PCB Glass Transition Temperature and Thermal Expansion Testing (TMA): A Complete Technical Guide

Glass transition temperature (Tg) and coefficient of thermal expansion (CTE) testing is essential for verifying the thermal reliability of PCB laminates, especially in lead-free assembly environments. This technical guide explains how Thermomechanical Analysis (TMA) is used to measure Tg, Z-axis expansion, and total thermal expansion of printed circuit board materials in accordance with IPC-TM-650 2.4.24. Read on to learn about sample preparation, test procedures, typical acceptance values, and how the results relate to plated through-hole reliability.

Technical Overview

Glass transition temperature (Tg) is the temperature region at which the amorphous resin system in a PCB laminate changes from a rigid, glassy state to a softer, more rubbery state. Below Tg, the cross-linked epoxy matrix is dimensionally stable and mechanically strong. Above Tg, molecular chain mobility increases sharply, the modulus of the resin drops, and the rate of thermal expansion accelerates dramatically. Although the laminate does not melt, virtually every mechanical and electrical property degrades once the material operates above its Tg.

Thermomechanical Analysis (TMA) is the standard technique for measuring both Tg and thermal expansion of PCB materials. In a TMA test, a specimen is placed on a quartz platform and a low-mass quartz probe rests on its surface while the sample is heated at a controlled rate. A linear variable displacement transducer continuously records the dimensional change of the specimen as a function of temperature. The resulting expansion curve yields the coefficient of thermal expansion below Tg (alpha 1), the coefficient above Tg (alpha 2), the Tg itself, and the total expansion percentage across a defined temperature range.

Z-axis expansion is the single most critical output of this test. Because glass fibers constrain the laminate in the X and Y directions but not through the thickness, expansion in the Z direction is typically an order of magnitude higher. Excessive Z-axis expansion stresses the copper barrels of plated through holes (PTH) during solder reflow and thermal cycling, causing barrel cracking, inner-layer interconnect separation, and latent field failures. Since the adoption of lead-free soldering, which exposes boards to peak temperatures of 245–260°C, Tg and CTE testing has become a mandatory qualification step for laminate materials and finished PCBs.

Testing Samples

Copper-Clad Laminate (CCL) Panels

Raw laminate sheets are the primary sample type for material qualification. Specimens are cut from production panels, and the copper foil is chemically etched away so that the measurement reflects the resin and glass fabric system rather than the composite with copper. Both single-sided and double-sided laminates of any thickness from 0.1 mm to 3.2 mm can be evaluated, including standard FR-4, high-Tg FR-4, halogen-free epoxies, and specialty blends.

Finished PCB Coupons and Production Boards

Test coupons removed from fabricated panels allow verification of the actual board as delivered. This is important because pressing cycles, lamination parameters, and multiple thermal excursions during fabrication can shift the Tg and expansion behavior of the cured system. Coupons with representative layer counts and resin contents should be selected, and exposed copper at the measurement site is normally etched or mechanically removed before testing.

High-Performance and Specialty Materials

• Mid-Tg and high-Tg FR-4 epoxies (Tg 130–180°C) used for multilayer and lead-free compatible boards
• Polyimide laminates (Tg 220–260°C) for high-temperature and aerospace applications
• BT-epoxy and blend systems for IC substrates and high-reliability packaging
• Halogen-free and phosphorus-based flame-retardant systems for environmental compliance
• Prepreg and resin-only samples for curing studies, using stacked plies pressed to a defined thickness

Testing Items

A complete TMA characterization of PCB material typically reports the following parameters:

• Glass transition temperature (Tg) by TMA, reported as the extrapolated onset and midpoint of the change in slope of the expansion curve
• CTE alpha 1 — coefficient of thermal expansion below Tg, typically 40–70 ppm/°C in the Z direction for FR-4
• CTE alpha 2 — coefficient of thermal expansion above Tg, typically 200–450 ppm/°C for FR-4, reflecting the softening of the resin
• Average CTE over a specified span such as 50°C to 260°C
• Total expansion percentage — the integral Z-axis growth from 50°C to 260°C, a key indicator of lead-free reflow compatibility (target typically below 3.0–4.0%)
• Delta Tg (ΔTg) — the difference between the Tg measured on a second thermal scan and the first scan; a large increase indicates incomplete cure in the as-received laminate
• Time to delamination (T260 / T288 / T300) — an associated TMA mode in which the specimen is held at a constant elevated temperature until delamination is detected

Testing Methods

Sample Preparation

Specimens are cut to approximately 6.4 mm × 6.4 mm (0.25 in × 0.25 in) with the full laminate thickness retained. Copper cladding is etched from both sides, and the sample is cleaned to remove residues. To eliminate absorbed moisture, which can depress the measured Tg and distort the expansion curve, specimens are baked at 105°C for 1 to 2 hours (or per the applicable specification) and cooled in a desiccator before testing. The specimen thickness is measured accurately at the probe contact point.

