Metal Particles: The existence of both observable and subvisible particulate matter can pose risks during the whole biologic drug development, packaging, and delivery process.
A lot of sources of possible particulate contamination exist. Inherent particles, similar to protein aggregates, originate from the formula itself. Substantial contamination risks may also stem from intrinsic sources such as filter fibers or metal fragments from processing equipment, or even glass processors from primary packaging.
Extrinsic sources such as clothing or hair fibers also pose a risk of contamination, with other potential contaminants including organic, inorganic, protein, and metal particulates.
Raman spectroscopy provides users the ability to chemically fingerprint organic, inorganic, and protein particles, LIBS may be used to conduct elemental analysis, identifying metals and inorganics.
The identification of particles is carried out by matching their elemental and compound spectra with incorporated LIBS and Raman reference databases. Alongside these built-in databases, it is also possible to include custom reference spectra, allowing identification to be tailored to a specific procedure.
LIBS was employed to identify the elemental composition of five separate inorganic contaminants: a bottle cap fragment, a metal crimp cap fragment, copper wire, a syringe needle fragment, and a glass shard. Additionally, a custom reference database of items found in the test laboratory was used to identify the specific source of the metal cap fragment.
Visible particles were created by cutting or pruning materials frequently found in a lab. Metal particles were obtained from four distinct sources: a bottle cap, a common syringe needle found in the laboratory, a metal crimp cap from Wheaton Industries, and aluminum jumper wire. A glass particle was sourced from lead-barium glass shards.
Metal Particles Distribution
Each sample was prepared by placing the particle onto a nitrocellulose membrane glued to an aluminum mesh backing to set up an adhesive round. The glue round was completely dried before particle identification took place.
The glue round included five particles, all of which were automatically analyzed using LIBS. Next, the adhesive round was set on the Hound tool and the entire sample area was imaged with a 10x scanning objective.
The majority of the particles present were identified via one LIBS measurement. Particles that owned an outer coating (in this case, copper wire) necessitated the use of repeat measurements at the identical spot — this enabled the device to burn through the outer coating prior to collecting a measurement that allowed particle identification.
Spectra from each particle were assessed with the built-in LIBS reference database, allowing an identification. A match position between the sample itself and the reference spectra was calculated by multiplying the Pearson correlation by 1000.
Two repeated LIBS measurements were necessary before the composition of the copper wire could be recognized. The initial measurement provided an unidentified result, but this was able to burn through the cable’s outer coating.
With an expandable customized reference database and LIBS, the operator can produce an extremely specific reference spectrum of any substances used within their procedure or lab. This powerful functionality enables users to properly identify the exact origin of a particle if contamination is detected in the future.