Technical Reference Guide

Structural Silicone &
IGU Glazing Science

From molecular bonding chemistry to field inspection — a complete guide for architects, facade engineers, and glazing contractors.

Section 01

How Aluminium Bonds to Glass

Both glass and aluminium have very different surface chemistries, yet silicone manages to create a durable, weather-resistant bridge between them. Understanding why requires looking at what each surface actually is at a molecular level.

ALUMINIUM SUBSTRATE (Al₂O₃ oxide layer ≈ 4–10 nm) –OH–OH–OH –OH–OH–OH –OH STRUCTURAL SILICONE (–Si–O–Si– backbone + organofunctional groups) –Si–OH–Si–OH–Si–OH –Si–OH–Si–OH–Si–OH GLASS (SiO₂ surface with –Si–OH silanol groups) Covalent / H-bond at interface Silicone bulk (Si–O–Si backbone) Glass (SiO₂)
Fig 1 — Molecular interface: Silicone creates hydrogen bonds & condensation-cured covalent bonds at both the aluminium oxide surface and the glass silanol surface.

🔷 Aluminium Surface Chemistry

Aluminium instantly oxidises in air, forming a thin but stable aluminium oxide (Al₂O₃) layer typically 4–10 nm thick. This oxide is covered with hydroxyl groups (–OH). Silicone primers react with these groups through condensation, forming an Al–O–Si covalent bond.

Al–OH + HO–Si(primer) → Al–O–Si(primer) + H₂O // Condensation at Al surface
Al₂O₃ surface energy ≈ 900 mJ/m² // High energy → good wetting by silicone
Primer (e.g. organosilane) MW ≈ 150–250 g/mol

🔷 Glass Surface Chemistry

Glass surfaces (SiO₂) are rich in silanol groups (–Si–OH). When silicone contacts glass, the silicone polymer chains form siloxane bonds (Si–O–Si) — the same chemistry as the glass surface itself. This is why silicone bonds exceptionally well to glass: it is chemically related to glass.

≡Si–OH + HO–Si≡ → ≡Si–O–Si≡ + H₂O // Siloxane bond formation
Bond energy Si–O–Si ≈ 452 kJ/mol // Very strong covalent bond
Glass surface energy ≈ 500–800 mJ/m²
💡
Why Primer Matters on Aluminium

Raw aluminium often has lubricant residues, anodising coatings, or paint. A silane-based primer deposits a mono-molecular layer that bridges between the oxide surface and the silicone, ensuring the bond is to the oxide — not to contamination. Without primer, adhesion failure (cohesive → adhesive) is almost certain within 1–2 years of UV/moisture exposure.

🔗

Covalent Bond (Primary)

Si–O–Si & Al–O–Si linkages. Bond energy 450+ kJ/mol. Permanent, unaffected by water or temperature when properly formed.

🧲

Van der Waals (Secondary)

Weak physical attractions between polymer chains and substrate. Important across the bulk of the contact area — strength in numbers.

💧

Hydrogen Bonding

Transient –OH interactions during cure. These become permanent covalent bonds after condensation reaction completes.

🎯

Mechanical Interlocking

Surface roughness (anodised finish) creates micro-keying. Increases effective contact area by 10–40%.

Section 02

IGU Glass-to-Glass Bonding

An Insulating Glass Unit (IGU) consists of two or more glass panes separated by a spacer bar, with the internal cavity filled with argon or dry air. The edge seal is a dual-sealant system critical to the unit's longevity.

OUTER GLASS INNER GLASS Argon / Dry Air λ = 0.016 W/m·K Aluminium / SS Spacer + Desiccant Aluminium / SS Spacer + Desiccant Edge Seal Cross-Section Spacer + Desiccant Spacer + Desiccant ① Primary Seal PIB Butyl — vapour barrier 0.2 g/m²·day MVTR ② Secondary Seal Polysulfide or Silicone Structural integrity of IGU ③ Spacer Bar Al, stainless, or warm-edge Molecular sieve desiccant
Fig 2 — IGU dual-seal system: PIB butyl primary seal for moisture vapour control, silicone/polysulfide secondary seal for structural integrity.

