Hexamethyldisiloxane of pharmaceutical grade is not formulated as an active pharmaceutical ingredient; in oral and injectable manufacturing it functions as a vapour-phase silylating agent, plasma polymerization precursor, and hydrophobic contact-surface modifier. The compound has a molecular weight of 162.38 g/mol, a boiling point of 100–101 °C at 101.3 kPa, and a density of approximately 0.764 g/cm³ at 25 °C. A pharma-grade certificate of analysis commonly includes chromatographic purity ≥99.0% by GC-FID, water ≤0.05% by Karl Fischer titration, and evaporation residue ≤0.005%; however, acceptance limits remain supplier-defined because no harmonized USP or Ph. Eur. monograph applies to HMDSO as a direct excipient. In a Type I borosilicate glass vial campaign, pre-dried vials are held at 60 °C for 30 min before HMDSO vapour is introduced at 120–160 °C for 2–4 h under 0.2–0.5 bar absolute nitrogen. The surface water contact angle shifts from below 10° to 95–110°, measured by sessile drop on a Krüss DSA100 or equivalent. The reaction consumes free silanol groups and releases trimethylsilanol, forming a covalently bound trimethylsilyl layer. Treated vials are evaluated using USP <1660> for glass delamination tendency and ISO 4803:2021 for inner surface hydrolytic resistance. The treatment is confined to the primary container contact surface; HMDSO is not listed in ICH Q3C(R8) class 1/2/3 residual solvent tables, so final-product screening by headspace GC-MS according to USP <467> is required during process qualification.

What Limits HMDSO Plasma Polymerisation on Cycloolefin Copolymer Syringe Barrels?

In cyclic olefin copolymer prefilled syringe production, HMDSO is introduced as a monomer in a 13.56 MHz RF capacitively coupled plasma chamber. The substrate rack is held at 40–80 °C and the chamber is evacuated to a base pressure of 0.005 mbar before monomer flow. Process pressure is maintained at 0.10–0.30 mbar with HMDSO flow of 10–50 sccm, and oxygen is added at an O₂:HMDSO flow ratio of 0.5–1.5. The power density is constrained to 0.05–0.15 W/cm². Below 0.05 W/cm², deposition is incomplete and wetting remains non-uniform; above 0.15 W/cm², ion bombardment converts the coating to brittle SiO₂, and the coefficient of thermal expansion mismatch with COP causes microcracking after autoclaving. The coating thickness is controlled at 20–80 nm by deposition rate monitoring with a quartz crystal microbalance. Adhesion is evaluated after steam sterilization at 121 °C for 30 min using a cross-hatch tape test adapted from ISO 2409, with acceptance criterion ≤1 on the 0–5 scale. Extractables are quantified by headspace GC-MS following USP <1663>; coating-derived siloxane extractables typically fall below the analytical evaluation threshold when the oxygen-to-monomer ratio exceeds 0.8. Published data for specific COP resin grades are limited, because crystallinity and surface oxidized layer thickness vary between injection moulding campaigns.

Elastomeric Pharmaceutical Closure Passivation and Adsorption Loss Reduction

Bromobutyl and chlorobutyl closures for parenteral vials receive HMDSO plasma treatment at platen temperatures below 60 °C to avoid rubber crosslink degradation. The plasma chamber is operated at low power density, 0.03–0.08 W/cm², with argon as the carrier gas and HMDSO vapour delivered through a heated line at 100–110 °C. Coating thickness is limited to 30–80 nm. At thickness above 100 nm, coating cohesive failure is observed after three autoclave cycles at 121 °C for 30 min, confirmed by SEM cracking and increased particulate counts in water for injection per USP <788>. The functional requirement is suppression of nonspecific adsorption of hydrophobic active substances on halogenated rubber. Adsorption loss is assessed by spiking a 1 mg/mL model peptide solution into coated and uncoated closure leachate cells and quantifying recovery after 24 h at 40 °C by UPLC. In production campaigns, the difference in recovery between coated and uncoated closures is typically 2–5% at the 10 µg/mL concentration, where published comparison data permit quantification. Surface passivation is also monitored by static water contact angle, with a target of 105–115° on cured elastomer surfaces. The process window is narrow: plasma power below 0.03 W/cm² yields non-coherent deposition and exposed rubber domains, while power above 0.08 W/cm² accelerates volatile siloxane release into the container headspace. Each closure formulation requires revalidation because zinc oxide and stearic acid concentrations in the rubber compound influence crosslink density and coating adhesion.

