IC Package Substrate Expansion Testing (CTE): Standards, Methods, and Reliability Applications

IC package substrate expansion testing measures the coefficient of thermal expansion (CTE) of package substrates, a critical parameter that governs warpage behavior, solder joint reliability, and long-term interconnect durability in semiconductor packaging. Using thermomechanical analysis (TMA) in accordance with IPC-TM-650 and ASTM E831, this test characterizes dimensional stability across the full operating and reflow temperature range of electronic assemblies. This article explains the test principle, sample preparation, measured parameters, testing procedures, and industry applications in detail.

Technical Overview

The IC package substrate is the electrical and mechanical interconnect platform between a semiconductor die and the printed circuit board (PCB). As a multilayer laminate composed of resin, glass fabric, and copper wiring, the substrate expands and contracts with temperature, and this dimensional change must be precisely controlled and characterized. Expansion testing—most commonly performed by thermomechanical analysis (TMA)—quantifies the coefficient of thermal expansion (CTE) of the substrate in the X, Y, and Z axes, and simultaneously identifies the glass transition temperature (Tg) at which the expansion behavior of the resin system changes abruptly.

The importance of CTE matching stems from the steep expansion gradient across a packaged assembly. A silicon die expands at approximately 2.6 to 3 ppm/°C, while a typical PCB laminate expands at 14 to 17 ppm/°C, and the package substrate sits between these two extremes. Any mismatch in expansion generates thermomechanical stress during reflow soldering and thermal cycling, which can drive solder joint fatigue, underfill delamination, low-k dielectric cracking in advanced nodes, and copper-plated via failure. In particular, Z-axis (thickness direction) expansion strains plated through-holes and microvias, making it one of the most closely watched reliability parameters in substrate qualification.

Testing Samples

Z-Axis (Thickness Direction) Samples

Z-axis expansion is measured on small coupons cut from finished substrates or raw laminate panels, typically 3 to 6 mm square. Because a single substrate layer may be too thin to produce a reliable signal, multiple coupons are often stacked to achieve a total thickness of 0.5 to 3 mm. Both faces of the sample must be flat, parallel, and free of copper features, solder mask, or burrs that would distort the probe contact and produce artificial displacement readings.

X-Axis and Y-Axis (In-Plane) Samples

In-plane expansion is measured on rectangular strips, commonly 6 to 10 mm long, 2 to 3 mm wide, and 0.1 to 1.5 mm thick. Because glass-fabric-reinforced substrates are anisotropic, separate strips must be cut along the machine direction (MD) and the transverse direction (TD) of the panel to capture the full in-plane expansion behavior. Edges should be cut cleanly, since rough or angled edges introduce measurement scatter.

Sample Conditioning and Preparation

• Moisture removal: samples are typically baked at 105 °C for 2 hours (or per the applicable material specification) to eliminate absorbed moisture, which would otherwise plasticize the resin and shift the Tg and CTE readings.
• Conditioning: after baking, samples are stabilized at 23 ± 2 °C and 50 ± 5 % relative humidity for at least 24 hours unless the specification requires an as-received condition.
• Dimensional recording: initial length or thickness is measured accurately at room temperature using a calibrated micrometer or the TMA itself, as this dimension is the reference for all CTE calculations.
• Stress relief: samples must be handled without bending, and any specimens showing delamination, warpage, or cracks are rejected before testing.

Testing Items

A complete expansion test on an IC package substrate characterizes the following parameters:

• CTE below Tg (alpha 1): the linear expansion coefficient in the glassy region, expressed in ppm/°C, which governs behavior during normal device operation.
• CTE above Tg (alpha 2): the expansion coefficient in the rubbery region, typically two to five times higher than alpha 1, which governs behavior during reflow and high-temperature processes.
• Glass transition temperature (Tg): identified from the inflection point of the dimension-versus-temperature curve, usually reported as the extrapolated onset temperature.
• Total expansion: the cumulative dimensional change across a defined range, such as 30 °C to 260 °C, expressed as a percentage of the initial dimension.
• In-plane anisotropy: the difference between MD and TD expansion, an indicator of glass cloth style and lamination quality.
• Delta CTE: the ratio or difference between Z-axis and in-plane expansion, used in warpage modeling and stress simulation.

Testing Methods

Thermomechanical Analysis (TMA) Principle

TMA is the industry-standard technique for substrate expansion testing. A flat-tipped quartz or silica probe rests on the sample under a small static force (typically 5 to 20 mN), and a linear variable displacement transducer continuously records the change in sample dimension while the furnace heats the specimen under a nitrogen purge of 20 to 50 mL/min. The resulting dimension-versus-temperature curve yields the CTE slopes and the Tg inflection directly.

