HK1118950B - Fretting and whisker resistant coating system and method - Google Patents

Description

Wear and whisker resistant coating system and method

Technical Field

The present invention relates to systems and methods for coating electrically conductive substrates, and more particularly to multi-layer systems and methods for coating electrically conductive substrates.

Background

In this application, "base" as used to define an alloy means that the alloy contains at least 50% by weight of the defined element, for example, "copper-based alloy" means that the weight percentage of copper in such an alloy is greater than 50%. In the electrical and electronics industries, copper and copper-based alloys (hereinafter collectively referred to as copper) are widely used in the fabrication of wire connectors, electrical wires, printed circuit boards, ball grid arrays (ball grid arrays), lead frames (leadframes), multi-chip assemblies, and the like. While copper has good electrical conductivity, it is highly susceptible to oxidation and rusting when exposed to high temperature, humidity or chemical environments. Oxidation and rusting of copper generally increases the contact resistance of the copper, thereby degrading the performance of the electrical device. In addition, oxidation and rusting of copper also reduces solder wettability, causing problems during soldering.

One method of avoiding or reducing oxidation and tarnishing of copper is to plate a copper substrate with a tin or tin-based alloy (hereinafter collectively referred to as tin) coating. This tin coating serves to prevent or reduce copper oxidation and thus can maintain the electrical properties of the copper substrate. However, there are still many problems with using tin as a coating for a conductive substrate. At room temperature (typically 25 ℃), the tin coating and the copper substrate interdiffuse to form a copper-tin intermetallic compound (IMC). These copper-tin intermetallic compounds reduce the thickness of the tin coating, resulting in an increase in the contact resistance of the copper substrate, reduced solderability, and a faster rate of such diffusion between the tin coating and the copper substrate at high temperatures.

Exemplary thermal change processes include a process requiring wire bonding or encapsulation in a polymer at 250 c for a few seconds, a reflow process at 300 c for a few seconds, and a process at 150 c for 8 to 168 hours to control the reduction in the thickness of the tin coating for friction reduction.

One method for reducing the effect of copper-tin intermetallic compound formation and maintaining low contact resistance is to increase the thickness of the tin coating. However, this approach not only increases the cost of the components, but also has an impact on functionality. When a tin coating is used in an electrical connector, the thick soft tin coating increases friction, which in turn increases the force required to insert the connector, making insertion and removal of the electrical connector difficult. Thick tin or tin alloy coatings are also undesirable for electronic devices since the trend is toward ultra-thinning and miniaturization. In addition, in the case of using a tin coating on a lead (lead) of an electronic device, the thick tin coating causes problems in the definition of the coplanarity and thickness of the lead.

Another way to reduce the effect of copper-tin intermetallic compound formation is to add a transition barrier layer between the copper substrate and the tin coating to prevent the formation of copper-tin intermetallic compounds. For example, U.S. patent No.4,441,118 reports that low rates of copper-tin intermetallic compound formation can be achieved using copper-nickel alloy substrates with nickel content between 15% and 30%.

In another example, p.j.kay and c.a.mackay discuss the use of various metals as transition barriers, published in 1979 "Transactions of the Institute of metal finishing" at 51, page 169. One embodiment of the article describes a silver barrier layer having a thickness of 1 micron. However, this example proved to be undesirable because the silver transition barrier layer was not able to actually reduce the diffusion rate between copper and tin. U.S. patent No.4,756,467 to Schatzberg discloses a solderable connector that includes a copper substrate, a thin silver layer, a silver-tin alloy layer, and an outermost tin layer. Wherein the silver-tin alloy layer is formed by diffusion annealing. Japanese patent No.2670348 to Furukawa electric limited (publication No. 02-301573) discloses a copper substrate coated with a nickel or cobalt barrier layer on which is a silver layer on which is a melt-solidified (melt-solidified) layer of tin or tin alloy.

The applicant's own U.S. patent application serial No. 10/930,316, filed 10/31 in 2004, which is a continuation of U.S. patent application serial No. 09/657,794, discloses a thin corrosion resistant layer disposed between a copper substrate and a tin coating. Metals disclosed as corrosion resistant layers are zinc, chromium, indium, phosphorus, manganese, boron, thallium, calcium, silver, gold, platinum, palladium, and combinations and alloys of these metals.

Other barrier layers are disclosed in U.S. patent nos. 5,780,172 and 5,916,695 to fisher et al, owned by the present applicant.

Another problem that can arise with the use of tin as a coating for a conductive substrate is that tin is prone to fretting corrosion. Fretting corrosion is the oxidation of the contact surfaces caused by the relative movement (rubbing) between two mating contact surfaces. Such friction-induced oxidation may result in an unacceptable increase in contact resistance. Certain metals, such as silver, are known to have excellent fretting corrosion resistance. However, silver is susceptible to tarnishing in air due to the presence of sulfur dioxide in the air, resulting in the formation of silver sulfides on the silver surface. This tarnishing is not only aesthetically unacceptable, but may also degrade the electrical contact properties of the silver.

Another problem with using tin or other metals, such as zinc, indium, antimony or calcium, as a coating for a conductive substrate is that tin or other metals as described above tend to produce whiskers. As tin ages, whiskers form and stresses begin to develop at the tin or tin/intermetallic compound (IMC) interface. Another cause of whisker formation is due to internal stresses generated during electroplating. To relieve these stresses, tin single crystals nucleate from the surface, forming whiskers. Each whisker continues to grow until the internal stresses are completely released. Whiskers may cause a number of different problems, including shortening the distance between adjacent electrical contact surfaces. A common way to reduce whisker growth is to add a small amount of lead (Pb) to the tin coating to form an alloy. However, for health and environmental reasons, many industrial processes are required to reduce or prohibit the use of lead.

Thus, it would be desirable to develop a coating system that maintains low contact resistance and good solderability after rubbing and heat exposure, while also having the characteristics of a lower coefficient of friction and/or reduced tin whisker growth.

Disclosure of Invention

In accordance with a first embodiment of the present invention, there is provided a coated conductive substrate having specific properties, the substrate having a plurality of closely spaced features and tin whiskers susceptible to short circuit formation. Such substrates include lead frames, terminal pins, circuit traces (circuit traces) on printed circuit boards and flexible circuits, including wires, lines and circuit traces. The conductive substrate has a plurality of conductive lines spaced apart a distance that can be bridged by tin whiskers, and a silver or silver-based alloy layer covering at least one surface of at least one of the plurality of conductive lines, a fine-grained tin or tin-based alloy layer directly overlying the silver layer.

