PROCESS FOR ELECTRICALLY CONNECTING CONTACT SURFACES OF ELECTRONIC COMPONENTS

Abstract

A process for electrically connecting contact surfaces of electronic components by capillary wedge bonding a round wire of 8 to 80 μm to the contact surface of a first electronic component, forming a wire loop, and stitch bonding the wire to the contact surface of a second electronic component, wherein the wire comprises a wire core having a silver or silver-based wire core with a double-layered coating comprised of a 1 to 50 nm thick inner layer of nickel or palladium and an adjacent 5 to 200 nm thick outer layer of gold.

Claims

1. A process for electrically connecting a contact surface of a first electronic component with a contact surface of a second electronic component comprising the subsequent steps: (1) capillary wedge bonding a wire having a circular cross-section with an average diameter in the range of 8 to 80 μm to the contact surface of the first electronic component, (2) raising the capillary wedge bonded wire to form a wire loop between the capillary wedge bond formed in step (1) and the contact surface of the second electronic component, and (3) stitch bonding the wire to the contact surface of the second electronic component, wherein the capillary wedge bonding of step (1) is carried out with a ceramic capillary having a lower face angle within the range of from zero to 4 degrees, wherein the wire comprises a wire core with a surface, the wire core having a double-layered coating superimposed on its surface, wherein the wire core consists of a material selected from the group consisting of pure silver, doped silver with a silver content of >99.5 wt.-% and silver alloys with a silver content of at least 89 wt.-%, and wherein the double-layered coating comprises a 1 to 50 nm thick inner layer of nickel or palladium and an adjacent 5 to 200 nm thick outer layer of gold.

2. The process of claim 1, wherein step (1) is carried out with a KNS-iConn bonder and wherein the capillary wedge bonding process parameters include at least one of (a′) to (g′): (a′) an ultrasonic energy in a range of 50 to 100 mA, (b′) a force in a range of 10 to 30 g, (c′) a constant velocity in a range of 0.3 to 0.7 μm/s, (d′) a contact threshold in a range of 60 to 70%, (e′) a bonding temperature in a range of 25 to 175° C., (f) a tail length extension in a range of 200 to 500 μm, (g′) an ultrasonic ramp in a range of 0 to 50%

3. The process of claim 1, wherein the wire core consists of pure silver consisting of 99.95 to 100 wt.-% of silver and up to 500 wt.-ppm of further components other than silver.

4. The process of claim 1, wherein the wire core consists of doped silver which consists of >99.5 to 99.997 wt.-% of silver, 30 to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver and the at least one doping element.

5. The process of claim 1, wherein the wire core consists of a silver alloy which consists of 89 to 99.50 wt.-% of silver, 0.50 to 11 wt.-% of at least one alloying element, up to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver, the at least one alloying element and the at least one doping element.

6. The process of claim 5, wherein the at least one alloying element is selected from the group consisting of palladium, gold, nickel, platinum, copper, rhodium and ruthenium.

7. The process of claim 5, wherein the at least one doping element is selected from the group consisting of calcium, nickel, platinum, copper, rhodium and ruthenium.

8. The process of claim 1, wherein the first electronic component is a substrate having a contact surface or a semiconductor having a contact surface in the form of a bond pad and the second electronic component is a substrate having a contact surface or a semiconductor having a contact surface in the form of a bond pad.

9. The process of claim 8, wherein the first electronic component is a semiconductor having a contact surface in the form of a bond pad and the second electronic component is a substrate having a contact surface.

10. The process of claim 8, wherein the first electronic component is a substrate having a contact surface and the second electronic component is a semiconductor having a contact surface in the form of a bond pad.

11. The process of claim 2, wherein the wire core consists of doped silver which consists of >99.5 to 99.997 wt.-% of silver, 30 to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver and the at least one doping element.

12. The process of claim 2, wherein the wire core consists of a silver alloy which consists of 89 to 99.50 wt.-% of silver, 0.50 to 11 wt.-% of at least one alloying element, up to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver, the at least one alloying element and the at least one doping element.

