Pharma glass defects - 3. Cracks

Hello everyone – welcome to Part 3 of an ongoing series about defects in pharma glass packaging.  As mentioned in Part 1 of this series, residual stress has been removed as a critical defect in the updated version of PDA Technical Report 43, that leaves us with Cracks as the only visual defect that is always considered Critical, irrespective of size or location on the container.

A Crack is a physical discontinuity produced by fracture that provides a continuous pathway from the exterior environment to the interior of the glass container (see Footnote 1).  Given the potential to be located anywhere on a container, cracks are considered a Critical defect because they can enable the ingress of microbes and/or chemical degradation through diffusion of reactive gases such as oxygen, carbon dioxide, and water into the drug product.  Multiple cases of drug products compromised by Cracks have been documented in the literature, some of which have resulted in patient injury or death (see Footnote 2).  Appreciating the significance of Cracks is easy enough, but understanding the root cause(s) and tactics for avoiding Cracks requires a little more discussion.

The creation of a Crack defect really means the creation of physically distinct surfaces by fracture.  What is the cause of this fracture?  Inorganic glasses of the sort used for pharmaceutical packaging are generally considered brittle materials.  The term “brittle” might imply that something is weak, but it has a specific meaning within the context of materials science.  Figure 1 shows basic “stress-strain curves” that illustrate the difference between brittle and ductile materials in response to an applied stress (see Footnote 3).  The application of a stress causes a change in the dimensions of a material that is measured as strain.  Assuming a tensile load, the slope of the stress-strain curve is the “elongation rate” – i.e., the rate at which the material lengthens in response to the tensile stress.  A brittle solid will generally show a linear elongation rate up to the point of fracture, whereas the elongation rate of ductile materials is nonlinear.  This means that a brittle material will return back to its original dimensions upon removal of the stress.  However, a ductile material will exhibit permanent deformation if it has experienced a stress beyond the initially linear portion of the stress-strain curve.  Upon reaching a given strain level, the atomic-/molecular-scale structure of a ductile material can change in a way that dissipates stored strain energy (see Footnote 4).  In contrast, the structure of brittle materials is relatively inflexible – strain energy continues to be stored up to the point where inter-atomic bonds are ruptured, resulting in fracture.

While stress-induced fracture is the general source of Cracks in glass containers, it’s worth noting that there are multiple pathways in practice to forming a Crack.  For convenience, I’ll broadly categorize them as fast or slow processes.  How fast is “fast”?  Terminal crack velocities in glass can range from about 700 to 2,500 m/s depending on glass composition and, by extension, properties of relevance such as glass density, elastic modulus, etc. (Footnote 5) – in other words, pretty damn fast.  The essentially instantaneous generation of a crack will occur when the applied stress is relatively high.  Such stresses can occur throughout the container life cycle, including but not limited to: thermo-mechanical stresses during vial conversion, shipping from the glass supplier to the fill-finish site, during the fill-finish process (e.g., heel impact of vials at transition points on the line, improperly set up cappers, etc.), and shipping from the fill-finish site to warehouses and healthcare providers.  A slow process to crack formation is likely linked to environmentally assisted stress corrosion, a process that I previously covered in my post on static fatigue of glass.  In brief, glass surface flaws (e.g., Check defects) can grow gradually over time to form a Crack due to the combined action of an applied tensile stress and chemically reactive water vapor.

What options are available for reducing the likelihood of Cracks in glass containers containing parenteral drugs?  One general approach is taking steps to decrease the magnitude of stresses encountered by the container throughout its life cycle.  Numerous tactics can be considered here – good handling practices by the glass supplier and the end user (see PDA Technical Report 87 for recommendations), proper set-up and maintenance of fill-finish equipment, optimization of container geometry, etc. External coatings can also help protect glass from typical sources of damage such as contact between containers on the fill-finish line.  Finally, various technologies for container-closure integrity testing and manual/semi-automated/automated inspection of visual defects can be used to detect cracks.  While these inspection technologies are generally quite sensitive, each approach has potential limitations that can allow Cracks to go undetected.  I would therefore strongly recommend not solely relying on inspection for crack prevention, particularly since Cracks can form further along the supply chain after inspection has already been performed.

Questions or comments? – please leave them below or feel free to directly contact me.

Footnotes

1.  I covered this in a previous post on strength, but it’s worth repeating – the definition that I just cited for a crack is specific to pharmaceutical packaging.  Outside of pharma, the term “crack” can take on a broader usage that also covers what I’ll generally call a flaw – i.e., a discontinuity that doesn’t necessarily fully penetrate the wall of a glass container.  This distinction is more than ontological fluff, given the potential impact of a cracked glass container on the safety and efficacy of a drug product.

2.  Selected references include:

·       Steinberg BL (1950).  Ampul contamination in spinal anesthesia. Anesthesiology, 11: 257-258.

·       Cope RW et al. (1952).  Letters to the Editors – Injection routine in operating-theatres. The Lancet, 259: 669.

·       Sack RA (1970).  Epidemic of Gram-negative organism septicemia subsequent to elective operation. American Journal of Obstetrics and Gynecology, 107: 394-399.

·       Roberston MH (1970).  Fungi in fluids – A hazard of intravenous therapy. Journal of Medical Microbiology, 3: 99-102.

·       Maddocks AC and Raine G (1972).  Contaminated drip fluids. British Medical Journal, 2: 231.

·       Elin RJ et al. (1975).  Evaluation of bacterial contamination in blood processing. Transfusion, 15: 260-265.

·       Daisy JA et al. (1979).  Inadvertent administration of intravenous fluids contaminated with fungus. Annals of Internal Medicine, 91: 563-565.

·       Boom FA, Paalman ACA (1981).  Barstjes in ampullen ten gevolge van spanning in het glass. Pharmaceutisch weekblad, 3: 35-38.  Translated title: Cracks in ampoules due to stress in the glass

·       Frechette CD (1990). Failure Analysis of Brittle Materials, American Ceramic Society, Westerville, OH, p. 116.

·       Bloomfield S et al. (2015). Lesser-known or hidden reservoirs of infection and implications for adequate prevention strategies: Where to look and what to look for. GMS Hygiene and Infection Control, 10: 1-14.

3.  Simply saying “applied stress” lacks details about the loading conditions that must be specified when performing real mechanical testing of materials.  Let’s assume in this case that we are performing a simple tensile test – i.e., we are attempting to pull the materials apart.  Also, the curves being shown in Figure 1 are for illustrative purposes only.  No particular significance should be placed on the relative slopes of the lines, the relative stresses at which the materials fracture, etc.

4.  The exact mechanism for dissipating stored strain energy depends on the class of material.  For example, plastic deformation in metals tends to occur by mechanisms such as slip movement, twinning, and grain boundary sliding.  The extent to which these mechanisms are active depend on the specific composition/structure of the material – e.g., while both are metals, cast iron is relatively brittle while copper is relatively ductile.  Environmental conditions are obviously important as well, and not always in an obviously intuitive manner.   Copper actually becomes more ductile with decreasing temperature, while steel alloys tend to become more brittle.

5.  Refer to the following excellent review article for more information on terminal crack velocities in glass:  Quinn GD (2018).  On terminal crack velocities in glass. International Journal of Applied Glass Science, 10: 7-16.