Coefficient of Thermal Expansion Analysis of PET Ceramic Substrates: Testing Methods, Standards, and

The coefficient of thermal expansion (CTE) is one of the most critical performance parameters for PET ceramic substrates, directly determining dimensional stability, interconnect reliability, and long-term durability in electronic packaging applications. This article provides a comprehensive technical analysis of CTE behavior in PET-based ceramic substrates, covering material fundamentals, sample types, standardized testing methods such as TMA and dilatometry, and practical guidelines for interpreting test results.

Technical Overview

PET ceramic substrates are composite materials that combine a polyethylene terephthalate (PET) polymer phase with ceramic constituents, either in the form of dispersed ceramic fillers within the PET matrix or as laminated structures in which PET films are bonded to ceramic layers. These materials are widely used in flexible printed circuits, membrane switches, LED packaging, and hybrid electronic modules because they offer a favorable balance of flexibility, dielectric performance, thermal conductivity, and cost efficiency.

The coefficient of thermal expansion (CTE), typically denoted as α and expressed in ppm/°C, quantifies the fractional dimensional change of a material per unit temperature change. For PET ceramic substrates, CTE control is essential for three reasons. First, the polymer phase exhibits a relatively high intrinsic CTE of approximately 20 to 25 ppm/°C below its glass transition, which increases sharply above the transition region. Second, ceramic fillers such as alumina, aluminum nitride, or silica possess significantly lower expansion coefficients, typically between 0.5 and 8 ppm/°C, creating a strong compositional dependence of the composite CTE. Third, in assembled devices, thermal cycling between ambient and operating temperatures generates cyclic thermomechanical stresses at interfaces with silicon chips, copper traces, and solder joints, and CTE mismatch is a leading cause of delamination, solder fatigue, and microcracking.

Because the CTE of PET ceramic substrates is highly sensitive to filler loading, filler morphology, lamination architecture, and temperature range, accurate and standardized CTE measurement is an indispensable part of material qualification, incoming inspection, failure analysis, and new product development in the electronics supply chain.

Testing Samples

Ceramic-Filled PET Composite Substrates

These samples consist of PET matrices loaded with ceramic powders, commonly alumina (Al2O3), aluminum nitride (AlN), fused silica, or barium titanate, at loading levels ranging from 20 to 70 percent by weight. Specimens are typically supplied as flexible sheets or panels with thicknesses between 0.1 mm and 1.0 mm. Test coupons should be cut along both the machine direction (MD) and transverse direction (TD) to capture any anisotropy introduced during film extrusion or calendering.

PET Film Laminated Ceramic Substrates

These samples comprise rigid ceramic substrates, such as 96 percent or 99.6 percent alumina plates with thicknesses of 0.25 mm to 1.0 mm, laminated or adhesive-bonded with PET insulating or cover films. Because the CTE of the laminate differs through the thickness compared with the in-plane directions, specimens must be prepared for both in-plane and z-axis expansion measurement. Edge delamination and interfacial integrity should be verified under magnification prior to testing.

Metallized PET Ceramic Substrates

Samples carrying sputtered, electroplated, or thick-film printed copper, silver, or nickel metallization represent the as-fabricated condition of many commercial products. The metallization layer modifies the effective CTE of the composite stack and influences residual stress. Specimens should retain the full multilayer construction, and the effective expansion of the complete stack should be measured rather than the uncoated base material alone.

Sample Preparation and Conditioning

• Cut specimens to the dimensions required by the selected method, commonly 5 mm to 10 mm in length for TMA and 20 mm to 50 mm for dilatometry or strain-gauge techniques.
• Ensure the two end faces of each specimen are flat, parallel, and perpendicular to the measurement axis, with a parallelism tolerance of better than 10 micrometers.
• Condition specimens at 23 ± 2 °C and 50 ± 5 percent relative humidity for a minimum of 24 hours, or dry to constant mass when moisture effects are to be excluded.
• Mark and record the measurement direction relative to the machine direction, filler orientation, and lamination layup.
• Measure initial dimensions with a calibrated micrometer or optical measurement system at the conditioning temperature before testing.

Testing Items

The complete CTE characterization program for PET ceramic substrates typically includes the following measured and derived parameters:

• Mean linear CTE below the glass transition (α1), measured over a specified low-temperature interval such as 25 °C to 60 °C.
• Mean linear CTE above the glass transition (α2), measured over a specified high-temperature interval such as 90 °C to 150 °C, where the polymer phase exhibits rubbery-state expansion.
• Glass transition temperature (Tg) determined from the inflection point of the expansion curve, since PET transitions typically occur near 75 °C to 85 °C.
• In-plane CTE anisotropy ratio between machine direction and transverse direction.
• Through-thickness (z-axis) CTE for laminated constructions, which is frequently two to three times higher than the in-plane value.
• Reversible dimensional change after thermal cycling, quantifying hysteresis between heating and cooling expansion curves.
• Residual permanent linear change after exposure to the maximum test temperature, indicating irreversible shrinkage or stress relief.

Testing Methods

Thermomechanical Analysis (TMA)

TMA is the preferred method for polymer-based substrates with thicknesses between 0.1 mm and 3 mm and is performed in accordance with ASTM E831 or ISO 11359-2. A specimen is placed on a quartz platform and contacted by a low-mass flat-tipped expansion probe, typically applying 0.05 N to 0.2 N of force. The furnace heats the specimen at a controlled rate of 3 °C/min to 5 °C/min over a range such as -50 °C to 200 °C under a nitrogen purge of 20 mL/min to 50 mL/min. The linear displacement signal is recorded continuously, and the mean CTE is calculated from the slope of the dimension-versus-temperature curve over defined intervals. TMA requires only small specimens, delivers excellent resolution of the glass transition, and permits separate determination of α1 and α2, but probe pressure and specimen flatness must be controlled carefully for thin flexible films.

