Precise Determination of the Coefficient of Thermal Expansion (CTE) of Substrates: Principles, Metho

The coefficient of thermal expansion (CTE) is one of the most critical thermomechanical properties of electronic substrates, directly governing solder joint reliability, plated through-hole integrity, and package warpage behavior. Precise CTE determination is typically performed using thermomechanical analysis (TMA) or dilatometry in accordance with internationally recognized standards such as IPC-TM-650, ASTM E831, ASTM E228, and ISO 11359-2. This article provides a comprehensive technical overview of substrate CTE measurement, covering sample types, test items, measurement methods, industrial applications, and frequently asked questions.

Technical Overview

The coefficient of thermal expansion (CTE) quantifies the fractional dimensional change of a material per unit temperature change, defined as α = (1/L₀) × (dL/dT), and is conventionally expressed in parts per million per degree Celsius (ppm/°C). For substrates used in printed circuit boards (PCBs) and integrated circuit (IC) packaging, CTE is a first-order design parameter because thermomechanical mismatch between the substrate, silicon die (approximately 2.6 ppm/°C), and copper conductors (approximately 17 ppm/°C) drives stress accumulation during thermal cycling. Excessive CTE mismatch accelerates solder joint fatigue, plated through-hole cracking, copper foil delamination, and interfacial failure in build-up layers.

Organic substrates such as FR-4 laminates are fiber-reinforced composites and therefore exhibit pronounced anisotropic expansion behavior. The Z-axis (through-thickness) CTE, which is dominated by the resin system, typically ranges from 40 to 70 ppm/°C below the glass transition temperature (Tg) and can rise to 200–400 ppm/°C above Tg. In contrast, the in-plane (X–Y) CTE is constrained by the glass fabric and typically falls between 10 and 20 ppm/°C. For this reason, precise CTE characterization must clearly distinguish between Z-axis and in-plane values, and between the pre-Tg expansion rate (α₁) and the post-Tg expansion rate (α₂). Ceramic substrates such as alumina (6–7 ppm/°C), aluminum nitride (4–5 ppm/°C), and silicon nitride (approximately 3 ppm/°C) are nearly isotropic but demand higher measurement resolution because their absolute expansion is small.

Accurate CTE determination requires rigorous control of instrumentation, calibration, sample preparation, conditioning, and thermal profiling. Accredited third-party laboratories performing CTE measurements routinely comply with ISO/IEC 17025 quality management requirements, ensuring traceable calibration and defensible measurement uncertainty for materials qualification, failure analysis, and finite element simulation inputs.

Testing Samples

Organic Rigid Laminates and Copper-Clad Laminates

This category covers FR-4, CEM-1/CEM-3, high-Tg epoxy, modified epoxy, BT (bismaleimide triazine), polyimide, PPO-based, and low-CTE engineered laminates. Copper cladding is normally removed by chemical etching prior to Z-axis measurement so that the expansion of the dielectric composite itself is characterized. Specimens are cut from panels at representative locations and conditioned to a stable moisture equilibrium before testing.

Ceramic Substrates

Ceramic substrates include alumina (Al₂O₃), aluminum nitride (AlN), silicon nitride (Si₃N₄), and co-fired multilayer structures in both thick-film and thin-film formats. Because these materials have low absolute expansion values, they require high-resolution instrumentation, careful baseline correction, and calibration against certified reference materials such as sapphire or fused silica.

Flexible Substrates and Polyimide Films

Flexible circuit base films, adhesive-coated polyimide films, adhesiveless cast laminates, and coverlay materials constitute another important sample group. Polyimide films typically exhibit in-plane CTE values near 20 ppm/°C, and measurement frequently requires film-extension fixtures or non-contact optical techniques instead of conventional compression probes.

Metal-Core and Metal-Based Substrates

Metal core PCBs based on aluminum (approximately 23 ppm/°C), copper (approximately 17 ppm/°C), and insulated metal substrate (IMS) constructions require CTE evaluation of both the metal base and the composite structure to predict thermal cycling durability in power electronics applications.

Glass and Semiconductor Substrates

Glass-core substrates, glass interposers, silicon wafers, and silicon interposers used in advanced 2.5D/3D packaging feature ultra-low CTE values in the range of 0.5–9 ppm/°C. Interferometric or optical dilatometric methods are often necessary to resolve such small dimensional changes with acceptable uncertainty.

Testing Items

• Z-axis (through-thickness) CTE below the glass transition temperature (α₁), typically reported over 30–50 °C up to the onset of Tg.
• Z-axis CTE above the glass transition temperature (α₂), captured from the steeper expansion slope above Tg.
• Glass transition temperature (Tg) determined by TMA, identified as the intersection point of the α₁ and α₂ expansion segments.
• In-plane (X–Y) CTE measured on flat specimens using horizontal dilatometry, strain gauge methods, or optical techniques.
• Full dimensional change (ΔL/L₀) versus temperature curves, including total expansion percentage over specified temperature windows.
• CTE comparison of materials before and after moisture preconditioning, thermal cycling, or simulated reflow exposure.
• CTE characterization of cured prepreg, resin-only samples, and build-up dielectric layers for packaging substrate development.

Testing Methods

Thermomechanical Analysis (TMA)

TMA is the industry-preferred method for Z-axis CTE measurement of organic substrates, performed in accordance with IPC-TM-650 method 2.4.24, IPC-TM-650 method 2.4.24.1, ASTM E831, and ISO 11359-2. A flat-ended quartz probe rests on the specimen surface under a small preload while the furnace applies a linear temperature ramp, typically 5 °C/min or 10 °C/min, under a nitrogen purge. CTE is calculated from the slope of the expansion curve, with α₁, α₂, and Tg extracted in a single run. Typical specimens measure 4–6 mm square with thickness between 0.5 mm and 3 mm; specimens are often stacked to achieve adequate thickness for reliable Z-axis resolution.

