Determination of Coefficient of Thermal Expansion Variation in PET Ceramic Substrates

The coefficient of thermal expansion (CTE) of PET ceramic substrates is a decisive parameter for dimensional stability, solder joint reliability, and long-term performance in electronic packaging and flexible circuit applications. This article presents a complete technical guide to measuring CTE variation in PET ceramic substrates, covering sample preparation, key testing items, standard measurement methods such as thermomechanical analysis (TMA) and dilatometry, typical applications, and answers to frequently asked questions.

Technical Overview

PET ceramic substrates are composite materials that combine a polyethylene terephthalate (PET) polymer phase with ceramic fillers or ceramic layers, producing laminates that offer the flexibility and processability of a polymer together with the thermal conductivity and dimensional stability of ceramics such as alumina, silica, or aluminum nitride. These materials are widely used in flexible printed circuits, hybrid electronics, thermal management modules, and insulated metal substrate constructions.

The coefficient of thermal expansion (CTE) quantifies the fractional dimensional change of the substrate per unit temperature change, defined as alpha = (dL / L0) / dT, and is expressed in ppm/°C (10^-6/K). For PET ceramic substrates, the CTE is not a single fixed value: it varies strongly with temperature, exhibits a marked transition at the glass transition of the PET phase (approximately 78–80 °C), and depends on filler loading, filler orientation, lamination structure, and the measurement direction relative to the machine direction (MD) and transverse direction (TD) of the film.

Accurate determination of CTE variation is essential because a mismatch between the substrate CTE and the expansion of silicon chips, copper conductors, solder joints, or ceramic heat spreaders generates thermomechanical stress during processing and service. Excessive mismatch leads to warpage, delamination, copper trace cracking, and solder fatigue failure. Below the glass transition temperature, ceramic-filled PET substrates typically show CTE values in the range of 15–40 ppm/°C, while above Tg the CTE of the polymer phase rises sharply and the effective expansion may increase two- to five-fold. Measuring this variation across the full operating temperature window is therefore a core requirement for material qualification and reliability engineering.

Testing Samples

Ceramic-Filled PET Composite Substrates

These samples consist of a PET matrix loaded with 30–70 wt% ceramic powder (alumina, crystalline silica, or aluminum nitride). Specimens are cut into rectangular strips, typically 3–10 mm in length for TMA measurement or 10–50 mm for dilatometry, with thickness between 0.1 mm and 3 mm. Cut surfaces must be flat, parallel, and free of burrs to ensure uniform load transfer during measurement.

PET Film Laminated with Ceramic Layers

Multilayer samples in which biaxially oriented PET film is laminated with cast ceramic green tape or coated ceramic layers must be tested in both MD and TD orientations, because biaxial stretching during film manufacture introduces strong in-plane anisotropy. At least three specimens per direction should be prepared to characterize scatter.

Metallized PET Ceramic Substrates

Copper-clad or otherwise metallized substrates should be tested both with and without the metal layer when possible, so that the contribution of the metal foil to the effective laminate CTE can be separated from the substrate contribution. Specimen conditioning before testing normally follows 23 °C / 50% RH for 24–48 hours, because absorbed moisture plasticizes PET and shifts both Tg and the measured expansion behavior.

Testing Items

A complete CTE characterization program for PET ceramic substrates typically includes the following items:

• Linear CTE below the glass transition temperature (alpha-1), normally reported between 30 °C and 60 °C.
• Linear CTE above the glass transition temperature (alpha-2), normally reported between 100 °C and 150 °C, ensuring the substrate remains below the melting range of PET.
• Glass transition temperature determined from the inflection point of the expansion curve, as a companion result to the CTE transition.
• Anisotropic expansion behavior, comparing machine direction and transverse direction specimens.
• Thermal shrinkage and irreversible dimensional change after the first heating scan, which reveals residual stress relaxation from film orientation and lamination.
• CTE hysteresis and repeatability over multiple thermal cycles, used to assess dimensional stability in cyclic service environments.
• Effective laminate CTE for metallized constructions, required for warpage simulation and reliability modeling.