In accordance with a second embodiment of the present invention, there is provided a coated conductive substrate having special properties for use where frictional wear debris oxidizes and resistivity increases, such as in a connector. An isolation layer is deposited on such a conductive substrate to prevent diffusion of the substrate into a plurality of subsequently deposited layers. These subsequently deposited layers include a sacrificial layer for forming an intermetallic compound with the tin; a metal layer capable of forming a low resistivity oxide (referred to herein as a low resistivity oxide metal layer) deposited on the sacrificial layer; the outermost tin or tin-based alloy layer is deposited directly on the low resistivity oxide metal layer.

In a second embodiment, the barrier layer is preferably nickel or a nickel-based alloy and the low resistivity oxide metal layer is preferably silver or a silver-based alloy.

When the coated substrate of the second embodiment is heated, a particular structure is produced having a copper or copper alloy substrate, an intermediate layer formed from a mixture of metals including copper and tin, and an outermost layer formed from a mixture of copper-tin intermetallics including phases and silver-rich phases.

The silver-rich phase is believed to minimize the increase in resistivity caused by oxidation of the frictional wear debris.

The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

Drawings

The present invention will become more fully understood from the detailed description given herein below and the accompanying drawings. In the drawings, like elements are represented by like reference numerals, wherein,

fig. 1 is a top plan view of a lead frame according to a first embodiment of the present invention prior to encapsulation and coating;

fig. 2 is a side plan view of the lead frame of fig. 1 after encapsulation and before coating in accordance with a first embodiment of the present invention;

fig. 3 is a cross-sectional view of the lead frame of fig. 1 after encapsulation and coating in accordance with a first embodiment of the present invention;

FIG. 4 is a cross-sectional view of a conductive strip coated in accordance with a second embodiment of the present invention;

FIG. 5 is a cross-sectional view of the conductive strip of FIG. 4 configured as a connector;

FIG. 6 is an enlarged cross-sectional view of a portion of the connector of FIG. 5 illustrating the effects of friction debris;

FIG. 7 is a schematic flow chart illustrating the fabrication of the first embodiment of the present invention;

FIG. 8 is a schematic flow chart illustrating the fabrication of a second embodiment of the present invention;

FIG. 9 is an illustration of interdiffusion between layers of a substrate coated with different layer combinations;

FIG. 10 is a photomicrograph of the surface of a coated substrate of the present invention heated at 150 ℃ for one week;

fig. 11 is a cross-sectional photomicrograph of the coated substrate shown in fig. 10.

Detailed Description

Referring to fig. 1, the lead frame includes a plurality of wires 10 formed of a conductive metal, such as copper or a copper-based alloy. Each of the plurality of wires 10 terminates at a wire inner end 12 forming a central window occupied by a die pad 14. Typically, a thin noble metal layer, such as silver, is applied to the inner tail 12 of the wire and to the die pad 14 to enhance die attachment and wire bonding. When silver is used as the coating metal, the thin layer is typically 3 to 6 microns thick and is deposited by electrodeposition. One or more Integrated Circuit (IC) devices 16, commonly referred to as semiconductor chips, are then attached to the chip-pads 14, such as by a low temperature metal solder or a thermally conductive polymer adhesive. Thin metal wires 18, or strips of conductive metal foil, connect the circuitry on the electrically active side of the integrated circuit device 16 to the wire inner tail end 12. The die pad 14, integrated circuit device 16, wire inner tail end 12, and wire intermediate portion 21 are then encapsulated with a molding resin generally along the perimeter indicated by dashed line 20.

Fig. 2 is a side view of the assembly of fig. 1, from which it can be seen that the wire 10 extends from the molding resin 22. The outer portions 23 of the conductors extending from the molding resin are typically soldered to external circuitry, such as traces on a printed circuit board. For optimum conductivity, the wire is typically formed of copper or a copper alloy, but non-copper metals such as iron-nickel and iron-nickel-cobalt may also be used. Copper and copper alloys are easily oxidized, and oxides formed on the surface affect the solderability of copper and copper alloys.

To prevent the formation of oxides, it is common practice to deposit an anti-corrosion layer on the copper wire. Such a material that is easy to solder as a rust resistant layer is tin or a tin-based alloy. However, when copper and tin are exposed to room temperature or higher, there is diffusion between the copper and tin. When a copper-tin intermetallic compound is formed on the surface of the layer, the tarnish resistance and solderability of the structure are reduced. In order to reduce the diffusion rate between copper and tin and to reduce the formation of intermetallic compounds, it is common practice to provide a barrier layer, such as nickel, between the substrate and the tarnish-resistant layer.

Tin whiskering is a property of tin that refers to the release of internal stresses through the growth of thin tin filaments. Referring back to fig. 1, the wires 10 are closely spaced and tin whiskers may lap in the gaps 24 between adjacent wires, forming a short circuit. In general, when the distance between two wires is 1 mm or less, the possibility of tin whiskers forming a lap joint between the two wires will be high. Although a number of methods have been proposed to prevent tin whisker formation, these methods still have limitations. It is known that alloying tin with another metal, such as lead, can reduce whisker formation, but lead is toxic. Another known method is to heat the tin above its melting point and to reduce whisker formation by a known reflow process. However, controlling the flow of liquid tin is difficult and often during reflow, a bond between the wires is formed.

According to a first embodiment of the present invention and referring to fig. 3, tin whiskers may be reduced by forming the conductor 10 on a substrate 26, coating the substrate 26 with a silver or silver-based alloy layer 28, and then depositing a tin fine particle layer 30 directly on the silver or silver-based alloy layer 28. By "direct" deposition is meant deposition next to the silver or silver-based alloy layer 28 without any intervening layer of other material. If the substrate 26 is formed of a metal other than copper or a copper-based alloy, a thin layer of copper, between 0.025 and 0.51 microns (i.e., 1-20 microinches) thick, may be deposited on the substrate prior to deposition of the silver layer 28. The silver layer 28 may be replaced with a silver-based alloy layer and the tin layer may be replaced with a tin-based alloy layer.