Test Procedure

• Measure and record the initial sample thickness or length at ambient conditions.
• Place the sample on the flat quartz stage and apply the expansion probe with the specified static force, ensuring full and stable contact.
• Equilibrate at the starting temperature (typically 25 to 30 °C) for 5 to 10 minutes, then heat at a constant rate—5 °C/min per IPC-TM-650, or 10 °C/min where the material specification allows.
• Ramp to the end temperature, typically 260 °C, 280 °C, or 300 °C, depending on the resin system and applicable specification sheet.
• Optionally perform a second heating run to erase thermal history and confirm repeatability, reporting data from the run specified by the standard.
• Calculate alpha 1 as the slope of the linear region below Tg (for example, 50 to 150 °C), alpha 2 as the slope above Tg (for example, Tg + 20 °C to the end temperature), and Tg as the intersection of the extrapolated alpha 1 and alpha 2 slope lines.

Calculation Method

The CTE is calculated as the dimensional change normalized to the initial dimension and divided by the temperature interval, expressed in ppm/°C. For a sample of initial length L0 undergoing a change of delta L over a temperature interval delta T, CTE equals (delta L / L0) divided by delta T, multiplied by one million. All calculations are performed on the linear portion of the curve, and regression analysis is used to minimize operator judgment.

Applicable Standards

• IPC-TM-650 method 2.4.24: TMA determination of CTE and Tg for copper-clad laminates and package substrate materials.
• ASTM E831: standard test method for linear thermal expansion of solid materials by TMA.
• ASTM E1545: determination of Tg by TMA, used to confirm the transition temperature reported alongside the CTE values.
• IPC-4101 and IPC-4103: material specifications that define the CTE and Tg acceptance limits for laminate and substrate grades.

Typical Reference Values

• In-plane CTE (X/Y) of reinforced core substrates: approximately 6 to 17 ppm/°C below Tg.
• Z-axis alpha 1 of build-up dielectric layers: approximately 20 to 50 ppm/°C.
• Z-axis alpha 2 above Tg: approximately 100 to 250 ppm/°C.
• Tg by TMA: typically 150 °C to 260 °C depending on the resin system, with higher values generally improving reflow stability.

These ranges are indicative only; contractual acceptance criteria are always defined by the customer specification sheet or the relevant IPC slash sheet for the specific material grade.

Applications

• Material qualification: verifying that new core materials, build-up films, and prepregs meet CTE and Tg requirements before release to substrate fabrication lines.
• Incoming inspection: screening supplier lots of raw laminate and dielectric film against specification limits to prevent reliability escapes.
• Warpage control: supplying accurate CTE data for finite element simulation of package warpage across the die attach, molding, and reflow processes.
• Reliability prediction: correlating substrate expansion with solder joint fatigue life, microvia cracking, and plated through-hole integrity in thermal cycling from -40 °C to 125 °C or wider ranges.
• Advanced packaging development: matching substrate expansion with silicon interposers and molding compounds in 2.5D and 3D architectures, flip-chip ball grid arrays, and chip-scale packages.
• Process troubleshooting: detecting under-cure, post-mold shift, or moisture-related degradation by comparing expansion curves against a known-good baseline.

FAQ

Why is Z-axis CTE more critical than in-plane CTE for package substrates?

Z-axis expansion directly strains copper-plated through-holes and microvias during reflow and thermal cycling. Excessive Z-axis expansion causes barrel cracking, corner cracking of via structures, and intermittent opens, whereas in-plane CTE primarily influences die-to-substrate stress and package warpage. Both must be controlled, but Z-axis failure modes are usually the first reliability limit reached.

What is the difference between alpha 1 and alpha 2?

Alpha 1 is the CTE below the glass transition temperature, where the resin is rigid and expansion is low; it dominates during normal device operation. Alpha 2 is the CTE above Tg, where the softened resin expands two to five times faster; it dominates during reflow and high-temperature processing. A large jump between alpha 1 and alpha 2 indicates higher risk of via damage during thermal excursions.

What temperature range is used for the test?

Most specifications heat from 30 °C to 260 °C or 280 °C at 5 °C/min, which covers normal operation through lead-free reflow temperatures. Materials intended for extreme applications may be evaluated over wider ranges as defined by the customer specification.

How does moisture affect the expansion test results?

Absorbed moisture plasticizes the resin system, lowering the measured Tg and distorting both CTE slopes. Samples are therefore baked before testing, and conditioning requirements are strictly controlled to ensure reproducible results.

Can dilatometry be used instead of TMA?

Push-rod dilatometers can measure bulk expansion, but TMA is strongly preferred for IC package substrates because it accommodates the small, thin coupons typical of substrate materials, applies precise and minimal probe force, and provides superior resolution for Tg detection on the same run.

How long does a single expansion test take?

A complete test, including loading, equilibration, and a heating ramp from 30 °C to 260 °C at 5 °C/min, typically takes 60 to 90 minutes per sample, with additional time required if a second heating run or replicate specimens are specified.