Thermal Stability Testing and Analysis of Underfill Adhesives

Thermal stability testing of underfill adhesives is essential for verifying the long-term reliability of flip-chip, BGA, CSP, and wafer-level packages exposed to reflow temperatures and sustained high-temperature operation. This article explains the sample types, test items, and standard test methods—including TGA, DSC, TMA, high-temperature storage, and thermal cycling—used by accredited third-party laboratories to evaluate the thermal endurance of underfill materials. Use this guide to understand test parameters, typical acceptance criteria, and application requirements for capillary, no-flow, and molded underfill systems.

Technical Overview

Underfill adhesives are low-viscosity polymer materials—typically one-component epoxy systems filled with fused silica—that are dispensed beneath flip-chip dies, ball grid arrays (BGAs), chip-scale packages (CSPs), and wafer-level packages to reinforce solder interconnections. Once cured, the underfill forms a rigid composite that mechanically couples the silicon die to the substrate, redistributes the thermomechanical stress generated by the coefficient of thermal expansion (CTE) mismatch between silicon, solder, and laminate, and protects the joints from fatigue, moisture, and mechanical shock.

Thermal stability describes the ability of the cured underfill to preserve its chemical structure, glass transition temperature (Tg), mass, adhesion, and mechanical performance both during processing—such as lead-free reflow peaks of up to 260 °C—and during long-term service at elevated temperature. Degradation mechanisms include residual post-cure reactions, chain scission, thermo-oxidative crosslinking, breakdown of the filler–matrix interface, and volatilization of low-molecular-weight species. Thermal stability testing therefore combines thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), thermomechanical analysis (TMA), long-duration heat aging, and thermal cycling to characterize the material from room temperature through decomposition, and to project service life in accordance with recognized international standards.

Testing Samples

Accredited third-party laboratories evaluate underfill adhesives both as cured bulk specimens (plaques, films, or powders) and as assembled packages, because geometry, cure profile, and filler sedimentation all influence the thermal response. Specimens are normally conditioned before baseline testing, for example 40 h at 23 °C and 50% relative humidity per ASTM D618. The principal material types submitted for testing include:

Capillary Underfill (CUF) Adhesives

Low-viscosity liquid epoxies that flow by capillary action beneath an already-attached die or package. Bulk plaques of 1–2 mm thickness are cast and cured using the supplier's recommended schedule, or material is extracted directly from populated boards so that the actual production cure state is assessed.

No-Flow Underfill (NUF) Adhesives

Fluxing, unfilled or low-filler formulations applied before die placement and cured simultaneously with the solder reflow. Testing focuses on whether the single-step cure achieves full conversion and whether the material survives 260 °C lead-free reflow peaks without mass loss, voiding, blistering, or degradation.

Molded Underfill (MUF) Compounds

Epoxy molding compounds with engineered silica filler size distribution that combine underfill and encapsulation in one transfer-molding step. Molded disks and bars are tested, with TGA residue used to verify filler loading and TMA used to confirm that the CTE matches flip-chip requirements.

Assembled Package Specimens

Complete flip-chip, BGA, or CSP units are subjected to high-temperature storage and thermal cycling, then examined by acoustic microscopy and cross-sectioning for underfill delamination, filler settling, cracking, and solder joint integrity.

Testing Items

Depending on the qualification program, the thermal stability evaluation of underfill adhesives typically covers the following items:

• Glass transition temperature (Tg) of the cured material and the Tg shift after post-cure or thermal aging
• Decomposition onset temperature and the 5% and 10% weight-loss temperatures (Td, T5%, T10%)
• Peak degradation temperature (Tmax) and residual mass at 600 °C, used to estimate silica filler content
• Percent mass loss after high-temperature storage, for example 1,000 h at 150 °C or 175 °C
• Coefficient of thermal expansion below and above Tg (α1 and α2)
• Degree of cure, residual cure exotherm, and total reaction enthalpy
• Reflow survival at 245–260 °C peak temperature, including mass change, blistering, and delamination
• Thermal endurance characterization, such as temperature index or relative thermal index determination per IEC 60216 and UL 746B
• Supporting measurements of storage modulus retention and adhesion or die shear strength after heat aging

Testing Methods

The thermal stability of underfill adhesives is assessed through a complementary set of instrumental and environmental methods, each targeting a different degradation or property-change mechanism. Unless otherwise specified, heating rates of 5–10 °C/min under an inert nitrogen purge are standard, with oxidative runs in air to reveal thermo-oxidative susceptibility.

Thermogravimetric Analysis (TGA)

Specimens of 5–20 mg, taken as powder or cut from cured plaques, are heated from ambient temperature to 600 °C under nitrogen and, in a second run, under air, following ASTM E1131, ISO 11358, ASTM D3850, and JIS K 7120. The test reports the onset of decomposition, T5% and T10% weight-loss temperatures, the peak degradation rate temperature, and the residual mass that corresponds closely to inorganic filler content. Rapid-heating screening variants at higher scan rates are sometimes used for incoming material comparison.