13. The process of claim 12, wherein the at least one alloying element is selected from the group consisting of palladium, gold, nickel, platinum, copper, rhodium and ruthenium.

14. The process of claim 12, wherein the at least one doping element is selected from the group consisting of calcium, nickel, platinum, copper, rhodium and ruthenium.

15. The process of claim 1, wherein step (1) is carried out with a Shinkawa bonder and wherein the capillary wedge bonding process parameters include at least one of (a″) to (h″): (a″) an ultrasonic energy in a range of 50 to 100 mA, (b″) a force in a range of 10 to 30 g, (c″) a constant velocity in a range of 0.3 to 0.7 μm/s, (d″) a contact threshold in a range of 60 to 70%, (e″) a bonding temperature in a range of 25 to 175° C., (f″) a cut tail length in a range of 85 to 110 μm, (g″) a sink amount in a range of −6 to −12 μm, (h″) an ultrasonic sloping in a range of 0 to 50%.

16. The process of claim 15, wherein the wire core consists of doped silver which consists of >99.5 to 99.997 wt.-% of silver, 30 to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver and the at least one doping element.

17. The process of claim 15, wherein the wire core consists of a silver alloy which consists of 89 to 99.50 wt.-% of silver, 0.50 to 11 wt.-% of at least one alloying element, up to <5000 wt.-ppm of at least one doping element and up to 500 wt.-ppm of further components other than silver, the at least one alloying element and the at least one doping element.

18. The process of claim 17, wherein the at least one alloying element is selected from the group consisting of palladium, gold, nickel, platinum, copper, rhodium and ruthenium.

19. The process of claim 17, wherein the at least one doping element is selected from the group consisting of calcium, nickel, platinum, copper, rhodium and ruthenium.

Description

EXAMPLES

[0105] Test Methods A. to H.

[0106] All tests and measurements were conducted at T=20° C. and a relative humidity RH=50%.

[0107] A. Salt-Solution Soaking Test of Capillary Wedge Bonds:

[0108] The wires were capillary wedge bonded to Al-0.5 wt.-% Cu bond pads. The test devices with the so-bonded wires were soaked in salt-solution at 25° C. for 10 minutes, washed with deionized (DI) water and later with acetone. The salt-solution contained 20 wt.-ppm NaCl in DI water. The number of lifted wedge bonds were examined under a low power microscope (Nikon MM-40) at 100× magnification. Observation of a higher number of lifted capillary wedge bonds indicated severe interfacial galvanic corrosion.

[0109] B. Moisture Resistance Test of Capillary Wedge Bonds:

[0110] The wires were capillary wedge bonded to Al-0.5 wt.-% Cu bond pads. The test devices with the so-bonded wires were stored at 130° C. temperature, 85% relative humidity (RH) for 8 hours in a highly accelerated stress test (HAST) chamber and later examined for the number of lifted wedge bonds under a low power microscope (Nikon MM-40) at 100× magnification.

[0111] Observation of a higher number of lifted capillary wedge bonds indicated severe interfacial galvanic corrosion.

[0112] C. Electrical Resistivity:

[0113] Both ends of a test specimen, i.e. a wire of 1.0 meter in length, were connected to a power source providing a constant current/voltage. The resistance was recorded with a device for the supplied voltage. The measuring device was a HIOKI model 3280-10, and the test was repeated with at least 10 test specimens. The arithmetic mean of the measurements was used for the calculations given below.

[0114] The resistance R was calculated according to R=V/I.

[0115] The specific resistivity ρ was calculated according to ρ=(R×A)/L, wherein A is the mean cross-sectional area of the wire and L the length of the wire between the two measuring points of the device for measuring the voltage.

[0116] The specific electrical conductivity σ was calculated according to σ=1/ρ.

