Underfill adhesive high-temperature expansion analysis is a critical thermomechanical characterization process that measures the coefficient of thermal expansion (CTE) and glass transition temperature (Tg) of underfill materials used in flip-chip, BGA, and wafer-level electronic packages. By quantifying dimensional changes at elevated temperatures through thermomechanical analysis (TMA) in accordance with ASTM E831 and ISO 11359-2, manufacturers can prevent solder joint fatigue, delamination, and warpage in high-reliability assemblies. This article examines the testing principles, sample requirements, evaluation items, methods, and applications of high-temperature expansion analysis for underfill adhesives.
Technical Overview
Underfill adhesives are polymer materials, most commonly silica-filled epoxy systems, that are introduced into the gap between a semiconductor die and its package substrate in flip-chip, ball grid array (BGA), and chip-scale package (CSP) assemblies. Once cured, the underfill mechanically couples the die to the substrate, redistributes thermomechanical stress away from the solder interconnects, and significantly improves resistance to temperature cycling, mechanical shock, and vibration. Because the cured underfill acts as a load-bearing element within the package stack-up, its dimensional behavior at elevated temperature directly governs long-term reliability.
The central engineering challenge is thermal expansion mismatch. Silicon expands at approximately 2.6 ppm/°C, lead-free solder alloys at roughly 21 to 25 ppm/°C, and organic substrates anywhere from 10 to 70 ppm/°C depending on material and orientation. The underfill must be formulated so that its expansion bridges this mismatch. Below the glass transition temperature (Tg), the coefficient of thermal expansion, designated α1, is typically targeted between 25 and 40 ppm/°C to match the solder joints and minimize shear strain during thermal excursions.
Above Tg, the cured network softens and the expansion rate rises steeply. The coefficient above the transition, designated α2, commonly reaches 80 to 200 ppm/°C, which is three to five times greater than α1. If a package operates or is processed near or above the underfill Tg, this accelerated expansion can produce die stressing, fillet cracking, interfacial delamination, and rapid solder fatigue. High-temperature expansion analysis therefore focuses on quantifying α1, α2, Tg, and total expansion across the complete service and processing temperature window.
- Solder joint fatigue driven by cyclic expansion mismatch during temperature cycling
- Underfill-to-die or underfill-to-substrate interfacial delamination
- Fillet cracking and moisture-induced popcorning during reflow soldering
- Package warpage, bulging, and loss of coplanarity at elevated temperature
- Cracking of low-k dielectric layers on advanced silicon nodes due to die stress
Testing Samples
Capillary Underfill (CUF)
Capillary underfills are low-viscosity liquids drawn into the die-to-substrate gap by capillary action, typically containing 50 to 70 weight percent silica filler. Bulk specimens are cast into molds or cured in shallow pans, then cut and surface-ground to achieve the flat, parallel faces required for reliable expansion measurement. The cure schedule used for specimen preparation must replicate the actual production profile to ensure the data reflect the true crosslink density.
No-Flow Underfill (NUF)
No-flow underfills are fluxing adhesives that cure simultaneously with solder reflow, so they are unfilled or only lightly filled to allow solder joint collapse. Because their properties are highly dependent on the reflow thermal history, expansion specimens must be cured using the same peak temperature and time-above-liquidus profile experienced by the assembly.
Molded Underfill (MUF)
Molded underfills combine encapsulation and underfill functions in a single highly filled molding compound applied by transfer or compression molding. Expansion evaluation is performed on molded plaques or discs, with attention to filler distribution and flow-induced orientation, which can cause anisotropic expansion behavior.
Film and Wafer-Level Underfill
B-staged film underfills are laminated at wafer level and fully cured during die attach. Specimens are prepared by laminating and curing films according to the supplier-recommended pressure, temperature, and time schedule, then stacked if necessary to reach the minimum testable thickness.
Specimen Conditioning
Cured bulk specimens are generally prepared at 2 to 10 mm thickness with parallel faces for thermomechanical analysis. Standard conditioning at 23 ± 2 °C and 50 ± 5% relative humidity for at least 24 hours, per ASTM D618 or ISO 291, is applied before testing. For hygrothermal studies, additional moisture preconditioning at 85 °C and 85% relative humidity is used to quantify the contribution of absorbed moisture to total expansion.
Testing Items
High-temperature expansion analysis of underfill adhesives typically encompasses the following measurable parameters, each of which maps directly to a specific reliability concern in the packaged assembly:
- Coefficient of thermal expansion below Tg (α1), the primary formulation control parameter, typically specified at 25 to 45 ppm/°C
- Coefficient of thermal expansion above Tg (α2), an indicator of high-temperature dimensional stability, typically 80 to 200 ppm/°C
- Glass transition temperature (Tg) determined from the expansion curve, commonly required to exceed 110 to 130 °C for consumer products and higher for automotive grades
- Total linear expansion percentage over defined spans, such as 25 to 100 °C for operation and 25 to 260 °C for lead-free reflow simulation
- Onset and extrapolated transition temperatures characterizing the breadth of the glass transition region
- Hygroscopic swelling contribution measured after moisture preconditioning at 85 °C and 85% relative humidity
- Package-level z-axis expansion and warpage behavior for fully assembled test vehicles
Acceptance criteria are usually established in material specifications and qualification guidelines. Typical requirements include α1 no greater than 40 ppm/°C, a controlled α2-to-α1 ratio, Tg above the maximum sustained operating temperature with adequate margin, and limited total expansion across the reflow profile to protect fragile low-k structures and maintain joint integrity.
