Richard J. Coyle

The influence of room temperature aging on ball shear strength and microstructure of area array solder balls

Solder ball shear strength is determined for ball grid array (BGA) packages in the as-received condition and after solder reflow preconditioning (exposure to multiple solder reflow profiles). Immediately after reflow, the solder shear strength is considerably higher than the solder shear strength measured on packages in the as-received condition. When the shear strength is monitored as a function of the time after solder reflow preconditioning, it is found to decrease significantly with time. Microhardness tests made in parallel with the shear measurements, provide quantitative confirmation of a degradation in mechanical properties of the solder over time. The observed changes in mechanical properties are the result of room temperature annealing, or age-softening of the solder. Cross sectioning and scanning electron microscopy are used to identify and track microstructural changes in the solder balls with room temperature aging. The microstructural changes are correlated to the decrease in shear strength and microhardness and in some cases, to a change in failure mode. These results are discussed in terms of accurate interpretation of shear test data and suggestions are made for test procedures that can minimize data variations induced by solder aging effects.

The effect of modifications to the nickel/gold surface finish on assembly quality and attachment reliability of a plastic ball grid array (peer review version)

Electrolytic and electroless Ni/Au are common pad surface finishes on area array (BGA or CSP) packages and printed wiring boards (PWB). The electroless nickel/immersion gold (ENIG) process often is implemented when there is insufficient space to allow bussing for the more common electrolytic Ni/Au plating. The ENIG process continues to be used despite evidence that it may cause catastrophic, brittle, interfacial solder joint fractures. In this investigation a plastic ball grid array (PBGA) test vehicle is used to compare quality and reliability of standard and experimentally modified ENIG surface finishes. The standard electrolytic Ni/Au surface finish is used as the control cell for the experiment. Ball shear tests and optical and scanning electron microscopy are performed on as-received and thermally preconditioned packages to evaluate package quality prior to assembly. Accelerated temperature cycling (0/+100/spl deg/C and -40/+125/spl deg/C) is used to evaluate solder joint attachment reliability. Detailed failure mode analysis is used to compare the fracture modes in the ball shear and thermal cycled samples in the electroless and electrolytic packages. The results are discussed in terms of the failure modes and the characteristics of the different Ni/Au surface finishes.

The influence of test parameters and package design features on ball shear test requirements

Ball shear testing is the most common test method used to assess solder ball attachment quality on area array packages. New package designs have a range of package pitch, pad sizes, solder ball diameters and surface finishes that can complicate basic ball shear testing and data analysis. The current area array standard, generalized for typical 1.27 mm pitch BGA packages, must be expanded to address issues and problems associated with evolving area array technology. One persistent problem is random occurrence of poor attachment quality, manifested as catastrophic brittle fracture. Although the failure mode has been added to ball shear acceptance criteria, optimum parameters to test for this defect are not defined or specified. Many studies recommend thermal preconditioning in conjunction with ball shear to expose latent brittle defects. Isothermal aging and exposure to multiple reflow profiles are used commonly for thermal preconditioning but there is no universal set of test parameters. This investigation identifies critical test parameters and package variables and evaluates their effect on ball shear test data. Shear tests are conducted as a function of the test displacement rate and two key thermal preconditioning variables: (1) the number of reflow cycles used for reflow preconditioning and (2) the time and temperature for isothermal aging. The package variables of interest are the bond attachment pad diameter and the solder ball diameter, which are related to package pitch. The results are discussed in terms of accurate interpretation of shear test data and suggestions are made for addition to existing ball shear test standards.

The influence of nickel/gold surface finish on the assembly quality and long term reliability of thermally enhanced BGA packages

Increasing numbers of chip carrier packages such as BGA and CSP use electroless nickel/immersion gold (ENi/IAu) surface finishes. However, the results of several studies attribute solder ball brittle failure to this surface finish. Brittle fractures have occurred in service and have been generated under various testing conditions. The typical failure mode is brittle interfacial fracture between the Ni plating and the intermetallic compound formed during soldering. Brittle interfacial failures are unpredictable and catastrophic but occur at relatively low defect levels. Discounting premature or infant mortality failures, ENi/LAu components can exhibit satisfactory long term reliability performance. This behavior is indicative of assembly quality or package quality defects, not true wear-out failures. This paper compares the results of quality and reliability testing of three sets of 560 I/O thermally enhanced BGA packages. Two sets were constructed with substrates using ENi/IAu surface finishes and the third is constructed with substrates using a typical electrolytic Ni/electrolytic Au surface finish. Ball shear tests, optical microscopy and SEM are performed on as-received and thermally preconditioned packages to evaluate package quality prior to assembly. Accelerated thermal cycling is used to evaluate long term reliability. Failure mode analysis is used to compare fracture modes in the ball shear and thermal cycled samples in the electroless and electrolytic packages. The results are discussed in terms of proposed mechanisms for brittle fracture and known characteristics of the thermally enhanced packages.