[0117] D. Electro-Migration Test of Wires:

[0118] Two wires of 75 μm diameter were kept parallel within a millimeter distance on a glass plate under the objective lens of a low power microscope Nikon MM40 model at 50× magnification. A water drop was formed by a micropipette between the two wires to be connected electrically. One wire was connected to a positive and the other to a negative pole and +5 V was given to the wires. The two wires were biased with +5 V direct current in a closed circuit, connected in series with a 10 kΩ resistor. The circuit was closed by wetting the two wires with a few drops of de-ionized water as an electrolyte. Silver electro-migrated from the cathode to the anode in the electrolyte forming silver dendrites, sometimes the two wires bridged. The rate of growth of silver dendrites strongly depended on the wires' coating layer and—in case of a silver alloy wire core—the alloying additions.

[0119] E. Vickers Micro-Hardness:

[0120] The hardness was measured using a Mitutoyo HM-200 testing equipment with a Vickers indenter. A force of 10 mN indentation load was applied to a test specimen of wire for a dwell time of 12 seconds. The testing was performed on the center of the wire or the FAB.

[0121] F. Capillary Wedge Bonding (1.sup.st Wedge) Process Window Area:

[0122] Measurements of the capillary wedge bonding process window area were done by a standard procedure. The test wires were capillary wedge bonded to a Al-0.5 wt % Cu bond pad of a silicon die using a KNS-iConn bonder tool (Kulicke & Soffa Industries Inc., Fort Washington, Pa., USA). The vital capillary wedge bonding process parameters were: ultrasonic energy of 75 mA, compressive force of 20 g, constant velocity of 0.5 μm/s, contact threshold of 65%, bonding temperature of 150° C., tail length extension of 350 μm, ultrasonic ramp of 25%. The process window values were based on a wire having an average diameter of 17.5 μm.

[0123] The four corners of the process window were derived by overcoming the two main failure modes:

[0124] (1) supply of too low force and ultrasonic energy lead to non-stick on pad (NSOP) of the wire, and

[0125] (2) supply of too high force and ultrasonic energy lead to short wire tail (SHTL).

[0126] G. Stitch Bonding (2.sup.nd Wedge) Process Window Area:

[0127] Measurements of the stitch bonding process window area were done by a standard procedure. The test wires were stitch bonded to a gold plated lead finger on a BGA (ball grid array) substrate using a KNS-iConn bonder tool (Kulicke & Soffa Industries Inc., Fort Washington, Pa., USA). The process window values were based on a wire having an average diameter of 17.5 μm.

[0128] The four corners of the process window were derived by overcoming the two main failure modes:

[0129] (1) supply of too low force and ultrasonic energy lead to non-stick on lead finger (NSOL) of the wire, and

[0130] (2) supply of too high force and ultrasonic energy lead to short wire tail (SHTL).

[0131] H. Elongation (EL):

[0132] The tensile properties of the wires were tested using an Instron-5564 instrument. The wires were tested at 2.54 cm/min speed, for 254 mm gauge length (L). The load and elongation on fracture (break) were acquired as per ASTM standard F219-96. The elongation was the difference in the gauge length (ΔL) of the wire between start and end of the tensile test, usually reported in percentage as (100.Math.ΔL/L), calculated from the recorded load versus extension tensile plot. The tensile strength and the yield strength were calculated from the break and yield load divided by the wire area. The actual diameter of the wire was measured by the sizing method, weighing a standard length of the wire and using the density of it.

[0133] Wire Samples 1 to 12

[0134] A quantity of silver (Ag), palladium (Pd) and gold (Au) of at least 99.99% purity (“4N”) in each case were melted in a crucible. Small amounts of silver-nickel, silver-calcium, silver-platinum or silver-copper master alloys were added to the melt and uniform distribution of the added components was ascertained by stirring. The following master alloys were used:

TABLE-US-00001 Master Alloy Composition Ag-0.5 wt.-% Ni 99.5 wt.-% Ag 0.5 wt.-% Ni Ag-0.5 wt.-% Ca 99.5 wt.-% Ag 0.5 wt.-% Ca Ag-0.5 wt.-% Pt 99.5 wt.-% Ag 0.5 wt.-% Pt Ag-0.5 wt.-% Cu 99.5 wt.-% Ag 0.5 wt.-% Cu

[0135] For the alloys of Table 1 the corresponding combination of master alloys were added.

