Chapter 3 Result and discussion
Test sample was fabricated by steps as above, several measurements were set up and done. XPS analysis to identify Au content on bonding pads after cleaning, bonding pressure condition to find minimum pressure requirement, SEM picture of Au-Au bonding adhesion test and cross section are included, contact resistance measurement and AFM analysis shown as below.
3.1 HDXPS Analysis
Au surface contamination is always concerned in thermocompression bonding process. It is believe that the higher surface cleanness improved, the higher bonding strength can be achieved. Since there is not formed significant oxide film on gold at room temperature, and solubility of oxygen, carbon, and hydrogen in gold are all negligibly small, it is believed that gold absorbed organic substance when exposed to the atmosphere [10]. To examine the effectiveness of surface cleaning by Sulfuric Acid and Hydrogen Peroxide mixed solution, XPS analysis was performed on Au surface is shown in Fig. 3.1.
It can be seen that after 240 seconds dipping into Sulfuric Acid and Hydrogen Peroxide mixed solution, the content of Au increase from 41.9% to 75.6% and organic contamination, such as carbon or oxygen decrease from Au surface.
Compare with cleaning by Ar/H2 plasma after 1 minuite treatment, surface content of Au increased from 37.5% to 80.5% [9].
10
0 5 10 15 20 25 30 35 40
0 10 20 30 40 50 60 70 80 90
0 100 200 300 400 500
C‐PC
O‐PC
Au‐PC
C‐SAB
O‐SAB
Au‐SAB
Piranha Clean time (s) SAB Clean Time (min)
SurfaceContent(%)
Figure 3.1: XPS Content analysis of bonding surface by Sulfuric Acid and Hydrogen Peroxide mixed treatment and compared with SAB process[9].
3.2 Bonding pressure condition
XPS analysis shows that Sulfuric Acid and Hydrogen Peroxide mixed cleaned after 240 seconds, Au contents rise up to 70%, which we chose as our typical clean time. To examine the failure mode of 40um Au bump, the test were carried out after bonding under the condition of different pressure. It was found that higher pressure bonding pressure lead chip to be bond after cleaning. In Tab.
3.2, chip and carrier bonded at 160℃, the minimum pressure requirement in this experiment is 100Mpa.
11
Pressure(Mpa) Bonded
30 ╳
50 ╳
100 ○
120 ○
Table 3.2: Pressure condition (○-bonded, ╳-fail)
3.3 Au-Au bonding adhesion test
Fig. 3.3 shows the SEM picture of chip and carrier forcefully separated bonding transition structure in the bonding condition of 160 100Mpa for℃ 180 seconds. We know that failure occurs at the weakest region of bonded interface.
In Fig. 3.3(a), copper wires and bonding pads on carrier were lifted by exerting external force as the carrier and chip are bonded together. Bonding pads were bonded tightly enough to lift off copper wire from carrier as the detail part of SEM picture shown in Fig. 3.3(b). In Fig. 3.3(c), shows the detail SEM image of lifted off bonding pad. By using EDX analysis, sputtered Ti seed layer adhere on carrier after separation. As the result, we found out that the sputtered Ti/Cu interface is the weakest region to be broken. We have discovered that by using mixture of Sulfuric Acid and Hydrogen Peroxide clean, bonding strength is strong enough to break the interface between Cu and Ti.
12
Figure 3.3: SEM picture of chip and carrier separated by external force.
3.4 Cross section
Fig. 3.4 shows the SEM picture of the entire cross section image of the bonding area in the condition of 160 , 100Mpa, 180 secon℃ ds. The upper and lower bump were connect together with high accuracy and no damage was found at interface. There is a lift off area between Si carrier and lower bump, was filled with broken pieces during polish process without unferfilling. As shown in figure, Several voids are found at both right and left ends of the bonding interface as the result of bump deformation under high pressure.
These voids may cause contact resistance increases.
(a)
Ti Bonding pad
(b) (c)
13
Figure 3.4: SEM picture of cross section of Au circular bonding interconnection.
3.5 AFM analysis
The roughness of the bonding pads play an important role in dominating the resistance between bonding pads. The decrease in the roughness leads to the increase in the bonding area. The RMS roughness values of electrodes surface, which are very large using in concept of SAB [11]. In comparison with SAB method mentioned in [9], roughness were in degree of 200nm to 400nm after 10 minutes treatment, By using mixture of Sulfuric Acid and Hydrogen Peroxide clean, RMS roughness was measured at 50nm as shown in Fig 3.5, would supply a good bonding surface to bond with lower contact resistance.
Upper Bump
Lower Bump
Si Chip
Si Chip
Voids Estimated Bonding
Interface
Ti/Cu seed layer
40um
Lift off area
14
Figure 3.5: AFM roughness measurement.
3.6 Contact resistance measurement
The contact resistance of each bonding pad can be derived by dividing the total resistance by the total number of bonding pads. The effect of different bonding temperature on electrical resistance is shown in Fig 3.6. Under the bonding condition of 100MPa, cleaned for 240 seconds, bonding temperature increased from 160℃ to 300℃, the contact resistance decreased from 45 to 17mΩ. The higher temperature applied on the contact area during the bonding process, the more Au surface can be welded, and contact area between the electrodes is greatly increased, the more resistance is decreased.
15
15 20 25 30 35 40 45 50
150 200 250 300 350
Temperature (℃)
Contactresistance (mΩ)
Figure 3.6: Contact resistance measurement for different Sulfuric Acid and Hydrogen Peroxide mixed treating time.
16