Liquid polysulfide prepolymers used in aerospace sealants are predominantly thiol-terminated poly(oxyalkylene disulfide) oligomers synthesized by aqueous suspension condensation of bis(2-chloroethyl) formal and sodium polysulfide, with 1,2,3-trichloropropane incorporated as a branching monomer at levels sufficient to raise the number-average thiol functionality from a difunctional baseline of approximately 2.0 to values between 2.1 and 2.5. The finished prepolymer is a moderately viscous liquid at 25 °C, with dynamic viscosity typically between 25 Pa·s and 80 Pa·s when measured by rotational viscometry in accordance with ISO 3219:2021 or ASTM D2196-20, and with a density near 1.27 g/cm³ to 1.35 g/cm³ under ASTM D792-20. Crosslink density in the cured sealant is an emergent network parameter that depends on thiol equivalent weight, branching monomer mole fraction, the stoichiometric ratio of oxidizing curing agent to thiol, cure temperature, and moisture availability during vulcanization. The number-average molecular weight between crosslinks, M_c, can be estimated from equilibrium solvent swelling by the Flory–Rehner equation using toluene or methyl ethyl ketone immersion data obtained under ISO 1817:2022, with the solvent-polymer interaction parameter χ determined from separate swelling experiments or taken from published compilations. The same resin variables also control the gel point, which for a condensation-type thiol oxidation network can be approximated by the Carothers relation p_c = 1/(f_av − 1), where f_av is the average thiol functionality; increasing the branching monomer content from 0.5 mol% to 2.0 mol% shifts the gel point to lower thiol conversion and increases the concentration of elastically active network chains at full cure.

Regulation of crosslink density therefore begins with prepolymer selection. A linear polysulfide of M_n approximately 4000 g/mol with negligible branching will cure to a loose network with high elongation and pronounced solvent uptake, whereas a branched prepolymer of similar M_n with 1.5 mol% to 2.0 mol% trifunctional branching yields a tighter network with lower equilibrium swell and higher modulus. In practice, aerospace sealant formulators use a combination of functionality, curing-agent index, and curative type rather than relying on a single variable. The thiol equivalent weight is obtained by iodometric titration, often following the method described in the manufacturer’s certificate of analysis and traceable to the stoichiometric equation 2 R–SH + MnO₂ → R–S–S–R + MnO + H₂O. Because the reaction is an oxidative coupling, each thiol group is a reactive site and each polysulfide chain carries at least two terminal thiols; trifunctional branch sites contribute additional terminal thiols when the branching monomer is incorporated during the initial polymerization. The presence of residual sodium polysulfide, cyclic disulfides, and terminal alkyl halides in the prepolymer must be controlled because these species alter the effective thiol functionality and can introduce dangling ends or non-reactive diluent fractions that reduce the measured crosslink density.

How Does the Stoichiometric Balance Between Thiols and Manganese Dioxide Determine Network Topology?

Manganese dioxide is the predominant ambient-temperature oxidizing curative for two-part polysulfide sealants intended for fuel tank and general aircraft sealing. The stoichiometric index is expressed as the mass of MnO₂ per 100 g of prepolymer or as a percentage of the theoretical thiol oxidation demand. When the index is below approximately 90%, the conversion of thiols is incomplete and the cured network contains unreacted mercaptan groups, pendant chain ends, and a measurable reduction in covalent crosslink density. The consequences are not limited to lower hardness; equilibrium solvent uptake in jet fuel increases, adhesion to chromic acid-anodized aluminium under ASTM D1002-10(2019) decreases, and compression set after 22 h at 70 °C rises above the values acceptable for Class B fuel-tank sealant. When the index exceeds approximately 105%, unreacted MnO₂ remains dispersed as particulate filler, and the network becomes more rigid through hydrodynamic reinforcement rather than through additional covalent crosslinks. This condition can increase tensile modulus but also produces a brittle cured film with reduced elongation at break and lower peel adhesion because the excess metal oxide can promote oxidative scission during long-term thermal aging. The optimized index for typical aerospace polysulfide systems lies between 95% and 102%, but the exact range shifts with filler loading, prepolymer functionality, and accelerator content.