Push-Rod Dilatometry

Vertical or horizontal dilatometry per ASTM E228 employs longer specimens, typically 10–50 mm, making it particularly suitable for in-plane CTE measurement and for ceramic, glass, and metal substrates. The technique offers excellent repeatability for low-expansion materials when combined with differential measurement and reference-material baseline subtraction.

Interferometric Dilatometry

Optical interferometric methods conforming to ASTM E289 provide displacement resolution at the nanometer or sub-nanometer level, enabling measurement uncertainty below 0.1 ppm/°C. These methods are indispensable for ultra-low-CTE materials such as glass cores, silicon, and glass-ceramic composites.

Non-Contact Optical and Strain-Based Methods

Digital image correlation (DIC), moiré interferometry, and high-temperature strain gauges capture full-field in-plane strain on substrates subjected to controlled heating. These methods are valuable for anisotropic materials, large panels, and warpage-related studies where local CTE variation must be mapped.

Calibration, Conditioning, and Quality Control

All measurement campaigns rely on instrument calibration and verification using certified reference materials such as fused silica, sapphire, and aluminum oxide standards. Specimens are conditioned per IPC-TM-650 practice, commonly at 23 ± 2 °C and 50 ± 5 % relative humidity, and pre-drying or moisture sensitivity preconditioning is applied when project requirements dictate. Measurement uncertainty is typically ±5 % of the reported value or ±1 ppm/°C, whichever is greater, and is documented in test reports issued under ISO/IEC 17025-accredited quality systems.

Applications

• Incoming material inspection and quality control of copper-clad laminates, prepreg, and packaging substrates at PCB fabrication facilities.
• Package design and material selection for BGA, flip-chip, and chip-scale packages, where matching substrate CTE to die and mold compound mitigates warpage.
• Reliability engineering inputs for solder joint fatigue life prediction using Coffin–Manson-type models and thermal cycling test planning.
• Plated through-hole and microvia integrity assessment, since Z-axis CTE above Tg is a primary driver of barrel cracking during reflow and rework.
• Finite element analysis (FEA) and warpage simulation, where accurate temperature-dependent CTE curves materially improve model correlation.
• Automotive, aerospace, and 5G/RF power electronics qualification programs requiring stable dimensions across wide operating temperature ranges.
• Failure analysis and supplier benchmarking of competing laminate systems, ceramic substrates, and build-up dielectric materials.

FAQ

What is the coefficient of thermal expansion and what units are used?

The CTE describes how much a material expands or contracts per degree of temperature change, normalized to its original dimension. It is expressed in ppm/°C (equivalent to 10⁻⁶/°C) for most substrate applications, and sometimes in µm/(m·°C).

What is the difference between Z-axis and in-plane CTE?

Z-axis CTE measures expansion through the thickness of the substrate, dominated by the resin system, while in-plane CTE measures expansion along the length and width, constrained by glass fabric or ceramic structure. In FR-4, Z-axis α₁ may be 40–70 ppm/°C whereas in-plane CTE is often 10–20 ppm/°C, so the distinction is essential for both reliability analysis and simulation.

What do α₁ and α₂ represent?

α₁ is the expansion coefficient below the glass transition temperature, and α₂ is the steeper coefficient above Tg where the resin softens. The transition between the two slopes defines the TMA-based Tg, and α₂ is widely used as a key reliability indicator because peak reflow temperatures usually exceed Tg.

Which standards govern substrate CTE measurement?

Widely applied standards include IPC-TM-650 method 2.4.24 for Z-axis CTE of PCB laminates by TMA, ASTM E831 for solid materials by TMA, ASTM E228 for dilatometry, ASTM E289 for interferometric methods, ISO 11359-2 for thermomechanical analysis of plastics, and IEC 61189 series methods for interconnection structure materials.

What specimen dimensions are required for TMA testing?

Typical TMA specimens are 4–6 mm square with a thickness of 0.5–3 mm; thin laminates are often stacked to reach sufficient thickness. Dilatometry generally requires longer specimens, typically 10 mm or more in the measurement direction. Edges should be flat, parallel, and free of burrs to ensure uniform probe contact.

What temperature range and heating rate are used?

For organic substrates, testing commonly runs from room temperature to 260 °C or 300 °C to capture α₁, Tg, and α₂ in a single scan, using heating rates of 5 °C/min or 10 °C/min under nitrogen. Ceramic and glass substrates may be evaluated over wider ranges, for example −50 °C to 500 °C, depending on the application profile.

How accurate are CTE measurements?

With proper calibration, baseline correction, and conditioned specimens, typical uncertainty is approximately ±5 % of the reading or ±1 ppm/°C, whichever is larger. Ultra-low-expansion materials measured interferometrically can achieve uncertainties below 0.1 ppm/°C. Reports from accredited third-party laboratories include measurement conditions, calculation windows, and stated uncertainties.

When should TMA be chosen over dilatometry?

TMA is preferred for Z-axis CTE, α₂, and simultaneous Tg determination on organic laminates with limited specimen thickness. Dilatometry is preferred when longer specimens are available, when in-plane CTE is needed, or when very low-expansion ceramics and glasses require high-precision push-rod or optical detection.