The strength of the contact surface of the two metals is smaller than that of the metal itself. Therefore, the portion of the lead 10 encapsulated by the molding resin is preferably not covered with the silver layer and the tin layer, which are coated only on the portion of the lead extending from the molding resin. The silver layer 28 has a thickness in the range of 0.025 microns to 3.05 microns (1-120 microinches). When the thickness is less than 0.025 μm (1. mu. inch), tin whiskers are not sufficiently suppressed. When the thickness exceeds 3.05 micrometers (i.e., 120 microinches), the cost will be increased. The preferred silver layer thickness is between 0.05 microns and 1.02 microns, i.e., 2-40 microinches, and the most preferred silver layer thickness is in the range of 0.13 microns to 0.51 microns, i.e., 5-20 microinches.

The tin layer 30 has a thickness of between 0.00025 microns and 10.2 microns, i.e., 0.01-400 microinches. When the thickness is less than 0.00025 μm, i.e., 0.01. mu. inch, both corrosion resistance and solderability are reduced. When the thickness exceeds 10.2 μm, i.e., 400. mu. inches, a lap joint is easily formed between adjacent wires. The preferred tin layer thickness is from 0.51 microns to 3.8 microns, i.e., 20-150 microinches. The most preferred tin layer thickness ranges from 0.51 microns to 2.03 microns, i.e., 20-80 microinches.

Tin obtained by means of electrodeposition is fine particles, as opposed to coarse particles, which are obtained by the reflow process described below. Typical average particle sizes are between 0.1 and 100 microns, with preferred sizes being between 0.5 and 5 microns, as opposed to the standard particles after reflow which are on the order of millimeters. The fine particles generally have better ductility so that the wire can be bent at large angles without cracking of the coating. While fine-grained tin is believed to be more prone to tin whisker generation, the silver underlayer in this embodiment allows the use of fine-grained tin.

Although the first embodiment of the present invention is described with respect to a lead frame having a plurality of closely spaced wires, the tin-free whisker coating of the present invention may also be applied to other structures, such as terminal pins, printed circuit boards and flexible circuits, which contain closely spaced other devices, such as wires and circuit traces.

The second embodiment of the present invention is applied to a connector. Unlike lead frames, most connectors are not affected by tin whiskers because adjacent connectors are typically far enough apart to avoid short circuits caused by tin whiskers. In addition, because the connectors are not as closely spaced as the wires in the lead frame, the reflow process can reduce internal stresses in the tin coating. Furthermore, diffusion between copper and tin is often used to reduce the thickness of free tin (tin), thereby reducing the friction and force required to insert the probe into the socket.

The connector typically has increased resistivity due to friction debris. The phenomenon of frictional wear occurs during small amplitude relative vibration of two surfaces. Frictional wear produces small removable particles on the surfaces that are in contact with each other. These small particles are then oxidized and oxide fragments accumulate at the connector interface. Since the resistivity of tin is about 0.12 micro ohm meters (μ Ω · m) at room temperature, and the resistivity of tin oxide is about 1 micro ohm meter at room temperature, frictional wear deteriorates the electrical properties of the connector.

Connectors formed in accordance with the second embodiment of the present invention may have reduced frictional wear. Referring to fig. 4, the substrate 26 is typically copper or a copper-based alloy, but other conductive metals may be used. When any of the other conductive metals is used as the substrate, a thin copper layer is deposited on the substrate as described above. A thin copper layer may also be deposited on a copper alloy substrate to provide a pure copper surface to facilitate subsequent deposition and adhesion of various layers.

Deposited on the copper or copper-based alloy substrate or thin copper layer is an isolation layer 32. The barrier layer may be any metal that prevents diffusion between the copper and other constituent substrate components, preferably a transition metal such as nickel, cobalt, iron, manganese, chromium, molybdenum or alloys thereof. The spacer layer has a thickness of between 0.051 microns and 2.03 microns, i.e., 2-80 microinches. If the thickness of the spacer layer is less than 0.051 microns (2 microinches), it is not effective in preventing diffusion. If the thickness of the spacer layer exceeds 2.03 microns (80 microinches), the electrical and mechanical performance of the connector may be adversely affected. Preferably, the thickness of the barrier layer is from 0.1 microns to 1.02 microns, i.e., 4-40 microinches. More preferably, the thickness of the barrier layer ranges from 0.1 micron to 0.51 micron, i.e., 4-20 microinches.

Deposited over the isolation layer 32 is a sacrificial layer 34. The sacrificial layer 34 is a metal that combines with silver and tin to form an alloy or intermetallic. To reduce friction, the free tin thickness of the outermost layer 36 is reduced. This thickness reduction may be achieved by heating the assembly to cause the sacrificial layer to bond with the inner portion of the outermost layer to form a higher hardness intermetallic. The preferred material for the sacrificial layer is copper or a copper-based alloy having a thickness of between 0.051 microns to 1.52 microns (2-60 microinches). The thickness of the sacrificial layer may be selected to retain at least on the order of 0.051 microns of free tin on the outer surface 38 of the outermost layer 36 when the sacrificial layer is consumed. The most preferred thickness range for the sacrificial layer of copper is 0.13 microns to 0.51 microns, i.e., 5-20 microinches, when the thickness of the outermost layer begins to be between 1.02 microns and 2.03 microns.

Deposited between the sacrificial layer 34 and the outermost layer 36 is a low resistivity oxide metal layer 40. The low resistivity oxide metal is a metal that forms an oxide at the intended operating temperature of the connector, and the resistivity of the metal oxide is lower than the resistivity of the tin oxide. Silver or silver-based alloys are preferred choices for the low resistivity oxide metal layer 40. While the resistivity of tin oxide is about 1 micro ohm-meter at room temperature, the resistivity of silver oxide is about 0.14 micro ohm-meter at room temperature. By adding silver oxide to the friction debris, the effect of frictional wear on the electrical resistivity of the mating insert is significantly reduced. The low resistivity oxide metal layer has a thickness of between 0.025 microns and 3.05 microns, i.e., 1-120 microinches. If the thickness is less than 1 micro-inch, the silver oxide is not sufficient to affect the electrical resistivity of the connector. If the thickness exceeds 3.05 micrometers (120 microinches), the cost will increase. Preferably, such low resistivity oxide metal layer is between 0.05 micron and 1.02 micron thick, i.e., 2-40 micro inches, and more preferably between 0.13 micron and 0.51 micron, i.e., 5-20 micro inches.

