micromachines

Article Effects of Bonding Parameters on Free Air Ball Properties and Bonded Strength of Ag-10Au-3.6Pd Alloy Bonding

Jun Cao 1,* , Junchao Zhang 1, John Persic 2 and Kexing Song 3 1 School of Mechanical and Power Engineering, Henan Polytechnic University, Jiaozuo 454000, China; [email protected] 2 Microbonds Inc, 7495 Birchmount Rd., Markham 60428, ON L3R 5G2, Canada; [email protected] 3 School of Materials Science and Engineering, Henan University of Science and Technology, Luoyang 471000, China; [email protected] * Correspondence: [email protected]

 Received: 8 July 2020; Accepted: 12 August 2020; Published: 14 August 2020 

Abstract: Free air ball (FAB) and bonded strength were performed on an Ag-10Au-3.6Pd alloy bonding wire (diameter of 0.025 mm) for different electronic flame-off (EFO) currents, times and bonding parameters. The effects of the EFO and bonding parameters on the characteristics of the FAB as well as the bonded strength were investigated using scanning electron microscopy. The results showed that, for a constant EFO time, the FAB of the Ag-10Au-3.6Pd alloy bonding wire transitioned from a pointed defined ball to an oval one, then to a perfectly shaped one, and finally to a golf ball with an increase in the EFO current. On the other hand, when the EFO current was constant and the EFO time was increased, the FAB changed from a small ball to a perfect one, then to a large one, and finally to a golf ball. The FAB exhibited the optimal geometry at an EFO current of 0.030 A and EFO time of 0.8 ms. Further, in the case of the Ag-10Au-3.6Pd alloy bonding wire, for an EFO current of 0.030 A, the FAB diameter exhibited a nonlinear relationship with the EFO time, which could be expressed by a quadratic function. Finally, the bonded strength decreased when the bonding power and force were excessively high, causing the ball bond to overflow. This led to the formation of neck cracks and decrease in the bonded strength. On the other hand, the bonded strength was insufficiently when the bonding power and force were small. The bonded strength was of the desired level when the bonding power and force were 70 mW and 0.60 N (for the ball bonded) and 95 mW and 0.85 N (for the wedge bonded), respectively.

Keywords: EFO current; EFO time; bonding power; bonding force; bonded strength

1. Introduction With the vigorous development of the information age, the electronic information industry has become an extremely important part of today’s society. Electronic products have been widely used in all aspects of social life, and promote the continuous development and technological innovation of the entire electronic industry [1,2]. The electronic and physical functions of and electronic devices can be transformed into the form suitable for machine systems and serve human society [3]. Electronic packaging technology and bonding materials directly affect the quality and service life of electronic products, so it is necessary to study the packing process. Wire bonding is a kind of electronic interconnection technology which uses fine metal wire to provide connection capability by heat, pressure and ultrasonic energy [4,5]. Wire bonding has the advantages of simple process, low cost and suitable for a variety of packaging forms, wire bonding is the most widely used interconnection technology in microelectronic packaging [6].

Micromachines 2020, 11, 777; doi:10.3390/mi11080777 www.mdpi.com/journal/micromachines Micromachines 2020, 11, 777 2 of 15

Among the costs required for all IC products, the cost proportion of packaging has become increasingly high recently. In recent years, this figure has even reached 40%. At the same time, more than 25% of the failure rate of IC products is caused by package failure [7,8]. Nowdays, there are many kinds of bonding materials, among which the base alloy bonding exhibit excellent mechanical properties, good oxidation resistance, and high reliability and are cost effective. As a result, they can limit light attenuation and improve the conversion rate in light-emitting (LED) packaging. Because of these advantages, these wires are employed widely in and LED packaging [9–16]. In microelectronic packaging, the chip and the outside world can achieve smooth input and output through the connection of external pins and integrated circuits, which is the core of the whole microelectronic packaging process [17,18]. The factors that may affect the bonding quality are bonding time, bonding power, bonding force and temperature [19,20]. With respect to the wire bonding process, the shape of the free air ball (FAB) has a determining effect on the strength of the joint formed between the lead and the substrate. The bonding wire melts under the action of the electronic flame-off (EFO), and the FAB is formed because of the surface tension and gravity. The size of the FAB is small (the diameter is approximately 35–55 µm), and it solidifies rapidly. The surface tension and solidification characteristics of the FAB during the melting process determine its shape [21,22]. Once the FAB is formed, it is bonded to the aluminum substrate of the chip, with the bonding parameters such as the bonding pressure, ultrasonic power, and bonding time, determining the bonding characteristics. A diffusion layer (of an intermetallic compound) is formed because of surface and dislocation diffusion, and it also affects the bonded strength [23]. The geometry of the FAB determines the initial bond strength of the ball joint. Further, the sphericity and roundness of the FAB directly affects the bonded strength between the chip and the substrate, which is the primary factor affecting the strength of the ball bonded. The composition of the Ag-10Au-3.6Pd alloy is more complex than that of pure metal. As a result, the geometry of the FAB is affected to a greater degree by the EFO parameters. On the other hand, in the case of the Ag-10Au-3.6Pd alloy bonding wire, the bonded strength, which determines the device life, is also affected by the bonding power and force. It is also the future development trend to optimize the bonding process parameters to improve the wire bonding quality. In recent years, there have been several experimental studies on the ball-forming properties of alloy wires. Chen et al. [24] reported that the EFO current and time are the two most important parameters related to the balling process. Tan et al. [25] performed an experimental study on the burning of balls of different diameters and determined the empirical formula for describing the diameter of the copper ball based on the ball-burning current and time. Xu et al. [26] studied the effects of the ultrasonic power and bonding pressure on the bonded strength during bonding using Au wire and concluded that a mismatch between the ultrasonic power and bonding pressure can significantly reduce the bonded strength. However, the above-described studies were mostly performed on wires of pure metals such as Cu and Au, and there has been little research on the effects of the ball-burning parameters on the ball-forming properties and bonded strength of the Ag-10Au-3.6Pd alloy bonding wire. In this study, the geometry of the FAB and the bonded strength of the Ag-10Au-3.6Pd alloy bonding wire was investigated for different EFO currents and times and bonding powers and forces, and the optimal bonding parameters for the wire were determined. The results of this study should serve as the theoretical basis for the use of the Ag-10Au-3.6Pd alloy bonding wire in microelectronic packaging.

2. Materials and Methods An Ag-10Au-3.6Pd alloy bonding wire with a diameter of 0.025 mm was employed in this study, shown in Figure1. The wire had average elongation of 16.4% and average tensile strength of 0.102 N (broken load). Micromachines 2020, 11, 777 3 of 16 flow rate of 0.6 L/min to protect the wire from oxidation. The geometry and size of the FAB and the bond geometry of the bonding wire (diameter of 0.025 mm) were investigated using a scanning electron microscopy (SEM) system, and the relationship between the EFO time and FAB diameter was determined. The bonded strength (ball shear/pull strength and wedge pull strength) were measuredMicromachines using2020, 11a ,DAGE 777 Series 4000 BS250 system at a test speed of 500 μm/s and shear height3 of 155 μm.