Push-Rod Dilatometry

Horizontal or vertical push-rod dilatometry, conducted per ASTM E228 or ISO 11359-1, measures the expansion of larger rigid specimens with lengths between 20 mm and 50 mm. The specimen pushes against a transducer through a fused silica or alumina push rod inside a temperature-controlled furnace. Dilatometry offers superior accuracy for low-expansion ceramic-rich samples and for reference-grade measurements, with measurement uncertainty typically below 0.1 ppm/°C, but it is less suitable for very thin, flexible, or soft polymer-rich materials. A system correction run using a certified reference material such as fused silica, sapphire, or platinum must be performed to compensate for apparatus expansion.

Strain Gauge and Optical Non-Contact Methods

Bonded high-temperature strain gauges following the practice of ASTM E289 or IPC-TM-650 method 2.4.24 style procedures are used for in-plane CTE evaluation of substrate coupons matching real circuit geometries. Non-contact optical techniques, including digital image correlation and laser interferometry, enable full-field expansion mapping and are particularly valuable for quantifying local CTE variations around filled regions, drilled holes, and metallization boundaries. These methods are increasingly applied during failure analysis of warped or cracked assemblies.

Procedure and Data Reduction Essentials

• Perform at least two thermal cycles; discard the first heating run if irreversible shrinkage from residual stress or moisture is observed, and report the second heating cycle as the stabilized CTE.
• Calculate the mean CTE as the slope of a least-squares linear regression of dimensional change versus temperature over each defined interval, normalized by the initial specimen length.
• Select temperature intervals that avoid the glass transition region itself, because curvature in this zone makes linear regression invalid.
• Report the test method, temperature intervals, heating rate, atmosphere, specimen orientation, and the number of replicates together with mean values and standard deviations, since CTE results are method- and interval-dependent.
• Verify measurement system suitability using certified reference materials traceable to national standards before each test campaign.

Typical Reference Values

• Unfilled PET film: approximately 20 to 25 ppm/°C below Tg and 60 to 120 ppm/°C above Tg.
• Alumina ceramic layers: approximately 6.5 to 7.2 ppm/°C between 25 °C and 300 °C.
• Aluminum nitride ceramic layers: approximately 4.2 to 5.0 ppm/°C.
• Ceramic-filled PET composites at 40 to 60 percent filler loading: typically 15 to 35 ppm/°C in-plane below Tg, depending on filler content and adhesion at the polymer-filler interface.
• Copper conductor: approximately 16.5 to 17.5 ppm/°C; silicon die: approximately 2.6 ppm/°C, both relevant as benchmark values for mismatch assessment.

Applications

CTE analysis data for PET ceramic substrates directly support design and qualification decisions across multiple electronic industry sectors:

• Flexible printed circuits and membrane switches, where CTE matching with copper foil prevents trace cracking and peel-strength degradation during reflow and thermal cycling.
• LED packaging and lighting modules, where controlled substrate expansion maintains phosphor layer integrity, solder joint life, and optical alignment over thousands of power cycles.
• Automotive electronics, including engine control units and sensor assemblies, where substrates must survive temperature cycling between -40 °C and 125 °C or higher in accordance with AEC-Q100 style qualification regimes.
• Power electronics and insulated gate bipolar transistor modules, where low-CTE ceramic layers paired with PET insulation balance thermal conduction with mechanical reliability under repeated heat pulses.
• High-frequency and 5G antenna substrates, where dimensional stability governs impedance consistency and antenna resonant frequency tolerances across operating temperatures.
• Materials research and development, where measured CTE versus filler loading relationships guide formulation optimization and validate micromechanical composite models such as the Turner and Kerner equations.

FAQ

What is a typical CTE value for a PET ceramic substrate?

In-plane CTE values below the glass transition generally fall between 15 and 35 ppm/°C for filled composites, while PET film laminated on alumina can approach effective values near 10 ppm/°C in the plane because the stiff ceramic layer constrains the polymer. Above the PET glass transition near 80 °C, the CTE can double or triple, so the operating temperature window of the end product must always be considered.

How does ceramic filler content influence the CTE?

Increasing ceramic loading progressively lowers the composite CTE because the low-expansion filler dilutes and mechanically constrains the high-expansion PET matrix. Improvement is nonlinear and depends on filler particle size distribution, aspect ratio, dispersion quality, and interfacial adhesion; poor adhesion can even increase apparent expansion due to debonding.

Why do CTE results differ between TMA and dilatometry?

Differences arise from probe force effects on soft specimens, different temperature intervals and heating rates, specimen geometry and orientation, and the mechanical constraint conditions of each apparatus. Comparing data is valid only when the test method, temperature range, and specimen preparation are matched.

What temperature range should be specified for CTE testing?

The range should bracket the intended service environment of the product, for example 25 °C to 60 °C for near-ambient applications and -40 °C to 125 °C for automotive applications. Because PET passes through its glass transition near 80 °C, separate intervals below and above the transition must always be reported rather than a single averaged value.

How can CTE mismatch failures be mitigated in assembly?

Effective strategies include raising ceramic filler content to shift the composite CTE toward that of copper, using compliant adhesives or stress-relief interlayers, symmetrical laminate layups to cancel bending moments, limiting stack thickness of high-CTE polymer layers, and validating the design with accelerated thermal cycling tests correlated to measured CTE data.

How many specimens should be tested for reliable results?

A minimum of three replicate specimens per direction is recommended for routine qualification, with five or more replicates for lot release testing or statistical process control, and both machine and transverse directions should always be characterized to capture anisotropy.