Substrate and Package Expansion Characteristics Testing: Methods, Standards, and Applications

Substrate and package expansion characteristics testing measures the coefficient of thermal expansion (CTE), glass transition behavior, and moisture-induced swelling of packaging substrates, molding compounds, and assembled packages. Using thermomechanical analysis (TMA), dilatometry, strain gauge methods, and dynamic warpage measurement in accordance with ASTM, ISO, IPC, JEDEC, and JIS standards, accredited third-party laboratories help semiconductor manufacturers prevent warpage, delamination, and solder joint failure caused by thermal expansion mismatch. This article explains the key test items, sample types, methods, and industrial applications of expansion characterization for electronic packaging materials.

Technical Overview

In semiconductor and electronics packaging, the substrate serves as the mechanical and electrical bridge between the silicon die and the printed circuit board. Because the package experiences wide temperature swings during reflow soldering, assembly, and field operation, the thermal expansion behavior of the substrate and surrounding packaging materials becomes a decisive reliability factor. When materials with dissimilar coefficients of thermal expansion (CTE) are joined, differential expansion generates thermomechanical stress at every interface, which can lead to solder joint fatigue, interconnect cracking, die attach failure, delamination of build-up layers, and package warpage.

A typical mismatch chain illustrates the problem: silicon expands at approximately 2.6 ppm/°C, copper at approximately 17 ppm/°C, ceramic substrates between 4 and 8 ppm/°C, while organic laminate substrates generally expand between 10 and 20 ppm/°C below the glass transition temperature and can exceed 40 ppm/°C above it. Epoxy molding compounds behave similarly, with a distinct jump in expansion rate after Tg. Characterizing these expansion properties accurately, both in-plane (X/Y) and through-thickness (Z-axis), is therefore essential for material selection, package design, finite element simulation, and reliability qualification.

Expansion characteristics testing covers the following core parameters:

• Coefficient of thermal expansion below the glass transition temperature (alpha 1, ppm/°C)
• Coefficient of thermal expansion above the glass transition temperature (alpha 2, ppm/°C)
• Glass transition temperature (Tg) determined from the expansion curve inflection point
• Total expansion percentage over a defined temperature range (for example, 50 to 260 °C)
• Hygroscopic swelling coefficient and moisture-induced dimensional change
• Dynamic warpage of packages as a function of temperature

Testing Samples

Organic Packaging Substrates

Bismaleimide triazine (BT) resin substrates, Ajinomoto-type build-up film substrates, high-Tg epoxy cores, modified epoxy glass fabrics, and polyimide flexible substrates used for BGA, flip-chip, and CSP packages. Samples may be tested as bare laminates, copper-clad panels, or etched substrates, and in-plane versus through-thickness orientations should be clearly distinguished because fiber reinforcement makes the laminate highly anisotropic.

Ceramic and Silicon Substrates

Alumina (Al2O3), aluminum nitride (AlN), silicon nitride, LTCC, and HTCC ceramic substrates used in power modules and RF packaging, as well as silicon interposers and wafer-level redistribution layers used in 2.5D and 3D integration. These materials exhibit low and nearly linear expansion, requiring high-precision dilatometry for accurate resolution.

Assembled Packages and Molding Materials

Finished BGA, LGA, QFN, and flip-chip packages for warpage and dimensional stability evaluation, plus epoxy molding compounds, underfill adhesives, die attach materials, and encapsulants in molded or cured plaque form for CTE and Tg measurement.

Copper-Clad Laminates and Raw Materials

CCL panels, prepreg sheets, build-up films, and dry film resist materials evaluated at the incoming inspection stage to verify supplier conformance to material specifications before lamination and substrate fabrication.

Testing Items

A comprehensive expansion characterization program typically includes the following test items:

• In-plane CTE (X and Y direction) over specified temperature ranges such as 25 to 125 °C or 25 to 150 °C
• Z-axis (through-thickness) CTE and total thickness expansion percentage from room temperature to reflow temperatures
• Alpha 1 and alpha 2 CTE segments separated by the Tg inflection point
• Glass transition temperature by TMA in accordance with ASTM E1545 or ISO 11359-2
• Thermal shrinkage and dimensional stability after thermal cycling or simulated reflow exposure
• Hygroscopic swelling coefficient (CME) after moisture preconditioning, expressed in ppm/% moisture uptake
• Time to delamination at 260 °C, 288 °C, and 300 °C (T260/T288/T300)
• Dynamic package warpage versus temperature profile, including maximum warpage at reflow peak
• Multilayer stack CTE matching assessment between core, build-up dielectric, and copper layers

Testing Methods

Thermomechanical Analysis (TMA)

TMA is the primary method for substrate and package expansion measurement. A low-mass probe rests on the specimen with a small static load while the furnace applies a controlled heating ramp, typically 5 °C/min in nitrogen, over ranges such as 25 to 300 °C. The dimension change versus temperature curve yields alpha 1, alpha 2, Tg, and total expansion. Governing standards include ASTM E831 for solid materials, ISO 11359-2, IPC-TM-650 method 2.4.24 for printed boards, and JIS K 7197 for plastics. Typical specimen sizes are 3 to 10 mm squares or strips, with thickness measured according to the applicable laminate test methods.