Differential Scanning Calorimetry (DSC)

DSC is performed per ASTM E1356 or ASTM D3418, ISO 11357, GB/T 19466, and JIS K 7122 using a heat–cool–heat program at 10–20 °C/min. The analysis delivers the midpoint Tg, the residual cure exotherm (degree of cure = residual enthalpy divided by total reaction enthalpy), and any degradation exotherm in oxidative atmosphere. A significant Tg increase between the first and second heating scans indicates incomplete cure, which correlates with higher mass loss and shorter thermal life in service.

Thermomechanical Analysis (TMA)

TMA is conducted per ASTM E1545 or ASTM E1640 for Tg and ASTM E831 for CTE, using a 5 × 5 mm specimen with a low-force expansion probe. Typical capillary underfills exhibit α1 values of roughly 25–40 ppm/°C below Tg, rising steeply above Tg. A large α1–substrate mismatch accelerates solder fatigue during thermal cycling, so TMA data are a core input to package reliability modeling.

High-Temperature Storage and Heat Aging

Steady-state dry heat exposure is carried out per IEC 60068-2-2, with extended oven aging at multiple temperatures such as 160, 180, 200, and 220 °C. Mass loss, Tg drift, CTE change, and delamination are re-measured after intermediate checkpoints (168, 500, and 1,000 h or longer). For lifetime projection, IEC 60216 thermal endurance testing applies Arrhenius analysis to an end-point criterion—commonly a defined percentage of mass or property loss—to derive the temperature index, while UL 746B procedures generate the relative thermal index for long-term rating.

Thermal Cycling and Thermal Shock

Change-of-temperature testing per IEC 60068-2-14 subjects assembled packages to cycles such as −40 °C to +125 °C or −55 °C to +125 °C with dwell times of 10–15 minutes, for 100 to 1,000 cycles. Because the silicon, underfill, solder, and substrate expand at different rates, this test complements steady-state aging by superimposing CTE-mismatch fatigue. Post-test acoustic microscopy, cross-sectioning, and electrical continuity checks confirm whether the underfill layer retained adhesion and protected the joints.

Applications

Thermal stability test data for underfill adhesives directly support design, qualification, and manufacturing decisions across the electronics industry:

• Flip-chip, micro-BGA, and CSP underfill qualification for consumer, computing, and mobile devices
• Wafer-level and panel-level packaging reliability assessment, where reflow exposure is repeated across multiple assembly steps
• Stacked-die, package-on-package, and flash memory modules such as eMMC and UFS constructions
• Automotive electronic control units, ADAS sensor modules, and domain controllers that require extended endurance at 125–150 °C ambient
• 5G RF front-end modules, system-in-package assemblies, and high-density interconnect substrates
• Power semiconductor modules where sustained operating temperatures approach the limits of organic adhesives
• Incoming material control, second-source comparisons, and supplier datasheet verification for packaging houses and contract manufacturers
• Failure analysis of field returns, where TGA and DSC evidence of underfill degradation helps identify thermal overstress root causes

By matching measured Tg, decomposition, and aging behavior to the actual temperature profile of the target application, engineers can select an underfill system with confidence and demonstrate compliance with customer and industry requirements.

FAQ

What temperatures can underfill adhesives withstand?

Well-formulated epoxy underfills typically show a cured Tg of 110–160 °C, a 5% weight-loss temperature above 300 °C, and the ability to survive brief lead-free reflow peaks of 245–260 °C. For continuous operation, most standard grades are rated for 125–150 °C, while high-Tg and enhanced-chemistry grades serve more demanding automotive and power applications.

What mass loss is acceptable after high-temperature storage?

Common internal acceptance criteria limit mass loss to approximately 1% or less after 1,000 hours at 150 °C, with no delamination, cracking, or Tg drift beyond the specified window. Final limits should always be agreed against the material datasheet and the end-product reliability requirements.

How much sample is required for thermal stability testing?

TGA generally requires 5–20 mg per specimen with three replicates; DSC uses 5–10 mg in sealed or open pans; TMA requires a flat plaque of roughly 5 × 5 mm and 1–2 mm thickness. High-temperature storage and thermal cycling are performed on whole packages or cured bars, typically with ten or more units per test condition to allow destructive inspection at intermediate readouts.

What is the difference between glass transition and thermal decomposition?

Glass transition is a reversible physical softening of the amorphous polymer regions, detected as a step change in heat flow or expansion by DSC and TMA, whereas thermal decomposition is an irreversible chemical breakdown—chain scission and oxidation—accompanied by measurable mass loss in TGA. A material must retain adequate margin between its Tg and its operating temperature, and between its operating temperature and its decomposition onset, to remain reliable.

How is long-term thermal endurance lifetime estimated?

Under IEC 60216, specimens are oven-aged at several elevated temperatures and a diagnostic property, such as mass or Tg, is monitored until it reaches a defined end-point. Arrhenius extrapolation of the resulting time-to-endpoint data yields a temperature index—the temperature at which the criterion is reached after the specified exposure time—providing a quantitative basis for long-term thermal ratings.

How is reflow compatibility verified in the laboratory?

Reflow simulation exposes cured underfill specimens to one or more profiles with a 245–260 °C peak and controlled time above 217 °C, or to a short isothermal hold at 260 °C in a TGA. Acceptable materials show negligible mass change, no blistering or voiding, and unchanged Tg and adhesion after multiple reflow passes.