The conductive strips shown in fig. 4 are formed in a connector, and a cross-sectional view of such a connector is shown in fig. 5. Such a connector includes a socket 42 and a probe 44. The socket is typically curved to form an effective point contact with the probe, and the probe is shaped to provide an internal stress in the socket effective to provide a positive pressure to keep the probe and socket in electrical contact at point 46.

Fig. 6 is an enlarged schematic view of the point contact defined by the dotted circle in fig. 5. Due to the vibration, the point 46 oscillates between a first contact point 48 and a second contact point 50. This wear produces metal oxide wear debris 52. A portion 54 of the wear debris covers the oscillation trajectory and affects the current flow between the point 46 and the probe 44.

The metal of the low resistivity oxide metal layer should be a noble metal, such as gold, platinum, palladium, whose oxide resistivity is lower than that of tin oxide (1 micro ohm-meter) or which is more difficult to form an oxide than silver. Table 1 lists a number of base metal oxides and gives their suitability as low resistance metal oxides. In Table 1, "O" stands for applicable, "X" stands for not applicable. Indium, iron, niobium, rhenium, ruthenium, vanadium, gold, platinum, palladium and zinc and mixtures of four of these metals may be used as silver substitutes.

TABLE 1

Fig. 7 is a flow diagram of a method of making the coated substrate shown in fig. 3 for an application involving only tin whiskers and not tin reflow to relieve internal stress. These applications include lead frames, closely spaced terminal pins (such as those found on pin grid array electronic package devices), and closely spaced circuit traces on printed circuit boards or flexible circuits. As shown in fig. 7, the first three steps are unique to the lead frame and in some embodiments the terminal pin apparatus. The remaining three steps are common to all of the above product categories.

The lead frame may be stamped or otherwise prepared by chemical etching on a substrate, typically copper or a copper-based alloy. The leadframe includes a centrally disposed die pad and a plurality of leads extending from at least one side of the die pad, typically four sides of the die pad. The lead frame is cleaned after fabrication, for example, by using commercial degreasers, alkaline cleaners such as Hubbard-Hall E-9354 (such cleaners are commercially available from Hubbard-Hall, Waterbury, Conn.). The use of an alkaline mixture in conjunction with anode/cathode electro-cleaning can generate oxygen or hydrogen bubbles to remove most of the impurities remaining on the substrate. The electro-cleaning process typically involves applying electricity at 20 degrees celsius to 55 degrees celsius for about 1 minute at a current density in the range of 93 to 465 amperes per square decimeter, i.e., 10 to 50 amperes per square inch.

Then, a metal, such as silver, having a thickness of between 3 microns and 6 microns, is applied 56 over the chip pad and the inner portion of the wire to enhance solderability and wire bonding. It is preferable to coat silver only on the innermost side of the wires for wire bonding and Tape Automated Bonding (TAB) technology. This is because in the next encapsulation step 58, the molding resin is required to be in direct contact with the copper substrate, providing a single interface for adhesion failure and moisture venting. Less preferred is to have the molding resin in contact with the silver layer and the silver layer in contact with the copper substrate to form two interfaces. The silver coating 56 may be any suitable method such as electrodeposition, electroless deposition, dip coating, chemical vapor deposition, or plasma deposition.

The integrated circuit device is then bonded to the die pad by conventional die attach methods 60, including soldering using low temperature solder, such as gold/tin eutectic, or adhesive bonding, such as metal-doped epoxy. Wire bonding is the use of small diameter wires or narrow strips of metal foil to electrically connect an integrated circuit with the wire inside portions of a lead frame. After die attach and wire bonding, the die pad, integrated circuit device, wire bonds, and inner portions of the leadframe leads are encapsulated in a thermosetting molding resin, such as an epoxy. The outer portion of the wire is then bent into a shape that can be connected to a printed circuit board or other external circuit.

A layer of silver or silver alloy 62 is then applied to the outer portion of the wire using methods such as electroplating, electroless plating, immersion plating, physical vapor deposition, chemical vapor deposition, plasma deposition or metal sputtering. The silver plating is between 0.025 microns and 3.05 microns, i.e. 1-120 microinches thick, and most preferably ranges from 0.051 microns to 0.51 microns, i.e. 2-20 microinches thick.

The preferred method of silver plating is electroplating with an aqueous solution of 31-56 grams per liter of silver cyanide, 50-78 grams per liter of potassium cyanide, 15-90 grams per liter of potassium carbonate, and a polishing agent. The plating temperature is at 20 degrees celsius to 28 degrees celsius and the current density is at 46.5 amps per square decimeter to 139 amps per square decimeter, i.e., 5-15 amps per square inch. An alternative silver plating method is the use of cyanide-free immersion plating, such as MacDermid Sterling from MacDermid, Inc. of Waterburri, Conneti^TMSilver (ii) in the presence of a silver compound.

A further layer of tin 64 is then applied over the silver plated outer conductor such that the tin has a thickness of 0.00015 microns to 10.2 microns, i.e., 0.006-400 microinches, and preferably a thickness of 0.5 microns to 2.03 microns, i.e., 20-80 microinches. A preferred method of tin plating is to use a solution containing a methane sulfonic acid based tin plating solution, such as Rohm and Haas Solderon^TMST200 (Philadelphia, Pa., Philadelphia, Rohm and Haas Company, Philadelphia, Pa., USA, Philadelid Sterling, Philadelphia, Pa., USA) AMAT^TMBright tin of AMAT. Typical conditions for use of the above electrolyte are a temperature of between 25 degrees celsius and 35 degrees celsius and a current density of between 46.5 amps per square decimeter and 465 amps per square decimeter, i.e. 5-50 amps per square inch.

The tin plated external leads are then soldered 66 to a printed circuit board or other external circuit, such as by using a tin/lead alloy solder or a suitable lead-free solder. The solder and soldering process are selected to ensure that the solder fuses to the tin layer without melting the tin layer. The tin layer is prevented from melting in order to prevent the liquid solder from bridging the wires.