① Primary Seal — Polyisobutylene (PIB) Butyl

The first seal is a hot-applied PIB (Polyisobutylene) strip placed between the spacer and glass. Its role is purely as a vapour barrier — it has extremely low moisture vapour transmission rate (MVTR ≈ 0.2 g/m²·day). PIB does not cure; it remains thermoplastic and relies on pressure adhesion.

PIB monomer: CH₂=C(CH₃)₂ → [–CH₂–C(CH₃)₂–]ₙ // Polymerised isobutylene
MW range: 300,000 – 1,000,000 g/mol
MVTR: 0.2 g/m²/day // Almost zero — best vapour barrier polymer

② Secondary Seal — Silicone or Polysulfide

The secondary seal provides structural strength to the IGU assembly. It bonds the two glass panes via the spacer, resisting wind loads, thermal cycling, and handling stresses. Two main chemistry types are used:

Property Polysulfide (PS) Silicone (Neutral Cure) Winner
UV ResistancePoor — must be shelteredExcellent — UV stableSilicone
Tensile Strength1.5–2.5 MPa0.5–1.5 MPaPolysulfide
Elongation at Break200–400%400–700%Silicone
Use in SSG❌ Not suitable✅ PreferredSilicone
Gas PermeabilityLowHigherPolysulfide
Service Temperature–40°C to +80°C–50°C to +150°CSilicone
Compatibility with PDMS siliconeGood with primerDirect bondSilicone
Section 03

Structural Silicone: 1-Part vs 2-Part

Structural silicone sealants come in two fundamental formulations. The choice between them affects application method, cure speed, joint geometry limits, and quality control requirements.

Parameter 1-Part (Moisture Cure) 2-Part (Chemical / RTV Cure)
Cure Mechanism Atmospheric moisture diffuses into joint; acetoxy, oxime or neutral by-products released Part A (base polymer) + Part B (crosslinker/catalyst) mixed; internal chemical reaction
Cure Depth Limit ~15 mm per pass (moisture cannot penetrate deeper in reasonable time) Unlimited depth — cures throughout regardless of thickness
Cure Time @ 23°C/50% RH 3–21 days for full cure (structural strength in 7–14 days) 24–72 hours to handle strength; full cure 3–7 days
Application Single cartridge, standard caulking gun — site friendly Dual-component equipment (pneumatic pump, static mixer) — factory/shop preferred
Quality Control Visual inspection; cure confirmation difficult without sampling Snap time / butterfly test performed every drum/cartridge change for confirmation
Joint Depth (Bite) Maximum 15 mm in single pass (deeper = multiple passes after partial cure) Up to 30–50 mm in single application
Typical Tensile Strength 0.8 – 1.4 MPa 1.0 – 2.0 MPa
Allowable Design Stress (ETAG) ≤ 138 kPa (20 psi) ≤ 138 kPa (20 psi)
Shelf Life 9–18 months (moisture sensitive) 12–24 months (Part A & B separately)
Temperature Sensitivity during Cure High — cure slows dramatically below 10°C Moderate — less affected by low RH
Primary Use Case Site-glazed SSG, re-glazing, repairs; weather sealing Factory-glazed IGU secondary seal, unitised curtain wall production
By-products during Cure Acetic acid (acetoxy), oximes, or alcohols (neutral) None — clean 100% solids reaction
Cost Lower (simpler formulation, less equipment) Higher (materials + equipment cost)
Common Products Dow 795, Sika 267, Ottoseal S 70 Dow 983, Sika SG-500, Momentive SSG4000
1-Part Moisture Cure –Si–OR + H₂O → –Si–OH + ROH –Si–OH + HO–Si– → –Si–O–Si– + H₂O 💧💧💧💧 Cure front moves inward from surface Cured → 2-Part RTV Cure Part A PDMS Base Part B Crosslinker + Uniform Cure Cures uniformly throughout — no depth limit
Fig 3 — Cure mechanism comparison: 1-part cures inward from surface (moisture-dependent), 2-part cures uniformly throughout (catalyst-driven).
Section 04

How Structural Strength is Achieved

The structural performance of a silicone joint comes from a combination of molecular network density, joint geometry, and substrate adhesion quality.