Micromachines 2020, 11, 777 3 of 16

flow rate of 0.6 L/min to protect the wire from oxidation. The geometry and size of the FAB and the bond geometry of the bonding wire (diameter of 0.025 mm) were investigated using a scanning electron microscopy (SEM) system, and the relationship between the EFO time and FAB diameter was determined. The bonded strength (ball shear/pull strength and wedge pull strength) were measured using a DAGE Series 4000 BS250 system at a test speed of 500 μm/s and shear height of 5 μm. Figure 1. Ag-10Au-3.6Pd alloy bonding wire. Figure 1. Ag-10Au-3.6Pd alloy bonding wire. A KAIJO FB-988 automatic wire bonder (KAIJO Co., LTD, Tokyo, Japan) was used for the wire bonding process, shown in Figure2. A packaging type 2835 LED was employed, and the capillary type used was SPT SU-50140-425E-ZU36-E. The bonding parameters are listed in Table1. During the bonding process, a mixture consisting of 95 vol% N2 + 5 vol% H2 was used as the forming gas at a flow rate of 0.6 L/min to protect the wire from oxidation. The geometry and size of the FAB and the bond geometry of the bonding wire (diameter of 0.025 mm) were investigated using a scanning electron microscopy (SEM) system, and the relationship between the EFO time and FAB diameter was determined. The bonded strength (ball shear/pull strength and wedge pull strength) were measured using a DAGE Series 4000 BS250Figure system 1. Ag-10Au-3.6Pd at a test speed allo ofy bonding 500 µm wire./s and shear height of 5 µm.

Figure 2. KAIJO FB-988 automatic wire bonder machine.

Table 1. Bonding parameters for Ag-10Au-3.6Pd alloy wire.

Ball Bond Wedge Bond Free Air Ball Impact Bond Force 0.65 0.75 Spark Voltage/V 5000 Force Ff/N F1/N Electronic Flame Bond Force 0.45/0.60/ US Power 0.015/0.020/0.025/0.0 60 Off(EFO) F/N 0.75 FigureFigure 2.P 2.KAIJO1 /mWKAIJO FB-988 FB-988 automaticautomatic wire wire bonder bonder machine. machine. 30/0.035 Current/A US Power Table 1.BondBonding Time parameters for Ag-10Au-3.6Pd alloy wire. 1.00/0.90/0.80/0.70/0. 55/70/85Table 1. Bonding parameters6 for Ag-10Au-3.6PdEFO Time/ms alloy wire. P/mW t1/ms 60 BallBall Bond Bond Wedge Wedge Bo Bondnd Free Air Free Ball Air Ball Bond Time Bond Force 0.70/0.85/ Impact ForceImpactFf/N 0.65 BondBond Force ForceF1/N 0.75 Spark Voltage/V 5000 8 TailElectronic Length/mm Flame O ff 0.15 Bondt/ms Force F/N 0.45/0.600.65/0.75 USF Power2/N P /mW1.000.75 60 Spark Voltage/V 0.015 5000/0.020 /0.025/0.030/0.035 Force Ff/N F1/N 1 (EFO) Current/A US Power P/mW 55/70/85US Bond Power Time t1 /ms 6ElectronicBonding EFO Flame Time /ms 1.00/0.90/0.80/0.70/0.60 Bond TimeBond t/ms Force 0.45/0.60/ 8 BondUS Power Force F / N80/95/105 0.70/0.85/1.00 Tail Length/mm0.015/0.020/0.025/0.0220 0.15 P2/mW 2 Temperature/°COff(EFO) US Power P /mW60 80/ 95/105 Bonding Temperature/ C 220 F/N 0.75 P1/mW2 ◦ 30/0.035 BondBond TimeTimet2/ ms 6 Current/A US Power Bond Time 6 1.00/0.90/0.80/0.70/0. 55/70/85 t2/ms 6 EFO Time/ms 3. ResultsP/mW t1/ms 60 Bond Time Bond Force 0.70/0.85/ 8 Tail Length/mm 0.15 3.1. Effectst of/ms EFO Current on GeometryF2/N of FAB of Ag-10Au-3.6Pd1.00 Alloy Bonding Wire US Power Bonding 80/95/105 220 The FAB geometries of the Ag-10Au-3.6PdP2/mW alloy bondingTemperature/°C wire for different EFO currents are shown in Figure3, the EFO timeBond was Time held constant at 0.8 ms. Figure3a shows that the FAB of the 6 bonding wire was sharp and pointedt2/ms when the EFO current was 0.015 A. Further, Figure3b shows that the geometry of the FAB changed to oval when the EFO current was increased to 0.020 A. Moreover, it can be seen from Figure3c that the FAB became round when the EFO current was increased further

Micromachines 2020, 11, 777 4 of 15 to 0.025 A. Next, Figure3d shows that the FAB geometry did not change when the EFO current was increased from 0.025 A to 0.030 A; however, its diameter did increase. Finally, Figure3e shows that the FAB became similar to a golf ball when the EFO current was 0.035 A. Further, the FAB was irregular, and thisMicromachines shape could 2020, 11 potentially, 777 lead to short circuits. 5 of 16

Figure 3. Cont.

Micromachines 2020, 11, 777 5 of 15 Micromachines 2020, 11, 777 6 of 16

FigureFigure 3. Geometries 3. Geometries of FAB of ofFAB Ag-10Au-3.6Pd of Ag-10Au-3.6Pd alloy alloy bonding bonding wire wire for for diff differenterent EFO EFO currents: currents: (a ()a 0.015) A, (b) 0.0200.015 A, A, (c ()b 0.025) 0.020 A, A, ((dc) 0.025 0.030 A, A, (d and) 0.030 (e) A, 0.035 and ( A.e) 0.035 A.

The3.2. bondingEffects of EFO wire Time melts on andGeometry forms of FAB an FAB of Ag-10Au-3.6Pd because of the Alloy surface Bonding tension Wire of the molten metal after the EFO process. As the FAB solidifies, heat is transferred from un-melted bonding wire to its neck The geometries of the FAB of the Ag-10Au-3.6Pd alloy bonding wire for different EFO times are (i.e., theshown heat-a in ffFigureected 4, zone). the EFO Hence, curren thet was FAB held solidifies constant fromat 0.030 its A. neck Figure toits 4 shows bottom, that and the itsFAB diameter is a is 2–3 timesregular that ball of when the wire. the EFO During time is the 0.40 EFO ms, process,however, ath highe diameter voltage of the is generatedFAB is smaller between than that the of EFOa tip and thestandard wiretail, FAB, because as shown of in which Figure the4a. The air withinFAB cont theinues gap to breakbe a regular downs, ball resulting when the inEFO the time transfer is of energyincreased to thewire from tail,0.40 ms which to 0.60 forms ms, however, the FAB. the After FAB thediameter EFO increases, process, theas shown temperature in Figure of4b. the The FAB is higherFAB than geometry the melting becomes point that of of the a standard wire. As ball a result, when the EFO wire time tail is melts increa further,sed to 0.80 leading ms, as to shown the growth of thein FAB. Figure If the 4c. EFOHowever, current the isFAB too geometry small (0.015 does ornot 0.020 change A), when the FABthe EFO bottom current solidifies is increased because further of a lack from 0.80 ms to 1.00 ms. On the other hand, the FAB diameter increases, as shown in Figure 4d. of energy, and the solidification sequence is broken, resulting in a pointed and subsequently oval ball, Finally, the FAB becomes a golf ball when the EFO time is increased to 1.20 ms, as shown in Figure as shown4e. in Figure3a,b, respectively. However, when the EFO current is increased from 0.020 A to 0.025 A, moreFigure heat 4a,b is show produced that, for because a fixed ofEFO the current higher of EFO0.030 current,A, the EFO and tip the produces wire tail enough melts energy completely. Consequently,to melt the the wire FAB tail, solidifies such that fromthe length its neck of th toe melted its bottom. wire tail Thus, becomes a regular shorter FAB as the is formed EFO time under is the surfacereduced tension, to 0.40 as shown and 0.60 in ms, Figure resulting3c. When in a smaller the EFO FAB. current However, is increased the FAB is further round and to 0.035 regular. A, theAs wire tail melts,can be and seen the from FAB Figure is formed 4c, the rapidlyFAB becomes because a standa of therd largeball when current. the EFO Further, time is the increased volume to of 0.80 the FAB is larger.ms. Conversely,Further, Figure when 4d shows the partthat the of thewire wire tail melts tail closestmore when to the the EFO EFO tiptime melts is increased faster to than 1.00 thatms; away the diameter of the FAB also increases. However, the FAB center remains unchanged. When the EFO from the tip, it causes the center of the FAB to deviate from that of the wire such that the former is current is kept constant (at 0.030 A), the EFO tip produces the same amount of energy irrespective of closerthe to thetime. EFO As a tip result, side. the As FAB a result, grows a in golf a regular ball FAB process is formed, from the as wire shown tail tip in Figureto the capillary3e. and exhibits a regular geometry. Moreover, its diameter is larger than that in the case of an EFO time of 3.2. Effects of EFO Time on Geometry of FAB of Ag-10Au-3.6Pd Alloy Bonding Wire 0.80 ms. On the other hand, a larger FAB result in an excessively large ball bonded, with the Theprobability geometries of a ofshort the circuit FAB of oc thecurring Ag-10Au-3.6Pd due to the bonded alloy bondingball increasing. wire for Further, different the EFObonding times are shownefficiency in Figure decreases4, the EFO when current the EFO was time held is increase constantd. When at 0.030 the EFO A. Figuretime is 4increased shows thatfrom the1.00 FAB ms is a to 1.20 ms, the FAB volume increases further, and the large FAB solidifies from its neck to its bottom, regular ball when the EFO time is 0.40 ms, however, the diameter of the FAB is smaller than that of with heat being transferred along this path. The temperature distribution of the FAB will be uneven a standardbecause FAB, of its as large shown diameter, in Figure and the4a. part The of FAB the FAB continues near its toneck be will a regular solidify ballfirst. when The other the parts EFO time is increasedsolidify fromin an 0.40upward ms todirection 0.60 ms, because however, of surface the FAB tension, diameter resulting increases, in the FAB as shown exhibiting in Figurethe 4b. The FABgeometry geometry of a becomescorrugated that ball. of aIn standard addition, ballthe whenlength theof the EFO heat-affected time is increased zone of to the 0.80 wire ms, neck as shown in Figureincreases4c. However, owing to the transfer FAB geometry of a greater does amount not change of heat from when the the ball EFO to the current wire. For is increasedinstance, the further from 0.80length ms increases to 1.00 ms. from On 30 theμm otherto 38 μ hand,m when the the FAB EFO diameter time is increases,increased from as shown 0.80 ms in to Figure 1.00 ms,4d. as Finally, the FABshown becomes in Figure a golf 5a,b. ball This when reduces the the EFO neck time strength. is increased Moreover, to 1.20 the melt ms, asrate shown of the wire in Figure tail closer4e. to the EFO tip side is higher than that for the opposite side, owing to the excessively long EFO time. The extent of melting of the wire tail closer to the EFO tip side is higher, causing the FAB center to deviate from that of the alloy wire. This results in the FAB becoming a golf ball, as shown in Figure 4e.