Fig. 8 is a flow diagram of a method for making the coated substrate shown in fig. 4, which method takes into account the effect of oxidized friction debris to reduce resistivity, as in an electrical connector. When the substrate material is not copper, or a copper alloy with a high alloy content, such as more than 2% by weight, a thin layer of copper may be deposited 68 on the substrate surface before the next layer is applied, as shown in fig. 7. The thin copper layer minimizes the effects of different metals on subsequent layer depositions, thereby allowing many different substrate materials to achieve more consistent product performance.

The copper layer has a minimum thickness of 0.13 microns, i.e., 5 microinches, and a typical thickness of between 0.51 microns and 1.02 microns, i.e., 20-40 microinches. While the copper layer and subsequent layers described below may be deposited by any suitable method, the preferred copper layer deposition 68 method is electroplating using an aqueous solution containing 20 to 70 grams per liter of copper ions and 50 to 200 grams per liter of sulfuric acid. The operating conditions are as follows: the temperature range is between 40 and 60 degrees celsius and the current density is between 186 and 929 amps per square decimeter, i.e., 20-100 amps per square inch.

A barrier layer 70 is next deposited. Suitable barrier layer materials include nickel, cobalt, chromium, molybdenum, iron and manganese and alloys or mixtures thereof deposited to a thickness in the range of 0.05 microns to 1.02 microns, i.e., 2-40 microinches, with a preferred thickness in the range of 0.1 microns to 0.51 microns, i.e., 4-20 microinches. A preferred method of depositing nickel layer 70 is electroplating using an aqueous solution nominally containing 300 grams per liter of nickel sulfamate, 6 grams per liter of nickel chloride and 30 grams per liter of boric acid. The operating conditions are as follows: the temperature is between 28 and 60 degrees celsius, the pH is between 3.5 and 4.2, and the current density ranges from 18.5 to 279 amperes per square decimeter, i.e., 2-30 amperes per square inch.

Next, a sacrificial layer 72 may be deposited using copper as the material to a thickness that will effectively bond with a portion of the tin under controlled thermal excursions to form a copper/tin intermetallic compound, such as Cu₃Sn、Cu₆Sn₅And (Cu alloy)_xSn_yWhile a layer of essentially pure tin (called free tin) remains on the surface. The free tin layer has a thickness of 0.051 mu m to 3.05 mu mI.e., 2-120 microinches, to provide a solderable, and tarnish-resistant layer. The intermetallic layer may be used to reduce friction by reducing the thickness of the soft free tin layer. For connectors, reducing friction can reduce the required insertion and extraction force.

After the sacrificial layer 72 is deposited, a metal 74, which forms a low resistivity oxide such as silver, is then deposited. The sacrificial layer is deposited to a thickness of 0.025 microns to 3.05 microns, i.e., 1-120 microinches, and preferably to a thickness of 0.13 microns to 0.51 microns, i.e., 5-20 microinches. A preferred method of depositing the sacrificial layer of silver is electroplating using an aqueous solution containing silver cyanide, or immersion plating using a pure cyanide solution as previously described. In addition to silver, indium, iron, niobium, rhenium, ruthenium, vanadium, gold, platinum, palladium and zinc, mixtures of these metals listed in table 1 can also be used as materials for the low-resistivity oxide metal layer.

After the sacrificial layer 74 is deposited, an outermost metal 76 is deposited, which has a melting point lower than any of the substrate, barrier layer, sacrificial layer and low resistivity oxide metal layer. Tin or a tin-based alloy is preferably used as the outermost layer. In most applications, lead is toxic and avoided, but lead-containing tin-based alloys are suitable for use in certain applications. The outermost layer 76 may be deposited using any of the methods previously described, or using special methods of tin deposition such as the HALT method (i.e., hot air leveling tin plating) and mechanical wiping. The outermost layer may be finished brightly or matte (matte) as desired. The matte finish may be electroplated with tin in a tin bath, as is known in the art for preparing this type of finish. Suitable electrolytes include the Solderon mentioned previously^TMST200 and StanTek^TM AMAT。

Tin reflow 78 is then performed, such as by heating the tin to a temperature above its melting point of 232 ℃. The preferred heat treatment is heating at 300 deg.C in air or in a protective atmosphere, such as nitrogen, for 1-10 seconds. The molten tin is then quenched to form a shiny tin surface.

The coated substrate is fabricated 80 into a desired component, such as a portion of a connector, either before or after reflow processing. The coated substrate may also be heated below the melting point of tin in air or nitrogen for about 1 to 168 hours at a temperature of from 150 c to 200 c in order to increase the intermetallic content while reducing the thickness of the free tin to the desired thickness, typically 0.051 microns to 0.51 microns, i.e., 2 microinches to 20 microinches.

Fig. 9A to 9D illustrate the mechanism for improving the coating of the present invention. Shown in fig. 9A is a tin-coated substrate 26 fabricated using the prior art. The substrate 26 is coated with a sacrificial layer 34 of copper, outermost of which is a tin layer 36. After one week of exposure to elevated temperatures, such as 150 degrees celsius, interdiffusion and bonding between the sacrificial layer 34 and the outermost layer 36 occurs, forming Cu near the substrate 26₃A Sn intermetallic layer 82 that diffuses up to the outermost surface 84. After high temperature irradiation, the outermost layer is Cu₃Sn intermetallic compound and Cu₆Sn₅A mixture of intermetallic compounds 86. These two copper intermetallic compounds are susceptible to oxidation, resulting in discoloration and an increase in resistivity.

FIG. 9B illustrates that after the substrate 26 is coated with the sacrificial layer 34, the silver layer 28, and the outermost tin coating 36 and heated at 150 degrees Celsius for one week, the substrate 26 is coated with an interposer 88, which is a mixture of copper and tin, and the outermost layer is comprised of Cu, in accordance with the present invention₃A mixture 90 of silver of Sn intermetallic compounds and a silver-rich phase 92. Silver-rich means that more than 50 atomic percent of silver is contained. Cu₃SnAg_xThe intermetallic compound provides a hard surface to reduce insertion and extraction forces and reduce frictional wear. The silver-rich phase provides corrosion resistance and reduces the increase in resistivity caused by corrosion from frictional wear debris.

