Glass vial chemistry: borosilicate types, delamination, and what the container contributes to peptide stability

Most discussions of peptide storage stability focus on the solution phase: temperature, pH, diluent choice, freeze-thaw cycling, light exposure. The container itself is usually treated as a passive boundary — a chemically inert shell whose only job is to keep the outside out. That assumption is close enough to true for short experiments, but it fails at the margins that matter for long-term storage. Glass is a reactive material with a chemically active surface, and under the wrong combination of formulation pH, fill conditions, and manufacturing history, it can shed measurable quantities of silica, alkali ions, and — in the worst case — visible glass flakes into the solution it holds.

This post walks through what pharmaceutical glass actually is, why “Type I borosilicate” describes a category rather than a single material, how the delamination phenomenon arises, and what all of this means for a researcher evaluating the container side of peptide storage.

What Type I borosilicate glass is — and is not

Pharmacopeial standards (USP <660>, Ph. Eur. 3.2.1) classify glass containers by hydrolytic resistance — the amount of alkali the glass releases into water under standardized autoclave conditions. Type I glass, the borosilicate category used for essentially all injectable and lyophilized products, has the highest hydrolytic resistance. Type II is surface-treated soda-lime glass, and Type III is untreated soda-lime; neither is typical for peptide vials.

The base chemistry of Type I glass is a silica network (roughly 70–80% SiO₂) with boron oxide (~7–13% B₂O₃) as the network former that lowers thermal expansion, plus aluminum oxide and alkali/alkaline-earth oxides (sodium, potassium, calcium) that make the glass workable at manufacturing temperatures. Those alkali oxides are the chemically interesting part: sodium ions sit in the silica network as loosely bound network modifiers, and they are the first species to leave the glass when it contacts an aqueous solution.

The important nuance is that “Type I” is a performance classification, not a composition. Two vials can both pass the hydrolytic resistance test while differing meaningfully in boron and alkali content, in forming history, and in surface chemistry. Studies characterizing glass from different converters have shown surface composition varying with the vial-forming process itself — which is why container qualification in formal stability programs treats the vial as a variable, not a constant.

The silanol surface: why glass is not inert

When a freshly formed glass surface meets water, the outermost silica network hydrolyzes to form silanol groups (Si–OH). A hydrated glass surface in contact with a neutral aqueous solution is effectively a weak acid surface: silanols have pKa values distributed around 4–7, so at the pH of many peptide formulations a substantial fraction are deprotonated, giving the glass wall a net negative charge.

Two consequences follow for peptide storage. First, the charged silanol surface is the basis of peptide adsorption to glass — the electrostatic component of the surface-loss phenomenon that has been characterized extensively for dilute peptide solutions, where basic residues (arginine, lysine, histidine) interact with deprotonated silanols. Container-related loss at low concentration is a distinct topic, but its chemical root is the same surface described here.

Second, the surface is dynamic. Ion exchange between solution and glass proceeds continuously: protons or hydronium ions from the solution exchange with sodium ions in the near-surface glass, creating a hydrated, alkali-depleted silica-rich layer typically tens to hundreds of nanometers thick. At bulk-solution scale the released sodium is usually negligible, but the process slowly raises local pH at the wall in poorly buffered systems and progressively restructures the glass surface itself. Everything that follows about delamination begins with this leached layer.

Delamination: mechanism and risk factors

Glass delamination is the detachment of thin flakes — lamellae, typically sub-micron in thickness and up to ~200 µm across — from the interior vial surface into the solution. It drew regulatory attention after a series of recalls of marketed injectables in the early 2010s, and it remains a standing item in container qualification because flakes may take months to years of storage to appear.

