Failure analysis (FA) is a term that product and quality mangers don’t often wish to articulate when considering the expense and potential complexity associated with the investigatory process. Unfortunately, it is a necessity more often than not, and part of the discovery process to isolate the root cause of electronic assembly failures. Properly utilized, FA can be a tool to confirm and identify the mechanism that instigates the causes of electrical failures, such as solder joint crack propagation, dendritic growth, ionic residue, and other sources of contamination that either impede or redirect current flow. The FA lab at the EMPF has documented many such modes of failure where the origin of the problem was traced directly to a faulty manufacturing process, or a material deficiency. In some cases the designs proved to be contributors to assembly failures by exceeding the limits of the manufacturing process to produce reasonable yields.

Failure analysis can be utilized for a number of different reasons:

• Direct Cause – an example of this may be a cracked die or component (Figure 6-1) where the obvious cause of failure has been isolated and identified.
  
  Another such example of direct cause can be the verification of incorrect metallurgy using X-Ray Fluorescence (XRF) analysis of a component, such as a ball grid array (BGA), resulting in cold solder joints.
  
  
• Indirect Cause – The cleanliness of a printed circuit board (PCB) is an example where the ionic concentration is greater than the allowable IPC specification (Table 6-1), but the actual area of failure has not been established.
  
  
• Process Indicators – One recent example seen at the EMPF indicated that a potential for de-wetting was beginning to occur as the BGA solder balls began to recede from the pad area, even though the BGAs retained a collapsed appearance. This indicated excessive time spent in the liquidus stage.
  
  
  
  
• Design Flaws – An example of this would be a “via in pad” board design for BGA placement, where the micro-via exhibited insufficient copper deposition due to process limitations. As a result, imbedded air pockets were lodged directly beneath the BGA ball, where insufficient solder wetting in the micro-via entrapped the air during reflow.
  
  Failure Detection Tools
  Analytical methods for detecting failure can be segregated into two categories:
  
  Destructive Methods – Where the sample for analysis will undergo physical or chemical changes rendering it inoperative for production use. This can be further divided into modes of detection:
  
  
  • Qualitative – usually identifies the nature of the failure and can incorporate such methods as:
    • Optical or Visual Assessments – the use of stereomicroscopes for assessing cross sectional samples. Can be a quantitative method as well as a qualitative method.
    • Spectroscopy – can include scanning electron microscopy (SEM) analysis for identification of intermetallic boundary
      layers in solder joints or dendrites on the surface between two adjacent conductors.
      
      
  • Quantitative – will ascertain the amount of the contaminant or source of failure, and in many cases identify its specific nature. Some examples are
  • Inductively Coupled Plasma (ICP) Spectroscopy – can
    identify and determine the specific quantities of a given
    element within the sample.
  • SEM/Energy Dispersive Spectroscopy (EDS) – can quantify the thickness of particular intermetallic layers, and in some cases identify the oxidation states.
  • Wetting Balance – uses the inherent nature of surface tension against various solders and pad finishes to calculate a force of retraction as a measure of wettability.

  Non-Destructive Methods – Can allow for re-use of the sample into production since the functionality of the assembly is retained. This needs to be a discretionary decision since the potential does exist for inadvertent damage through the course of analysis. Some examples of non-destructive testing are:

  • Fourier Transfer Infrared (FTIR) – a spectroscopic method of detecting organic contamination on the surface of a substrate or assembly.
    
    
  • Ionograph – an electrochemical method of detecting ions and measuring the total amount extracted from an assembly. Values are given as a relative quantity of NaCl.
    
    
  • XRF – another spectroscopic method of determining alloy
    constituents. This is a useful tool for screening component finish and PCB metallic thickness without destructive cross-sections.
    
    
  • Ion Chromatography (IC) – will quantify and identify specific ions in assemblies. Usually used as a process qualifier for flux and cleaning systems, where the results can be compared against the IPC specifications for board cleanliness.
    
    
  • X-Ray – a critical diagnostic tool for contract manufacturers who assemble BGAs as a means of detecting shorts, opens and incomplete collapse of the solder balls during thermal reflow.
    
    
  • Automated Optical Inspection (AOI) – an instrument used as a process tool for inspection and an indication of good paste deposition, component placement, and optical detection of surface anomalies on assemblies.

The EMPF training center offers courses in failure analysis that avail the students the opportunity to experience the use of many of these analytical tools. More importantly, the course is geared to assist the students in identifying the root causes of potential failures in bare boards, populated assemblies, and components, while utilizing that knowledge to select the best method of verification, and more critically, the recommended corrective actions to fix the problem.
For more information on FA training, please contact the Registrar at 610-362-1295 or email registrar@empf.org.