FIG. 9C illustrates the substrate 26 being coated with the interposer 96 after the substrate 26 has been coated with the barrier layer 32, the sacrificial layer 34, the silver layer 28, and the outermost tin coating 36 and heated at 150 degrees Celsius for one week in accordance with the present inventionThe interlayer is a mixture of nickel, copper and tin. The layer adjacent to layer 96 is a second layer 98 which is a mixture of nickel, copper, silver and tin. The outermost layer is composed of Cu as the first component₆Sn₅The intermetallic compound, excess tin, a small amount of silver, and the second component are a mixture of silver-rich phases 92.

Fig. 10 is a 2000 x photomicrograph of the outermost surface 84 of the coated substrate of fig. 9C after one week of heating at 150 ℃. The surface is a mixture of a copper-silver-tin phase 98 and a silver-rich phase 92, with the dark areas in the micrograph being the copper-silver-tin phase 98 and the bright areas being the silver-rich phase 92. Fig. 11 is a microscope photograph at 20000 magnification of the coating structure shown in fig. 9C and fig. 10.

Fig. 9D illustrates that after the substrate 26 is coated with the isolation layer 32, the silver layer 28, and the outermost tin layer 36, and heated at 150 c for one week, the substrate 26 is coated with a first interposer 100 that is a mixture of nickel, copper, tin, and a small amount of silver. Coated on the first interposer 100 is a second layer 102, the second layer 102 being a mixture of nickel, copper, tin, and silver. The second layer 102 extends to the surface of the outermost layer, which is primarily the silver-rich phase 92.

The examples given below will clearly illustrate the advantages of the coating system provided by the present invention. The following examples may be used to illustrate the present invention, but do not limit the scope of the present invention.

Examples

Example 1 tin whiskering

Samples of 51 mm x 12.7 mm x 0.25 mm, i.e., 2 inch x 0.5 inch x 0.010 inch, size were cut from the copper alloy C194 bars. The copper alloy C194 comprises 2.1-2.6 wt% of iron, 0.05-0.20 wt% of zinc, 0.015-0.15 wt% of phosphorus, and the balance of copper. The samples were cleaned by placing them in commercial alkaline cleaning solution at 50 degrees celsius for one minute with a cathodic current density of 139 amps per square decimeter, i.e., 15 amps per square inch.

Referring to Table 2, by electroplatingDepositing a nickel layer. The nickel plating solution is a nickel plating solution containing about 60 to 75 grams per liter of nickel, such as nickel sulfamate, and about 6 to 8 grams per liter of NiCl₂And about 38 to 53 grams per liter of an aqueous solution of 53 degrees celsius boric acid at a pH between 3.5 and 4.2. Nickel plating conditions were energized for about 60 seconds with a current density of 279 amps per square decimeter, i.e., 30 amps per square inch.

When depositing copper by electroplating, an aqueous solution is used containing about 20 to 70 grams per liter of copper, about 50 to 200 grams per liter of H at 40 to 60 degrees Celsius₂SO₄. The plating conditions were to be energized with a current density of 372 amps per square decimeter, i.e., 40 amps per square inch, for about 40 seconds.

When depositing a silver layer by electroplating, an aqueous solution is used containing 31 to 56 grams per liter of silver cyanide, 50 to 78 grams per liter of potassium cyanide, 15 to 90 grams per liter of potassium carbonate and a brightener. Operating conditions are a temperature of between 20 and 28 degrees celsius and a current density of between 46 and 139 amps per square decimeter, i.e., 5 to 15 amps per square inch.

When depositing the tin layer by electroplating, the matte tin deposit is formed by MacDermid Stantek^TMThe AMAT solution produces bright tin deposits using MacDermid Stantek^TMStellite (Stellite) 100 solution. The plating conditions were to be energized at 25 to 40 degrees celsius for about 50 to 400 seconds with a current having a current density of 279 amps per square decimeter, i.e., 30 amps per square inch.

Accelerated tin whisker testing was performed on samples bent and constrained in circular grooves with a radius of 76 mm, i.e., 3 inches. In this way, a constant bending stress can be generated on the tin coating, promoting tin whisker formation. The formation of tin whiskers on the compressed edge (concave surface) of the exemplary and comparable samples can be observed periodically under an optical microscope at 500X.

As can be seen from table 2, the inclusion of a silver layer in direct contact with the tin coating substantially eliminates tin whisker formation regardless of whether the outermost layer is coated with matte tin or bright tin. For the case where the outermost layer is coated with matte tin, the results of sample 2, sample 3, sample 4, and sample 5 can be compared. For comparison, sample 1 was a commercial thick tin product. For the case where the outermost layer is coated with bright tin, sample 20 and sample 21 can be compared.

Example 2 Effect of frictional wear on contact resistance

The samples were made using the copper alloys C194 and C7025 listed in table three, wrought tin monoliths and wrought silver monoliths, and had dimensions of 152 mm x 31.8 mm x 0.13 mm, i.e., 6 inches x 1.25 inches x 0.005 inches. The C7025 comprises 2.2 to 4.2 weight percent of nickel, 0.25 to 1.2 weight percent of silicon, 0.05 to 0.3 weight percent of magnesium, and the balance of copper.

Copper alloy samples were coated with an interposer and matte tin as in example 1, but the silver layer was MacDermid Sterling^TMIs deposited by immersion plating with SnSO containing 20 to 80 grams per liter of tin ions₄50 to 200 grams per liter of sulfuric acid and organic additive in a sulfate solution.

The effect of frictional wear on the contact resistance was determined by rotating a bump having a diameter of 6.4 mm, i.e. 0.25 inch, at a frequency of 5 hz, along a 20 micron long circle on the contact surface to be tested, for up to 20000 turns. A normal force of 100g was applied to the bump and contact resistance data was collected during bump movement. The number of turns recorded is the number of turns required to obtain a determined contact resistance. Higher number of turns indicates better abrasion resistance of the sample.

