In mineral-filled thermoplastics and thermosets, the equilibrium moisture content of the filler phase often exceeds the matrix resin by factors of 10 to 50 under identical relative humidity conditions, creating a concentration gradient that drives interfacial water accumulation during humid ageing and reflow soldering. The phenomenon is most pronounced in fillers with high specific surface areas, such as fumed silica with 200 m²/g to 380 m²/g by ISO 9277, precipitated silica, aluminium trihydroxide with BET surface areas of 2 m²/g to 12 m²/g, and magnesium hydroxide. These materials carry surface silanol groups and metal hydroxyl groups that hydrogen-bond water molecules; adsorbed water is not only a source of voids, delamination, and hydrolytic polymer degradation, but also a contributor to dielectric loss and ionic conductivity in electrical insulation. Untreated glass fibre and mineral fillers are routinely dried at 80 °C to 120 °C for 2 h to 6 h before compounding or moulding, yet thermal drying alone does not eliminate silanol groups and does not prevent re-adsorption when the filler is re-exposed to ambient air. Vapour phase silylation addresses this limitation by reacting organosilanes or silazanes directly from the gas phase with surface hydroxyl groups, replacing the strongly hygroscopic silanol or metal hydroxyl population with covalently bonded organosiloxane species. Unlike aqueous silane treatment, which deposits oligomeric siloxane layers and leaves residual water in the filler, vapour phase silylation operates at controlled water partial pressures and can produce a hydrophobic monolayer without bulk solvent or a post-drying step. The resulting moisture uptake, measured according to ISO 62 or ASTM D570, is typically reduced by 70 % to 95 % for high-surface-area silicas, while the absolute effect is smaller for low-surface-area mineral fillers where the moisture content is already below 0.2 wt%. Industrial interest in vapour phase silylation is concentrated in electrical and electronic applications, where moisture-induced dielectric breakdown and component warpage are critical failure modes, as well as in low-smoke zero-halogen cable compounds, where hydrated mineral fillers dominate the formulation. The treatment is also relevant for glass fibre sizing replacement or pre-treatment when retained tensile strength after hydrothermal conditioning is the primary acceptance criterion.
At a silanol density of 4.6 SiOH/nm² on a fully hydroxylated fumed silica surface, the outermost layer behaves as a two-dimensional array of strong hydrogen-bond donors and acceptors capable of adsorbing 0.5 mg/m² to 1.5 mg/m² of water at 23 °C and 50 % relative humidity. The actual moisture uptake of a filler depends less on its bulk chemical composition than on the concentration, distribution, and accessibility of surface silanol groups. Isolated silanols at 3750 cm⁻¹ in diffuse reflectance infrared Fourier transform spectroscopy are less hygroscopic than vicinal silanols hydrogen-bonded at 3540 cm⁻¹ to 3300 cm⁻¹, and geminal silanols on amorphous silica contribute disproportionately to physisorbed water because they present two adjacent hydroxyl groups capable of forming a stable water cluster. Thermogravimetric analysis of untreated fumed silica shows a mass loss between 105 °C and 400 °C that is commonly assigned to physisorbed water and surface hydroxyl condensation, but the relative contributions must be separated by coupled mass spectrometry or by titration with lithium aluminium hydride. After vapour phase silylation, the residual silanol density can be reduced to below 0.5 SiOH/nm² when hexamethyldisilazane is used at sufficient dosage, and the water contact angle on a pressed filler plate typically increases from below 20° for untreated silica to above 120° for the trimethylsilyl-modified surface. The thermodynamic affinity of the modified surface is best evaluated by dynamic vapour sorption rather than by single-point moisture content, because the shape of the water sorption isotherm at relative humidity values of 20 % to 85 % reveals whether residual silanol islands remain accessible. Low moisture absorption in the finished composite therefore requires not only high silane conversion but also a uniform spatial distribution of organosilane on the filler surface, since locally untreated regions can act as condensation nuclei for interfacial water accumulation.