TMA Test Procedure per IPC-TM-650 Method 2.4.24

The test is performed on a calibrated thermomechanical analyzer as follows:

• Place the specimen on the flat quartz platform and lower the expansion probe at its center, applying a static load of typically 5 g
• Purge the furnace with nitrogen at approximately 25–50 mL/min to prevent oxidative degradation
• Equilibrate at the starting temperature (typically 25–35°C), then heat at 5°C/min to the specified endpoint, commonly 260°C for lead-free compatibility assessment
• Record the dimension-versus-temperature curve continuously through the transition region
• Optionally perform a second heating run on the same specimen to obtain the fully cured Tg and calculate ΔTg
• For time-to-delamination measurements, heat rapidly to the isothermal setpoint (260°C, 288°C, or 300°C) and hold until the dimensional trace shows the abrupt expansion signature of delamination

Data Interpretation

The expansion curve is divided into two linear segments joined by a transition knee. The slope of the low-temperature segment, divided by the specimen length or thickness, gives alpha 1; the slope of the high-temperature segment gives alpha 2. Tg is reported where the extrapolated low-temperature and high-temperature expansion lines intersect (onset) or at the midpoint of the transition knee. Total expansion is the cumulative dimensional change expressed as a percentage between two defined temperatures. A well-cured, high-quality FR-4 shows a sharp, reproducible knee and a ΔTg of less than about 5°C between the first and second scans; a smeared or shifting transition suggests incomplete cure, moisture, or material non-uniformity.

Comparison with DSC and DMA Methods

Tg can also be measured by Differential Scanning Calorimetry (DSC, IPC-TM-650 2.4.24C) and Dynamic Mechanical Analysis (DMA, IPC-TM-650 2.4.24.2). DSC detects the change in heat capacity and typically reports the lowest Tg value; DMA detects the drop in storage modulus and reports the highest; TMA values fall in between and are preferred whenever CTE data is required, because a single test delivers both Tg and expansion behavior. Because the three methods respond to different physical phenomena, results should only be compared within the same technique.

Applications

• Laminate qualification against IPC-4101 slash sheet requirements for minimum Tg, Z-axis CTE, and time to delamination
• Incoming material inspection and supplier lot-to-lot consistency verification for PCB fabricators
• Lead-free and multiple reflow process compatibility assessment for SMT assembly houses
• Plated through-hole reliability prediction, since Z-axis expansion directly drives PTH barrel fatigue in thermal cycling
• Degree-of-cure verification for multilayer press cycles through ΔTg evaluation
• Failure analysis of delamination, measling, barrel cracking, and interconnect defects in returned boards
• New product development of high-Tg, low-CTE, and halogen-free resin systems

FAQ

Q: What is a typical Tg value for PCB materials?

A: Standard FR-4 laminates have a TMA Tg of approximately 130–140°C, mid-range systems around 150–160°C, and high-Tg FR-4 systems 170–180°C. Polyimide materials exceed 220°C. The required value depends on the IPC-4101 slash sheet and the maximum assembly or operating temperature of the end product.

Q: Why is Z-axis expansion more important than X-Y expansion?

A: The glass fabric mechanically constrains the laminate in the plane of the board, limiting X-Y CTE to roughly 12–16 ppm/°C. In the thickness direction there is no such constraint, so Z-axis CTE below Tg is typically 40–70 ppm/°C and accelerates several-fold above Tg. This differential expansion stresses PTH copper barrels and is the dominant mechanism behind barrel cracking during reflow and thermal cycling.

Q: Is a higher Tg always better?

A: Not necessarily. Tg only defines where the resin softens; total expansion, decomposition temperature (Td), and time to delamination are equally important. A laminate with moderate Tg but very low alpha 1 and alpha 2 may outperform a higher-Tg material with aggressive expansion above transition. Total expansion from 50°C to 260°C is generally the best single indicator of lead-free assembly robustness.

Q: What does a large ΔTg indicate?

A: If the second-scan Tg is significantly higher than the first-scan value (typically more than 5–10°C), the laminate was not fully cured when received. Incomplete cure reduces chemical resistance, bond strength, and thermal reliability. A well-processed board should show first- and second-scan Tg values within a few degrees of each other.

Q: How should specimens be conditioned before testing?

A: Specimens should be baked at 105°C for 1 to 2 hours and stored in a desiccator until testing. Absorbed moisture lowers the apparent Tg, distorts the expansion curve, and can cause apparent delamination artifacts, so proper drying is essential for reproducible results.

Q: What heating rate and probe load are used?

A: The standard condition per IPC-TM-650 2.4.24 is a heating rate of 5°C/min with a static probe load of approximately 5 g under nitrogen purge. Faster rates or heavier loads shift the measured Tg and must be reported if non-standard conditions are used, since Tg values are method-dependent.