PET Ceramic Substrate Coefficient of Thermal Expansion (CTE) Test Plan: Experimental Scheme, Standar

This article presents a complete coefficient of thermal expansion (CTE) test plan for PET ceramic substrates, covering sample preparation, testing items, instrument configurations, and data evaluation in line with ASTM E831, ISO 11359-2, and related standards. It provides material and reliability engineers with a practical experimental scheme for characterizing dimensional stability in flexible electronics, LED packaging, and thermal management applications. The protocol also explains how CTE results are used to predict warpage, delamination, and solder joint stress in downstream assembly processes.

Technical Overview

A PET ceramic substrate is a composite material in which polyethylene terephthalate (PET) serves as a polymer matrix or carrier film while ceramic fillers, such as alumina (Al2O3), silica (SiO2), or aluminum nitride (AlN), provide dimensional reinforcement and improved thermal conductivity. Because the neat polymer expands at roughly 20 to 50 ppm/K below its glass transition while typical ceramic phases expand at only 0.5 to 8 ppm/K, the resulting composite exhibits a strongly composition-dependent coefficient of thermal expansion (CTE).

CTE mismatch is one of the dominant failure drivers in electronic packaging. When a PET ceramic substrate is laminated to silicon dies (approximately 2.6 ppm/K), copper foils (approximately 17 ppm/K), or solder joints, differential expansion during reflow and thermal cycling generates interfacial stress, warpage, cracking, and delamination. A rigorous CTE test plan is therefore required for material qualification, formulation screening, finite element simulation inputs, and incoming quality control.

This experimental scheme defines sample requirements, measured quantities, instrumentation, and data treatment for CTE characterization of PET ceramic substrates from -50 to 200 degrees Celsius, which spans the practical service window of PET-based materials without approaching the polymer melting point near 250 degrees Celsius.

Testing Samples

Neat PET Reference Film

Biaxially oriented PET film of the same grade and thickness used in the substrate is tested as a baseline, typically 50 to 250 micrometers thick. This group establishes unreinforced matrix behavior and isolates the contribution of the ceramic phase.

Ceramic-Filled PET Composite Substrates

Injection-molded, cast, or calendared composite plates containing defined ceramic loadings (for example, 30, 50, and 70 percent by weight) are sectioned into specimens. Each batch documents filler type, particle size distribution, surface treatment, and processing history, because all four factors measurably shift the CTE.

Ceramic-Coated PET Laminates

Multilayer samples consisting of a PET core with a ceramic coating or metallization on one or both sides are tested to evaluate laminate-level expansion and potential warpage artifacts. Specimens are cut along the machine direction (MD) and transverse direction (TD) to quantify in-plane anisotropy introduced by film orientation.

Specimen Geometry and Conditioning

• Expansion-mode TMA specimens: 3 to 10 mm in length and 3 to 5 mm in width, with faces ground flat and parallel within 10 micrometers.
• Tension-mode film attachment specimens for thin films: strips 10 to 15 mm long and 2 to 5 mm wide.
• Push-rod dilatometer specimens: 10 to 25 mm long with a 4 to 10 mm square or circular cross section.
• A minimum of five replicate specimens per material code and per test direction.
• Conditioning at 23 plus or minus 2 degrees Celsius and 50 plus or minus 5 percent relative humidity for 24 hours, plus a dried comparison group held 24 hours at 80 degrees Celsius in vacuum, since PET absorbs roughly 0.4 percent moisture and plasticizes.

Testing Items

• Linear CTE in the in-plane (X and Y) directions for both machine and transverse orientations.
• Linear CTE in the through-thickness (Z) direction, which typically exceeds the in-plane value in laminated structures.
• CTE below the glass transition (alpha 1), measured over a regression interval such as 30 to 70 degrees Celsius.
• CTE above the glass transition (alpha 2), measured over a regression interval such as 90 to 130 degrees Celsius before significant softening begins.
• Glass transition temperature from the inflection point of the dimension-versus-temperature curve, expected near 75 to 85 degrees Celsius for PET-rich compositions.
• Anisotropy ratio between in-plane and through-thickness expansion.
• Residual thermal shrinkage and dimensional hysteresis across heating and cooling cycles.

