Cationic Photoinitiator CPI-820 is supplied as a mixed triarylsulfonium hexafluorophosphate salt dissolved in propylene carbonate. The product functions as a photoacid generator for cationic ring-opening polymerization of epoxide, oxetane, and vinyl ether resins under ultraviolet exposure in the 320 nm to 390 nm band. Representative liquid specifications include active content of 50 wt%, Brookfield viscosity at 25 °C of 250–450 mPa·s, and density of 1.22–1.28 g/cm³ at 20 °C. Residual water is maintained below 0.15 wt% to limit acid scavenging before cure. The product differs from free-radical Type I and Type II photoinitiators in propagation mechanism: photogenerated Brønsted acid initiates polymerization that continues after irradiation ceases, whereas radical chain ends terminate rapidly through bimolecular recombination and oxygen quenching. The lack of oxygen inhibition allows cure of thin exposed surfaces without nitrogen blanketing, but the same acid chemistry introduces sensitivity to basic fillers, amines, and ambient moisture that is not observed with conventional radical systems.

What Limits Through-Cure in Titanium Dioxide-Filled Cationic Coatings?

When titanium dioxide is present at 15 wt% based on total resin solids, the through-cure window narrows because titanium dioxide scatters incident UV below 400 nm and competes for photons with the triarylsulfonium chromophore. At 50 µm dry film thickness, a 120 W/cm medium-pressure mercury lamp delivering 800 mJ/cm² UVA dose produces surface cure but may leave the lower 10 µm under-converted if the pigment dispersion is not controlled. A three-roll mill with a 5 µm back gap is required to prevent pigment agglomerates larger than 5 µm from creating shadow zones. Through-cure monitored by FTIR-ATR at the oxirane band near 910 cm⁻¹ should exceed 85% conversion before the part is released from the line. The threshold aligns with supplier technical bulletins for similar triarylsulfonium salts. On open-faced coating lines, addition of 2 wt% CPI-820 to a bisphenol-A epoxy resin of epoxy equivalent weight 185–192 g/eq results in tack-free surfaces after 800 mJ/cm² UVA dose. Through-cure at 50 µm film thickness is obtained under the same dose because the propagating cationic species is not quenched by atmospheric oxygen, unlike acrylate radical systems where surface conversion at 50 ppm oxygen can remain below 60% without nitrogen inerting. In roll-coater applications, the product is pre-dissolved in a low-viscosity oxetane diluent before addition to a high-viscosity epoxy backbone to prevent localized acid-rich domains. A Brookfield RVT spindle 6 at 20 rpm is used for incoming viscosity checks; acceptance ranges from 3,800 mPa·s to 12,000 mPa·s at 25 °C depending on application method.

Residual Acid and Substrate Corrosion Boundaries

Residual photogenerated acid remains active after irradiation and can migrate to uncoated metal interfaces. On bare aluminum 1050 and cold-rolled mild steel, formulations containing CPI-820 produce measurable surface oxidation under 40 °C and 90% RH exposure for 72 h. Adhesion loss on untreated metal occurs when post-cure free acid values exceed 5 mg KOH/g. For such substrates, a post-bake of 80 °C for 30 min is required to complete conversion and reduce free acid migration. The product is not recommended for direct contact with zinc-rich primers because zinc oxide neutralizes the photogenerated acid and suppresses cure depth. Tertiary amine additives and amine-functional silanes are incompatible; even 0.1 wt% triethanolamine based on resin solids can consume a measurable portion of the photogenerated acid and produce undercured interlayers. Dark cure behavior after initial UV exposure distinguishes CPI-820 from free-radical systems. In a 30 µm epoxy clear coating, epoxy ring conversion measured by FTIR-ATR increases from 70% immediately after 600 mJ/cm² to 88% after 24 h at 25 °C. The process is temperature-dependent; storage of the same panel at 40 °C accelerates post-cure and reaches 93% conversion within 6 h. Production lines exploit this by moving coated or laminated parts to a heated staging area instead of relying on continuous UV exposure. The magnitude of dark cure is formulation-specific and should be validated by in-process FTIR or photo-DSC when batch-to-batch resin hydroxyl value changes by more than 5%.

When UV-LED Sources Replace Medium-Pressure Mercury Lamps

Absorption maxima of CPI-820 are located near 245 nm, 280 nm, 325 nm, and 360 nm. Medium-pressure mercury lamps deliver the broad emission required for through-cure, whereas a 365 nm UV-LED array initiates surface cure but may leave depth conversion incomplete in pigmented formulations because the 365 nm band has lower molar absorptivity and is strongly scattered by titanium dioxide. Formulators using LED-only systems often add a photosensitizer such as 2-isopropylthioxanthone at 0.2 wt% to 0.5 wt%, shifting useful sensitivity to 385 nm and 395 nm. The working distance and irradiance homogeneity should be verified with a calibrated radiometer to maintain a minimum peak irradiance of 1.5 W/cm² across the part. Published data for this specific configuration is limited for black pigmented cationic adhesives; qualification is performed using the actual LED array and conveyor speed rather than relying on mercury-lamp dose equivalencies. Compared with diaryliodonium hexafluorophosphate salts, CPI-820 exhibits greater thermal stability at 35 °C storage and lower decomposition at 120 °C lamination preheat. However, its absorption in the 320–360 nm region can cause surface cure at the expense of through-cure in highly filled systems, whereas iodonium salts with thioxanthone photosensitizers can be tuned below 300 nm for deep lamination. Unlike free-radical systems, cationic cure continues in shadow zones after exposure, which permits bonding of opaque substrates if the initial UV dose reaches the interface. The product also differs from free-radical Type II systems in that no amine co-initiator is required; this avoids the yellowing and migration associated with aromatic amine synergists.