Testing Methods
Thermomechanical Analysis (TMA) as the Primary Method
Thermomechanical analysis performed in accordance with ASTM E831 or ISO 11359-2 is the industry-standard technique for measuring high-temperature expansion of underfill adhesives. A flat specimen is placed on a quartz platform and contacted by an expansion probe applying a negligible static force, typically 0.01 to 0.1 N. The furnace is purged with nitrogen, and the specimen is heated at a controlled rate, most commonly 5 °C/min, over a range such as -50 °C to 250 or 300 °C so that both the sub-Tg and post-Tg regions are captured. A second heating run is frequently performed to remove the influence of thermal history and physical aging on the measured expansion behavior.
Data Reduction and Reporting
The coefficient of thermal expansion is calculated as CTE = ΔL / (L0 × ΔT), where L0 is the initial specimen length. Linear regression is applied to the quasi-linear segments of the expansion curve below and above the transition to extract α1 and α2, while Tg is determined from the inflection point of the dimensional-change curve or from the intersection of extrapolated tangents. Reports should include α1, α2, Tg, interval expansion percentages, heating rate, specimen orientation, and the number of replicates, with a minimum of three specimens recommended for statistical validity.
Complementary Techniques
- Dynamic mechanical analysis per ASTM E1640 or ISO 6721-11 to cross-verify Tg through mechanical stiffness change
- Differential scanning calorimetry per ASTM E1356 or ISO 11357-2 to confirm Tg and verify completeness of cure before expansion testing
- Fused-silica dilatometry per ASTM E228 to validate bulk expansion results on larger specimens
- Package-level thermomechanical profiling and optical warpage methods such as shadow moiré or digital image correlation for assembled units
- Temperature cycling per JEDEC JESD22-A104 and moisture sensitivity classification per JEDEC J-STD-020 to correlate expansion data with end-use reliability outcomes
Method Quality Control
Reliable results depend on rigorous quality control, including instrument length and temperature calibration using certified reference materials such as alumina or sapphire, verification of specimen face parallelism, control of preconditioning environment, and periodic verification with standard reference materials. Accredited third-party laboratories routinely apply these controls to ensure interlaboratory reproducibility, and procedures aligned with IPC-TM-650 Method 2.4.24 provide an established framework for z-axis expansion and Tg determination in electronic packaging materials.
Applications
Expansion analysis of underfill adhesives supports material selection, formulation development, qualification, and failure analysis across the entire advanced packaging ecosystem. The data are used to match underfill properties to die size, substrate type, solder alloy, and the thermal environment of the end product.
- Flip-chip ball grid array (FCBGA) packages for high-performance computing, where large die sizes amplify expansion mismatch stresses
- Chip-scale packages and wafer-level packages in mobile and handheld devices subjected to repeated reflow and drop events
- Fan-out wafer-level packaging and 3D stacked-die assemblies with through-silicon vias, where multi-layer expansion mismatch accumulates
- Automotive electronics qualified to AEC-Q100 Grade 1 or Grade 0 exposure profiles, requiring verified expansion stability up to 125 °C and beyond
- Power modules and high-brightness LED assemblies that sustain high junction temperatures for extended periods
- Stacked memory modules and advanced node devices employing fragile low-k interlayer dielectrics sensitive to compressive stress
FAQ
Why does underfill expansion increase sharply above the glass transition temperature?
Below Tg, the cured epoxy network is glassy and dimensionally rigid, so expansion is restrained primarily by the polymer backbone and filler. Above Tg, the network transitions to a rubbery state, molecular segments gain mobility, and the free volume increases rapidly, causing the expansion coefficient to jump to α2 values that are typically three to five times higher than α1.
What coefficient of thermal expansion values are considered acceptable for underfill adhesives?
Most specifications require α1 between approximately 25 and 40 ppm/°C so that the cured underfill expands in harmony with the solder joints rather than fighting against them. Above Tg, α2 values between 80 and 120 ppm/°C are generally preferred, and the Tg itself should sit comfortably above the maximum sustained operating temperature of the application.
How does absorbed moisture affect high-temperature expansion measurements?
Moisture absorbed by the epoxy matrix produces hygroscopic swelling that adds directly to thermal expansion, and during reflow the trapped moisture can vaporize and cause popcorning or delamination. Evaluations that precondition specimens at 85 °C and 85% relative humidity, with classification guidance from JEDEC J-STD-020, quantify this combined hygrothermal expansion behavior.
What heating rate and temperature range are used in TMA expansion testing?
A heating rate of 5 °C/min under nitrogen purge is the most widely applied condition under ASTM E831 and ISO 11359-2, with acceptable variation between 3 and 10 °C/min. The temperature range should extend from below the lowest service temperature, commonly -40 or -50 °C, up to 250 or 300 °C to fully capture α1, the transition region, and α2.
Can expansion testing be performed on actual packaged assemblies instead of bulk specimens?
Yes. Thermomechanical analysis can be applied directly to cross-sectioned or whole packages to measure z-axis expansion, and optical warpage techniques such as shadow moiré or digital image correlation map package deformation across temperature. These package-level measurements complement bulk material data and validate the interaction between underfill, die, and substrate in the real configuration.
How are expansion results linked to field reliability?
Expansion data feed into finite-element stress models and into accelerated test planning. Underfill exhibiting excessive α2 or low Tg will accumulate higher shear strain in solder joints during cycling per JEDEC JESD22-A104, shortening fatigue life, so expansion characterization is a proven early predictor of thermal cycling performance and a core element of material qualification programs conducted by accredited third-party laboratories.