Because silylation proceeds through nucleophilic attack of a surface silanol oxygen on the silicon centre of hexamethyldisilazane, the rate is highly sensitive to the presence of physisorbed water and to the local concentration of acidic or basic catalysts. In the vapour phase reaction between hexamethyldisilazane and silica, two surface silanols are consumed per silazane molecule to form two trimethylsilyl groups and one molecule of ammonia, which is removed continuously under vacuum or with a dry nitrogen sweep. The reaction follows pseudo-first-order kinetics at low silane partial pressures, with an apparent activation energy typically reported in the range of 35 kJ/mol to 60 kJ/mol depending on the silica hydroxyl distribution and the presence of trimethylchlorosilane as a catalyst. For trialkoxysilanes such as octyltriethoxysilane or propyltrimethoxysilane, vapour phase deposition first requires hydrolysis of the alkoxy groups by surface-adsorbed water, producing ethanol or methanol as a byproduct. This leads to a critical process window: if the relative humidity in the reactor is below 10 %, the hydrolysis rate may be too low to achieve practical conversion within 2 h, while if the relative humidity exceeds 40 %, the silane can undergo homogeneous oligomerization in the gas phase or on the filler surface, forming a loosely attached polysiloxane layer that increases extractables and reduces thermal stability. Process development therefore requires active control of water partial pressure, typically by pre-drying the filler to a defined moisture content of 0.1 wt% to 0.5 wt% before introducing the silane, and by monitoring the reactor dew point with a chilled-mirror sensor. The optimum dosage for a given filler is usually determined by measuring residual silanol content with diffuse reflectance infrared Fourier transform spectroscopy or by titration, and by confirming monolayer coverage through the absence of nitrogen surface area loss in excess of the geometrical silane footprint.
When a vapour-phase silylated filler is incorporated into a polymer matrix on a twin-screw extruder with a barrel length-to-diameter ratio of 40:1 and a specific mechanical energy input above 0.20 kWh/kg, the hydrophobic monolayer is exposed to combined thermal and shear stress that can partially degrade the organosilane before the filler reaches the final moulded part. Short-chain trimethylsilyl groups on silica fillers are generally stable at polyolefin compounding temperatures of 190 °C to 230 °C, as shown by isothermal thermogravimetric analysis in nitrogen at 220 °C for 60 min, where mass loss from the treated filler remains below 0.1 wt%. However, the same treatment can begin to degrade at polyamide 66 processing temperatures of 280 °C to 300 °C, particularly in the presence of oxygen introduced through an unvented feed throat or a worn barrel seal. Thermal oxidation of methyl-terminated silane monolayers proceeds through radical abstraction at the terminal methyl group, followed by the formation of silanol, aldehyde, and carboxylic acid species that partially restore hydrophilicity. The practical consequence is that vapour-phase silylation of fillers for high-temperature engineering polymers requires alkylsilanes with longer chains or aromatic substitution, such as octyltriethoxysilane or phenyltrimethoxysilane, because the additional organic mass and lower volatility reduce the rate of oxidative degradation. In a twin-screw compounding trial, the residence time distribution, screw configuration, and local melt temperature at the mixing elements determine whether the silylated filler surface survives intact; severe kneading blocks operating at 1200 rpm can generate local temperatures 20 °C to 40 °C above the set barrel temperature. Published data for the specific combination of vapour-phase octylsilane-treated glass fibres in high-temperature polyphthalamide is limited, but the general oxidative stability of alkylsilanes on silica is well documented by thermogravimetric analysis coupled with mass spectrometry under air. Consequently, the use of vapour-phase silylation in compounds destined for injection moulding at melt temperatures above 290 °C should be accompanied by post-compounding moisture absorption testing and by surface energy measurements on extracted fillers to verify that the hydrophobic layer has not been lost during processing.