[0136] Then a wire core precursor item in the form of 8 mm rods was continuous cast from the melt. The rods were then drawn in several drawing steps to form a wire core precursor having a circular cross-section with a diameter of 134 μm. The wire core precursor was intermediate batch annealed at an oven set temperature of 500° C. for an exposure time of 60 minutes and then electroplated with a double layer coating consisting of an inner palladium (or nickel) layer and an outer gold layer and thereafter further drawn to a final diameter of 17.5 μm and a final palladium or nickel layer thickness within the range of 1 to 4 nm and a final gold layer thickness within the range of 10 to 18 nm, followed by a final strand annealing at an oven set temperature of 220° C. for an exposure time of 0.6 seconds, immediately followed by quenching the so-obtained coated wires in water containing 0.07 vol.-% of surfactant.

[0137] By means of this procedure, several different samples 1 to 12 of coated silver and silver-based wires and an uncoated reference silver wire of 4N purity (Ref) were manufactured. Table 1 shows the composition of wires having a diameter of 17.5 μm. The composition was determined by ICP.

TABLE-US-00002 TABLE 1 Coating contribution of the Total coating's constituents Core Chemistry Au + Pd in wt.-% of coated wire wt.-ppm wt.-% wt.-% of Sample Au Pd Ni Ni Ca Pt Au Pd coated wire 4N Ag — — — 2 2 2 0.0002 0.0002 — (Ref) 1 0.5 0.05 — 2 2 2 1 1 2.55 2 0.5 0.05 — 2 2 2 1 3 4.55 3 0.5 0.05 — 2 30 2 1 3 4.55 4 0.5 0.05 — 10 20 10 1 3 4.55 5 1 0.05 — 2 2 2 1 3 5.05 6 1 0.05 — 2 2 2 1 3 5.05 7 0.5 — 0.04 2 2 2 1 1 2.5 8 0.5 — 0.04 2 2 2 1 3 4.5 9 0.5 — 0.04 2 30 2 1 3 4.5 10 0.5 — 0.04 10 20 10 1 3 4.5 11 1 0.05 — 2 2 2 1 3 5.05 12 1 0.1  — 2 2 2 1 3 5.1

[0138] Table 2 below shows certain test results. All tests were carried out with wires having a diameter of 17.5 μm, except for the electro-migration test which was performed with wires of 75 μm diameter.

TABLE-US-00003 TABLE 2 Sample Ref 1 2 3 4 5 6 7 8 9 10 11 12 Mechanical Elongation 7.4 4.1 4.4 4.0 3.9 3.8 4.0 4.2 4.4 4.3 4.1 4.1 4.2 properties (%) Tensile 183 470 481 485 483 486 481 483 480 482 484 481 481 strength (MPa) Yield 130 270 275 292 278 275 282 280 275 291 283 283 281 strength (MPa) Micro- 58 70 76 80 80 71 73 72 75 80 78 73 72 hardness, HV (10 mN/12 s) Electrical Resistivity 1.6 3.31 3.32 3.32 3.32 3.32 3.31 3.32 3.32 3.32 3.32 3.32 3.32 properties (μΩ .Math. cm) Salt-solution % capillary 80 1 0 0 0 0 0 1 0 0 0 0 0 soaking test wedge lift Moisture % capillary 50 2 0 0 0 0 0 1 0 0 0 0 0 resistance wedge lift test Electro- Rate of 25 0 0 0 0 0 0 0 0 0 0 0 0 migration growth of test silver dendrites (μm/s) Bonding 1st bond 50 250 440 440 430 430 410 390 500 450 440 440 400 process (capillary window wedge bond) (mA .Math. g) 2nd bond 225 75 140 130 130 130 130 125 150 135 135 135 130 (stitch bond) (mA .Math. g)