PET ceramic substrates combine a rigid ceramic body with PET insulating films, adhesives, and metallization layers, and their coefficient of thermal expansion (CTE) is one of the most critical parameters governing long-term reliability. Cold–hot thermal shock testing exposes these substrates to rapid transitions between temperature extremes to reveal cracking, delamination, adhesion loss, and CTE drift. This guide explains the sample types, test items, methods, standards, and acceptance criteria used for PET ceramic substrate CTE thermal shock testing.
Technical Overview
A PET ceramic substrate is a multilayer construction that joins a rigid ceramic base plate, typically 96% or 99.6% alumina (Al2O3) or aluminum nitride (AlN), with PET (polyethylene terephthalate) insulating films, adhesive layers, and patterned copper conductors. Because every material in this stack expands at a different rate when heated, the coefficient of thermal expansion of each layer, and of the composite as a whole, determines the stress state at each interface and ultimately the service life of the assembly.
The coefficient of thermal expansion, denoted as alpha, is defined as the fractional change in length per unit change in temperature, calculated as the change in length divided by the initial length multiplied by the temperature change, and is expressed in ppm/°C. Typical reference values are approximately 6.5 to 7.5 ppm/°C for 96% alumina, 4.3 to 5.0 ppm/°C for aluminum nitride, approximately 17 ppm/°C for copper, and roughly 15 to 25 ppm/°C for PET film below its glass transition temperature, which lies near 75 to 80 °C. Above the glass transition, the PET layer can expand several times faster, sharply increasing interfacial stress.
Cold–hot thermal shock testing, also called thermal cycling or temperature shock testing, accelerates the failure mechanisms driven by CTE mismatch. A specimen is transferred rapidly between a cold chamber and a hot chamber so that the full stack experiences temperature swings of 165 to 215 °C within seconds. Repeated cycles convert the differential expansion into cyclic shear and peel stress at the ceramic-to-adhesive, adhesive-to-PET, and copper-to-ceramic interfaces, revealing weak bonding, microcracks, and latent CTE instability within days instead of years.
- Baseline CTE characterization before stress exposure
- Cyclic thermal shock between programmable temperature extremes
- Post-shock CTE drift measurement and failure analysis
- Correlation of expansion behavior with delamination, cracking, and adhesion loss
Testing Samples
Bare PET Ceramic Substrates
Unclad ceramic substrates, commonly 50 mm by 50 mm or 100 mm by 100 mm panels with thicknesses of 0.25 mm to 1.0 mm, are tested first to establish the intrinsic CTE of the ceramic body as fired. These coupons provide the reference expansion curve against which all laminated and assembled variants are compared. Edges are ground and surfaces are cleaned so that measured expansion is not influenced by chips, burrs, or contamination.
Copper-Clad and PET-Laminated Ceramic Substrates
Production-representative laminates with PET film, adhesive, and copper foil stacks are tested to capture the true composite expansion behavior. Peel strength coupons with 3.2 mm wide copper traces and insulation resistance patterns are fabricated on the same panels so that adhesion and electrical performance can be tracked on identical material. Both X-Y in-plane and Z-axis through-thickness expansion are characterized, since the layered structure is strongly anisotropic.
Assembled Substrates and Test Vehicles
Fully populated test vehicles with soldered chip resistors, power diodes, and leaded or leadless packages are included to evaluate how CTE mismatch affects solder joints and interconnects under shock. These samples are instrumented with fine-gauge thermocouples to verify that the internal mass actually reaches the specified extremes within the programmed dwell time.
All samples are preconditioned by baking at 105 °C for 24 hours to remove absorbed moisture, visually inspected and photographed under magnification, and documented with baseline CTE, peel strength, and insulation resistance values before shock exposure begins.
Testing Items
The test program evaluates both the expansion properties of the material stack and the physical and electrical consequences of repeated thermal shock. The core items are listed below.
- In-plane and Z-axis coefficient of thermal expansion, including alpha-1 below the glass transition and alpha-2 above it for the PET layer
- Glass transition temperature of the PET and adhesive layers measured by TMA
- Thermal shock survivability, including ceramic cracking, PET film splitting, and interfacial delamination
- Copper foil and PET layer adhesion, expressed as peel strength retention after cycling
- Insulation resistance at 500 VDC and dielectric withstand voltage after shock exposure
- Warpage, twist, and dimensional stability of the substrate panel
- Solder joint integrity, plated via continuity, and conductor resistance drift
- Microstructural examination of cross sections by optical microscopy and SEM
Testing Methods
CTE Measurement by TMA and Dilatometry
Coefficient of thermal expansion is measured by thermomechanical analysis in accordance with ASTM E831 and ISO 11359-2, using specimens typically 3 to 10 mm in length heated at 5 °C/min under nitrogen with a low probe force of approximately 0.05 N. Ceramic substrates are scanned from room temperature to 300 °C or higher, while PET-containing composite samples are scanned from -40 °C to 150 °C to bracket the glass transition of the polymer layers. Larger coupons and the ceramic base itself may be characterized by push-rod dilatometry per ASTM E228 for improved accuracy at full panel scale. Results are reported as secant CTE over defined intervals, with alpha-1 and alpha-2 clearly separated for the polymer layers.