Dichromate-cured polysulfides are still encountered in legacy airframe applications, but their use is restricted under REACH and other chemical management regulations because of hexavalent chromium content. Sodium dichromate or ammonium dichromate curatives produce a rapid, highly crosslinked network at ambient temperature; however, the resulting sealant is prone to hardening on thermal aging and presents occupational exposure limits that are incompatible with modern aerospace production. Manganese dioxide systems, often accelerated with diphenylguanidine or other basic accelerators, provide a more controlled vulcanization kinetics profile. Addition of an accelerator before high-shear dispersion increases the risk of premature crosslinking because the thiol oxidation reaction is base-catalyzed; therefore, production mixing sequences typically add the accelerator as a separate catalyst masterbatch after the filler and curative have been dispersed. The degree of crosslinking can also be modified by substituting part of the MnO₂ with calcium peroxide or cumene hydroperoxide, but published data for these binary curative systems in aerospace sealants is limited and each substitution must be requalified against the relevant specification such as AMS 3265 or MIL-PRF-81733.

Charging a production-scale 40 L double-planetary mixer with a mixed filler package of precipitated calcium carbonate, hydrophobic fumed silica, and titanium dioxide requires staged addition to avoid agglomeration and to maintain a temperature below 35 °C. The mixer is typically operated at blade speeds from 20 rpm to 40 rpm with vacuum deaeration at 10 mbar to 20 mbar absolute. The base component is produced by dispersing the filler mixture into liquid polysulfide prepolymer under high-shear conditions using a dissolver blade at approximately 1000 rpm to 1500 rpm before vacuum mixing, because direct addition of all fillers at once produces pseudoplastic pastes with poor wet-out and entrapped air. Filler moisture must be controlled to below 0.1% by weight; otherwise the water competes with thiol groups at the MnO₂ surface, retards disulfide bond formation, and produces a tacky, undercured interlayer in thick-section applications. Batch-to-batch variation in filler surface area is a recognized production bottleneck. A change in precipitated calcium carbonate specific surface area from 6 m²/g to 12 m²/g can increase the base viscosity beyond the application limit for pneumatic Semco cartridges and alter the post-cure hardness even when the curing-agent index remains unchanged. In such cases, the formulation must be adjusted by reducing the hydrophobic fumed silica loading or by increasing the dilution of the accelerator masterbatch; this change is validated by measuring the mixed viscosity under ISO 3219:2021 and the pot life under the applicable AMS 3265 or MIL-PRF-81733 test method.

Solvent Swell, Tensile Modulus, and the Flory–Rehner Network Chain Density

Crosslink density in polysulfide sealants is most directly measured by equilibrium solvent swelling rather than by rheological methods, because the filled network contains carbonates, silicas, and metal oxides that interfere with dynamic mechanical analysis. A cured specimen of known thickness is immersed in methyl ethyl ketone or toluene at 23 °C to 25 °C until constant mass, and the volume fraction of rubber in the swollen gel, V_r, is calculated from the mass difference using the polymer and solvent densities determined under ASTM D792-20 and ASTM D891-20. The Flory–Rehner equation gives the network chain density ν_e as ν_e = −[ln(1 − V_r) + V_r + χV_r²] / [V_s(V_r^1/3 − V_r/2)], where V_s is the molar volume of the swelling solvent and χ is the Flory–Huggins interaction parameter. The same network parameter can be estimated from uniaxial tensile data using the neo-Hookean relationship E = 3ν_eRT, where E is the tensile modulus at low strain, R is the gas constant, and T is the absolute temperature. For a typical aerospace polysulfide sealant with a Shore A hardness between 30 and 50, the measured ν_e is approximately 1 × 10⁻⁴ mol/cm³ to 3 × 10⁻⁴ mol/cm³, corresponding to M_c values from roughly 3000 g/mol to 10000 g/mol depending on filler content and curative index.

The tabulated values are representative of trends reported in supplier technical data and qualification test reports; they are not a substitute for batch-specific certification. Tensile strength and elongation are determined simultaneously on die-cut specimens according to ASTM D412-16 or ISO 37:2017, with Type C or Type 2 dumbbells punched from a cured sheet. The changes in tensile modulus at 100% elongation distinguish formulations whose network chains are short and highly constrained from those with longer, more extensible disulfide segments. Tear strength determined by ASTM D624-00(2020) Die C or ISO 34-1:2022 Method B is also sensitive to crosslink density but follows a non-monotonic trend; tear resistance often passes through a maximum at intermediate M_c because excessive crosslinking suppresses energy dissipation through chain orientation and produces a sharp crack tip stress concentration. Hardness measured by ASTM D2240-15(2021) with a Shore A durometer is a rapid field indicator but it cannot separate covalent crosslinks from hydrodynamic filler reinforcement; two sealants with the same Shore A value can have different solvent swell and peel adhesion if one contains a high loading of carbon black or calcium carbonate.