Micromachines 2020, 11, 777 6 of 15 Micromachines 2020, 11, 777 7 of 16

Figure 4. Cont.

Micromachines 2020, 11, 777 7 of 15 Micromachines 2020, 11, 777 8 of 16

Figure 4. Geometries of FAB of Ag-10Au-3.6Pd alloy bonding wire for different EFO times: (a) 0.4 ms, Figure 4. Geometries of FAB of Ag-10Au-3.6Pd alloy bonding wire for different EFO times: (a) 0.4 ms, (b) 0.6 ms, (c) 0.8 ms, (d) 1.0 ms, and (e) 1.2 ms. (b) 0.6 ms, (c) 0.8 ms, (d) 1.0 ms, and (e) 1.2 ms. Figure4a,b show that, for a fixed EFO current of 0.030 A, the EFO tip produces enough energy to melt the wire tail, such that the length of the melted wire tail becomes shorter as the EFO time is reduced to 0.40 and 0.60 ms, resulting in a smaller FAB. However, the FAB is30μm round and regular. As can be seen from Figure4c, the FAB becomes a standard ball when the EFO time is increased to 0.80 ms. Further, Figure4d shows that the wire tail melts more when the EFO time is increased to 1.00 ms; the diameter of the FAB also increases. However, the FAB center remains unchanged. When the EFO current is kept constant (at 0.030 A), the EFO tip produces the same amount of energy irrespective of the time. As a result, the FAB grows in a regular process from the wire tail tip to the capillary and exhibits a regular geometry. Moreover, its diameter is larger than that in the case of an EFO time of 0.80 ms. On the other hand, a larger FAB result in an excessively large ball bonded, with the probability of a short circuit occurring due to the bonded ball increasing. Further, the bonding efficiency decreases when the EFO time is increased. When the EFO time is increased from 1.00 ms to 1.20 ms, the FAB volume increases further, and the large FAB solidifies from its neck to its bottom, with heat being transferred along this path. The temperature distribution of the FAB will be uneven because of its large diameter, and the part of the FAB near its neck will solidify first. The38μm other parts solidify in an upward direction because of surface tension, resulting in the FAB exhibiting the geometry of a corrugated ball. In addition, the length of the heat-affected zone of the wire neck increases owing to the transfer of a greater amount of heat from the ball to the wire. For instance, the length increases from 30 µm to 38 µm when the EFO time is increased from 0.80 ms to 1.00 ms, as shown in Figure5a,b. This reduces the neck strength. Moreover, the melt rate of the wire tail closer to the EFO tip side is higher than that for the opposite side, owing to the excessively long EFO time. The extent of melting of the wire tail closer to the EFO tip side is higher, causing the FAB center to deviate from that of the alloy wire. This results in the FAB becoming a golf ball, as shown in Figure4e.

3.3. Relationship between Diameter of FAB of Ag-10Au-3.6Pd and EFO Time FigureFor alloy 5. Length bonding of heat-affected wire, at zone the of moment Ag-10Au-3.6P whend alloy electronic bonding flame wire fo or ffdifferent(EFO) EFO is generated,times: the surrounding(a) 0.8 ms and ( airb) 1.0 is ms. ionized, and the alloy wire at the end melts, forming a metal melt free air ball (FAB) under the combined action of gravity and surface tension. The higher EFO power and 3.3.the Relationship longer EFO between time, the Diameter bonding of FAB wire of absorbs Ag-10Au-3.6Pd more energy and EFO and Time the larger FAB diameter formed. However,For alloy if the bonding parameters wire, are at too the large, moment the FAB when size willelectronic be too flame large, andoff the(EFO) neck is strengthgenerated, of jointthe surroundingformed by bonding air is ionized, will be tooand small,the alloy which wire will at leadthe end to the melts, fracture forming of joint a metal neck. melt In addition, free air ball the (FAB)EFO power under and the EFO combined time should action be of matched gravity withand eachsurface other. tension. When The the higher EFO power EFO ispower increased, and the longerEFO time EFO should time, bethe appropriately bonding wire prolonged absorbs more to fully energy absorb and the the energy larger and FAB improve diameter the bondingformed. However,quality [26 if]. the parameters are too large, the FAB size will be too large, and the neck strength of joint formed by bonding will be too small, which will lead to the fracture of joint neck. In addition, the EFO power and EFO time should be matched with each other. When the EFO power is increased, the

Micromachines 2020, 11, 777 8 of 16

Figure 4. Geometries of FAB of Ag-10Au-3.6Pd alloy bonding wire for different EFO times: (a) 0.4 ms, Micromachines 2020, 11, 777 8 of 15 (b) 0.6 ms, (c) 0.8 ms, (d) 1.0 ms, and (e) 1.2 ms.

