Topics

Dgeba Vs Novolac Vs Bisphenol F Epoxy Which Backbone Fits Your Application

DGEBA vs Novolac vs Bisphenol F Epoxy: Which Backbone Fits Your Application

In epoxy resin formulation, backbone selection is not a single-variable decision. The reference diglycidyl ether of bisphenol A resin, commonly abbreviated as DGEBA, delivers a broad balance of ambient-temperature handling, cured modulus, and adhesion. Commercial liquid DGEBA grades typically exhibit an epoxide equivalent weight of 182–192 g/eq when titrated by ASTM D1652 and a viscosity of 11,000–14,000 mPa·s at 25°C by ASTM D2196. The isopropylidene bridge and pendant methyl groups of bisphenol A impart chain rigidity and useful thermal stability after cure, but they also raise room-temperature viscosity and increase the risk of cold-crystallization during unheated storage. Bisphenol F diglycidyl ether, DGEBF, replaces the isopropylidene bridge with a methylene bridge. This reduces steric hindrance and gives unmodified liquid DGEBF resins a viscosity between 2,500 and 4,500 mPa·s at 25°C, with epoxide equivalent weights near 165–180 g/eq. Epoxy novolac resins, including epoxy phenol novolac and epoxy cresol novolac grades, carry multiple glycidyl ether groups along a phenolic novolac backbone. Average oxirane functionality can range from 2.2 to 6.0 groups per molecule. The high aromatic ring density of the novolac backbone raises cured glass transition temperature and chemical resistance, but it also produces semi-solid room-temperature handling and often requires heated storage, preheated piping, and heated static mixers.

Representative uncured resin backbone parameters
ParameterTest methodDGEBADGEBFEpoxy phenol novolac
Epoxide equivalent weight (g/eq)ASTM D1652182–192165–180170–190
Viscosity at 25°C (mPa·s)ASTM D219611,000–14,0002,500–4,500Semi-solid; 20,000–60,000 at 52°C
Average oxirane functionalityEpoxide titration plus number-average molecular weight2.02.02.2–6.0

For vacuum infusion and filament winding at ambient temperature, the mixed viscosity of the resin-hardener system rather than the neat resin viscosity controls injection pressure and fabric wet-out. A DGEBA resin mixed with a modified cycloaliphatic amine may produce a mixed viscosity of 600–1,000 mPa·s at 25°C. Substituting DGEBF or a 30–50% DGEBF blend with the same hardener commonly reduces mixed viscosity to 250–500 mPa·s under identical ratio and test conditions. The lower mixed viscosity reduces the pressure gradient required to wet a 600 g/m² biaxial glass fabric over a 3 m flow path by approximately 30–50% compared with a 900 mPa·s DGEBA system. This difference lowers the probability of mold deflection, dry-spot formation, and premature gel before complete wet-out in closed-mold processes. The same viscosity reduction permits higher filler loading in self-leveling coatings and grouting compounds. A DGEBF-based trowelable system can accept 65–75 wt% silica filler while retaining low-slump handling, whereas an equivalent DGEBA system often requires reactive diluent or solvent to reach comparable workability.

What processing window does a low-viscosity Bisphenol F system actually provide?

Bisphenol F changes more than flow behaviour. Because DGEBF has a lower epoxide equivalent weight than DGEBA, stoichiometric hardener demand increases by approximately 5–10%. The higher reactive-group concentration accelerates ambient gel development and raises exothermic peak temperature in thick sections. A 200 g cup mixed with a polyamide hardener at 25°C may show gel time of 35–55 min for DGEBA and 28–50 min for DGEBF, depending on catalyst level and moisture. Thermocouple monitoring in the same cup can record a 8–15°C higher exothermic rise for DGEBF. For thick-section potting and casting, the lower initial viscosity improves air release and reduces bubble entrapment, but the higher exotherm can exceed the developing glass transition temperature and produce thermal stress cracking or colour shift. Dynamic mechanical analysis by ASTM E1640 after full cure often shows DGEBF networks to have a slightly lower or comparable glass transition temperature to DGEBA networks when cured with the same hardener because the methylene bridge reduces rotational barriers; however, the lower epoxide equivalent weight offsets this by increasing crosslink density. The net effect is application-specific and should be verified with a defined cure schedule rather than inferred from neat resin viscosity alone.

In ambient-temperature adhesive and chemical-anchoring work, DGEBF can improve penetration into dust-contaminated drill holes and narrow cracks. Its low initial mixed viscosity is beneficial under low-pressure injection, but the same property can produce sag on vertical or overhead surfaces. Fumed silica at 2–4 wt% is frequently added to recover thixotropy without returning the system to DGEBA-level viscosity. DGEBA-based anchoring adhesives accelerated with tertiary amines generally tolerate substrate moisture up to approximately 4% by mass under common installation conditions, but room-temperature-cured systems may show glass transition temperatures of only 45–55°C without post-cure. Novolac epoxy-based anchoring formulations are uncommon for ambient concrete repair because their high viscosity complicates cartridge dispensing and requires heated resin delivery. Published data for this specific configuration is limited compared with field data for DGEBA and DGEBF anchor systems.