Thermal Aging and Compression Set Expose Incomplete Crosslink Development at Low Cure Temperatures

The vulcanization kinetics of polysulfide oxidation are strongly temperature dependent, and crosslink density cannot be considered stable until the sealant has reached full cure after the specified conditioning period. Aerospace production specifications typically condition test coupons at 23 °C ± 2 °C and 50% ± 5% relative humidity for 72 h to 168 h before mechanical testing. If the ambient temperature during cure falls below approximately 10 °C, the rate of thiol conversion decreases sharply and the apparent crosslink density after 24 h can be 40% to 60% lower than the value achieved at 23 °C, as inferred from solvent swelling and modulus measurements. The same phenomenon is observed in thick sections where exothermic reaction heat is dissipated rapidly and moisture cannot reach the interior; a 6 mm or thicker bead may require a post-cure at 40 °C to 60 °C for 24 h to complete crosslinking. Thermal aging at 70 °C or 100 °C after full cure can cause additional disulfide metathesis and oxidative chain scission, depending on the curative system and antioxidant package; the net effect on crosslink density may be an increase or decrease with time, so compression set testing according to ASTM D395-16e1 Method B or ISO 815-1:2019 is used to detect networks that are insufficiently crosslinked.

Compression set in polysulfide sealants is generally higher than in many other elastomers because the disulfide bond is exchangeable under load at elevated temperature. A low crosslink density formulation with M_c above approximately 8000 g/mol may exhibit compression set values above 80% after 22 h at 70 °C, while a tighter network with M_c near 3000 g/mol can remain below 50%. The AMS 3265 and MIL-PRF-81733 qualification protocols typically impose maximum compression set and maximum fuel swell requirements; meeting those requirements simultaneously requires balancing crosslinker content and curative index. Incomplete crosslinking at low cure temperature cannot be corrected by adding excess accelerator because the residual accelerator can migrate to the sealant-metal interface and contribute to corrosion; the proper correction is to maintain the required curing environment and to extend the cure time.

When Polysulfide Sealant Is Applied in Fuel Tank Faying Surfaces at Thicknesses Below 1.5 mm

In thin bond line applications, the crosslink density of the confined sealant is affected by moisture ingress from the edges, by the catalytic effect of metal ions, and by the oxygen available at the free surface. A faying-surface application below 1.5 mm thickness cures more rapidly at the edges than in the centre, producing a gradient in conversion and network density that is measurable by microindentation or solvent swelling of microtomed sections. Adhesion is measured by lap shear according to ASTM D1002-10(2019) or by peel according to ASTM D413-98(2013); both tests are specified in aerospace sealant qualification programs to ensure that the primed aluminium surface retains adhesion under fuel immersion and thermal cycling. The use of chromic acid anodized aluminium panels is standard because the anodic oxide morphology provides mechanical interlocking and chemical sites for polysulfide anchorage. The crosslink density at the interface must remain high enough to prevent adhesive failure under repeated fuel pressure cycles, but it must not become so high that the sealant can no longer accommodate differential thermal expansion between aluminium skins.

Production-scale application equipment includes pneumatic guns operating at pressures from 0.2 MPa to 0.6 MPa with static mixing tips of 24 to 32 elements; the pressure drop across the mixer tip is a critical process parameter because high pressure can separate the filler from the prepolymer and alter the local curing-agent ratio. Premature gelation inside the static mixer occurs when the ambient temperature exceeds 30 °C or when the pot life is reduced by amine-containing surface primers; therefore, mixing tips are sized to match the formulation pot life and application rate. Amine-based additives are generally incompatible with polysulfide sealants because they accelerate the thiol oxidation reaction and can produce localized crosslink density at the interface that leads to brittle failure. Moisture pre-drying of the substrate is required at relative humidity above 60% because water films interfere with adhesion and can cause microporosity in the cured sealant.