30μm

38μm

Figure 5. LengthLength of of heat-affected heat-affected zone zone of of Ag-10Au-3.6P Ag-10Au-3.6Pdd alloy alloy bonding bonding wire wire fo forr different different EFO EFO times: times: (a) 0.8 ms and (b) 1.0 ms. (a) 0.8 ms and (b) 1.0 ms. During the wire bonding process, the FAB geometry and diameter are the primary factors 3.3. Relationship between Diameter of FAB of Ag-10Au-3.6Pd and EFO Time determining the strength of the ball bonded. A large FAB diameter is tolerated in the case of large pad chips,For alloy as the bonding bonded wire, area increasesat the moment with the when FAB diameter,electronic resulting flame off in a(EFO) higher is bonded generated, strength. the surroundingHowever, for air smaller is ionized, die pad and chips, the analloy excessively wire at the large end FAB melts, can forming cause the a bondedmetal melt ball free to overflow, air ball (FAB)resulting under in an the increase combined in the action potential of gravity for short and circuits. surface tension. The higher EFO power and the longerThe EFO experimental time, the databonding for the wire FAB absorbs diameter more and EFOenergy current andtime the forlarger a wire FAB diameter diameter of 0.025 formed. mm wereHowever, fitted, if the the results parameters are shown are too in Figurelarge, 6the (the FAB EFO size current will be was too set large, at 0.030 and A). the It neck can bestrength seen from of joint the formedfigure that by thebonding FAB diameter will be too and small, EFO time which exhibited will lead an nonlinearto the fracture relationship. of joint Usingneck. theIn addition, least-squares the EFOmethod power for and data EFO fitting, time the should following be matched quadratic with equation each other. was When obtained the forEFO the power relationship is increased, between the the FAB diameter and EFO time:

4 2 3 D = 2.61905 10− + 0.22851 0.2881t + 0.13542t (1) − × − where D is the FAB diameter (mm) and t is the EFO time (ms). Further, using the least-squares analysis method, the correlation coefficient, r, between the two variables was calculated and found to be 0.99698, with the confidence interval being 0 < t < 1.20. Thus, it was confirmed that the obtained equation was a reasonable one and thus can be used by the packaging industry for determining the EFO time for Ag-10Au-3.6Pd alloy bonding wires with a diameter of 0.025 mm. Micromachines 2020, 11, 777 9 of 16

EFO time should be appropriately prolonged to fully absorb the energy and improve the bonding quality [26]. During the wire bonding process, the FAB geometry and diameter are the primary factors determining the strength of the ball bonded. A large FAB diameter is tolerated in the case of large die pad chips, as the bonded area increases with the FAB diameter, resulting in a higher bonded strength. However, for smaller die pad chips, an excessively large FAB can cause the bonded ball to overflow, resulting in an increase in the potential for short circuits. The experimental data for the FAB diameter and EFO current time for a wire diameter of 0.025 mm were fitted, the results are shown in Figure 6 (the EFO current was set at 0.030 A). It can be seen from the figure that the FAB diameter and EFO time exhibited an nonlinear relationship. Using the least-squares method for data fitting, the following quadratic equation was obtained for the relationship between the FAB diameter and EFO time:

𝐷 = −2.61905 × 10 + 0.22851 − 0.2881𝑡 + 0.13542𝑡 (1) where D is the FAB diameter (mm) and t is the EFO time (ms). Further, using the least-squares analysis method, the correlation coefficient, r, between the two variables was calculated and found to be 0.99698, with the confidence interval being 0 < t < 1.20. Thus, it was confirmed that the obtained equation was a reasonable one and thus can be used by the Micromachines 2020 11 packaging industry, , 777 for determining the EFO time for Ag-10Au-3.6Pd alloy bonding wires with9 ofa 15 diameter of 0.025 mm. 0.10

0.09

0.08 Free Air Ball Dia. 0.07 Fitting curve

0.06

0.05 Free Air Ball Dia. / mm Ball Air Free 0.04

0.03

0.02

0.01

0.00

-0.01 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Current Time / ms . Figure 6. Fitted curve for relationship between FAB diameter (D) and EFO time (t) for Ag-10Au-3.6Pd Figure 6. Fitted curve for relationship between FAB diameter (D) and EFO time (t) for Ag-10Au-3.6Pd alloy bonding wire with diameter of 0.025 mm. alloy bonding wire with diameter of 0.025 mm. 3.4. Effects of Bonding Power and Force on Ball Bonded Strength of Ag-10Au-3.6Pd Alloy Wire 3.4. Effects of Bonding Power and Force on Ball Bonded Strength of Ag-10Au-3.6Pd Alloy Wire During the bonding process, the FAB is deformed because of the bonding power, bonding force and the heatDuring applied, the as bonding shown process, in Figure the7. FAB The extentis deformed of deformation because of of the the bonding FAB depends power, onbonding the bonding force and the heat applied, as shown in Figure 7. The extent of deformation of the FAB depends on the power and bonding force. The bonded geometry of the Ag-10Au-3.6Pd alloy wire for a bonding power of bonding power and bonding force. The bonded geometry of the Ag-10Au-3.6Pd alloy wire for a 85 mW and bonding force of 0.75 N is shown in Figure8. In this case, the bonded ball overflows, and neck bonding power of 85 mW and bonding force of 0.75 N is shown in Figure 8. In this case, the bonded cracks develop in it. This phenomenon can cause the device to short circuit and fail. On the other hand, ball overflows, and neck cracks develop in it. This phenomenon can cause the device to short circuit the bonded ball is regular when the bonding power is 70 mW and the bonding force is 0.60 N, as shown and fail. On the other hand, the bonded ball is regular when the bonding power is 70 mW and the in Figurebonding9. force However, is 0.60 when N, as theshown bonding in Figure power 9. However, is 55 mW when and the the bonding bonding force power is 0.45is 55 N,mW the and bonded the ballbonding become force smaller, is 0.45 as shown N, the in bonded Figure ball10. Thebecome ball shearsmaller, strength as shown for different in Figure bonding 10. The parameters ball shear is shownstrength in Figure for different11. The mechanism bonding parameters of ball shear is strength shown testingin Figure is shown 11. The in Figuremechanism 12. It canof ball be seen shear from thestrength figure that testing the ballis shown shear in strength Figure 12. is the It can lowest be seen when from the the bonding figure that power the isball 55 shear mW andstrength the bonding is the forcelowest is 0.45 when N. Thus, the bonding these conditions power is 55 are mW unsuitable, and the asbonding the shear force strength is 0.45 doesN. Thus, not meetthese theconditions requirements. are However,unsuitable, when as the shear bonding strength power does is 85 not mW meet and the the requirements. bonding force However, is 0.75 N,when the the shear strength power is higheris 85 than mW the and standard the bonding value. force But, is in 0.75 this N, case, the the ball bonded shear strength exhibits ais stress higher concentration than the standard and can value. readily causeBut, a in short this circuit, case, the resulting bonded in exhibits device failure.a stress Theconcentration ball pull strength and can of readily the bonding cause wirea short for circuit, different Micromachines 2020, 11, 777 10 of 16 bondingresulting power in device and force failure. is shown The ball in Figurepull strength 13, the of mechanism the bonding of ballwire pull for strengthdifferent testing bonding is shownpower in

Figureand 14force, the is broken shown is in B. Figure It is evident 13, the from mechanism the figure of thatball thepull distribution strength testing of the is ball shown pull strengthin Figure is 14, wide whenthe thebroken bonding is B. It power is evident is 85 from mW the and figure the bonding that the force distribution is 0.75 N. of However,the ball pull in strength some cases, is wide the ballwhen pull strengththe bonding is very power low, in is these 85 mW cases, and device the bonding failure may force occur. is 0.75 For N. a However, bondingpower in some of cases, 70 mW the and ball bonding pull forcestrength of 0.60 is N, very the balllow, pull in these strength cases, is stable device and failure meet may the requirements. occur. For a bonding On the other power hand, of 70 for mW a bonding and powerbonding of 55 force mW of and 0.60 bonding N, the ball force pull of 0.45strength N, the is distributionstable and meet of the the ball requirements. pull strength On is the wide. other Finally, hand, the ballfor bond a bonding is regular power and theof 55 bonding mW and shear bonding strength force is asof high0.45 N, as the desireddistribution value of when the ball the pull bonding strength power is 70is wide. mW and Finally, the bonding the ball bond force is regular 0.60 N. and the bonding shear strength is as high as the desired value when the bonding power is 70 mW and the bonding force is 0.60 N.