In tablet and granule manufacturing, HMDSO is applied as a vapour-phase hydrophobizing agent for anhydrous granules containing moisture-sensitive active ingredients, including hydrates and lyophilized powders. The treatment is performed in a fluid-bed dryer fitted with a side-stream vapour generator, at product temperature 60–100 °C for 1–3 h. The HMDSO vapour is carried by nitrogen at 0.5–2.0 L/min per kg of granulate, with reactor humidity maintained below 30% RH to prevent competitive adsorption of water on surface silanol sites. The resulting surface contact angle rises from 20–35° to 100–120°, measured by sessile drop on compressed granule compacts. Dynamic vapour sorption at 25 °C and 60% RH shows a mass uptake reduction of 0.2–0.5% over 6 h when compared with untreated granules of equivalent particle size distribution. The processing boundary is set by the thermal stability of the granulating binder: hydroxypropyl cellulose and povidone-containing matrices must remain below their glass transition temperature, which is typically 110–130 °C for dry povidone. Residual HMDSO is monitored in the final blend by headspace GC-MS using a limit of ≤25 ppm unless the drug product specification defines a stricter value. Because HMDSO is not listed in ICH Q3C(R8) class 1/2/3 solvent tables, release must be justified through USP <467> or a validated in-house method. This application is confined to processing aid status and requires removal by subsequent drying or coating steps; direct addition of liquid HMDSO to tablet formulations as an excipient is not supported by current compendial references.

When Residual Water Activity in HPMC Capsule Shells Exceeds 60% RH

Hard hydroxypropyl methylcellulose capsule shells for hygroscopic oral powders are plasma-coated in a rotating drum reactor using HMDSO as the monomer. The drum is operated at 20–40 rpm, with process pressure 0.05–0.20 mbar, power density 0.02–0.06 W/cm², and monomer flow 5–15 sccm. The coating is deposited at 20–50 nm thickness to avoid shell brittleness. Treated capsules are conditioned at 22–25 °C and 50% RH for 24 h before filling trials. Moisture vapour transmission rate of the treated shell is measured by the desiccant method of USP <671> using an aluminium test fixture; comparative data in the same fixture show MVTR reduction of 18–35% relative to untreated HPMC shells, depending on shell wall thickness and residual water activity. The key process conflict is oxygen contamination: an oxygen-to-HMDSO ratio below 0.5 produces an organic silicone-like film with lower barrier performance, while a ratio above 1.5 raises surface energy and reduces powder flow on the capsule-filling machine. Coated shells must be checked for static charge build-up during high-speed filling at ≥80,000 capsules/h; if charging exceeds 1.0 kV/cm, the film may attract fine particles and impair shell closure. The working window for residual moisture is tight: residual moisture above 8% in the shell before plasma treatment yields blisters under the coating after drying, observed by optical microscopy at 100× magnification. Published data for HPMC capsule plasma coating are limited; each shell formulation and bed load must be qualified by destructive adhesion testing after coating.

On injectable formulation lines, HMDSO is used as a temporary anti-adhesion coating on 316L stainless steel intermediate bulk containers after electropolishing and passivation. The vessel is heated to 80–100 °C under vacuum, and HMDSO vapour is introduced at 0.1–0.3 bar for 30–90 min. The resulting monomolecular layer reduces product loss of low-concentration protein solutions by preventing adsorption to surface oxides. Cleanability is tested by swabbing a spiked riboflavin solution and measuring recovery at 470 nm, with acceptance criterion ≥90% compared with uncoated controls. The coating is removed by cleaning-in-place with 0.5 M sodium hydroxide at 60 °C for 20 min; reapplication is required after each cleaning cycle. This use is confined to product-contact equipment, and leachable evaluation is performed according to USP <1663> before process validation.