Two-Chamber Air-to-Air Thermal Shock Testing
The primary shock method uses a dual-chamber system following JEDEC JESD22-A104, IEC 60068-2-14, and MIL-STD-883 Method 1010. Specimens dwell at each extreme, commonly 10 to 15 minutes or until the monitored sample temperature stabilizes within 2 °C, and are mechanically transferred between chambers in less than 10 seconds for military-grade profiles or within one minute for commercial profiles. Typical profiles include -40 °C to +125 °C for automotive-grade substrates and -55 °C to +150 °C or -65 °C to +150 °C for harsh-environment applications, with cumulative cycle counts of 100, 300, 500, and 1000 cycles evaluated in sequence.
Liquid-to-Liquid Thermal Shock Testing
For the most aggressive screening, liquid-to-liquid shock is performed per MIL-STD-883 Method 1011, transferring samples between a cold bath near 0 °C and a heated bath near 100 °C in under 10 seconds. This method achieves heating and cooling rates exceeding 100 °C per minute and is reserved for qualification of thin, fully sealed laminates, since bath fluids must not be allowed to wick into exposed PET edges or drilled holes.
Post-Stress Characterization and Failure Analysis
After each cycle checkpoint, samples are re-measured for CTE drift by TMA, with a change of more than 5 to 10 percent from baseline typically flagged as degradation. Peel strength is retested per IPC-TM-650 2.4.8 with retention below 80 percent treated as failure, insulation resistance must remain above 100 megohms at 500 VDC, and warpage is checked against limits defined in IPC-TM-650 2.4.22. Nondestructive inspection by X-ray and scanning acoustic microscopy locates internal delamination, while dye penetration, cross sectioning, and SEM with EDS confirm crack paths and interface separation at the ceramic, adhesive, and PET boundaries.
Applications
PET ceramic substrate CTE thermal shock testing supports qualification and routine reliability monitoring across industries where bonded ceramic-polymer stacks face severe cyclic temperature excursions in service.
- Electric vehicle power modules, on-board chargers, and DC-DC converters using IGBT and SiC MOSFET die on ceramic substrates
- Automotive engine controllers, transmission electronics, and under-hood sensor platforms
- LED chip-on-board substrates and high-power lighting drivers
- Thick-film and PET-film flexible heaters for seats, mirrors, mirrors, and battery preheating
- Medical diagnostic equipment, wearable sensor modules, and portable instrumentation
- Aerospace, avionics, and defense electronics qualified to MIL-STD-883 environments
- 5G and RF communication hardware requiring dimensionally stable, low-loss substrates
FAQ
What temperature range and cycle count should be selected? Automotive-grade PET ceramic substrates are usually validated at -40 °C to +125 °C, while harsh-environment and military programs use -55 °C to +150 °C or -65 °C to +150 °C. Cycle counts range from 100 cycles for screening to 500 or 1000 cycles for full qualification, with interim electrical and dimensional checks at each checkpoint.
How large a CTE mismatch is acceptable? Interfacial stress scales roughly with the product of the CTE difference, the temperature swing, and the elastic modulus of the adjoining layers. As a practical guideline, keeping the effective CTE mismatch between adjacent stack layers within about 2 to 4 ppm/°C over the operating range, for example by matching ceramic grade and adhesive formulation, substantially reduces delamination risk under thermal shock.
Why does the PET layer show two CTE values? PET is a semicrystalline polymer with a glass transition near 75 to 80 °C. Below the transition the film expands at alpha-1 of roughly 15 to 25 ppm/°C, while above it alpha-2 can be several times higher. Both values must be reported and considered, because the alpha-2 regime governs stress during high-temperature dwell in every shock cycle.
What is the difference between thermal shock and temperature cycling? Thermal shock uses rapid transfer between fixed-temperature zones, with sample transition times under 10 seconds in two-chamber or liquid systems, producing steep gradients and high instantaneous stress. Temperature cycling uses slower programmed ramps, typically 10 to 15 °C per minute, which stresses the stack more uniformly. Both are complementary and are frequently run in sequence during qualification.
How is CTE drift judged after shock exposure? Identical TMA scans are performed before and after cycling, and the secant CTE over the same temperature intervals is compared. A shift of more than 5 to 10 percent, any new inflection point, or expansion hysteresis between heating and cooling curves indicates polymer degradation, interface slip, or microcracking, and is treated as a failure even if the sample appears visually intact.
Which standards govern the testing? Key references include ASTM E831 and ISO 11359-2 for TMA-based CTE, ASTM E228 for dilatometry, JEDEC JESD22-A104 and IEC 60068-2-14 for air-to-air temperature shock, MIL-STD-883 Methods 1010 and 1011 for military-grade cycling and liquid-to-liquid shock, and IPC-TM-650 methods 2.4.8 and 2.4.22 for peel strength and warpage evaluation.