Search results

The purpose of this paper is to attempt to study the failure mechanism of BGA (ball grid array) Cu wire bond ball lift and specifically focused on substrate outgassing’s impact on Cu wire bonding quality and reliability.

The Galvanic corrosion theory has been widely adopted in explaining the failure mechanism of Cu ball bond lift issue during reliability test or field application in the presence of moisture. In this study, ion chromatography was performed on BGA substrate halogen analysis. EDX (energy-dispersive X-ray spectroscopy) was also used to detect the contaminant’s element at the bottom surface of a window clamp. Further FTIR (Fourier transform infrared spectroscopy) analysis verified that the contamination is from substrate outgassing during wire bonding. A new window clamp design proved effective in reducing the negative impact from substrate outgassing during wire bonding.

The solder mask in a fresh substrate contains a chlorine element. The chlorine can be detected in the BGA substrate outgassing during wire bonding by FTIR and EDX analyses, which have a negative impact on the Cu wire bonding. The window clamp with a larger opening can reduce the negative impact of the Cu wire bonding from the BGA substrate outgassing.

Because of the limitation of time and resources, bonding pad surface contamination from substrate outgassing and its correlation with Cu bonding ball lift failure after reliability test will be studied in depth later.

The BGA substrate outgassing has negative impacts on Cu wire bondability. A window clamp with a larger opening can reduce the negative impact from substrate outgassing.

This paper aims to investigate the root causes of and implement the improvements for the inter layer dielectric (ILD) crack for LQFP C90FG (CMOS90 Floating Gate) wafer technology devices in copper wire bonding process.

Failure analysis was conducted including cratering, scanning electron microscopy inspection and focus ion beam cross-section analysis, which showed ILD crack. Root cause investigation of ILD crack rate sudden jumping was carried out with cause-and-effect analysis, which revealed the root cause is shallower lead frame down-set. ILD crack mechanism deep-dive on ILD crack due to shallower lead frame down-set, which revealed the mechanism is lead frame flag floating on heat insert. Further investigation and energy dispersive X-ray analysis found the Cu particles on heat insert is another factor that can result in lead frame flag floating.

Lead frame flag floating on heat insert caused by shallower lead frame down-set or foreign matter on heat insert is a critical factor of ILD crack that has never been revealed before. Weak wafer structure strength caused by thinner wafer passivation1 thickness and sharp corner at Metal Trench (compared with the benchmarking fab) are other factors that can impact ILD crack.

For ILD crack improvement in copper wire bonding, besides the obvious factors such as wafer structure and wire bonding parameters, also should take other factors into consideration including lead frame flag floating on heat insert and heat insert maintenance.

The purpose of this paper was to attempt to confirm the root cause of wafer damage issue by heavy Al wire wedge bonding and propose some permanent solutions for it.

The infra red–optical beam-induced resistance change (IR-OBIRCH) analysis defines the position of an abnormal hotspot. A cross section and an scanning electron microscope (SEM) confirmed the wafer damage issue and its position. Based on the position of wafer damage, the wedge tool with different life and Al buildup was checked found to be on the wedge tool. Finite element analysis (FEA) modeling analysis and simulation experiment guarantee the Al buildup, and low wedge deformation thickness (WDT) can cause the wafer damage issue. Finally, design of experiment (DOE) experiments are designed to optimize wedge tool dimension and wedge-bond parameters to eliminate wafer damage issue.

Wafer damage issue caused the Vpwr-OUTPUT leakage issue by IR-OBIRCH analysis. Al buildup was found on wedge tool with different life and its size gets larger along with the increase in wedge tool life. Low WDT and bigger Al buildup can cause the wafer damage. Designing new wedge tool and parameters optimization can increase WDT.

Because of the limitation of time and resources, finite element method (FEM) modeling and wedge tool dimension could not be studied more deeply.

This paper sets an example on how to find out the root cause of wafer damage by a step-by-step analysis and put forward a quick solution accordingly for the issue.

Nanocomposites represent a new prospective branch in the huge field of polymer materials science and technology. It has been shown that an overall enhancement of properties of polymers can be achieved under certain conditions by the addition of nanoparticles. To examine the influence of microstructure on the tribological performance of nanocomposites, different ways of compounding were used in this study. It was found that the friction and wear behavior of polymeric nanocomposites under sliding environment was rather sensitive to the dispersion states of the nanoparticles. When the microstructural homogeneity of the nanocomposites was improved, their wear resistance could be increased significantly. The present work demonstrates the importance of TiO₂‐nanoparticles dispersion in an epoxy resin matrix, on the materials’ tribological properties, when sliding against a smooth steel counterpart.

The present study aims to find out the best polymer/polymer pair in electrical insulating applications. Moreover, the effects of different polymer counterpart and applied load on the friction and wear behaviour of PA 46 + 30%GFR and unfilled PA 66 thermoplastic polymers are to be studied.

Friction and wear tests vs PA 46 + 30%GFR and PPS + 30%GFR polymer composites were carried out on a pin‐on‐disc arrangement and at a dry sliding conditions. Tribological tests were performed at room temperature under 20, 40 and 60 N loads and at 0.5 m/s sliding speed.

The results showed that, the coefficient of friction decreases with the increasing of load (up to 40 N) for PA 46 + 30%GFR composite and polyamide (PA) 66 polymer used in this study. However, above 40 N applied load the coefficient of friction increases. The specific wear rate for PA 46 + 30%GFR and PA 66 against PPS + 30%GFR polymer composite counterpart are about in the order of 10⁻¹³ m²/N while the specific wear rate for PA 46 + 30%GFR and PA 66 against PA 46 + 30%GFR polymer composite counterpart are in the order of 10⁻¹⁴ m²/N. For PA 46 + 30%GFR composite and unfilled PA 66 polymers tested the specific wear rate values increased with the increment of load. The highest specific wear rate is for unfilled PA 66 against PPS + 30%GFR with a value of 2.81 × 10⁻¹³ m²/N followed by PA 66 against PA 46 + 30%GFR with a value of 2.26 × 10⁻¹³ m²/N. The lowest wear rate is PA 46 + 30%GFR polymer composite against PA 46 + 30%GFR polymer composite counterpart with a value of 3.19 × 10⁻¹⁴ m²/N. The average specific wear rates for unfilled PA 66 against PA 46 + 30%GFR is 80 times higher than PA 46 + 30%GFR wear rate while specific wear rates for unfilled PA 66 against PPS + 30%GFR is 100 times higher than that of PA 46 + 30%GFR wear rate. From point view of tribological performance, PA 46 + 30%GFR is a more suitable engineering thermoplastic composite materials for electrical contact breaker applications.

In the present work, tribological tests were performed only at room temperature under three different loads and a sliding speed. This is the limitation of the work.

This work is easily used for industrial polyamides to check their tribological behaviours.

This is an original and experimental study and it will be useful both for academicians and for industrial sides.