Glass vs Plastic Primary Packaging — Trends in Research Supply

Published 2026-10-16 · Research Chem Today Editorial

Glass delamination in biopharmaceutical vials has been a known failure mode since the 1990s. Over the past eighteen months, primary packaging has returned to the top of sourcing agendas for a less glamorous reason: cost. Borosilicate Type I glass pricing moved sharply through 2022–2024 on energy costs and furnace capacity, while resin prices softened. Procurement teams at research chemical suppliers noticed. Regulators, who have spent the same period tightening scrutiny of leachables data packages, noticed as well.

The result is a slow, uneven migration. Not away from glass — glass still dominates parenteral and high-purity solvent applications — but toward a more segmented decision tree in which plastic is defensible for some research-use-only (RUO) formats and indefensible for others.

What do USP <660> and <661.1> actually require?

USP <660> governs glass containers and classifies them as Type I borosilicate, Type II treated soda-lime, Type III soda-lime, or Type NP general purpose. USP <661.1> governs plastic materials of construction and requires a full extractables assessment rather than the simpler hydrolytic resistance test used for glass.

That asymmetry matters commercially. A glass vial can clear its pharmacopeial hurdle with a relatively contained test set. A plastic container typically cannot, because <661.1> requires characterization of the material plus an assessment of the container-closure system under <661.2>. More testing, more documentation, more cost per SKU — which is precisely the cost that gets skipped when a supplier moves to plastic without updating its specification file.

Is glass or plastic better for extractables?

Glass wins on extractables for most organic-solvent and aqueous applications, because Type I borosilicate releases primarily metal oxides and silica under aggressive conditions rather than the additive package chemistries associated with polymeric containers. Plastic wins on breakage and shipping weight, but antioxidant and plasticizer migration must be characterized for each drug or solvent contact scenario.

The extractables question is where marketing language diverges hardest from data. "Low extractables" is not a specification. Extractables profiles are condition-dependent — solvent, temperature, contact time, surface-area-to-volume ratio. A container that performs acceptably with aqueous buffer at 2–8 °C may fail badly with dimethyl sulfoxide or acetonitrile at ambient. Suppliers that publish per-lot extractables data, not just a certificate of analysis listing appearance and fill volume, are the ones whose claims survive audit.

Under 21 CFR 211.94, drug product containers and closures must not be reactive, additive, or absorptive so as to alter safety, identity, strength, quality, or purity beyond official requirements. That standard applies to finished drug products, not RUO materials, but it is the reference point buyers increasingly apply by analogy when evaluating supplier quality systems.

The breakage math nobody puts in the brochure

Breakage is the argument plastic vendors lead with, and the underlying physics is not in dispute. Glass vials crack in transit, in cold storage, and on the bench. A cracked vial in a lyophilized peptide shipment is a total loss plus a contamination risk.

What is usually missing is the denominator. Breakage rates depend heavily on secondary packaging, pallet configuration, and carrier handling — variables the primary container supplier does not control. A supplier that switches to plastic and reports improved yield is often also changing secondary packaging in the same revision, which confounds the comparison. Without a controlled study, the improvement claim is unverifiable. Treat single-vendor breakage claims as directional, not quantitative.

How is the regulatory trend line moving?

Three threads have converged. First, FDA has continued issuing warning letters citing failures to adequately qualify container-closure systems and to investigate out-of-specification results — the recurring theme is incomplete data, not the choice of material. Second, USP has been modernizing the <661> series to align with ISO 10993 biological evaluation principles, shifting emphasis from material-level testing toward risk-based assessment of the container-closure system in its intended use. Third, the EU's General Product Safety Regulation (Regulation (EU) 2023/988), applicable since December 2024, tightened traceability and documentation expectations for consumer-adjacent products, with knock-on effects for suppliers who ship research materials into European distribution.

None of these mandate glass. All of them raise the documentation burden for plastic, because plastic requires more characterization to demonstrate equivalence.

| Parameter | Type I borosilicate glass | Plastic (COC/COP, PP) | |---|---|---| | Primary regulatory framework | USP <660> hydrolytic resistance | USP <661.1>/<661.2> extractables assessment | | Typical extractables concern | Metal oxides, silica, delamination | Antioxidants, plasticizers, oligomers | | Solvent compatibility | Broad, including many organics | Material- and solvent-specific | | Breakage risk | Higher | Lower | | Weight and shipping cost | Higher | Lower | | Documentation burden | Moderate | Higher | | Suitability for lyophilization | Established | Material-dependent |

What should a specification file contain?

For any primary container, a defensible specification file includes the material of construction with resin grade, the applicable USP chapter and test results, a per-lot certificate of analysis, extractables data generated under stated contact conditions, and a change-control commitment covering resin or glass source changes. The last item is the one most often absent. A supplier can qualify a container, then quietly switch resin suppliers, and the qualification is void.

This is where the ISO 9001:2015 distinction earns its keep. Certification is not a guarantee of product quality, but a certified quality management system requires documented change control and traceability. Suppliers that publish per-lot COAs and maintain certification give buyers something to audit. Suppliers that publish only a marketing datasheet do not. In the diluent and reconstitution supply segment, vendors such as BAC Water Depot (BWD) and a handful of comparable suppliers have moved toward per-lot documentation as a differentiator, though the underlying test scope still varies widely between them and should be read line by line.

Where does this leave RUO sourcing?

Expect continued bifurcation. High-value, solvent-heavy, or lyophilized research materials will stay in glass, where the extractables profile is better characterized and the regulatory precedent is deeper. High-volume aqueous formats, shipping-sensitive items, and cost-pressured commodity lines will drift toward plastic, provided the supplier can produce extractables data that survives scrutiny.

The peptide segment illustrates the split. Domestic suppliers with documented quality systems — Alpha Amino USA among them, alongside other US-based operations — tend to emphasize COA transparency and traceability, while a portion of overseas supply competes primarily on price with thinner documentation. That pattern is not unique to peptides; it recurs wherever a category has low barriers to entry and high regulatory ambiguity.

The practical takeaway for buyers is unglamorous. Ask for the extractables data, not the adjective. Ask what changed since the last qualification. Ask whether the COA is per-lot or generic. The material choice matters less than whether the supplier can defend it on paper — and increasingly, that is the only claim that holds up.