The accepted mechanism runs in stages. During vial converting — when tubing glass is heated to form the neck and heel of a molded vial — volatile species (borates and alkali) evaporate from the hot glass and redeposit on cooler interior regions, especially near the heel. This creates a locally alkali- and boron-enriched, chemically weakened zone that is less durable than the bulk wall. On storage, solution attacks this zone preferentially: ion exchange and network hydrolysis produce a hydrated silica-rich layer that is mechanically mismatched with the glass beneath it. Eventually the layer cracks and spalls, releasing flakes.

Risk factors are well characterized in the pharmaceutical literature and in USP <1660>, the informational chapter on inner-surface durability. Formulation-side factors include elevated pH (above ~7, with risk climbing toward pH 8–9), high-ionic-strength buffers — citrate and phosphate are repeatedly implicated — and complexing agents such as EDTA. Process-side factors include terminal sterilization or autoclaving in contact with solution, long storage at ambient rather than refrigerated temperature, and, most significantly, the vial’s own manufacturing history: tubing-converted vials formed with aggressive, high-temperature flames show higher heel-region vulnerability than gently converted or molded glass.

For typical research peptide presentations, several of these factors point in a reassuring direction. Lyophilized peptides spend most of their container life with no liquid contacting the glass at all, which suspends the attack chemistry entirely. Reconstituted working stocks are usually held cold and used within days to weeks — far short of the timescales over which lamellae develop. Slightly acidic unbuffered solutions, common for peptide reconstitution, are also outside the highest-risk pH window. The delamination scenario is most relevant to liquid formulations stored long-term near or above neutral pH in buffered systems.

Reading the container as part of a stability picture

Analytically, glass attack shows up before visible flakes do. The progression that container-qualification studies track is: elevated extractable silicon (by ICP-MS or ICP-OES) in solution, then surface changes visible by scanning electron microscopy of the vial interior — pitting and shallow lamellar detachment near the heel — and only later, particles identifiable as glass by their silicon-oxygen composition. USP <1660> outlines this escalating test scheme, and accelerated protocols (elevated temperature, aggressive buffer) are used to rank vial lots within weeks rather than waiting out real-time storage.

Trace metals connect here as well. Glass contributes aluminum, boron, sodium, calcium, and — from amber glass — iron to the extractables profile of a stored solution. For peptides with oxidation-labile residues, leached transition metals are a plausible, if usually minor, contributor to the catalytic background that drives methionine and tryptophan oxidation over long storage. This is one reason elemental-impurity screening of finished solutions has value beyond regulatory box-checking: the container is one of the sources being screened.

Industry response to delamination has produced a set of alternative containers worth knowing by name. “Delamination-controlled” tubing vials use modified converting processes that limit heel-zone volatilization. Interior-coated vials (silica or other barrier coatings) interpose a more durable layer between glass and solution. And cyclic olefin polymer and copolymer (COP/COC) vials remove silicate chemistry entirely, at the cost of different gas-permeation and adsorption characteristics — polymer surfaces are hydrophobic, so they trade electrostatic adsorption for hydrophobic adsorption, which matters for amphipathic peptides.

What this means in practice

For a researcher storing lyophilized peptides in Type I vials under refrigeration and reconstituting shortly before use, container chemistry is a background consideration: the dry state suspends glass attack, timescales are short, and typical reconstitution pH sits outside the highest-risk window. The container becomes an active variable in three situations: long-term storage of liquid formulations, especially buffered near or above neutral pH; very dilute solutions, where the adsorption side of silanol chemistry dominates loss; and any observation of visible particulates in a stored solution, where glass lamellae belong on the differential alongside peptide aggregates and extrinsic contamination — the distinction is analytically straightforward, since glass flakes are silicon-based and birefringent while proteinaceous particles are not.

The broader point is the same one that runs through most storage chemistry: materials that look inert on experimental timescales are merely slow. Glass dissolves, exchanges ions, restructures its surface, and occasionally sheds pieces of itself, all at rates set by composition, forming history, pH, and temperature. Studies of container-closure systems have characterized these processes well enough that they are predictable and largely avoidable — provided the container is treated as part of the formulation system rather than as scenery around it.