Underfill Adhesive Low-Temperature Expansion Testing: Methods, Standards, and Reliability Assessment

Low-temperature expansion testing of underfill adhesives measures the coefficient of thermal expansion (CTE) and dimensional stability of cured encapsulants at subzero temperatures, typically from -70°C to room temperature. Using thermomechanical analysis (TMA) in accordance with ASTM E831, ISO 11359-2, and JIS K 7197, accredited third-party laboratories help semiconductor packagers verify that flip-chip and ball grid array (BGA) assemblies survive cold-environment service without solder joint cracking, delamination, or warpage. This article explains the test principle, specimen requirements, measured parameters, methodology, and typical applications of underfill adhesive low-temperature expansion evaluation.

Technical Overview

Underfill adhesives are thermosetting polymer materials, most commonly epoxy-based formulations filled with silica, that are dispensed or molded into the gap between a silicon die and its organic substrate in flip-chip, chip-scale package (CSP), and BGA assemblies. Once cured, the underfill mechanically couples the die to the substrate, redistributes thermomechanical stress away from fragile solder interconnects, and dramatically improves fatigue life under thermal cycling. The effectiveness of this stress redistribution depends directly on how closely the expansion behavior of the underfill matches the adjacent materials across the entire service temperature range.

Many electronic products must operate or survive at subzero temperatures: automotive electronics face -40°C cold-start and winter storage conditions, aerospace and defense systems are specified to -55°C or lower, and outdoor telecom infrastructure experiences severe seasonal cooling. At low temperatures, the mismatch between the low CTE of silicon (approximately 2.6 ppm/°C), the higher CTE of organic laminates (14-17 ppm/°C), and the intermediate CTE of the cured underfill generates shear and peel stresses at the bump level and at material interfaces. If the underfill contracts excessively or exhibits anomalous low-temperature behavior, solder joints crack, underfill-to-die or underfill-to-substrate interfaces delaminate, and package warpage becomes unacceptable.

Low-temperature expansion testing quantifies this risk by measuring the linear dimensional change and coefficient of thermal expansion of the cured adhesive from cryogenic or deep subzero temperatures up through the glass transition region. The primary measurement technique is thermomechanical analysis (TMA), performed per ASTM E831, ISO 11359-2, or JIS K 7197, supplemented by quartz dilatometry per ASTM D696 and cold-exposure conditioning per IEC 60068-2-1 and GB/T 2423.1. Testing is commonly carried out by accredited third-party laboratories as part of package qualification, material selection, incoming inspection, or failure analysis programs.

Testing Samples

Cured Bulk Underfill Specimens

The baseline specimen for TMA testing is a free-standing cured coupon of the underfill adhesive. Two-part or premixed frozen adhesives are degassed, cast into molds, and cured according to the adhesive supplier's recommended time-temperature profile, often followed by a post-cure to ensure full conversion. Specimens are typically cut and lapped to 2 to 10 mm in length with flat, parallel, and polished end faces perpendicular to the measurement axis. Uniform thickness and freedom from voids, bubbles, and embedded contaminants are essential, since entrapped air artificially inflates apparent expansion values.

Assembled Package Cross-Sections

To evaluate real product configurations, cross-sectioned flip-chip or BGA specimens containing the cured underfill between die and substrate can be prepared. Polished cross-sections allow the TMA probe to act on the composite stack, revealing effective expansion behavior of the actual material combination, while microscopy of sections before and after cold exposure reveals delamination, fillet cracking, and bump deformation.

Film and Preform Specimens

B-stage underfill films and preforms used in wafer-level and panel-level packaging are tested as cured laminates. These specimens are cured between release films under controlled pressure, then die-cut to the required geometry so that flow-induced filler orientation is captured in the measurement direction.

Conditioning States

Specimens are tested in at least two conditions: a dry as-cured state after desiccation, and a moisture-preconditioned state, for example after 85°C/85% RH exposure, because absorbed water plasticizes the epoxy network and can shift both the low-temperature expansion slope and the glass transition temperature. Standard laboratory atmospheres of 23°C and 50% relative humidity per ISO 291, with alternative conditioning per ASTM D618, are applied unless a specification dictates otherwise.

Testing Items

The low-temperature expansion test program for underfill adhesives typically reports the following parameters:

• Linear coefficient of thermal expansion below the glass transition (alpha-1) over subzero intervals such as -55°C to 25°C or -40°C to 23°C, expressed in ppm/°C
• Glass transition temperature (Tg) determined from the inflection in the expansion curve
• Linear coefficient of thermal expansion above Tg (alpha-2) for comparison and completeness
• Total linear expansion or contraction, delta L/L0, between room temperature and the lowest test temperature
• Dimensional change hysteresis between cooling and heating branches, indicating residual stress or physical aging
• CTE shift after moisture preconditioning and after simulated reflow exposure
• Dimensional stability and interface integrity after cold dwell per IEC 60068-2-1 or GB/T 2423.1, verified by microscopy or acoustic microscopy
• Sub-transition relaxation events, such as secondary beta transitions in the -80°C to -20°C region, which affect low-temperature toughness