Parameter	Test condition / equipment	Cationic CPI-820	Free-radical Type I
Surface cure in air	800 mJ/cm², 120 W/cm medium-pressure mercury lamp, 50 µm clear film	Tack-free	Surface tack unless nitrogen inerting used
Post-irradiation conversion change	FTIR-ATR, oxirane 910 cm⁻¹, 1 h at 25 °C	+8% to +15%	+1% to +3%
Linear shrinkage	Gas pycnometer, 25 µm film	1.2–2.5%	7–10%
Dark cure after 600 mJ/cm²	24 h, 25 °C, epoxy conversion	70% to 88%	No significant increase
Moisture sensitivity	60% RH, open application	High; requires pre-drying	Moderate

Shelf Stability Requires Strict Moisture Exclusion

Storage of opened containers at 60% RH leads to moisture uptake and reduces acid-generating efficiency. Headspace nitrogen blanketing or desiccant-equipped bungs are recommended after partial discharge. At 25 °C in sealed original drums, the product retains at least 90% of initial photoacid activity after 12 months; opened containers should be consumed within 4 weeks unless moisture analysis by Karl Fischer coulometry confirms water content below 0.20 wt%. Pre-drying of resin, pigment, and substrate is required when ambient relative humidity exceeds 60%, because water competes with the epoxide ring opening and contributes to tacky surfaces. The cured film water absorption under ISO 62:2008 is typically 1.2–2.4 wt% for clear epoxy systems at 24 h immersion, depending on oxetane diluent content and post-bake temperature. In cationic UV-curable inks used on food-contact board, migration of residual photoinitiator and propylene carbonate must be controlled. Extraction testing according to EN 1186-1:2002 migration framework or analogous single-sided contact protocols is required before commercial use. The product is not inherently food-contact approved; any use under FDA 21 CFR 175.300 requires specific migration data for the cured ink and packaging construction. The polymer matrix must be fully cured or a barrier layer is required to limit residual species below applicable specific migration limits. Basic calcium carbonate coatings in recycled board neutralize photogenerated acid and reduce cure; the board surface pH should be below 8.0 before application.

Balancing Dark Cure and Dimensional Growth in Thick Composite Lamination

Thick glass-reinforced epoxy composites formulated with CPI-820 require staged irradiation to overcome attenuation through glass mat and core layers. A first pass of 1,000 mJ/cm² UVA is applied per side, followed by 60 min at 70 °C to drive dark cure through the 2 mm laminate. The post-bake is necessary because light does not penetrate the inner glass rovings; cationic active centers must migrate by chain transfer and diffusion. Without post-bake, the laminate core retains uncured resin and exhibits tensile shear strength below 50% of fully cured values when tested under ISO 14130:1997 short-beam shear. This limitation does not apply to radical systems because radical cure also cannot proceed in shadow, but radical systems typically use peroxide post-cure under nitrogen. In stereolithography using cationically curable epoxy-oxetane resins, CPI-820 is used at 1.5–3.0 wt% with 385 nm laser or LED imaging. The depth of cure is controlled by the absorbance of the photoacid generator and pigmented dyes. A line depth of 100 µm per pass at 50 mW/cm² requires compensation for the dark growth that occurs between layers; this growth can reach 20 µm over 30 min unless a weak base inhibitor is added. The use of 0.05 wt% triphenylphosphine as a stabilizer suppresses post-exposure growth but reduces photospeed by 15–25%. Such stabilizer additions must be qualified because they lower the number of propagating oxonium ions and can shift the mechanical properties of the green part before thermal post-cure. Production qualification of CPI-820-cured coatings uses a staged test matrix before full release. Pencil hardness is measured by ISO 15184:2020 on glass panels at 24 h post-irradiation. Cross-cut adhesion to treated aluminum is tested by ISO 2409:2020 after 72 h at 23 °C/50% RH. Solvent resistance is checked with methyl ethyl ketone double rubs under ASTM D5402-19, with a minimum of 100 double rubs for a 20 µm clear film. If the coating is exposed to sunlight, accelerated weathering is conducted under ISO 4892-2:2013 cycle 1, with color change measured by ASTM D2244-21 and adhesion retained above 4B. These are standard industrial release criteria, not product-specific claims, and each formulation must be tested independently because pigment type and substrate surface energy alter cure response. In black adhesives containing carbon black at 3 wt%, shadow cure after 1 s of 365 nm LED exposure reaches bondable handling strength only if the LED array delivers at least 1.0 J/cm² total UVA dose and the adhesive spread is controlled to 100 µm or less. The dark-cure component then raises the lap shear strength on degreased aluminum 2024 to approximately 80% of final value within 24 h at 25 °C. This behavior reduces the need for continuous conveyor UV exposure on complex three-dimensional parts. In contrast, a radical acrylic adhesive with comparable filler loading requires line-of-sight irradiation to overcome oxygen inhibition and cannot develop strength in shadow zones. The operational boundary is strict: initial UV must reach at least the outer 20 µm of the joint, or dark cure cannot propagate into the bulk interface.