Push-Rod Dilatometry

For ceramic substrates, silicon interposers, and other low-expansion materials requiring superior resolution over wider temperature ranges, dilatometry per ASTM E228 provides displacement accuracy superior to TMA. A calibrated reference material such as sapphire or fused silica is run to correct system expansion, enabling confident measurement of expansion rates below 10 ppm/°C and elevated-temperature testing up to 1000 °C or beyond for ceramic and metal stack evaluations.

Strain Gauge (Resistance Change) Method

For in-plane CTE of full-size laminate panels, bonded metallized foil strain gauges measure resistance change with temperature. This approach, described in IPC-TM-650 method 2.4.24.1, evaluates representative production-size material rather than small coupons, capturing the effect of glass fabric weave and process history on X/Y expansion, and is widely used for copper-clad laminate qualification.

Dynamic Warpage Measurement

Package flatness as a function of temperature is measured using shadow moire interferometry, digital image correlation, or laser profilometry while the specimen passes through a simulated reflow profile, following JEDEC JESD22-B112 for temperature-dependent flatness measurement. Results identify the peak warpage temperature critical to solder ball collapse and coplanarity compliance with JEDEC JESD22-B106.

Moisture Sensitivity and Hygroscopic Swelling

Specimens are baked dry, then preconditioned at 85 °C and 85% relative humidity per JEDEC J-STD-020 moisture sensitivity level procedures. Dimensional changes measured by TMA before and after soaking, combined with gravimetric moisture uptake per IPC-TM-650 method 2.6.2, allow calculation of the coefficient of moisture expansion. This separates hygroscopic swelling from purely thermal expansion, which is essential for accurate warpage prediction in humid environments.

Time to Delamination and Reliability Correlation

TMA or dedicated thermal platforms hold specimens at isothermal 260 °C, 288 °C, or 300 °C until an expansion step indicates delamination, per IPC-TM-650 method 2.4.24.1. Results correlate with performance in temperature cycling per JEDEC JESD22-A104 (-40 to 125 °C or -55 to 125 °C) and solder joint reliability testing per IPC-9701, linking material expansion behavior to end-product durability.

Applications

Expansion characteristics testing supports the entire electronics packaging development and manufacturing lifecycle:

• Substrate material selection and CTE matching for flip-chip, BGA, and chip-scale packages to minimize die stress and interconnect fatigue
• 2.5D/3D advanced packaging development, where silicon interposer, build-up dielectric, and molding compound expansion must be balanced to control warpage
• Input material properties for finite element analysis and warpage simulation before tape-out and tooling commitments
• Power module design using ceramic direct-bonded copper substrates, where CTE grading between die, ceramic, and baseplate governs solder layer life
• Incoming quality control of copper-clad laminates, prepreg, and build-up films against procurement specifications
• Reliability qualification for automotive electronics per AEC-Q100 and industrial and consumer product certification programs
• Failure analysis of delamination, cracking, and open solder joints by comparing expansion data of failed versus conforming materials
• Process window optimization for lamination curing, reflow profiling, and conformal coating operations

FAQ

What CTE range is considered suitable for flip-chip substrates?

Most organic flip-chip substrates target an in-plane CTE of 10 to 17 ppm/°C to sit between silicon (about 2.6 ppm/°C) and FR-4 printed circuit boards (about 14 to 17 ppm/°C). The optimum value depends on die size, ball pitch, and mold compound selection, and is usually validated by warpage simulation and board-level temperature cycling data.

Why do alpha 1 and alpha 2 differ so much?

Below Tg, the polymer matrix is in a glassy state with restricted molecular mobility, producing low expansion. Above Tg, the matrix enters a rubbery state with substantially higher free volume increase, so the CTE can double or triple. This jump is why Z-axis expansion and package stress rise sharply during reflow, which peaks near or above Tg for many materials.

How does absorbed moisture affect the measurement?

Moisture plasticizes the resin, lowering the effective Tg and adding hygroscopic swelling that can be mistaken for thermal expansion. Best practice is to test both a baked-dry specimen and a moisture-preconditioned specimen, then determine the coefficient of moisture expansion separately so the two contributions can be modeled independently.

What sample sizes and preparation are required for TMA testing?

Through-thickness measurements typically use specimens of 3 to 6 mm square with parallel, burr-free surfaces, while in-plane measurements use strips of 6 to 10 mm length cut along and across the glass weave direction. Specimens must be flat, fully cured, and conditioned consistently, and etched samples are preferred when the copper contribution must be excluded.

Which method should be selected: TMA, dilatometry, or strain gauge?

TMA suits polymers, laminates, and molding compounds with moderate expansion rates and delivers CTE plus Tg from one run. Dilatometry is preferred for ceramics, silicon, and metals demanding higher precision or extreme temperatures. The strain gauge method best represents in-plane behavior of production-size panels. Accredited third-party laboratories often combine all three to give a complete anisotropic expansion picture for design simulation.

How do T260, T288, and T300 results relate to reliability?

These time-to-delamination values indicate how long the material stack survives at lead-free reflow temperatures before interfacial separation. Longer times reflect stronger adhesion and better resistance to moisture-driven popcorning during assembly, and they correlate with lower delamination risk during multiple reflow passes and rework operations.