FigureFigure 7.7. SchematicSchematic of ball bonded. bonded.

Figure 8. Geometry of ball bonded for bonding power of 85 mW and bonding force of 0.75 N.

Figure 9. Geometry of ball bonded for bonding power of 70 mW and bonding force of 0.60 N.

Micromachines 2020, 11, 777 10 of 16 Micromachines 2020, 11, 777 10 of 16 and force is shown in Figure 13, the mechanism of ball pull strength testing is shown in Figure 14, theand broken force is is shown B. It is evidentin Figure from 13, thethe figuremechanism that th ofe distribution ball pull strength of the ball testing pull is strength shown isin wide Figure when 14, the brokenbonding is powerB. It is evidentis 85 mW from and the the figure bonding that forcethe distribution is 0.75 N. However, of the ball inpull some strength cases, is the wide ball when pull strengththe bonding is very power low, is in85 thesemW andcases, the device bonding failure force may is 0.75 occur. N. However,For a bonding in some power cases, of the70 mWball pulland bondingstrength forceis very of low,0.60 N,in thethese ball cases, pull strengthdevice failure is stable may and occur. meet Forthe requirements.a bonding power On theof 70other mW hand, and forbonding a bonding force powerof 0.60 ofN, 55the mW ball and pull bonding strength force is stable of 0.45 and N, meet the the distribution requirements. of the On ball the pull other strength hand, isfor wide. a bonding Finally, power the ball of 55bond mW is andregular bonding and the force bond of ing0.45 shear N, the strength distribution is as high of the as ballthe desiredpull strength value whenis wide. the Finally, bonding the power ball bond is 70 is mW regular and and the thebonding bond ingforce shear is 0.60 strength N. is as high as the desired value when the bonding power is 70 mW and the bonding force is 0.60 N.

Micromachines 2020, 11, 777 Figure 7. Schematic of ball bonded. 10 of 15 Figure 7. Schematic of ball bonded.

Figure 8. Geometry of ball bonded for bonding power of 85 mW and bonding force of 0.75 N. Figure 8. GeometryGeometry of ball bonded for bonding powe powerr of 85 mW and bonding force of 0.75 N.

Micromachines 2020, 11, 777 11 of 16

MicromachinesFigure 2020 9., 11Geometry, 777 of ball bonded for bonding power of 70 mW and bonding force of 0.60 N. 11 of 16 Figure 9. Geometry of ball bonded for bonding power of 70 mW and bonding force of 0.60 N. Figure 9. Geometry of ball bonded for bonding power of 70 mW and bonding force of 0.60 N.

Figure 10. Geometry of ball bonded for bonding power of 55 mW and bonding force of 0.45 N. Figure 10. GeometryGeometry of of ball ball bonded bonded for bonding powe powerr of 55 mW and bonding force of 0.45 N.

80 80 80 80 70 70 ） 70 70

） 60 60 0.01 N 0.01 60 （ 60 50 0.01 N 0.01 50 （ 50 50 40 40 40 40 30

30 /g strength Ballshear 30

30 /g strength Ballshear 20 20 Ball shear strength / / strength Ballshear 20 20 Ball shear strength / / strength Ballshear 10 10 Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW 10 10 Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Figure 11. Shear strength of ball bonded. Figure 11. Shear strength of ball bonded. Figure 11. Shear strength of ball bonded.

Figure 12. Mechanism of ball shear strength testing [27]. Figure 12. Mechanism of ball shear strength testing [27].

20.0 20.0 20.0 20.0 17.5 17.5 17.5 17.5 15.0 15.0 0.01 N) 0.01 15.0 （ 15.0

0.01 N) 0.01 12.5 12.5 （ 12.5 12.5 10.0 10.0 10.0 10.0 7.5 7.5 Ballpull strength /g 7.5 7.5 Ballpull strength /g Ball pull strength / strength Ball pull 5.0 5.19 5.0 5.19 4.554.59 4.554.59 3.69 3.69 Ball pull strength / strength Ball pull 5.0 5.19 5.0 5.19 4.554.59 2.5 4.554.59 2.5 3.69 Force0.75N/Power85mW3.69 Force0.60N/Power70mW Force0.45N/Power55mW Force0.75N/Power85mW Force0.60N/Power70mW Force0.45N/Power55mW 2.5 2.5 Force0.75N/Power85mW Force0.60N/Power70mW Force0.45N/Power55mW Force0.75N/Power85mW Force0.60N/Power70mW Force0.45N/Power55mW Figure 13. Pull strength of ball bonded. Figure 13. Pull strength of ball bonded.

Micromachines 2020, 11, 777 11 of 16

Micromachines 2020, 11, 777 11 of 16

Figure 10. Geometry of ball bonded for bonding power of 55 mW and bonding force of 0.45 N. Figure 10. Geometry of ball bonded for bonding power of 55 mW and bonding force of 0.45 N. 80 80

8070 8070 ） 7060 7060 ） 0.01 N 0.01

（ 6050 6050 0.01 N 0.01 （ 5040 5040

4030

4030 /g strength Ballshear

20 3020 Ball shear strength /g strength Ballshear Ball shear strength / / strength Ballshear 30

2010 2010 Ball shear strength / / strength Ballshear Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW 10 10 Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Force0.75N\Power85mW Force0.60N\Power70mW Force0.45N\Power55mW Figure 11. Shear strength of ball bonded. Micromachines 2020 11 , , 777 Figure 11. Shear strength of ball bonded. 11 of 15

Figure 12. Mechanism of ball shear strength testing [27]. Figure 12. Mechanism of ball shear strength testing [27]. Figure 12. Mechanism of ball shear strength testing [27]. 20.0 20.0

20.017.5 20.017.5

17.515.0 17.515.0 0.01 N) 0.01 （ 15.012.5 15.012.5 0.01 N) 0.01

（ 12.510.0 12.510.0

7.5 10.07.5

10.0 Ballpull strength /g

Ball pull strength / strength Ball pull 7.55.0 5.19 7.55.0 5.19 4.554.59 Ballpull strength /g 4.554.59 3.69 3.69

Ball pull strength / strength Ball pull 5.19 2.5 5.19 5.02.5 5.0 4.59 Force0.75N/Power85mW4.554.59 Force0.60N/Power70mW Force0.45N/Power55mW Force0.75N/Power85mW4.55 Force0.60N/Power70mW Force0.45N/Power55mW 3.69 3.69 Micromachines2.5 2020, 11, 777 2.5 12 of 16 Force0.75N/Power85mW Force0.60N/Power70mWFigureForce0.45N/Power55mW 13. Pull strength ofForce0.75N/Power85mW ball bonded. Force0.60N/Power70mW Force0.45N/Power55mW Figure 13. Pull strength of ball bonded. Figure 13. Pull strength of ball bonded.

