IC Substrate TMA Thermal Expansion Testing: Principles, Methods, and Applications

IC substrate TMA thermal expansion testing is a critical analytical procedure used to measure the coefficient of thermal expansion (CTE) and glass transition temperature of IC package substrates, ensuring dimensional stability and long-term reliability in semiconductor packaging. This article provides a comprehensive technical overview of TMA testing for IC substrates, covering sample types, testing items, standard test methods such as ASTM E831 and IPC-TM-650 2.4.24, and typical applications in material qualification and packaging reliability engineering.

Technical Overview

IC substrates, also known as package substrates, serve as the electrical and mechanical interconnection bridge between the semiconductor die and the printed circuit board. Because these substrates experience repeated thermal cycling during chip attachment, reflow soldering, and field operation, their thermal expansion behavior directly determines package warpage, solder joint fatigue life, and interfacial adhesion integrity.

Thermomechanical Analysis (TMA) is the standard laboratory technique for characterizing the thermal expansion properties of IC substrate materials. In a TMA test, a specimen is placed on a quartz platform and contacted by a probe applying a small static force. As the specimen is heated at a controlled rate under an inert atmosphere, a precision displacement sensor records the dimensional change in the thickness direction as a function of temperature.

From the resulting expansion curve, engineers extract the coefficient of thermal expansion (CTE) below the glass transition temperature (α1), the CTE above Tg (α2), and the Tg itself, identified as the inflection point where the expansion rate increases sharply. For organic IC substrates, typical z-axis α1 values range from 15 to 50 ppm/°C, which must be carefully matched against silicon (approximately 2.6 ppm/°C) and copper (approximately 17 ppm/°C) to minimize thermomechanical stress in the finished package.

Testing Samples

BT Resin Substrates

Bismaleimide triazine (BT) substrates are widely used for memory chips, CSP packages, and RF modules. TMA testing of BT materials focuses on verifying α1, α2, and Tg consistency between production lots, since BT glass fabrics exhibit relatively low in-plane expansion but higher out-of-plane expansion.

ABF Build-up Film Substrates

Ajinomoto Build-up Film (ABF) based substrates dominate high-density flip-chip applications such as FCBGA and advanced CPU/GPU packaging. Because build-up dielectric layers are thin and resin-rich, their z-axis CTE and Tg strongly influence via reliability and package flatness. Both raw film coupons and finished multilayer substrates can be evaluated by TMA.

Epoxy Glass Core Laminates

The rigid core of an IC substrate is typically a glass-fabric-reinforced epoxy or modified epoxy laminate, often copper-clad. Specimens may be tested with copper etched away to isolate the dielectric response, or with copper retained to evaluate the composite behavior relevant to the actual product.

Ceramic and Specialty Substrates

Alumina (Al2O3) and aluminum nitride (AlN) ceramic substrates, as well as glass, silicon interposers, and embedded capacitor materials, can also be characterized by TMA. These low-expansion materials usually exhibit linear expansion curves without a Tg transition, and testing confirms compliance with tight CTE specifications for hermetic and high-power packaging.

Testing Items

A complete IC substrate TMA thermal expansion test program typically includes the following measurable parameters:

• Z-axis CTE below Tg (α1), reported in ppm/°C over a defined temperature interval
• Z-axis CTE above Tg (α2), typically several times larger than α1 for organic materials
• Glass transition temperature (Tg) determined from the expansion curve inflection point
• Total percentage thickness expansion over a specified range, such as 50 °C to 260 °C
• Delta CTE or expansion step behavior across the transition region
• In-plane (X-Y) CTE evaluation using horizontal dilatometry or TMA lateral fixtures when required for warpage modeling
• Comparative CTE analysis of core material, build-up layers, and finished composite stacks

Testing Methods

Applicable Standards

The most widely referenced procedures include ASTM E831 for linear thermal expansion of solid materials by TMA, IPC-TM-650 method 2.4.24 for Tg and z-axis thermal expansion of printed boards, and ISO 11359-2 for thermomechanical analysis of plastics. Customer-specific specifications based on these standards are common in the IC substrate industry.