In semiconductor packaging, epoxy moulding compounds are silica-filled to 70 wt% to 90 wt% with fused silica particles having a particle size distribution that typically spans 0.5 µm to 75 µm. The equilibrium moisture content of a moulding compound is governed primarily by the filler loading, the resin polarity, and the presence of silane coupling agents, but the interface between the epoxy matrix and the silica particle is the dominant pathway for moisture diffusion and the site where condensate accumulates during high-temperature reflow. A fused silica filler that has been vapour-phase silylated with an alkylsilane such as propyltrimethoxysilane exhibits reduced water sorption when measured by ASTM D570 at 85 °C and 85 % relative humidity for 168 h; typical moulding compounds based on untreated silica may absorb 0.6 wt% to 0.9 wt% moisture, whereas silylated silica formulations often absorb 0.2 wt% to 0.4 wt% under the same conditions. The improvement in moisture absorption translates directly into a reduced probability of popcorn cracking during lead-free reflow at 260 °C because the saturated water vapour pressure at the silica-polymer interface exceeds the fracture strength of the compound when moisture content is high. Qualification testing according to JEDEC J-STD-020 at moisture sensitivity level 3 or level 2a requires specific floor life, reflow simulation, and electrical verification after package cracking; vapour-phase silylated fillers have been used in packages targeting these classifications. The effect on dielectric properties is equally important for automotive and industrial electronics: a reduction in interfacial water reduces the dielectric constant and dissipation factor measured according to IEC 62631-3-1, and increases the volume resistivity determined by ASTM D257 under humid conditions. Because the filler surface after vapour-phase silylation is hydrophobic, the epoxy matrix may still absorb moisture through its polar functional groups, but the elimination of the hydrophilic silica interface removes the fastest diffusion channel and the most probable location for delamination. The exact moisture absorption values vary with filler particle size distribution, epoxy formulation, and post-mould curing schedule, and published data for specific vapour-phase silylated fused silica in production moulding compounds is limited because most commercial data remain proprietary.
When continuous glass fibre roving is treated by an aqueous γ-aminopropyltriethoxysilane size, the resulting siloxane interphase is typically 20 nm to 100 nm thick and contains adsorbed water, oligomeric silanols, and residual surfactants. The aqueous size is essential for fibre protection, strand integrity, and matrix compatibility, but it also introduces a hydrophilic interphase that can absorb moisture and reduce the retention of mechanical properties after hydrothermal conditioning. Vapour phase silylation of glass fibres, by contrast, deposits a covalently bonded monolayer of organosilane directly on the glass surface, leaving a thickness in the range of 1 nm to 3 nm and eliminating the solvent-laden oligomeric network. The process can be implemented as a pre-treatment before the application of a film former or as a post-treatment on heat-cleaned glass fabric, but continuous roving production at line speeds above 600 m/min presents a significant mass-transfer challenge because the silane vapour must contact the moving glass surface within a residence time of less than 1 s. Laboratory-scale vapour phase treatment of glass fibres at 120 °C for 30 min in a sealed chamber has demonstrated reduced water adsorption and improved interfacial shear strength in epoxy matrices, but scaling this to a continuous forming line requires a closed deposition chamber with carefully controlled silane concentration, temperature, and exhaust pressure. The absence of a lubricating film former after vapour silylation can lead to fibre breakage during subsequent weaving or prepregging operations, which is why the vapour phase step is usually combined with a reduced-thickness size or a non-aqueous film former. Ignition loss testing according to ASTM D4963/D4963M is used to verify the organic content of the treated glass fabric, and tensile strength retention of rovings after water immersion is assessed according to ASTM D2343. The comparative moisture absorption of glass fibre laminates made from vapour-phase silylated fabric versus aqueous-sized fabric shows improved dry and wet flexural strength retention, but the overall laminate moisture uptake remains dominated by the resin matrix, so the benefit is most visible in interfacial failure modes rather than in total weight gain.