Figure 14. Mechanism of ball pull strength and wedge pull strength testing [27]. Figure 14. Mechanism of ball pull strength and wedge pull strength testing [27]. During bonding process, the bonding power and bonding force are applied on the bonding wire During bonding process, the bonding power and bonding force are applied on the bonding wire through the capillary to ensure that the bonding wire and pad are in close contact. The bonding wire through the capillary to ensure that the bonding wire and pad are in close contact. The bonding wire and pad surface exhibit high-frequency vibrations because of the bonding power, this removes all the and pad surface exhibit high-frequency vibrations because of the bonding power, this removes all contaminants present on the wire and pad surface. At the same time, a large number of dislocations the contaminants present on the wire and pad surface. At the same time, a large number of are produced within the bonding wire and pad. This results in the formation of fast diffusion channels dislocations are produced within the bonding wire and pad. This results in the formation of fast between the bonding wire and substrate. For the investigated ball bonded, when the bonding power diffusion channels between the bonding wire and substrate. For the investigated ball bonded, when and force are 55 mW and 0.45 N, respectively, the energy delivered and pressure applied are not the bonding power and force are 55 mW and 0.45 N, respectively, the energy delivered and pressure enough to completely remove all the contaminants and oxides present on the bonding wire and pad applied are not enough to completely remove all the contaminants and oxides present on the bonding surface. As a result, the number of dislocations in the bonding wire and pad remain small, and the wire and pad surface. As a result, the number of dislocations in the bonding wire and pad remain diffusion rate between the bonding wire and pad is also low. As a result, the ball shear strength is small, and the diffusion rate between the bonding wire and pad is also low. As a result, the ball shear small and does not meet the requirements. However, when the bonding power and force are 85 mW strength is small and does not meet the requirements. However, when the bonding power and force and 0.75 N respectively, the FAB undergoes deformation owing to the large bonding force. This leads are 85 mW and 0.75 N respectively, the FAB undergoes deformation owing to the large bonding force. to the overflowing of the ball, even the formation of chips and cracks can be found. At the same time, This leads to the overflowing of the ball, even the formation of chips and cracks can be found. At the the excessive power increases the stress in the neck and may cause it to crack. This, in turn, decreases same time, the excessive power increases the stress in the neck and may cause it to crack. This, in the pull strength, as shown in Figure 13. turn, decreases the pull strength, as shown in Figure 13. 3.5. Effects of Bonding Power and Force on Wedge Bonded Strength of Ag-10Au-3.6Pd Alloy Wire 3.5. Effects of Bonding Power and Force on Wedge Bonded Strength of Ag-10Au-3.6Pd Alloy Wire The geometries of wedge bonds formed using the Ag-10Au-3.6Pd alloy wire are shown in Figure 15, the bondingThe geometries power and of wedge force used bonds are formed 105 mW using and the 1.0 N,Ag-10Au-3.6Pd respectively. Italloy can wire be seen are thatshown the in extent Figure of 15, the bonding power and force used are 105 mW and 1.0 N, respectively. It can be seen that the extent of deformation of the fish tail is high and that the pad is also significantly deformed. The geometry of the bond formed for a bonding power of 80 mW and bonding force of 0.7 N is shown in Figure 16. It can be seen that, in this case, the fish tail is short and asymmetric. Finally, the geometry for a bonding power of 95 mW and bonding force of 0.85 N is shown in Figure 17. It can be seen that the geometry of the bond formed in this case is excellent.

Figure 15. Geometry of wedge bonded formed for bonding power of 105 mW and bonding force of 1.00 N.

Micromachines 2020, 11, 777 12 of 16

Figure 14. Mechanism of ball pull strength and wedge pull strength testing [27].

During bonding process, the bonding power and bonding force are applied on the bonding wire through the capillary to ensure that the bonding wire and pad are in close contact. The bonding wire and pad surface exhibit high-frequency vibrations because of the bonding power, this removes all the contaminants present on the wire and pad surface. At the same time, a large number of dislocations are produced within the bonding wire and pad. This results in the formation of fast diffusion channels between the bonding wire and substrate. For the investigated ball bonded, when the bonding power and force are 55 mW and 0.45 N, respectively, the energy delivered and pressure applied are not enough to completely remove all the contaminants and oxides present on the bonding wire and pad surface. As a result, the number of dislocations in the bonding wire and pad remain small, and the diffusion rate between the bonding wire and pad is also low. As a result, the ball shear strength is small and does not meet the requirements. However, when the bonding power and force are 85 mW and 0.75 N respectively, the FAB undergoes deformation owing to the large bonding force. This leads to the overflowing of the ball, even the formation of chips and cracks can be found. At the same time, the excessive power increases the stress in the neck and may cause it to crack. This, in turn, decreases the pull strength, as shown in Figure 13.

3.5. Effects of Bonding Power and Force on Wedge Bonded Strength of Ag-10Au-3.6Pd Alloy Wire Micromachines 2020, 11, 777 12 of 15 The geometries of wedge bonds formed using the Ag-10Au-3.6Pd alloy wire are shown in Figure 15, the bonding power and force used are 105 mW and 1.0 N, respectively. It can be seen that the deformationextent of deformation of the fish tailof the is high fish andtail thatis high the padand isthat also the significantly pad is also deformed. significantly The geometrydeformed. of The the bondgeometry formed of the for bond a bonding formed power for a ofbonding 80 mW power and bonding of 80 mW force and of bonding 0.7 N is shownforce of in 0.7 Figure N is shown16. It can in beFigure seen 16. that, It can in thisbe seen case, that, the fishin this tail case, is short the fi andsh tail asymmetric. is short and Finally, asymmetric. the geometry Finally, forthea geometry bonding powerfor a bonding of 95 mW power and bondingof 95 mW force and ofbonding 0.85 N isforce shown of 0. in85 Figure N is shown 17. It canin Figure be seen 17. that It can the be geometry seen that of the bondgeometry formed of the in thisbond case formed is excellent. in this case is excellent.

Micromachines 2020, 11, 777 13 of 16

Micromachines 2020, 11, 777 13 of 16 Figure 15. Geometry of wedge bonded formed for bonding power of 105 mW and bonding force of 1.00 N. Figure 15. Geometry of wedge bonded formed for bonding power of 105 mW and bonding force of 1.00 N.

Figure 16. Geometry of wedge bonded formed for bonding power of 80 mW and bonding force of 0.7 N.Figure 16. Geometry of wedge bonded formed for bonding power of 80 mW and bonding force of 0.7 N. Figure 16. Geometry of wedge bonded formed for bonding power of 80 mW and bonding force of 0.7 N.

