A Case Study Using CT Inspection Data

By: Joel F. Flumerfelt, Ph.D., Metallurgist

A client who manufactures copper heat sinks had one returned from their customer that displayed three surface discontinuities, Figure 1. Scanning Electron Microscopy (SEM) examinations conducted on Discontinuity 1 and Discontinuity 2 showed them to be arc spots created by an external energy source, such as may occur during an electrical discharge. The SEM examinations of Discontinuity 3 showed features that were ambiguous. A Scanning Acoustic Microscopy (SAM) inspection performed on Discontinuity 3 indicated an internal void, Figure 1, partially enveloped by delamination, i.e. cracks, within an alumina dielectric material between the top copper layer and the copper substrate, Figure 2.

A Computerized Tomography (CT) inspection of Discontinuity 3 facilitated an additional non-destructive characterization of this internal discontinuity, Figure 3. The permanent, digital CT records provided the possibility to review virtual cross-sections of the void after a destructive analysis. The CT slices showed the void had a spheroidal geometry. The CT slices and clipped isoview images also indicated the presence of a metal “uvula” hanging from the top copper layer and metal “nuggets” laying on the copper substrate. The dielectric material was not apparent in these images due to its low radiopacity relative to the surrounding copper.

The CT inspection, coupled with the complementary SAM data, provided a priori knowledge about how best to orient Discontinuity 3 in a metallographic cross-section to optimize further portrayals of the discontinuity’s physical nature. Without this non-destructive data, an otherwise random cross-section may have destroyed internal features that would truncate an understanding of how and why the void formed.

A cross-section plane through the center of the void showed evidence of solid-state plastic deformation within the grains of the copper metal surrounding the void. The cross-section examinations also revealed the “uvula” and “nuggets” were copper, Figure 4 to Figure 6. The porosity present in these particles was due to them making a transition from a molten state to their solid form. All these features observed in the cross-section were the result of a high energy, explosive event that caused a localized temperature spike in excess of the melting temperature of copper, which is 1,085°C.

A postulated mechanism for the explosive event was an electrical short circuit that took place between the two copper layers. The potential cause of the short circuit path was the presence of copper particles unintentionally embedded in the alumina dielectric material during the fabrication of the heat sink. Evidence to support this opinion was found in the form of another group of copper particles entrenched in the alumina dielectric material, observed elsewhere within the heat sink after peeling away the top layer of copper, Figure 7.

Figure 1

Figure 1 – Physical, non-destructive characterizations of surface discontinuities on a heat sink. Optical image scale is 0.5-mm / division, and the SEM micrograph scales are 100-µm. SAM image shows a void (black phase) partly surrounded by delamination (white phase) within an alumina dielectric material between two layers of copper.

Figure 2 image

Figure 2 – Cross-section of heat sink showing a layered construction of a thin copper top layer over an alumina dielectric material (white layer) on top of a copper substrate. The crack in the dielectric material is the delamination identified in the SAM inspection.

Figure 3

Figure 3 – CT images of Discontinuity 3 in the form of orthogonal views, depicted by the gray scale image slices, and a clipped, isoview image portraying a 3D rendering of the void.

Figure 4

Figure 4 – Etched metallographic cross-section of Discontinuity 3 showing plastic deformation within the grains of the top copper layer and copper substrate. A copper “uvula” hangs from the top copper layer, and copper “nuggets” rest on the copper substrate. The dark phase between the copper layers is the alumina dielectric material.

Figure 5

Figure 5 – Magnified view of copper “uvula” shown in Figure 3 and Figure 4.

Figure 6

Figure 6 – Magnified view of copper “nuggets” shown in Figure 3 and Figure 4.

Figure 7

Figure 7 – Copper particles found in alumina dielectric material remote from Discontinuity 3 after peeling away the top copper layer from the body of the heat sink.


New USP Testing Methods

By Matt Koval, Aspen Research Corporation, August 28, 2017

In the past, the USP listed a “Heavy Metals as Lead” specification for compounds and specified a wet chemical method, USP<231>, that was introduced in 1908.  This method is a color comparison test that was dependent on the visual acuity of the analyst and relied on the interaction of elements present in a sample with sulfide.  This is more of a screening test and is not element specific.  In fact, the response of the elements in question and the sensitivity of their response to the sulfide reagent varies by element and often under-reported.

In 2009, the International Conference on Harmonization (ICH) began the process of updating the assessment of elemental impurities in drug products.  As a result, on January 1, 2018, USP<232> and <233> will be implemented.  USP<232> takes a toxicological approach to elemental impurities of a drug product that specifies elements and their daily exposure limits based on route of entry of the drug.  USP<233> provides the guidelines for the instrumental methodology to be used and gives method validation criteria for the test methods.  Companies will have 36 months after the implementation to apply the new rules to existing drugs.  Methods specified in USP<233> are inductively coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled plasma mass spectroscopy (ICP-MS).  For the preparation of samples for testing, a closed vessel system is recommended to mitigate the loss of volatile elements.

The limits in USP<232> apply specifically to the drug product.  Unless specified by a monograph, they do not apply to drug substances or excipients.  However, elemental impurities may come from anything that is involved in the manufacture, storage, transport and administration of a drug.  Because of this, drug manufactures will require information from their suppliers regarding the elemental impurities of reagents, APIs, excipients, processing equipment, packages, delivery devices, etc.