Effect Of RF Plasma Process Gas Chemistry And Electrode Configuration On The Removal Of Copper Lead Frame Oxidation

By Daniel Chir and Johnson Toh

Lead frame surface oxidation can lead to surface delamination after molding or wire bonding issues. The application of plasma treatment has been proven to be safe and effective solution to address these issues. However, the effectiveness of plasma treatment for removing oxide is dependent on the correct use of recipe parameters, gas chemistry and electrode configuration.

In this paper, analytical techniques such as contact angle measurement, high magnification optical inspection and SEM-EDX are carried out on copper lead frames to evaluate the impact of using different plasma gas chemistries and electrode configurations. It is concluded that the use of Ar/H₂ is better than Ar gas chemistry in removing oxide from copper lead frames. Another conclusion is the placement of copper lead frames on ground electrode is showing higher oxide removal rate than the placement on powered electrode.

Introduction

Copper alloy is a major substrate material used in IC packaging due to its good thermal and electrical performance, good manufacturability, and low cost. However, it has very high affinity with oxygen in the atmosphere, which can cause oxidation during the assembly process. The risk of oxidation on copper surface increases during manufacturing process that require heating, especially during the wire bonding step [1]. The oxidation can lead to higher probability of delamination or bonding issues between the mold and metal surface during post mold curing [2] [3]. The optimum plasma configuration is known to be used for removing oxide from the copper surface and can increase the surface energy by up to 84% of its original value [4]. Placement of a product to be treated on a ground electrode subjects it to a more isotropic treatment than when it is placed on a powered electrode, which gives a more anisotropic treatment. Using the powered electrode will produce higher temperature surface and give lower treatment uniformity compared to treatment on the ground electrode [5].

Active plasma species which include ions, electrons, radicals and photons are formed during ionization of the process gases by RF energy. Ions will physically collide with the target surface, breaking bonds and releasing the surface material in the process known as sputtering. The H₂ radical formed in plasma is a good reducing agent that reacts with oxygen atoms on the oxidized surface to form water vapor. It is a useful process gas for removing oxides from the metallic surfaces.

Methodology

Material used: Two types of commercial copper alloy lead frames (bare copper and rough copper) were tested. The lead frames were exposed to temperatures that simulate the die attach, glue curing and wire bonding steps before they were sent for plasma treatment. This was done to simulate the oxidation that the lead frame will encounter during a typical assembly process.

Plasma equipment: The Nordson MARCH plasma system uses 13.56 MHz RF to generate plasma. The equipment had a H₂ safety kit that allows the use of up to 100% H₂ gas flow, but for this evaluation premixed gas of 95% Ar and 5% H₂ (Ar/H₂) was used. The strip is placed directly on the electrode in a configuration that is known as OTE (On-The-Electrode). This configuration is proven to give better uniformity and focuses the plasma energy on the top surface of the lead frame since the back of the lead frame sits flat on the electrode base and is not exposed to plasma. The process parameters of 600W RF power, 180s plasma time were used for this evaluation. The variables were the different electrode configurations and the gas chemistry.

Analysis method: The analysis of the copper lead frame surface was performed using two types of equipment: SEM-EDX and contact angle goniometer. The SEM-EDX was utilized in this study to analyze the oxide concentration on the sample surface The goniometer was used to analyze the wettability and surface energy.

Results

Contact angle

• Rough copper gives higher CA than bare copper lead frame
• Lower CA reading on the surface after plasma treatment process
• Ar Plasma gives a higher CA than Ar/H₂ plasma
• No significant difference if Ar/H2 chemistry was used and material was placed on either power or ground electrode

Contact Angle result of different configuration and material.

SEM-EDX

• Rough copper lead frame detected with higher oxygen concentration on the surface than the bare copper lead frame after plasma treatment
• Ar/H₂ plasma exhibits lower oxygen concentration than the Ar plasma when lead frame are treated on powered electrode
• The use of Ar/H₂ on ground electrode detected with the lowest oxygen concentration on the surface
• The CA result does not directly correlate with the oxide thickness on the surface based on this result analysis

SEM-EDX result on the copper lead frame.