Figure 17. Geometry of wedge bonded formed for bonding power of 95 mW and bonding force of 0.85 N. Figure 17. Geometry of wedge bonded formed for bonding power of 95 mW and bonding force of 0.85 N. GenerallyFigure 17. Geometry speaking, of if thewedge ultrasonic bonded powerformed isfor too bond high,ing the power ultrasonic of 95 mW softening and bonding effect willforce become of obvious.Generally0.85 N. But if thespeaking, ultrasonic if the power ultrasonic is too high, power it will is causetoo high, work the hardening ultrasonic [28 ,softening29]. In order effect to havewill becomelarger friction obvious. energy But if during the ultrasonic bonding, power larger is bonding too high, pressure it will cause is generally work hardening used. However, [28,29]. tooIn order high tobonding haveGenerally larger pressure friction speaking, will energy limit if thethe during degree ultrasonic bonding, of slip power between larger is bondingtoo interfaces high, pressure the and ultrasonic reduce is generally the softening friction used. energy, effectHowever, thuswill tooreducingbecome high obvious. bonding the bonded But pressure if strength the ultrasonic will of limit solder power the joints degree is [ 30too, 31 ofhigh,]. slip it between will cause interfaces work hardening and reduce [28,29]. the Infriction order energy,to have thuslarger reducing friction theenergy bonded during stre bonding,ngth of solder larger joints bonding [30,31]. pressure is generally used. However, too highDuring bonding the bonding pressure process, will limit the themetal degree atoms of diffuse slip between rapidly interfaces because of and the reduce application the friction of the bondingenergy, thus power reducing and force, the bondedresulting stre inngth bonded of solder form ation.joints When[30,31]. the bonding power is 105 mW and bondingDuring force the is bonding 1.00 N, process,the energy the metaland contac atomst diffusestress betweenrapidly becausethe bonding of the wire application and pad of arethe excessivelybonding power high, and because force, of resulting which the in bondedpad is defo formrmed.ation. This When decreases the bonding the bonded power strength. is 105 mW In andthis case,bonding the distributionforce is 1.00 of N, the the wedge energy pull and strength contac ist wide, stress and between in some the cases, bonding the wedge wire pulland strengthpad are isexcessively lower than high, the becausestandard, of as which shown the in pad Figure is defo 18,rmed. the mechanism This decreases of wedge the bonded pull strength strength. testing In this is showncase, the in distribution Figure 14, the of brokenthe wedge is D. pull In thesestrength cases, is wide, the device and in life some will cases, be decreased. the wedge In pulladdition, strength the useis lower of excessively than the standard, large bonding as shown power in Figureand force 18, thecan mechanismcause significant of wedge stress pull concentration strength testing in the is bondedshown in area, Figure decreasing 14, the broken the bond is D.reliability. In these cases,Moreov theer, device when lifethe willvalues be ofdecreased. the bonding In addition, parameters the areuse tooof excessivelylarge, the fish large tail bonding length is power reduced and and force the caneffective cause contact significant area stressbetween concentration the bonding in wire the andbonded pad area,becomes decreasing small and the hence bond the reliability. bonded Moreovstrengther, decreases. when the Finally, values there of the is bonding stress concentration parameters inare the too fish large, tail, the which fish cantail causelength the is reducedbonded toand br theeak. effective On the othercontact hand, area a between bonding thepower bonding of 80 wiremW and padbonding becomes force small of 0.70 and N henceare insufficient the bonded to strengthremove alldecreases. the contaminants Finally, there and is oxides stress present concentration on the wirein the surface fish tail, and which pad. canIn this cause case, the the bonded number to ofbr dislocationseak. On the otherformed hand, within a bonding the wire power is insufficiently, of 80 mW thenand bonding resulting force in the of formation0.70 N are insufficientof fewer diffusion to remo channelsve all the between contaminants the wire and and oxides pad present surface. on As the a result,wire surface the rate and of pad.diffusion In this between case, the the number wire and of dislocations pad surface formed is redu withinced, this the decreases wire is insufficiently, the strength ofthen the resulting wedge bonded, in the formation as shown ofin fewerFigure diffusion 16. For the channels Ag-10Au-3.6Pd between alloythe wire bonding and padwire, surface. the bonded As a strengthresult, the was rate found of diffusion to be high between for a bonding the wire powerand pad of surface95 mW andis redu bondingced, this force decreases of 0.75 N.the strength of the wedge bonded, as shown in Figure 16. For the Ag-10Au-3.6Pd alloy bonding wire, the bonded strength was found to be high for a bonding power of 95 mW and bonding force of 0.75 N.

Micromachines 2020, 11, 777 13 of 15

During the bonding process, the metal atoms diffuse rapidly because of the application of the bonding power and force, resulting in bonded formation. When the bonding power is 105 mW and bonding force is 1.00 N, the energy and contact stress between the bonding wire and pad are excessively high, because of which the pad is deformed. This decreases the bonded strength. In this case, the distribution of the wedge pull strength is wide, and in some cases, the wedge pull strength is lower than the standard, as shown in Figure 18, the mechanism of wedge pull strength testing is shown in Figure 14, the broken is D. In these cases, the device life will be decreased. In addition, the use of excessively large bonding power and force can cause significant stress concentration in the bonded area, decreasing the bond reliability. Moreover, when the values of the bonding parameters are too large, the fish tail length is reduced and the effective contact area between the bonding wire and pad becomes small and hence the bonded strength decreases. Finally, there is stress concentration in the fish tail, which can cause the bonded to break. On the other hand, a bonding power of 80 mW and bonding force of 0.70 N are insufficient to remove all the contaminants and oxides present on the wire surface and pad. In this case, the number of dislocations formed within the wire is insufficiently, then resulting in the formation of fewer diffusion channels between the wire and pad surface. As a result, the rate of diffusion between the wire and pad surface is reduced, this decreases the strength of the wedge bonded, as shown in Figure 16. For the Ag-10Au-3.6Pd alloy bonding wire, the bonded strengthMicromachines was 2020 found, 11, 777 to be high for a bonding power of 95 mW and bonding force of 0.75 N. 14 of 16

17.5 17.5

15.0 15.0 0.01 N) 0.01 12.5 12.5 （

10.0 10.0

7.5 7.5 Wedge pullstrength /g pullstrength Wedge 5.0 5.0 Wedge pullstrength /

2.5 2.5 Force1.0N/Power105mW Force0.85N/Power95mW Force0.70N/Power80mW Force1.0N/Power105mW Force0.85N/Power95mW Force0.70N/Power80mW Figure 18. Pull strength of wedge bonded. Figure 18. Pull strength of wedge bonded. 4. Conclusions 4. Conclusions In this study, the effects of different EFO and bonding parameters on the bonds formed using an Ag-10Au-3.6PdIn this study, the alloy effects bonding of different wire were EFO analyzed. and bonding The parameters primary conclusions on the bonds of the formed study using can be a summarizedAg-10Au-3.6Pd as follows:alloy bonding wire were analyzed. The primary conclusions of the study can be summarized as follows: (1) For the Ag-10Au-3.6Pd alloy bonding wire (diameter of 0.025 mm), for a constant EFO time, (1) Forthe the geometry Ag-10Au-3.6Pd of the FAB alloy changes bonding from wire a pointed (diameter ball of to 0.025 an ovalmm), ball, for a round constant ball EFO and time, golf ballthe geometrywhen EFO of current the FAB is increased.changes from On thea pointed other hand, ball to for an a oval constant ball, EFO round current, ball and the golf FAB ball geometry when EFOchanges current from is a increased. small round On ball the to other a regular hand, round for a ball constant and finally EFO tocurrent, a golfball. the FAB The FABgeometry was a changesregular roundfrom a ball small when round the ball EFO to time a regular was 0.80 round ms ball and and EFO finally current to was a golf 0.030 ball. A. The FAB was a (2) regularFor the round investigated ball when bonding the EFO wire, time at an was EFO 0.80 current ms and of 0.030EFO current A, the FAB was diameter 0.030 A. and EFO time

(2) Forexhibited the investigated a nonlinear bonding relationship, wire, at which an EFO could current be expressed of 0.030 A, by the a quadratic FAB diameter function. and EFO time exhibited a nonlinear relationship, which could be expressed by a quadratic function. (3) For the Ag-10Au-3.6Pd alloy bonding wire, the ball bonded is regular and the ball shear strength (3) For the Ag-10Au-3.6Pd alloy bonding wire, the ball bonded is regular and the ball shear strength meets the requirements when the bonding power is 70 mW and the bonding force is 0.60 N. meets the requirements when the bonding power is 70 mW and the bonding force is 0.60 N. On On the other hand, the geometry of the wedge bonded is excellent when the bonding power is the other hand, the geometry of the wedge bonded is excellent when the bonding power is 95 95 mW and the bonding force is 0.85 N. mW and the bonding force is 0.85 N.

Author Contributions: Theoretical analysis, J.C., K.S.; geometry analysis, J.C., K.S.; experimental operation, J.Z.; investigation, J.C., J.C., J.Z.; J.Z.; data data curation, curation, J.C., J.C., J.Z.; J.Z.; writing writing original original draft draft preparation, preparation, J.Z., J.Z., J.P.; J.P.; supervision, supervision, K.S. K.S.All All authors have read and agreed to the published version of the manuscript. authors have read and agreed to the published version of the manuscript.

Funding: This research was funded by the key scientific and technological projects in HeNan Province China, grant number 192102210007.

Conflicts of Interest: The authors declare no conflict of interest.