Testing Methods

Thermomechanical Analysis (TMA)

TMA is the reference method. A flat-tipped quartz expansion probe resting on the specimen with a small static load tracks dimensional change while a furnace programmed with liquid nitrogen cooling sweeps the temperature. A typical low-temperature protocol equilibrates the specimen at -70°C, holds for thermal stabilization, then heats at 3 to 5°C/min under nitrogen purge to 150°C or beyond, comfortably past the Tg of most underfills. The instantaneous CTE at any temperature is the slope of the dimension-versus-temperature curve; alpha-1 and alpha-2 are computed by linear regression over user-defined temperature windows using a two-tangent intersection method for Tg. Calibration with a reference material of known expansion, such as fused silica or an aluminum standard, precedes each measurement campaign per ASTM E831 and ISO 11359-2.

Quartz Dilatometry

For larger coupons or when a classic averaging method is required, a quartz-tube dilatometer per ASTM D696 measures expansion over the standard -30°C to +30°C interval or extended subzero ranges. This method provides an independent check on TMA data and suits specimens with slight surface roughness.

Cold Exposure with Dimensional Metrology

Complementary to intrinsic CTE measurement, whole packages or cured plaques are subjected to steady-state cold per IEC 60068-2-1 or GB/T 2423.1 at -40°C or -55°C for a defined dwell, followed by measurement of warpage, fillet geometry, and interface integrity using profilometry, optical microscopy, and scanning acoustic microscopy. Combined with thermal cycling per IEC 60068-2-14, this validates that low-temperature contraction does not initiate delamination or solder damage.

Data Interpretation and Reporting

The test report identifies the adhesive lot, mix ratio, cure schedule, conditioning history, specimen geometry, test method and standard, temperature program, alpha-1 and alpha-2 values with regression confidence, Tg, expansion curves, and any observed anomalies such as step changes, cracks, or probe slippage. Acceptance is typically judged against the package design window, in which the underfill alpha-1 should fall between the die and laminate CTE values, commonly 25 to 45 ppm/°C for highly filled snap-cure and no-flow underfills.

Applications

Low-temperature expansion testing of underfill adhesives supports reliability engineering across the electronics industry:

• Automotive electronic control units, ADAS sensor modules, and EV power electronics qualified for -40°C cold-crank and winter operation
• Aerospace and defense avionics specified to -55°C and below, where mission-critical solder integrity is mandatory
• Outdoor 5G base stations, microwave radio units, and telecom infrastructure exposed to seasonal subzero environments
• Industrial controllers, energy storage systems, and consumer devices shipped or operated in cold climates
• Flip-chip, CSP, fine-pitch BGA, wafer-level, fan-out, and 2.5D/3D advanced packaging qualification programs
• Adhesive supplier research and development for new formulations, filler loading optimization, and incoming material lot control
• Failure analysis, where anomalous low-temperature expansion is correlated with field cracking or delamination complaints

FAQ

Why does low-temperature expansion behavior matter more than room-temperature CTE for underfills?

Polymers are stiffest and most brittle at low temperatures, and the CTE mismatch stress between silicon, underfill, and laminate peaks during cold excursions. A material with acceptable room-temperature CTE can still over-contract at -40°C, transferring damaging stress to solder bumps and interfaces precisely when the adhesive has the least capacity to relax it.

What temperature range is used in the test?

Most programs sweep from -70°C or -55°C up to 150°C to 250°C in a single TMA run, capturing alpha-1 in the subzero region, the glass transition, and alpha-2 above it. The reported subzero window, such as -55°C to 25°C or -40°C to 23°C, is selected to match the product's specified operating envelope.

What CTE values are considered acceptable below Tg?

Highly filled capillary-flow underfills typically target 25 to 40 ppm/°C, while no-flow and fast-cure formulations may reach 45 to 60 ppm/°C. Acceptance depends on die size, bump pitch, substrate CTE, and the qualification thermal cycle, so values are judged against the specific package design margin rather than a universal limit.

How does moisture affect the results?

Absorbed moisture lowers the effective Tg and can alter the subzero expansion slope through water-phase transitions within the polymer. Testing both dry and 85°C/85% RH preconditioned specimens reveals the worst-case condition for field reliability.

Can the test be performed on actual assembled packages instead of bulk coupons?

Yes. Cross-sectioned assemblies can be measured to obtain effective composite expansion, and whole packages can be cold-conditioned per IEC 60068-2-1 with before-and-after warpage and acoustic microscopy to detect delamination, giving a product-level complement to the intrinsic material data.

How are alpha-1 and alpha-2 distinguished in reporting?

Alpha-1 is the linear regression slope of the expansion curve below the glass transition, and alpha-2 is the steeper slope above it. The TMA software locates Tg at the intersection of the two tangent lines, and both slopes are reported with the temperature windows over which they were calculated.