TABLE 3

Sample (I)	Substrate	Interposer	Surface layer	Surface layer thickness (microinches)	Number of turns corresponding to 10 milliohm contact resistance	Number of turns corresponding to 10 ohm contact resistance
							1	C194	Nickel/copper	Tin with rough surface	0.51 micron (20)	61	3269
2	C194	Nickel/copper/0.13 micron (5 microinch) silver	Tin with rough surface	0.51 micron (20)	79	4400
3	C194	Is free of	Tin with rough surface	1.02 micron (40)	116	2269
							4	C194	0.13 micron (5 microinch) silver	Tin with rough surface	1.07 micron (42)	490	＞5000
							5	Tin oxide	Is free of	Is free of	N/A	253	6530
6	Tin oxide	Is free of	Is free of	N/A	＞20000	＞20000

^*Indicating that the test terminated after 5000 cycles

Comparing sample 1 with sample 2 made using the present invention, it can be seen that the 0.13 micron (i.e., 5 micro-inches) silver layer applied to sample 2 effectively reduces the resistance generated by frictional wear on the substrate, and compared to sample 1, sample 2 has about 30% more turns to achieve a 10 milliohm contact resistance and about 35% more turns to achieve a 10 ohm contact resistance.

Comparing sample 3 with sample 4 made using the present invention, it can be seen that the 0.13 micron (i.e., 5 micro-inches) silver layer coated on sample 4 effectively reduces the resistance generated by frictional wear on the substrate, and compared to sample 3, sample 4 has an increase of approximately 322% in the number of turns required to achieve a 10 milliohm contact resistance and an increase of more than 120% in the number of turns required to achieve a 10 ohm contact resistance.

The properties of the monolithic cured silver (i.e., sample 6) were better than either of the coated copper substrate samples, but were not suitable for use in the fabrication of electrical connectors due to cost and rust. The entire block of wrought tin (i.e., sample 5) may have better wear resistance due to the presence of a large amount of free tin or increased hardness due to rotation, but is not suitable for connector fabrication due to low strength.

Example 3 coefficient of friction

A copper alloy C194 sample having dimensions of 152 mm x 31.8 mm x 0.13 mm (i.e., 6 inches x 1.25 inches x 0.005 inches) was coated with an interposer and matte tin as described above. After heating the sample to 350 degrees celsius in air and quenching in water, a reflowed tin surface is formed.

The coefficient of friction was determined as the ratio of the drag force generated by a bump having a diameter of 6.4 millimeters (i.e., 0.25 inches) on a tin-coated flat surface at a rate of 3 millimeters per second for 10 revolutions to the normal force, i.e., R/N. The normal force is the dead weight loaded and no lubricant is used between the tin coated flat and the bump. The resistance is measured as the relative sliding of the bump on the flat surface of the sample. The values reported are the average of all 10 cycles. A lower R/N indicates a lower coefficient of friction. Table 4 lists the average values measured for 10 revolutions of the cam.

TABLE 4

As the R/N decreases, the insertion force required to insert the probe into the receptacle decreases. Comparing sample 3 with samples 1 and 2, it can be seen that coating with a 5 micron silver layer can reduce the R/N value of the outer matte tin by 14%. Comparing sample 3 with sample 4, it can be seen that further increasing the thickness of the silver layer not only does not have a significant beneficial effect on the coefficient of friction, but also results in increased costs.

A comparison of sample 5 and sample 6 shows that better results are achieved when reflow tin is used as the outermost coating, and the R/N can be reduced by about 45%.

Example 4 interlayer interdiffusion

Tables 5 through 8 list compositional measurements of the structures shown in figures 9A through 9D to illustrate the formation of a silver-rich phase on the outermost surface of a substrate coated in accordance with the present invention. The thickness of the sample was measured in microinches before the temperature reached 150 degrees celsius after heating for one week using XRF (i.e., x-ray fluorescence spectroscopy). The composition and atomic percent of the heated sample were determined by EDX (i.e., energy dispersive x-ray fluorescence spectrometer).

Watch 5 (FIG. 9A)

Reference numerals in fig. 9A	Components	Thickness micron (micro-inch)		Reference numerals in fig. 9A	Components	Atomic percent
26	C194	N.A.		26	C194	N.A.
							34	Copper (Cu)	0.51-1.02(20-40)		82	Copper tin	75％25％
36	Tin (Sn)	1.02-2.03(40-80)		86	Copper tin	56％44％

Watch 6 (fig. 9B)

Reference numerals in fig. 9B	Components	Thickness micron (micro-inch)		Reference numerals in fig. 9B	Components	Atomic percent
26	C194	N.A.		26	C194	N.A.
							34	Copper (Cu)	0.51-1.02(20-40)		88	Copper tin	79％21％
28	Silver (Ag)	0.13-0.25(5-10)		90	Copper tin silver	74％23％3％
							36	Tin (Sn)	1.02-2.03(40-80)		92	Silver tin copper	56％25％19％

Table 7 (FIG. 9C)

Reference numerals in fig. 9C	Components	Thickness micron (micro-inch)		Reference numerals in fig. 9C	Components	Atomic percent
26	C194	N.A.		26	C194	N.A.
							32	Nickel (II)	0.13-0.51(5-20)		96	Copper nickel tin	42％32％26％
34	Copper (Cu)	0.18-0.46(7-18)		98	Copper tin nickel silver	50％41％7％2％
							28	Silver (Ag)	0.13-0.25(5-10)		94	Tin, copper and silver	77％17％6％
36	Tin (Sn)	1.02-2.03(40-80)		92	Silver tin copper	56％31％13％

Table 8 (fig. 9D)

Reference numerals in fig. 9D	Components	Thickness micron (micro-inch)		Reference numerals in fig. 9D	Components	Atomic percent
26	C194	N.A.		26	C194	N.A.
							32	Nickel (II)	0.13-0.51(5-20)		100	Tin-nickel-copper-silver	41％34％24％1％
28	Silver (Ag)	0.13-0.25(5-10)		102	Tin silver copper nickel	35％27％23％15％
							36	Tin (Sn)	1.02-2.03(40-80)		92	Silver tin copper	64％26％10％

It should be noted that the EDX analysis results may deviate by a few percent due to the broadening of the x-rays and the thickness of the penetration. However, for comparison purposes, the above results can effectively distinguish the samples.

Although the invention has been shown and described with respect to the illustrated embodiments, it should be understood that the foregoing and various other changes, omissions and additions in the form and detail thereof may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (31)

1. An electrically conductive material covered with a plurality of layers (32, 34, 40, 36), comprising:

a conductive substrate (26);

an isolation layer (32) deposited on the substrate (26) for preventing diffusion of a component of the substrate (26) into the plurality of layers (32, 34, 40, 36);

a sacrificial layer (34) deposited on the isolation layer (32) for forming an intermetallic with tin;

a low resistivity oxide metal layer (40) deposited on the sacrificial layer (34), the metal of the low resistivity oxide metal layer being a metal whose oxide has a lower resistivity than tin oxide; and

an outermost layer (36) of tin or a tin-based alloy deposited directly on the low resistivity oxide metal layer (40).

