spent” stripping treatment facility.
solution containing 15 g/L of palladium and 54 g/L of free
cyanide salt had an etch rate of 0.32 µm/min after being used In-plant recovery of palladium can be effected primarily
intermittently for 15 days. The solution was replenished by chemical or electrochemical methods; possibly also by a
twice during this period, with 70 g/L salt mixture, 100 mL/L combination of them. Because of the extreme stability of the
oxidant solution, and 5 mL inhibitor solution. After each palladium cyanide complex, cyanide ions must be removed,
replenishment, the etch rate was found to be 0.65 µm/min, precipitated or destroyed in all cases. A chemical method has
and 0.6 µm/min, respectively, demonstrating a recovery of been found to precipitate the bulk of the complexed palla-
80 percent of the initial rate. The results are presented dium. The solids can then be filtered, dewatered and shipped
graphically in Fig. 4. for refining as a dry cake, which would ease hazardous waste
The stripping solution was also found to be quite stable in management and reduce shipping costs. The resultant filtrate
storage, with only marginal losses in etch rate after 40 days, is then treated with sodium hypochlorite solution, a well-
which is particularly advantageous. The results are listed in known reagent for cyanide destruction,24 which precipitates
Table 3. the remaining nickel as oxide and destroys the free cyanide.
Any trace of metal left in the filtrate is then removed with
Production Scale Application of the Stripper sodium borohydride, after which Pd, Ni and cyanide are
The stripping solution described above was successfully reduced to undetectable levels. The resulting filtrate can then
applied in refining approximately 4,000 lb of low-grade, be evaluated, further treated, (e.g. neutralized) and discarded
precious-metal scrap etched in 50-lb batches for the purpose as per local environmental regulatory agency guidelines.
of precious metal recovery. The parts had been plated with 20
to 50 µin. of Pd-Ni alloy. The solution volume was 160 gal Summary
(approx. 600 L) and the temperature was maintained between A reliable stripping solution for rework and metal recovery
40 and 50 °C. The parts were loaded into a stainless steel has been developed, comprising a three-component system.
wire-mesh basket prior to immersion in the solution. Solution This formulation provides satisfactory and economical pro-
agitation was approximately 50 gal/min, and the exit nozzle cessing of palladium, gold-flash palladium or palladium-
from the pump was aimed directly at the bottom of the basket nickel finishes. The system has been demonstrated to be
to ensure exposure of all parts to the etching solution. After exceptionally selective for palladium and palladium-nickel
15 min, the innermost parts were spot-tested for Ni underplate vs. copper. Electrolytic recovery of the precious metal, using
exposure, using nitric acid. A drop of concentrated nitric acid standard equipment, is described. The relationships between
placed on a well-rinsed part turns green immediately if the the etch rate, dissolved metal and free cyanide concentration
nickel underplate is exposed. If Pd or Pd-Ni alloy remains, the were also established, along with analytical methodologies
spot turns yellow after approximately 30 sec. If the test useful for process control.
indicated that Pd-Ni remained on the surface, the basket was
immersed for an additional 15 min and retested. The parts Editor’s note: Manuscript received, August 1995.
were rinsed twice with water before being discarded for base
metal refining. The results are shown in Figs. 5 and 6. References
The conducting salts and oxidant solution were replen- 1. Technical Memorandum No. 82-52312-18, J.A. Abys,
ished only once after 2,000 lb had been processed. Replenish- “A Unique Palladium Electrochemistry - Characteristics
ment was required because, after 30 min of etching, the spot and Film Properties,” AT&T Bell Laboratories, June 10,
test still showed Pd remaining on the surface of the parts. The 1982.
inhibitor solution was not replenished, because the parts were 2. I.V. Kadija et al., Proc. AESF SUR/FIN ’90, Session N
not destined for reuse. (1990).
3. H.K. Straschil et al., Proc. AESF SUR/FIN ’90, Session
Metal Recovery R (1990).
Choice of Recovery Methods 4. J.A. Abys, et al., Metal Fin., 89, 43 (July, 1991).
Most of the precious metal in the spent stripping solution 5. J.A. Abys, et al., “The Electrodeposition of Pure Palla-
(including gold) is present in the form of its cyanide dium with a Palladium Hydroxide Replenishing System.
complex(es). The simplest way to recover palladium from A Non-Cyanide, Low Ammonia, Low Maintenance Proc-
spent stripping solutions is to return the solutions to the ess”, Proc. 14th AESF/EPA Conf., January, 1993.
manufacturer for refining. Existing regulations for the ship- 6. E.J. Kudrak, et al., Proc. Connectors ’93 Conf., May,
ment of “hazardous waste,” however, may render that option 1993.
financially unattractive. In such cases, it is preferable to 7. E.J. Kudrak, et al., Plat. and Surf. Fin., 78, 57 (March,
recover the precious metal as a medium- to high-purity metal 1991).
or sludge that is more easily transported. The decision whether 8. H.K. Straschil, Metal Fin., 90, 42 (Jan., 1992).
to ship the solutions to the refinery as such, or to send the 9. E.J. Kudrak, et al., Plat. and Surf. Fin., 79, 49 (Feb.,
recovered precious metal for refining, depends on a variety of 1992).
economic and engineering factors that must be weighed, such as:

68 PLATING & SURFACE FINISHING


10. E.J. Kudrak and J.A. Abys, Interconnection Technol., 9, cally controlled Hull cell, and the solderability and wire-
18 (June, 1993). bondability of palladium and its alloys. In addition to numer-
11. E.J. Kudrak, J.A. Abys and F. Humiec, Proc. AESF ous scientific publications, he holds more than 50 patents.
SUR/FIN, Session H (June, 1993). Frank Humiec is a member of the
12. Patent application pending. Electroplating Development Group at
13. E.B. Saubestre, Stripping of Electrodeposits, in Modern AT&T Bell Laboratories. His work in-
Electroplating, 3rd Ed., F.A. Lowenheim, Ed., John cludes plating, process improvements,
Wiley & Sons, Inc., New York, NY, 1963; p. 765. analytical and technical services. He
14. Nathaniel Hall, G.B. Hogaboom, Jr., and J.B. Mohler, holds a BS in chemistry from Niagara
Stripping Metallic Coatings, in Metal Finishing Guide- University and an MS in chemistry from
book & Directory, 48th Ed., Vol. 78, No. 1A, 1980; p. Fairleigh Dickinson University. He is
507. co-author of several papers in the areas
15. T.A. Shankoff, U.S. patent 3,839,110, (1974) of inorganic and pharmaceutical chemistry.
16. Richard Springer and W.R. Meyer, U.S. patent 2,649,361
(1949). Dr. Joseph A. Abys is the business
17. John J. Grunwald, U.S. patent 3,242,090 (1966). manager of the Electroplating Chemi-
18. Laurence R. Fountain, U.S. patent 3,819,494 (1974). cals & Services Group, AT&T Bell
19. H.M. Solidum and R.G. Zobbi, U.S. patent 3,935,005 Laboratories. He received a PhD in in-
(1976) organic chemistry from Brown Univer-
20. Augustus Fletcher and W.L. Moriarty, U.S. patent sity. He has been working in the field of
4,548,791 (1985). electrodeposition for 15 years, with con-
21. Mark Weisberg and F.P. Butler, U.S. patent 3,102,808 centration on plating of palladium and
(1963). its alloys. Recently the group’s efforts
22. Elena H. Too and D.R. Marx, U.S. patent 4,483,739 have expanded in the areas of gold-tin
(1984). alloys, oxide anodes, electroforming and fundamental work
23. C.-H. Huang, Plat. and Surf. Fin., 73, 96 (June, 1986). on the hydrodynamically controlled Hull cell. In addition to
24. L.J. Durney, Ed., Electroplating Engineering Hand- many scientific publications, Dr. Abys holds more than 25
book, 4th Ed., Van Nostrand Reinhold Co., Inc., New patents in the U.S. and abroad.
York, NY, 1984; pp. 212-214.

About the Authors


Dr. H.K. Straschil is a member of the
Electroplating Development Group at
AT&T Bell Laboratories, 600 Moun-
tain Ave., Murray Hill, NJ 07974-0636.
He is involved in research and process
development of palladium, its alloys,
and gold-tin alloys. He received his
doctor of technical science degree (in-
organic and physical chemistry) from
the Technical University, Vienna, Austria.
Joseph J. Maisano, CEF, is a member
of the technical staff at AT&T Bell
Laboratories. He holds a BS in bio-
chemistry/biophysics from the Univer-
sity of Connecticut and an MS in chem-
istry from Rutgers University. He pro-
vides analytical support for product
development efforts, specializing in
analytical methods development. He is
a co-author of numerous scientific pub-
lications and holds nine U.S. patents.
Dr. Igor V. Kadija is currently a mem-
ber of the technical staff with the Elec-
troplating Chemicals & Services Group
at AT&T Bell Laboratories. He holds a
PhD in chemical engineering from the
University of Belgrade. His experience
includes more than 25 years in surface
finishing for electronics and the chlor-
alkali industry. Recently his work has
been in electrodeposition of palladium
and its alloys, plus fundamental work on the hydrodynami-
Free Details: Circle 146 on postpaid reader service card.

January 1996 69

You might also like