GLASS / Al SILICONE Joint W (Bite) T (Thickness) FRAME ETAG 002 Formula B_min = F / (2L · σ_allow) B = Structural bite width (mm) F = Design wind load (N) L = Joint length (mm) σ = Allowable stress ≤ 138 kPa Aspect ratio: 1 ≤ W/T ≤ 3
Fig 4 — SSG joint geometry: Bite width (B) and joint thickness (T) must satisfy ETAG 002 with aspect ratio W/T between 1 and 3.

Strength Source 1: Crosslink Network Density

During cure, crosslinking reactions create a 3D polymer network. Higher crosslink density = higher modulus but lower elongation. Structural silicones are formulated for a balance: stiff enough to transfer loads, flexible enough to accommodate thermal movement (±15% joint movement capability).

PDMS chain: –[Si(CH₃)₂–O]ₙ– MW per repeat unit = 74 g/mol
Crosslinker: Si(OR)₄ or methyltrimethoxysilane
Typical bulk MW of cured network: 50,000 – 500,000 g/mol (network, not linear)
Shore A hardness after cure: 20–50
Young's Modulus (E): 0.5 – 2.0 MPa

Strength Source 2: Adhesion Interface

True structural performance requires cohesive failure (failure within the silicone bulk) rather than adhesive failure (failure at the interface). Cohesive failure confirms the substrate bond is stronger than the material itself — the design goal.

Cohesive Failure (Good)

Silicone tears through its own bulk. Substrate surface retains silicone residue. Bond is confirmed stronger than material. Design is structurally sound.

Adhesive Failure (Bad)

Clean peeling at the silicone–substrate interface. No silicone residue on substrate. Indicates surface contamination, missed primer, or incompatible substrate. Reglazing required.

Strength Source 3: Joint Geometry (Bite & Thickness)

The ETAG 002 / ASTM C1401 design rules limit allowable design tensile stress to 138 kPa (20 psi) regardless of the silicone's higher tested strength. This safety factor accounts for long-term creep, UV degradation, thermal cycling, and workmanship variability. The joint aspect ratio (Bite W ÷ Thickness T) must stay between 1 and 3.

⚠️
Dead Load Consideration

Glass weight exerts a permanent static shear stress on the silicone. Most manufacturers allow 11,000–14,000 Pa for dead load shear. Glass must be supported on setting blocks; silicone alone should not carry dead load in standard SSG systems — only in specially designed full-structural systems.

Section 05

Material Compatibility & Laboratory Tests

Before any structural silicone is applied on a project, a series of tests must be conducted to verify that the silicone, primers, substrates, and adjacent materials are chemically compatible. This is non-negotiable per ETAG 002, ASTM C1401, and most manufacturer warranties.

Why Compatibility Matters

Many materials commonly used in facades — EPDM gaskets, foam tapes, PVC sealants, bitumen, certain coatings — contain plasticisers, oils, solvents, or reactive chemicals that migrate into silicone during or after cure. This can cause:

Staining / discolouration Inhibited cure Loss of adhesion Swelling of silicone Reduced tensile strength Crazing / cracking