Testing Methods

Thermomechanical Analysis (TMA) per ASTM E831 and ISO 11359-2

TMA is the primary method for thin, polymer-rich substrates. A flat-tipped quartz probe applies a small preload of 0.02 to 0.05 N, and a displacement transducer records dimensional change during heating at 5 degrees Celsius per minute under nitrogen purge at 30 to 50 mL/min. The program runs from -50 to 200 degrees Celsius with a 5-minute isothermal hold at the start to achieve equilibrium and erase minor thermal history.

Film and Tension Mode for Thin PET Substrates

Specimens thinner than 200 micrometers are clamped in the film attachment with a small static tensile force of 0.01 to 0.02 N. This configuration prevents probe-penetration artifacts and yields in-plane CTE directly. A preload sweep verifies that the measured slope is force-independent and free of creep.

Push-Rod Dilatometry per ASTM E228

Composite plates thicker than 3 mm are measured with a push-rod dilatometer for lower uncertainty at low expansion levels. The furnace ramps at 3 to 5 degrees Celsius per minute, and expansion is referenced to a high-purity fused silica holder whose contribution is subtracted during analysis.

Data Processing and CTE Calculation

The mean CTE is computed as the change in length per unit initial length per degree, obtained by least-squares linear regression over the specified temperature intervals. Reporting includes the regression interval, correlation coefficient, mean value, standard deviation, and uncertainty, with a target repeatability within plus or minus 0.5 ppm/K or 5 percent, whichever is greater. Two heating runs are performed, and second-run data are reported when first-run stress relaxation is observed.

Calibration and Quality Control

• Temperature calibration with indium (156.6 degrees Celsius) and tin (231.9 degrees Celsius) melting point standards.
• Length-change calibration and baseline verification using fused silica or sapphire reference materials.
• Duplicate runs with repositioned specimens to confirm flatness and proper probe seating.
• A control specimen in every test batch to monitor drift between calibration cycles.

Applications

CTE data generated by this test plan directly support design and manufacturing decisions across the electronics supply chain. The results feed warpage simulation, solder joint fatigue lifetime models, and laminate stack-up optimization for chip-on-film and flexible circuit assemblies.

• Flexible printed circuits and chip-on-film assemblies where substrate expansion must be reconciled with silicon die and copper foil.
• LED packaging substrates requiring matched expansion between the ceramic layer, phosphor coating, and heat sink interface.
• Automotive sensor and control modules subjected to -40 to 125 degrees Celsius thermal cycling qualification.
• High-frequency antenna and 5G module substrates where dimensional stability preserves impedance tolerance.
• Composite formulation screening, where CTE versus filler loading curves guide ceramic content and coupling agent selection.
• Incoming inspection and statistical process control for substrate manufacturers supplying consumer and industrial electronics.

FAQ

What CTE range is typical for a PET ceramic substrate?

Depending on filler loading and orientation, in-plane CTE generally falls between 8 and 25 ppm/K below Tg, compared with 20 to 50 ppm/K for neat PET and 0.5 to 8 ppm/K for the ceramic phase. Through-thickness values can be 20 to 50 percent higher in laminated structures.

Which standard should be referenced in the test report?

Cite ASTM E831 or ISO 11359-2 for TMA, ASTM E228 for dilatometry, and IPC-TM-650 method 2.4.24 when the material is evaluated as a copper-clad laminate. The report must state the standard, temperature interval, heating rate, and specimen orientation.

Why is the test capped near 200 degrees Celsius?

PET softens progressively above Tg and melts near 250 degrees Celsius. Extension beyond 200 degrees Celsius risks probe penetration, creep under preload, and irreversible degradation, which corrupts the data and can damage the instrument.

How do moisture and thermal history affect the results?

Absorbed moisture plasticizes PET, lowering Tg and raising alpha 1 by several ppm/K, while residual orientation relaxes during the first heating run. Testing both conditioned and dried specimens, and reporting second-heating data, removes these artifacts.

Why must in-plane and through-thickness CTE be measured separately?

Film orientation and layered ceramic coatings create strong anisotropy. Warpage and delamination predictions require directional values rather than an isotropic assumption, particularly for z-axis expansion during reflow.

How many specimens are required for a statistically valid result?

A minimum of five specimens per material code and direction is specified, with mean and standard deviation reported. Additional replicates are recommended when discriminating formulations whose CTE differs by less than 1 ppm/K.