Surface finish effect on reliability of SAC 305 soldered chip resistors

Purpose – The purpose of this paper is to assess the long‐term reliability of lead‐free Sn96.5Ag3.0Cu0.5 (SAC305) under accelerated temperature cycling (ATC) conditions. Test vehicles consisted of commercial 2512 ceramic chip resistors soldered to printed circuit boards (PCBs) using three different Pb‐free surface finishes: organic solderability preservative (OSP), immersion silver (IAg) and electroless nickel immersion gold (ENIG).Design/methodology/approach – Two populations of solder joints were monitored continuously during a thermal cycle of 0°C to 100°C with a ramp rate of 10°C/min and a 30 min dwell at the temperature extremes. One population was cycled to 2,500 cycles, the other population was cycled to 8,250 cycles. Failures were defined in accordance with the IPC‐9701A industry test guidelines and failure data are reported as characteristic life, η. Microstructural evolution was characterised using metallographic techniques and back‐scattered scanning electron microscopy (SEM). Fractography was ...

Ball shear versus ball pull test methods for evaluating interfacial failures in area array packages

In this investigation, a ball pull (tensile) test is investigated as an alternative to the ball shear test for evaluating the solder joint integrity of area array packages. The relative effectiveness of the pull and shear methods is compared using BGA packages with documented susceptibility to brittle interfacial failure during accelerated temperature cycling tests or isothermal aging. Accelerated temperature cycling is used typically to measure long term solder joint attachment reliability in various use environments and isothermal aging is used to measure susceptibility to degradation following high temperature storage. The shear and pull tests are conducted on packages in the as received condition and after thermal preconditioning. Metallographic failure analysis and scanning electron microscopy with energy dispersive X-ray analyses are used to characterize the solder joints and fracture modes. The ball shear and ball pull results are compared and discussed in terms of the ability to predict susceptibility to interfacial failures in area array packages.

Laser weld process improvements for optical isolator assembly

Optical isolators are used in lightwave systems to reduce feedback that could be propagated in the signal fiber. Several methods can be used to assemble and join the various components which comprise the completed device. Since the single mode fiber core is 8 microns in diameter, it is desirable to utilize a robust assembly technique which will produce a minimum of distortion during the assembly process. Process induced distortion could randomly change the device coupling which may have an adverse effect on operation yield. With the development of laser welding systems, devices can be manufactured with localized welds providing small amounts of thermal distortion. This paper compares two isolator assembly techniques: resistance welding and laser welding. A comparison of device power losses after the weld sequence and thermal cycling data shows that laser welding is a more robust and stable technique for isolator manufacture.

The effect of variations in nickel/gold surface finish on the assembly quality and attachment reliability of a plastic ball grid array

Electrolytic and electroless Ni/Au are common pad surface finishes on area array (BGA or CSP) packages and printed wiring boards (PWB). The electroless nickel/immersion gold (ENIG) process often is implemented when there is insufficient space to allow bussing for the more common electrolytic Ni/Au plating. The ENIG process continues to be used despite evidence that it may cause or contribute to catastrophic, brittle, interfacial solder joint fractures. In this investigation a plastic ball grid array (PBGA) test vehicle is used to compare quality and reliability of four variations of the ENIG surface finish. The standard electrolytic Ni/Au surface finish is used as the control cell for the experiment. Ball shear tests and optical and scanning electron microscopy are performed on as-received and thermally preconditioned packages to evaluate package quality prior to assembly. Accelerated temperature cycling (0/+100/spl deg/C and -40/+125/spl deg/C) is used to evaluate solder joint attachment reliability. Detailed failure mode analysis is used to compare the fracture modes in the ball shear and thermal cycled samples in the electroless and electrolytic packages. The results are discussed in terms of the failure modes and the characteristics of the different Ni/Au surface finishes.

Application of an Excimer Laser Machining Technique in the Construction of an Optical Fiber Bus

An excimer laser technique was developed to machine precision tap openings in a polymer‐clad silica optical fiber for use in the construction of an optical backplane. In this technique, a telescope was used to image the laser beam onto the surface of the fiber at the desired tap location. The cladding then was removed using a prescribed number of pulses from the excimer laser. Single optical fiber taps were constructed using an ultraviolet‐curable acralate to attach a tap fiber to a laser‐machined opening in a central fiber. Tap measurements were made on single taps constructed from 650 × 350 micron elliptical openings and 250 micron diameter circular openings. Tap measurements also were made on a triple tap assembly which was fabricated using elliptical tap openings. The elliptical openings, with the major axis oriented along the longitudinal axis of the fiber, were determined to be more suitable for tap construction. The results of this study indicate that excimer laser machining is a promising techniqu...

Excimer laser machining of optical fiber taps

Precision openings for construction of an optical backplane have been machined in an optical fiber using an excimer laser operating at a wavelength of 193 nm. The openings were made by imaging the laser beam onto the polymer fiber cladding with a telescope, then ablating the cladding with a sufficient number of pulses to expose the core. Circular openings measuring 250 and 625 microns and elliptical openings measuring 650 X 350 microns have been made in the cladding of a 1 mm polymer-clad silica fiber. Examination by scanning electron microscopy reveals that the best quality openings are obtained with either the smaller circular geometry or the elliptical geometry. For various reasons, elliptical openings, with the major axis oriented along the longitudinal axis of the fiber, appear more suitable for tap construction. Individual optical fiber taps have been constructed by attaching a tap fiber to a laser machined opening in a central fiber using an ultraviolet-curable acralate. Individual tap measurements were made on the elliptical and the 250 micron circular openings. In addition, a triple tap assembly was made using elliptical tap openings. These results indicate that the excimer laser machining technique may be applicable to the construction of a linear tapped bus for optical backplanes.