Standard Laboratory Compatibility Tests

Test Standard Method Pass Criterion Duration
Adhesion (Peel) ASTM C794 / ISO 8510 180° peel test on substrate at 50 mm/min ≥ 80% cohesive failure 7 days cure + 7 days immersion
Adhesion-in-Peel (with aging) ASTM C794 Specimens aged in water at 70°C for 7 days, then peel No loss of cohesive failure after water immersion 21 days total
Staining Test ASTM C510 / EOTA TR012 Sealant applied to stone/masonry and assessed visually after UV exposure No visible staining beyond 3 mm from edge 7 + 28 days
Compatibility (Migration) EOTA TR012 / ASTM C1087 Adjacent material placed in contact with silicone in curing state; cure quality assessed Full cure within expected snap time; no inhibition 48–72 hrs
Tensile Adhesion ASTM C1135 / EN ISO 8339 H-bar specimens pulled at 50 mm/min (12×12×50 mm joint) ≥ 0.14 MPa; ≥ 75% cohesive failure 21 days cure
Modulus of Elasticity ASTM C1135 H-bar tensile at 12.5% and 25% elongation Per manufacturer spec (low mod preferred) 21 days cure
Accelerated UV / Weathering ASTM G154 / ISO 4892 Xenon arc or UV-B lamp, 1000–2000 hours ΔE colour < 5; tensile strength retention > 70% 6–10 weeks
Thermal Cycling EOTA TR010 –20°C to +70°C × 100 cycles, joint under 12.5% tension No cracking, delamination or > 10% strength loss 4–5 weeks

Substrate-Specific Requirements

SubstratePreparationPrimerSpecial Notes
Float Glass (clear)IPA wipe × 2Usually not required with neutral siliconeTin side vs air side — test both
Coated Glass (Low-E, frit)IPA wipe, consult coating supplierOften requiredSome coatings incompatible — must test
Mill-finish AluminiumMEK or IPA cleanSilane primer (e.g. AP-133)Remove mill oil completely
Anodised AluminiumIPA cleanRequired (porous surface)Primer penetrates anodising micro-pores
PVDF-coated AlMEK cleanSpecific primer per brandTest each coating brand separately
Stainless SteelIPA cleanRequiredEnsure passive oxide layer intact
Structural Foam TapeN/AN/ARun compatibility test before use as backer
Section 06

The Butterfly Test

The butterfly test (also called the snap time test or bead fold test) is the primary in-shop quality control check performed before every shift and every drum/pail change of 2-part structural silicone. It verifies that Part A and Part B are mixing correctly and that the material will cure within specification.

STEP 1 Extrude 200–300 mm bead onto paper Immediately after extrusion Check colour mix ratio STEP 2 Fold paper in half over the bead Press flat and immediately unfold ✅ PASS — Mixed Uniform tear / cohesive failure Snap within specified time window ❌ FAIL — Issues • Sticky / not snapping = wrong ratio • Marbled streaks = incomplete mixing • Too fast snap = high catalyst • No snap = Part B exhausted / empty 📋 LOG RECORD • Date & Time • Operator name • Drum / batch number • Snap time (minutes) • Mix ratio (visual) • Pass / Fail Spec snap time: Typically 15–75 min @ 23°C / 50% RH STOP if FAIL — purge and retest before apply
Fig 5 — Butterfly test procedure: A folded bead that snaps cleanly confirms correct A:B mixing ratio and cure kinetics.

Butterfly Test Frequency Requirements

  1. Start of each shiftFirst bead of the day, before any production glass is sealed.
  2. Every drum / pail changePart A and Part B are changed independently — risk of mismatched ratio increases at every changeover.
  3. After any equipment stoppage > 30 minMaterial sitting in static mixer may partially cure and block the mixer, altering ratio of next extrusion.
  4. Change in temperature > 10°CCold material is more viscous — pump rates may shift, altering the mix ratio.
  5. Snap time check vs specIf snap time is outside the manufacturer's specified window, stop production and investigate.
🚨
Failed Butterfly Test — Actions Required

If the butterfly test fails, all units glazed since the last passing test must be quarantined and re-tested (peel adhesion on witness samples). Material in the line must be purged for a minimum of 3× the static mixer volume before the next test. Document all actions in the QC log.

Section 07

Curing Chemistry — Molecular Level

Silicone cures by condensation crosslinking (1-part moisture cure or 2-part RTV). Here is what happens at the molecular scale:

🔗
Uncured PDMS

Linear poly(dimethylsiloxane) chains with –OCH₃ or –OH terminal groups. MW ~15,000–50,000 g/mol. Viscous liquid.

💧
Hydrolysis

–Si–OCH₃ + H₂O → –Si–OH + CH₃OH. Silanol groups generated on chain ends & crosslinker. Moisture-triggered (1-part) or catalyst-triggered (2-part).