References

1. Chuan-zhi, L.I. Analysis on the Countermeasures of high-end electronic information industry in China.Econ. Probl.2019, 1, 12. 2. Sun, Y. Development status and Prospect of electronic information engineering. Commun. World 2018, 25, 65. 3. Minbo, T. Electronic Packaging Engineering Beijing. Tsinghua University Press: Beijing, China, 2003; Volume 2. 4. Chao, Y.; Zhaojian, Y.; HailingQiao, Z. Progress in wire bonding technology. Electron. Technol. 2007, 28, 205–210. 5. Tian, H.E. Current situation and development trend of wire bonding technology. Spec. Equip. Electron. Ind. 2004, 33, 12–14. 6. Guo, R.; Gao, L.; Mao, D.; Li, M.; Wang, X.; Lv, Z.; Chiu, H. Study of free air ball formation in Ag-8Au-3Pd alloy wire bonding. Microelectron. Reliab. 2014, 54, 2550–2554. 7. Zhang, J.; Yang, L.J.; Xu, L.C.; Han, S.L. Development of microelectronic bonding technology. Weld. Technol. 2007, 36, 5–8.

Micromachines 2020, 11, 777 14 of 15

Funding: This research was funded by the key scientific and technological projects in HeNan Province China, grant number 192102210007. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Li, C. Analysis on the countermeasures of high-end electronic information industry in China. Econ. Probl. 2019, 1, 12. 2. Sun, Y. Development status and Prospect of electronic information engineering. Commun. World 2018, 25, 65. 3. Minbo, T. Electronic Packaging Engineering Beijing; Tsinghua University Press: Beijing, China, 2003; Volume 2. 4. Chao, Y.; Yang, Z.; Qiao, H. Progress in wire bonding technology. Electron. Technol. 2007, 28, 205–210. 5. Tian, H.E. Current situation and development trend of wire bonding technology. Spec. Equip. Electron. Ind. 2004, 33, 12–14. 6. Guo, R.; Gao, L.; Mao, D.; Li, M.; Wang, X.; Lv, Z.; Chiu, H. Study of free air ball formation in Ag-8Au-3Pd alloy wire bonding. Microelectron. Reliab. 2014, 54, 2550–2554. [CrossRef] 7. Zhang, J.; Yang, L.J.; Xu, L.C.; Han, S.L. Development of microelectronic bonding technology. Weld. Technol. 2007, 36, 5–8. 8. Zhou, W.-Y.; Pei, H.-Y.; Luo, J.-Q.; Wu, Y.-J.; Chen, J.-L.; Kang, F.-F.; Kong, J.-W.; Ji, X.-Q.; Yang, G.-X.; Yang, A.-H. Influence of Pd content on microstructure and performance of Au-coated Ag alloy wire. Chin. J. Mech. Eng. 2020, 31, 6241–6247. 9. Cao, J.; Ding, Y.T.; Guo, T.B. Effect of copper properties and bonding parameters on bonding quality. Mater. Sci. Technol. 2012, 20, 76–79. 10. Cao, J.; Wu, W.X. Effects of Au coated Ag bonding wire properties on bonded quality. Mater. Sci. Technol. 2018, 26, 96–100. 11. Czerny, B.; Mazloum-Nejadari, A.; Khatibi, G.; Weiss, L.; Zehetbauer, M. Fatigue testing method for fine bond wires in an LQFP package. Microelectron. Reliab. 2016, 64, 270–275. [CrossRef] 12. Gross, D.; Haag, S.; Reinold, M.; Schneider-Ramelow, M.; Lang, K.D. Heavy copper wire-bonding on silicon chips with aluminum-passivated Cu bond-pads. Microelectron. Eng. 2016, 156, 41–45. [CrossRef] 13. Martin, S.R.; Christian, E. The reliability of wire bonding using Ag and Al. Microelectron. Reliab. 2016, 63, 336–341. 14. Qin, W.; Cohen, I.M.; Ayyaswamy, P.S. Ball size and HAZ as Functions of EFO parameters for Bonding wire. Adv. Electron. Packag. 1997, 1, 391–398. 15. Cao, J.; Fan, J.; Gao, W. Effect of Ag-10Au-3.6Pd alloy bonding wire properties and structure on bond strengths and reliability. Intern. Conf. Electron. Packag. 2016, 4, 82–87. 16. Hang, C.J.; Song, W.H.; Lum, I.; Mayer, M.; Zhou, Y.; Wang, C.Q.; Moon, J.T.; Persic, J. Effect of electronic flame off parameters on copper bonding wire: Free-air Ball deformability, heat affected zone length, heat affected zone breaking force. Microelectron. Eng. 2009, 86, 2094–2103. [CrossRef] 17. Wu, J.; Rockey, T.; Yauw, O.; Shen, L.; Chylak, B. Bonding of Ag-alloy wire in LED packages. In Proceedings of the 35th IEEE/CPMT International Electronic Manufacturing Technology Symposium (IEMT), Ipoh, Perak, Malaysia, 6–8 November 2012; pp. 1–4. 18. Gu, L.; Chen, Q.; Li, J.; Chen, Z.; Zhou, J.; Du, M.; Chung, M. Comparison of Ag wire and Cu wire in memory package. J. Funct. Mater. Devices 2013, 19, 240–244. [CrossRef] 19. Yu, Z. Process Research on Improving Bonding Strength of Hot-Pressed Ultrasonic Ball Solder Joint. Master’s Thesis, Southwest Jiaotong University, Chengdu, China, 2010. 20. Xue, R. Research on the Influence of Ultrasonic Power and Bonding Pressure on the Bonding Quality of Gold Wire. University of Electronic Science and Technology of China: Chengdu, China, 2016. 21. Tseng, Y.W.; Hung, F.Y.; Lui, T.S.; Chen, M.Y.; Hsueh, H.W. Effect of annealing on the microstructure and bonding interface properties of Ag-2Pd alloy wire. Microelectron. Reliab. 2015, 55, 1256–1261. [CrossRef] 22. Tseng, Y.W.; Hung, F.Y.; Lui, T.S. Tensile and electrical properties of gold-coated silver bonding wire. Microelectron. Reliab. 2015, 55, 608–612. [CrossRef] 23. Tanna, S.; Pisigan, J.L.; Song, W.H.; Halmo, C.; Persic, J.; Mayer, M. Low cost Pd coated Ag bonding wire for high quality FAB in air. In Proceedings of the IEEE Electronic Components & Technology Conference, San Diego, CA, USA; 2012; Volume 282, pp. 1103–1109. Micromachines 2020, 11, 777 15 of 15

24. Chen, J.L.; Lin, Y.C. A New approach in Free Air ball formation process parameters analysis. In Proceedings of the IEEE Transom Electronic Packaging Manufacturing, Las Vegas, NV, USA; 2000; Volume 23, pp. 116–122. 25. Tan, J.; Toh, B.H.; Ho, H.M. Modeling of free air ball for copper wire bonding. In Proceedings of the 6th Electronics Packaging Technology Conference, Singapore; 2004; pp. 711–717. 26. Xu, H.; Liu, C.; Silberschmidt, V.V.; Pramana, S.S.; White, T.J.; Chen, Z.; Acoff, V.L. New mechanisms of void growth in Au–Al wire bonds: Volumetric shrinkage and intermetallic oxidation. Scr. Mater. 2011, 65, 642–645. [CrossRef] 27. Xu, X. Effect of Geometric Parameters of Ceramic Capillary on Bonding Strength and Reliability; HeNan Polytechnic University, Jiaozuo: China, 2020; pp. 16–20. 28. Langenecker, B. Effect of sonic and ultrasonic radiation on safety factors of rockets and missiles. AIAA J. 1963, 1, 80–83. [CrossRef] 29. Wang, F.; Han, L.; Zhong, J. Effect of ultrasonic power on the wire bonding strength. Chin. J. Mech. Eng. 2007, 43, 107–111. 30. Mayer, M. Microelectronic Bonding Process Monitoring by Integrated Sensors; ETH Zurich: Zürich, Switzerland, 2000. 31. Liang, Z.N.; Kuper, F.G.; Chen, M.S. Concept to relate wire bonding parameters to bondability and ball bond reliability. Microelectron. Reliab. 1998, 38, 1287–1291. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).