Production-scale vapour silylation units for particulate fillers are usually configured as batch rotary cone dryers or continuous fluidised bed reactors, with the selection governed by filler bulk density, particle size, and the required silane dosage. A rotary cone vacuum dryer used for filler silylation typically consists of a heated conical vessel with a rotating screw that lifts and circulates the powder, a jacket capable of maintaining 80 °C to 140 °C, a vacuum system reaching 1 kPa to 10 kPa, and a vaporiser that introduces silane at a controlled rate of 0.5 g/min to 2.0 g/min per kilogram of filler. The powder is first dried under vacuum to remove physisorbed water, then the silane is admitted while the vessel rotates at 10 rpm to 30 rpm; the reaction is allowed to proceed for 60 min to 180 min before residual ammonia, methanol, or ethanol is stripped under vacuum at the end of the cycle. Fluidised bed reactors are preferred for high-surface-area fillers because the gas flow provides continuous contact and avoids mechanical shear that could fracture brittle particles. The superficial gas velocity is maintained at 1.5 to 3 times the minimum fluidization velocity, and the silane vapour is introduced through a heated distributor plate to prevent condensation. Process control in both reactor types relies on online dew point measurement at the gas outlet, infrared monitoring of silane concentration, and temperature measurement at multiple points in the bed or vessel. Safety considerations are significant because silanes are flammable, chlorosilanes release hydrogen chloride upon contact with moisture, and the byproducts ammonia, methanol, and ethanol require scrubbing or thermal oxidation. Equipment for vapour phase silylation of fillers intended for electrical applications is generally designed to comply with ATEX Directive 2014/34/EU for explosive atmospheres, and the reactor exhaust must meet local volatile organic compound limits. The main process conflict in rotary cone vacuum drying is balancing the need for high vacuum to remove water against the need to retain enough residual moisture for alkoxysilane hydrolysis; operators often target a reactor dew point between -10 °C and 0 °C to obtain reproducible monolayer coverage without excessive silane oligomerization.
Low-smoke zero-halogen cable compounds containing 150 phr to 220 phr of aluminium trihydrate or magnesium dihydroxide are particularly sensitive to filler moisture because water vapour generated at the screw barrel can produce surface roughness, internal porosity, and screw slippage. Aluminium trihydrate releases its three moles of crystalline water at temperatures above 180 °C, and this endothermic decomposition is the source of its flame-retardant performance, but the same filler also adsorbs free moisture on its particle surface at ambient conditions. Magnesium dihydroxide has a higher decomposition temperature near 340 °C, making it suitable for polypropylene-based compounds but also more difficult to predry because the crystallisation water is retained during normal drying at 80 °C. Vapour phase silylation of these metal hydroxide fillers with alkylsilanes reduces the adsorbed surface water without altering the bulk crystal structure or the endothermic decomposition enthalpy measured by differential scanning calorimetry at 180 °C to 220 °C for aluminium trihydrate and 320 °C to 360 °C for magnesium dihydroxide. The hydrophobic surface layer reduces moisture pickup during storage and improves flowability, but the presence of the organic layer can also change the ignition behaviour slightly, and cone calorimetry according to ASTM E1354 should be used to confirm that the peak heat release rate and total heat release remain within the formulation limits. For wire and cable applications, flame retardance is also evaluated by vertical burn tests under UL 94 and by oxygen index measurement according to ISO 4589-2; the addition of 0.5 wt% to 1.0 wt% of silylation agent on the filler does not usually shift these values by more than experimental error. The primary benefit in cable extrusion is the reduction of surface defects and the elimination of pre-drying at high relative humidity, which is especially valuable in high-humidity manufacturing locations where untreated filler can absorb more than 1 wt% moisture within 12 h of exposure. A predrying step at 60 °C for 4 h is still recommended before extrusion when the ambient relative humidity exceeds 60 %, even for vapour-phase silylated fillers, because storage and handling equipment may introduce moisture into the compound from other raw materials.