Condensation

–Si–OH + HO–Si– → –Si–O–Si– + H₂O. New siloxane bonds form, linking chains. Network begins to grow and viscosity rises sharply.

🕸️
Gelation (Snap)

Critical crosslink density reached. Material transitions from viscous liquid to elastic gel. This is the "snap" in the butterfly test. Tack-free surface forms.

💪
Full Cure

3D network fully developed. All reactive groups consumed. Shore A hardness stabilises. Design tensile and shear strength achieved. No further significant property change.

// Condensation cure overview (neutral cure / 2-part)

Step 1 — Activation:
  R–Si–(OCH₃)₂ + 2 H₂O → R–Si–(OH)₂ + 2 CH₃OH

Step 2 — Condensation:
  ≡Si–OH + HO–Si≡ → ≡Si–O–Si≡ + H₂O

Step 3 — Network growth (repeat n times):
  [–Si(CH₃)₂–O–]ₙ crosslinked via trifunctional silane nodes
  Mn of network segment ≈ 1,000–5,000 g/mol between crosslinks

// Key parameters
Tin catalyst (2-part): dibutyltin dilaurate, [Sn(Bu)₂(OOCC₁₁H₂₃)₂], MW = 631 g/mol
Platinum catalyst (addition cure): ~20–50 ppm of Karstedt's catalyst
Crosslink density (ν): ν = ρ / Mc where Mc = MW between crosslinks

Molecular Weight & Strength Relationship

ParameterValueEffect on Properties
PDMS backbone MW (uncured)15,000–50,000 g/molHigher MW → better elongation but slower cure
Crosslinker MW (e.g. MTMS)136 g/molLower MW = more crosslink sites per gram
MW between crosslinks (Mc)3,000–10,000 g/molLower Mc = stiffer, higher modulus
Network strand density (ν)0.1–1.0 mol/LHigher ν = stronger but less flexible
Filler (fumed silica) MW~60 g/mol per SiO₂ unitReinforces network; increases tensile 3–5×
Si–O bond dissociation energy452 kJ/molThermal stability to >200°C
C–C bond (organic polymer)347 kJ/molSilicone more UV/heat stable than organics
🔬
Why Silicone Outperforms Organics

The Si–O–Si backbone has a bond angle of 143° (vs 111° for C–C), giving silicone exceptional conformational flexibility at low temperatures. The high bond energy (452 kJ/mol) and the large Si–O bond length (1.64 Å) make silicone resistant to UV photodegradation at wavelengths down to 280 nm. Combined with the non-polar methyl groups that resist water absorption, silicone is the only polymer suitable for 50+ year structural glazing applications.

Section 08

Deglazing Procedure

Deglazing is the controlled removal of glass bonded with structural silicone from a facade or IGU assembly. It requires careful methodology to avoid glass breakage, substrate damage, and personal injury.

Safety First

Deglazing from height requires scaffold or elevated work platform. Full PPE: safety glasses, cut-resistant gloves (EN 388 Level 4+), hard hat, and fall arrest equipment. Two-person operation minimum for panels > 0.5 m².

Tools Required

🔪

Oscillating Multi-Tool

Bi-metal or diamond blade for cutting cured silicone close to the glass edge. Most controlled for thin joints.

🔧

Lame / Knives

Hook-bladed lame knife or razor scraper for final separation and silicone removal from frames.

🪝

Suction Cups

Mechanical glass lifters rated for the panel weight × 3 safety factor. Used to support glass during cutting.

🛡️

Tape & Protection

Apply duct tape or protection film grid to glass face before cutting to contain potential breakage.