2. The conductive material of claim 1, wherein a copper layer is disposed between the conductive substrate (26) and the isolation layer (32).

3. The conductive material of claim 1, wherein the isolation layer (32) is selected from the group consisting of nickel, cobalt, iron, manganese, chromium, molybdenum, and alloys of these metals.

4. The conductive material of claim 3, wherein the barrier layer (32) is nickel or a nickel-based alloy.

5. The conductive material of claim 4, wherein the nickel or nickel-based alloy barrier layer (32) has a thickness in the range of 0.051 microns to 2.03 microns, i.e., 2 microinches to 80 microinches.

6. The conductive material of claim 4, wherein the sacrificial layer (34) is copper or a copper-based alloy.

7. The conductive material of claim 6, wherein the copper or copper-based alloy sacrificial layer (34) has a thickness in the range of 0.051 microns to 1.52 microns, i.e., 2 microinches to 60 microinches.

8. The conductive material of claim 7, wherein said low resistivity oxide metal layer (40) is selected from the group consisting of silver, indium, iron, zinc, niobium, rhenium, ruthenium, vanadium, gold, platinum, palladium and alloys of these metals.

9. The conductive material of claim 8, wherein the low resistivity oxide metal layer (40) has a thickness in the range of 0.025 micron to 3.05 micron, 1 micro inch to 120 micro inches.

10. The conductive material of claim 9, wherein the low resistivity oxide metal layer (40) is silver or a silver-based alloy.

11. The conductive material of claim 10, wherein the silver or silver-based alloy layer (40) has a thickness ranging from 0.05 microns to 1.02 microns, i.e., from 2 microinches to 40 microinches.

12. The conductive material of claim 10, wherein the outermost tin or tin-based alloy layer (36) has an average particle size in a range of 0.1 microns to 100 microns.

13. An electrical connector comprising a socket (42) and a probe (44), and at least one of the socket (42) or the probe (44) comprising:

a conductive substrate (26);

an isolation layer (32) deposited on the substrate (26) for preventing diffusion of a component of the substrate (26) into the plurality of layers (32, 34, 40, 36);

a sacrificial layer (34) deposited on the isolation layer (32) for forming an intermetallic with tin;

a low resistivity oxide metal layer (40) deposited on the sacrificial layer (34), the metal of the low resistivity oxide metal layer being a metal whose oxide has a lower resistivity than tin oxide; and

an outermost layer (36) of tin or a tin-based alloy deposited directly on the low resistivity oxide metal layer (40).

14. The electrical connector of claim 13, wherein a copper layer is disposed between said conductive substrate (26) and said isolation layer (32).

15. The electrical connector of claim 13, wherein said barrier layer (32) is selected from the group consisting of nickel, cobalt, iron, manganese, chromium, molybdenum and alloys of these metals.

16. The electrical connector of claim 15, wherein said insulating layer (32) has a thickness in the range of 0.1 micron to 1.02 microns, i.e. 4 microinches to 40 microinches.

17. The electrical connector of claim 16, wherein said barrier layer (32) is nickel or a nickel-based alloy.

18. An electrical connector as in claim 17, wherein said sacrificial layer (34) is copper or a copper-based alloy.

19. The electrical connector of claim 18, wherein said low resistivity oxide metal layer (40) is selected from the group consisting of silver, indium, iron, zinc, niobium, rhenium, ruthenium, vanadium, gold, platinum, palladium and alloys of these metals.

20. The electrical connector of claim 19, wherein said low resistivity oxide metal layer (40) is silver or a silver-based alloy.

21. The electrical connector of claim 20, wherein said outermost tin or tin-based alloy layer (36) has an average particle size in the range of 0.1 microns to 100 microns.

22. A method for making an element resistant to increased resistivity due to frictional wear debris, comprising the steps of:

(a) covering (70) the conductive substrate with an isolating layer for preventing diffusion of a component of the conductive substrate through the isolating layer at an intended operating temperature of the component;

(b) covering (72) the isolation layer with a sacrificial layer for forming an intermetallic compound with an outermost covering layer of the component;

(c) covering (74) the sacrificial layer with a low resistivity oxide metal or metal alloy layer, the metal of the low resistivity oxide metal or metal alloy layer being a metal whose oxide has a lower resistivity than the tin oxide;

(d) directly overlaying (76) said low resistivity oxide metal or metal alloy layer with an outermost layer having a melting point lower than the melting point of any of said conductive substrate, said barrier layer, said sacrificial layer, said low resistivity oxide metal or metal alloy layer;

(e) heating (78) the overlying conductive substrate to a temperature effective to melt and reflow the outermost layer.

23. The method of claim 22, wherein the conductive substrate is coated with a layer of copper prior to step (a).

24. The method of claim 23, wherein the barrier layer is selected from the group consisting of nickel, cobalt, iron, manganese, chromium, molybdenum, and alloys and mixtures of these metals.

25. The method of claim 24, wherein the sacrificial layer is copper or a copper-based alloy.

26. The method of claim 25 wherein said low resistivity oxide metal layer is selected from the group consisting of silver, indium, iron, niobium, rhenium, ruthenium, vanadium, gold, platinum, palladium, and zinc.

27. The method of claim 25, wherein said low resistivity oxide metal layer is silver or a silver-based alloy deposited (74) by electroplating or immersion plating, the layer having a thickness in the range of 0.025 to 3.05 microns, i.e. 1 to 120 microinches.

28. The method of claim 27 wherein said steps (d) (76) and (e) (78) are combined into a single step comprising a hot air leveling tin plating HALT process.

29. The method of claim 28, wherein the outermost layer is tin or a tin-based alloy deposited (76) by an electroplating process.

30. The method of claim 29, wherein the covered conductive substrate is heated to a temperature below the melting point of the tin or tin-based alloy to reduce the thickness of free tin before or after step (e).

31. A method as claimed in claim 22, including the step of forming (80) said covered conductive substrate into socket and probe elements forming a connector.