Testing Methods

Thermomechanical Analysis (TMA)

TMA is the preferred method for thin PET ceramic substrates and is performed in accordance with ASTM E831 or ISO 11359-2. A specimen with flat, parallel faces is placed on a quartz platform and contacted by a low-mass probe with a static force of typically 0.01–0.05 N to avoid indentation and creep artifacts. The sample is heated at 3–5 °C/min under nitrogen purge from -50 °C or room temperature up to 150–200 °C. The expansion curve (dL/L0 versus temperature) is recorded continuously, and alpha-1 and alpha-2 are obtained by linear regression over defined intervals below and above the Tg inflection.

Push-Rod Dilatometry

For thicker, rigid, or highly filled substrates, push-rod dilatometry per ASTM E228 provides higher precision over longer specimens. The sample is heated in a furnace while a fused silica push-rod transmits length change to a displacement sensor, allowing measurement from room temperature to 300 °C or higher. Correction for the expansion of the sample holder and push-rod is applied using calibration runs with reference materials such as fused silica, copper, or sapphire.

Optical and Non-Contact Methods

For ultra-thin free-standing films below 100 µm, non-contact techniques such as laser interferometry or digital image correlation avoid probe-loading errors. These methods are particularly valuable when the substrate is too compliant for reliable contact measurement above the glass transition region.

Data Processing and Reporting

The mean CTE over a temperature interval is calculated from the slope of the least-squares line fitted to the dL/L0 curve. Reports should state the temperature range for each reported value, the heating rate, the specimen orientation, the conditioning history, and whether the result is from the first or second heating scan. Comparing first and second scans is strongly recommended for oriented PET materials, since the first scan often shows relaxation shrinkage superimposed on true thermal expansion.

Applications

Measured CTE data for PET ceramic substrates directly support the following engineering activities:

• Material selection and incoming quality control for flexible printed circuit and rigid-flex substrate manufacturing.
• Warpage prediction and finite element simulation of multilayer assemblies during reflow soldering and thermal cycling.
• Solder joint fatigue life assessment, where cyclic strain driven by CTE mismatch is a dominant failure driver.
• Thermal management design for LED modules and power devices, where the substrate must bridge the expansion gap between ceramic heat spreaders and organic circuit layers.
• Automotive and under-hood electronics qualification, where operating temperatures routinely cross the PET glass transition and the alpha-1 to alpha-2 transition must be controlled.
• Research and development of new ceramic-filled formulations, using measured CTE versus filler content to validate rule-of-mixtures and micromechanical expansion models.

FAQ

What is a typical CTE value for a PET ceramic substrate? Highly filled grades generally exhibit 15–30 ppm/°C below Tg, rising to 40–80 ppm/°C or higher above Tg, depending on filler type, loading, and orientation. Unfilled PET film is markedly higher, especially above its glass transition.

Why does the CTE of PET ceramic substrates change with temperature? The PET matrix undergoes a glass transition near 78–80 °C, above which chain mobility increases sharply and the polymer phase expands much more rapidly. The ceramic filler has a low and nearly constant CTE, so the composite expansion jumps at Tg but remains moderated relative to neat PET.

Which standard should be used for measurement? ASTM E831 and ISO 11359-2 govern TMA measurement of plastics and composites, ASTM E228 governs push-rod dilatometry, and ASTM D696 applies to dilatometric measurement at near-ambient conditions. The choice depends on specimen thickness, stiffness, and the temperature range of interest.

What heating rate is appropriate? A rate of 3–5 °C/min is standard. Faster rates can shift the apparent transition region and inflate the measured expansion slope, while slower rates improve equilibrium but extend test time and can promote moisture loss or annealing effects.

Why do first and second heating scans give different results? Oriented PET and laminated substrates contain residual processing stresses. During the first scan these stresses relax and produce irreversible shrinkage superimposed on reversible thermal expansion. The second scan reflects the relaxed material and is often used as the reference CTE, provided the substrate is not re-oriented in service.

How is measurement direction handled for anisotropic substrates? Specimens must be cut along the machine direction and the transverse direction and tested separately. Reporting a single CTE value for an anisotropic PET ceramic substrate without direction information is insufficient for reliable warpage and stress analysis.

Does moisture affect the results? Yes. Moisture plasticizes PET, lowers the effective glass transition, and alters expansion behavior. Standard conditioning at 23 °C / 50% RH, or documented dry-state conditioning, should always be specified together with the test result.