The treatment of nanoscale fumed silica in a vapour phase reactor imposes narrower operating limits than the treatment of ground mineral fillers because the bulk density of as-produced fumed silica is often below 50 g/L and the surface silanol density is high. Fluidised bed treatment of fumed silica requires careful control of gas velocity because the ultrafine particles form agglomerates that can channel or defluidise at velocities only slightly above the minimum fluidization velocity. In practice, continuous vapour silylation of fumed silica is often performed in a mechanically agitated contacting vessel rather than a conventional fluidised bed, since the mechanical energy helps break up loose agglomerates and exposes fresh surface to the silane vapour. The dosage of hexamethyldisilazane required to achieve a residual silanol density below 0.5 SiOH/nm² is typically 0.15 g to 0.30 g per gram of silica, depending on the initial hydroxyl density and the degree of pre-drying. If the dosage is too low, residual silanol islands remain and the moisture uptake measured by dynamic vapour sorption at 50 % relative humidity remains above 0.3 wt%. If the dosage is too high or the reactor temperature exceeds 130 °C, the excess silane can condense as a liquid film inside the agglomerates, causing irreversible aggregation and a loss of the high-surface-area structure that is essential for reinforcement. The specific surface area after silylation is checked by ISO 9277, and the carbon content is determined by elemental analysis to verify that the silane loading is in the expected range for a monolayer. Published data for vapour-phase silylation of nanoscale fumed silica in production-scale reactors is limited, but laboratory studies consistently show that the optimal reaction temperature is between 80 °C and 120 °C and that the water partial pressure must be held below the saturation value corresponding to a reactor dew point of 0 °C to avoid uncontrolled hydrolysis of silazane to silanol and ammonia. The processed silica is then subjected to high-shear dispersion in the target polymer, where the degree of agglomerate breakdown is strongly influenced by the hydrophobicity of the treated surface and the compatibility of the organic silane with the matrix.
| Filler grade | Silane | BET surface area | Residual silanol density | Moisture uptake at 50% RH | Test method |
|---|---|---|---|---|---|
| Fumed silica, untreated | None | 300 m²/g | 4.2 SiOH/nm² | 1.2 wt% | ISO 9277, dynamic vapour sorption |
| Fumed silica, vapour HMDS | Hexamethyldisilazane | 280 m²/g | 0.4 SiOH/nm² | 0.06 wt% | ISO 9277, dynamic vapour sorption |
| Fumed silica, vapour octylsilane | Octyltriethoxysilane | 270 m²/g | 0.8 SiOH/nm² | 0.12 wt% | ISO 9277, dynamic vapour sorption |
Polyamide 66 glass-filled connector housings conditioned at 23 °C and 50 % relative humidity reach an equilibrium moisture content of approximately 1.8 wt% when using standard glass fibre, but the interfacial region contributes disproportionately to tensile strength loss because water attacks the glass-polymer interphase and causes microcracking. A connector housing moulded from polyamide 66 containing 30 wt% vapour-phase silylated glass fibre exhibits improved retention of tensile strength after conditioning at 85 °C and 85 % relative humidity for 500 h when compared with the same compound containing aqueous-sized fibre, even though the total water absorption of the two compounds differs by less than 0.1 wt% under the same conditions. The explanation lies in the localisation of moisture at the interface: aqueous-sized glass fibre contains hygroscopic sizing components that swell and plasticise the interphase, while vapour-phase silylation removes the water-swellable layer and replaces it with a more stable hydrophobic monolayer. Injection moulding of these compounds is performed at barrel temperatures of 260 °C to 290 °C and a clamp force that depends on the projected area of the connector housing; for a typical multi-cavity mould with 32 cavities and a shot size of 120 g, a clamp force of 1200 kN to 1600 kN is often specified. The silylated glass fibre must be predried at 80 °C for 4 h before moulding if the ambient relative humidity exceeds 60 %, because the polyamide resin itself is highly hygroscopic and will absorb water during open storage. Tensile properties are measured according to ISO 527-2 at a test speed of 5 mm/min for conditioned dry specimens, and notched Charpy impact strength is determined according to ISO 179-1/1eA. Published property comparisons for vapour-phase silylated glass fibre in polyamide 66 are limited, and the available data indicate that the main benefit is not a large reduction in total moisture uptake but a reduction in moisture-induced interfacial failure.