Specimen Preparation

Specimens are generally cut to a square or circular footprint of 6 to 10 mm, with thickness representative of the actual substrate construction, commonly 0.1 mm to 1.5 mm. The top and bottom surfaces must be flat, parallel, and free of burrs or residues. Copper may be selectively etched depending on whether dielectric-only or composite data is required. Prior to testing, samples are typically baked at 105 °C to 125 °C for 1 to 2 hours to remove absorbed moisture, then conditioned at 23 °C and 50% relative humidity if specified.

Instrument Calibration and Test Conditions

Instrument calibration is performed using a certified reference material such as sapphire or alpha alumina for expansion correction, and high-purity indium (melting point 156.6 °C) for temperature accuracy. Typical test conditions include:

• Temperature range: ambient to 260 °C or 300 °C, covering reflow and operational extremes
• Heating rate: 3 to 5 °C per minute for organic substrates, with slower rates improving Tg resolution
• Static probe force: approximately 0.02 N to 0.1 N to ensure contact without specimen indentation
• Atmosphere: high-purity nitrogen purge at 20 to 50 mL/min to prevent oxidative degradation
• Optional pre-thermal scan or erase cycle to relieve residual stress from lamination and cure

Data Analysis

Tangential CTE is calculated as the slope of the dimensional change curve divided by initial specimen length, evaluated over defined windows such as 50 to 150 °C for α1 and 200 to 250 °C for α2. Tg is reported as the extrapolated onset or midpoint of the transition inflection. Results are documented in a test report containing the expansion curve, calculated coefficients, test conditions, and specimen description.

Applications

TMA thermal expansion data for IC substrates supports nearly every stage of the semiconductor packaging lifecycle:

• Incoming material qualification of core laminates, build-up films, and finished substrates against supplier specifications
• Warpage simulation and prediction for flip-chip and BGA assemblies, where through-thickness CTE drives package curvature during reflow
• Solder joint reliability assessment, since CTE mismatch between substrate, silicon die, and mold compound drives thermal cycling fatigue
• Via and interconnect integrity evaluation, because excessive z-axis expansion above Tg stresses copper vias and triggers cracking
• Process development for lamination, curing, and build-up cycles, verifying that processing achieves the targeted degree of cure and dimensional stability
• Advanced packaging design, including 2.5D/3D integration, chiplets, and fan-out architectures requiring exceptionally flat and dimensionally stable substrates
• Failure analysis support, where abnormal expansion behavior helps identify under-cured resin, moisture uptake, or material substitution

FAQ

Why is z-axis CTE so critical for IC substrates?

Through-thickness expansion dominates stress on plated through-holes, microvias, and solder joints. Excessive z-axis CTE above Tg can cause via cracking, interfacial delamination, and solder joint failure during reflow and thermal cycling, so it is a primary acceptance criterion for substrate materials.

Should the copper foil be removed before testing?

It depends on the objective. Removing copper isolates the dielectric material behavior for material specification purposes, while retaining copper reflects the actual composite structure relevant to package-level warpage. Both approaches are accepted when clearly documented.

What heating rate should be used?

A heating rate of 5 °C/min is the industry default aligned with IPC-TM-650 2.4.24, while 3 °C/min offers improved resolution of the Tg inflection for thin build-up films. The selected rate must be consistent between comparison tests because faster rates shift the apparent Tg upward.

How does TMA differ from dilatometry and DMA?

TMA measures dimensional change under a small static load and is ideal for z-axis CTE and Tg of thin substrate materials. Dilatometry suits larger specimens and broader temperature ranges, while DMA characterizes mechanical modulus transitions. Tg values from different methods are not directly interchangeable and should be labeled by technique.

What are typical acceptance limits?

Specifications vary by product, but common requirements for high-end organic substrates include α1 below 30 ppm/°C, Tg above 150 °C by TMA, and total expansion over a reflow-like range below approximately 3 to 4 percent. Exact limits are defined jointly by the substrate manufacturer and end customer.

Can finished multilayer substrates be tested directly?

Yes. Finished substrates can be cut to coupon size and tested to evaluate the combined effect of core, build-up layers, and patterned copper, provided the surfaces are flat and the thickness is sufficient for reliable displacement measurement. This approach is valuable for production monitoring and failure investigation.