Step-by-Step Deglazing Procedure

  1. Survey & documentPhotograph all four sides. Record joint width, depth, and any visible adhesion failures before cutting. Confirm temporary support plan.
  2. Apply protection film & suction cupsFilm protects glass face. Secure suction cups and attach lifting equipment before cutting begins. Never start cutting unsupported glass.
  3. Score the weatherseal (if present)Remove any cap bead or weatherseal covering the structural silicone using a utility knife. Clean away loose material.
  4. Cut structural silicone — first passUsing oscillating tool with appropriate blade, cut along the glass edge at a shallow angle. Do not force or lever — cut smoothly.
  5. Cut all four sides progressivelyCut each side partially, then complete each. Avoid cutting one side fully while other sides are still fully bonded — creates bending stress.
  6. Controlled removalOnce all cuts are complete, use suction cups to gently pull glass away from frame with a rolling motion — do not yank or jerk.
  7. Remove silicone residueClean all residual silicone from frame and glass using razor scrapers, then wipe with MEK or IPA. All old silicone must be removed before resealing.
  8. Inspect surfacesExamine frame for corrosion, damage, or failed coatings. Document failure mode (adhesive vs cohesive) and photograph for records.
  9. Prepare for re-glazingApply fresh primer per manufacturer specifications. Allow primer to dry to tack-free. Apply new silicone within the primer open time window.
Section 09

Physical Inspection — What to Look For

After deglazing, a thorough visual and tactile inspection of the removed silicone and substrates reveals the bond quality and failure mode. Here is a field guide:

✅ Signs of a Good, Properly Cured Bond

Cohesive Failure

Silicone rubber residue remains on BOTH surfaces after separation
The tear is through the silicone body — not at the interface
Confirms bond strength exceeded silicone tensile strength

Full Joint Fill

No voids, air pockets or gaps visible in cross-section
Silicone completely fills the designed bite width and thickness
Consistent bead profile along full perimeter

Good Cure Depth

Silicone is uniformly elastic throughout thickness (no uncured core)
No sticky, liquid, or tacky zones within joint cross-section
Shore A hardness consistent through depth

Correct Colour

Uniform colour throughout — no marbling or streaks (2-part)
No discolouration (yellowing = UV degradation of early silicone)
No chalking or surface powdering

❌ Signs of Problems

Adhesive Failure

Clean interface — no silicone on one substrate
Causes: contamination, missed primer, incompatible coating, expired primer
Action: Full reglaze with fresh preparation

Uncured Core

Sticky or liquid silicone in the centre of thick joints
1-part: bite width exceeded 15 mm without multiple passes
2-part: failed butterfly test, wrong mix ratio

Voids / Bubbles

Air pockets visible in joint cross-section
Causes: poor tooling, too-rapid extrusion, cold material
Reduces effective bonded area — structural risk

Cracking / Crazing

Surface or through-cracks in silicone body
Causes: incompatible adjacent material, thermal overstress, UV degradation
Indicates loss of elasticity — replace immediately

Delamination at Glass

Silicone peels cleanly from coated glass (Low-E, silk-screen)
Coating incompatibility or silicone applied on wrong face
Check glass specification — some coatings are not bondable

Undersized Bite

Bite width less than calculated minimum (often < 6 mm)
Results from glass misalignment or tape/spacer taking up bite
Must be supplemented with mechanical retainers
📸
Documentation is Critical

Photograph every deglazed unit showing the failure mode in cross-section with a scale reference. Mark cohesive vs adhesive failure areas as a percentage. This data feeds back into the root-cause analysis and any warranty or insurance claims.

Section 10

🖩 Structural Silicone Bite Width Calculator

Based on ETAG 002 / ASTM C1401 methodology. Calculates minimum structural bite width and joint thickness for a rectangular glass panel under wind load and dead load, using allowable stress design.

Calculation Results
Glass Panel Area
Glass Self-Weight
Total Wind Force on Panel
Min Structural Bite Width (Wind)
Min Bite (Dead Load Shear)
Governing Bite Width
Recommended Joint Thickness (T)
Aspect Ratio (B/T)
Silicone Volume Required (perimeter)
Status
⚠️
Disclaimer

This calculator provides preliminary estimates for guidance only, based on simplified ETAG 002 / ASTM C1401 methodology. Results must be verified by a qualified facade engineer before use in construction. Always consult the silicone manufacturer's technical data and project-specific structural calculations.