Appearance of bare copper lead frame before and after plasma treatment.

High magnification scope for bare copper lead frame after plasma treatment.

Appearance of rough copper lead frame after plasma treatment.

High magnification scope for rough copper lead frame after plasma treatment.

The pre-plasma treated lead frame was observed to be discolored after the heat exposure in the different process steps. Discoloration was still found on samples treated with Ar on powered electrode and with Ar/H₂ on a powered electrode setup. However, there was no discoloration observed on the lead frame after treatment with Ar/H₂ on the ground electrode setup.

Appearance of the rough copper lead frame after plasma treatment shows different discoloration from the treated bare copper lead frame. This could be the result of the thicker oxide layer on the rough copper compared to the bare copper lead frame. On both types of lead frames there was a trend of reduction in discoloration intensity with reducing oxide thickness. There is more discoloration remaining on the rough copper compared to the bare copper lead frames that were processed with Ar/H₂ on ground electrode. Further recipe optimization is required to remove the remaining oxide and attain lower discoloration on the surface.

Why does the use of Ar/H₂ and ground electrode work better in removing the oxide from the surface?

• H₂ plasma can lower the activation energy of Cu2O (from 114.53 to 75.64 kJ mol-1 ) due to H₂ radicals present in the H₂
• Higher possibility of adsorption of the H₂ radical on copper surface and promotes reaction with the oxygen atoms to form H₂O and follow by desorption of H₂O as volatile byproduct.
• The use of anisotropic reaction is not favored to remove oxide easily.

Conclusion

In summary, the use of H₂ and the placement on the ground electrode can increase the rate of oxide removal from the copper surface. It has produced the lowest contact angle, the most effective removal of the discoloration and the lowest oxide concentration on the surface based on the result obtained in this study. This is explained by the lower activation energy required to reduce copper oxide to copper when H₂ radicals are formed in the plasma. In contrast, with a physical process on the powered electrode where ion sputtering is dominant, the removal of oxide is less effective.

The key takeaway from this report is that a correct gas chemistry and a suitable electrode configuration are critical to obtain an optimum plasma process that requires a shorter process time and lowers the risk of overtreatment or heat related issues. From a manufacturing perspective, this results in higher production throughputs and better yields.

References

1. C.-L. Chuang, J.-N. Aoh and R.-F. Din, “Oxidation of copper pads and its influence on the quality of Au/Cu bonds during thermosonic wire bonding process,” Microelectronics Reliability, Vols. 46(2-4), pp. 449-458, 2006.
2. C. T. Clhong, A. Leslie, E. Len and L. ‘Thiam, “Investigation on the Effect of Copper Leadframe Oxidation,” 1995 Proceedings. 45th Electronic Components and Technology Conference, 1995.
3. M.-F. Shu, K. Chen, B. Yang, W. Liu and Y.-H. Tseng, “Lead-Frames Copper Oxidation Effect for a QFN Package Delamination Improvement,” Conference: 2016 6th Electronic System-Integration Technology Conference (ESTC), 2016.
4. L. Y. M. Wong Jun Hao, “A Comparative Analysis on Physical and Chemical Plasma Cleaning Effects on Surfaces,” 2013 IEEE 15th Electronics Packaging Technology Conference (EPTC 2013), 2013
5. J. Getty, L. Wood and C. Nordson, “Critical Plasma Processing Parameters for Improved Strength of Wire Bonds,” 2016.
6. K. C. Sabat, R. K. Paramguru and B. K. Mishra, “Reduction of Copper Oxide by Low-Temperature,” Hydrogen Plasma. Plasma Chem Plasma Process 36, pp. 1111-1124, 2016.

Johnson Toh is a senior application supervisor in RF Plasma at Nordson MARCH.