Novolac backbones raise crosslink density but impose viscosity and mixing penalties

Epoxy phenol novolac and epoxy cresol novolac resins introduce more than two oxirane groups per molecule. When cured with an aromatic amine such as 4,4′-diaminodiphenyl sulfone, the resulting network exhibits shorter inter-crosslink chain lengths and higher aromatic ring concentration than a DGEBA or DGEBF network at equivalent cure. Dynamic mechanical analysis by ASTM E1640 typically shows glass transition temperatures 20–40°C higher than a DGEBA/aromatic amine comparison system. The same network tightening reduces tensile elongation. A DGEBA/amine casting tested by ASTM D638 may show elongation at break of 3–6%, whereas an epoxy novolac/amine casting frequently falls to 1–2%. This lower elongation reduces compliance and can cause cracking in thick castings or in bonded joints subjected to thermal cycling.

Processing an epoxy novolac system involves heated resin drums, heated hardener loops, and preheated static mixers. At 52°C, many epoxy phenol novolac resins remain above 20,000 mPa·s; at 25°C they are semi-solid and cannot be reliably metered through standard gear pumps or piston dispensers. Transfer molding grades based on novolac resins and silica fillers operate at mold temperatures of 160–180°C and clamping pressures of 20–80 bar. These conditions are not comparable to ambient laminating or coating lines. The high crosslink density is nevertheless useful in solvent-exposed linings, high-temperature adhesives, semiconductor encapsulants, and printed circuit board laminates where long-term thermal and chemical resistance outweigh handling difficulty.

Chemical resistance data support the novolac advantage for solvent immersion. Immersion in methyl ethyl ketone at 25°C for 7 days according to ASTM D543 can produce mass gains in the range of 2–4% for a DGEBA/aromatic amine network and 0.8–1.5% for an epoxy novolac/aromatic amine network. Immersion in 10% sulfuric acid and 10% sodium hydroxide solutions shows smaller differences but still favours the higher-functionality novolac system. This behaviour follows from reduced free volume and restricted segmental mobility, not from a change in the chemical nature of the ester or ether linkages. It should not be interpreted as equivalent to oxidative stability. Above 180°C, all three backbone types undergo progressive oxidative chain scission and carbonyl formation, and the hardener chemistry often dominates the ageing response.

If DGEBA crystallization interrupts ambient-temperature metering

Standard liquid DGEBA stored below 15–18°C can crystallize without agitation or seeding. Crystallization converts the clear resin into a slushy or opaque mass that blocks drum suction lines, gear pumps, flow meters, and static mixer elements. The solid phase does not fully redissolve unless the entire drum contents reach 45–50°C and are held under low-shear agitation for several hours. Automatic meter-mix lines with a 0.1 mm static mixer element can experience backpressure spikes when small crystalline particles accumulate at the mixing element. DGEBF is less prone to spontaneous crystallization at low temperature because the mixed methylene-bridged bisphenol F isomers disrupt regular crystal packing. Where DGEBA must be used in an unheated warehouse, drum heaters set at 40–50°C and a recirculation loop through a 0.25 mm filter are standard. Prolonged storage of DGEBA at 50°C can slowly advance the epoxide group concentration and darken the resin; a weekly titration by ASTM D1652 and a colour check against a control sample are typical control measures.

High-temperature service life is not determined by backbone selection alone. Heat deflection temperature measured by ASTM D648 at 1.82 MPa identifies the temperature at which a specimen deflects by a defined amount under load. DGEBA systems often fall between 100–140°C, DGEBF systems may fall between 90–125°C, and novolac-based systems can exceed 160–200°C after full cure. Glass transition temperature measured by ASTM E1640 or ISO 11357-2 is a better indicator of changes in network mobility. Continuous exposure above 150°C in air causes carbonyl formation and chain scission in DGEBA and DGEBF networks, and the higher aromatic density of novolac systems slows but does not eliminate oxidative embrittlement. Thermogravimetric analysis under nitrogen and air by ISO 11358-1 can identify the temperature of 5% mass loss, but mass loss alone does not quantify mechanical embrittlement or adhesion loss.

Electronics and electrical insulation applications impose additional requirements for ionic purity, moisture absorption, and flammability. DGEBA-based anhydride-cured systems may show volume resistivity above 10^14 Ω·cm by IEC 62631-3-1, and novolac systems can maintain similar resistivity after moisture ageing because their lower equilibrium water uptake reduces conductive path formation. Water absorption after 24 h immersion by ASTM D570 is typically 0.1–0.3% for DGEBA/anhydride systems and 0.2–0.5% for DGEBF/anhydride systems, although filler content and cure schedule influence the measured values. In printed circuit board laminates, DGEBA or DGEBF resins are often advanced with brominated bisphenol A to meet UL 94 V-0 flammability and then reinforced with woven glass. Epoxy novolac resins are added to raise glass transition temperature and lower moisture sensitivity in high-layer-count boards and semiconductor packaging. The same novolac structure is used in epoxy molding compounds for integrated circuit encapsulation because it balances high filler loading, low melt viscosity under transfer mold conditions, and low ionic contamination.

For solvent-resistant tank linings and secondary containment coatings, novolac epoxy systems are specified where continuous contact with methyl ethyl ketone, methanol, toluene, or mixed BTX streams is expected. The high crosslink density limits solvent swelling and maintains coating hardness under immersion. Bisphenol F epoxy systems are selected for low-temperature application and low-VOC compliance when crews must spray at 10–15°C. A DGEBF system can be applied by airless spray with mixed viscosity below 800 mPa·s under those conditions, while DGEBA may require heated plural-component pumps and heated hose. In contrast, DGEBA remains the default backbone for general-purpose adhesives, castings, and coatings that require a balance of cost, ambient handling, and mechanical strength without extreme solvent or high-temperature service. The selection is therefore determined by the processing sequence, the end-use thermal envelope, and the solvent exposure severity rather than by any single data sheet value.

Top