IC Substrate XY-Direction Expansion Testing: In-Plane CTE Measurement and Analysis

IC substrate XY-direction expansion testing measures the in-plane coefficient of thermal expansion (CTE) of IC package substrates, a critical parameter that governs solder joint reliability, warpage behavior, and fine-line registration in advanced semiconductor packaging. This article explains the testing principles, sample requirements, measurement methods, industry standards, and practical applications of XY-direction expansion testing for IC substrates.

Technical Overview

IC substrates function as the mechanical and electrical bridge between the silicon die and the printed circuit board, and their dimensional stability under thermal load directly determines package reliability. The XY direction, also referred to as the in-plane direction, covers the length and width of the substrate, while the Z direction corresponds to thickness. Because the in-plane CTE of a substrate must be matched against the silicon chip (approximately 2.6 ppm/°C) and the assembly board (typically 14–17 ppm/°C), any excessive XY expansion generates thermomechanical stress in solder joints, bump interconnects, and underfill layers.

XY-direction expansion testing quantifies the dimensional change of the substrate material as a function of temperature. The result is expressed as the coefficient of thermal expansion in ppm/°C, with values reported both below the glass transition temperature (alpha1) and above it (alpha2). The test is commonly performed on BT (bismaleimide triazine) resin cores, epoxy-glass laminates, ABF (Ajinomoto Build-up Film) layers, polyimide films, and finished package substrates used in BGA, FCCSP, and FCBGA products. Established procedures such as IPC-TM-650 Method 2.4.24 and ASTM E831 govern specimen preparation, heating profiles, and data calculation.

Testing Samples

Core Materials and Copper-Clad Laminates

Raw core materials and copper-clad laminates (CCL) are the most frequently tested sample type. Glass-fabric-reinforced materials such as BT and FR-4 class epoxy laminates exhibit strong in-plane CTE anisotropy because the woven glass fibers constrain expansion in X and Y while leaving the resin-dominated Z direction free to expand. Specimens are usually etched to remove copper foil so that the resin system response is measured independently of the copper constraint.

Finished Package Substrates

Complete substrates with build-up layers, microvias, through holes, and patterned copper can also be tested to evaluate the effective expansion of the finished product. These samples reflect the combined behavior of the core, dielectric build-up films, and copper structures, providing data that closely represents real package conditions.

Specimen Preparation and Conditioning

Representative preparation and conditioning steps include:

• Cutting rectangular strips along the machine direction (X) and transverse direction (Y), typically 3 to 6 mm wide and 10 to 25 mm long.
• Etching away surface copper when the resin or laminate CTE is the target parameter.
• Edge sanding to ensure parallel, burr-free faces for uniform load contact in the tester.
• Conditioning at 23 ± 2 °C and 50 ± 5% relative humidity for a minimum of 24 hours before measurement.
• Marking orientation and thickness measurements at multiple points to define the original length L0 accurately.

Testing Items

The standard measurement scope for XY-direction expansion testing includes the following parameters:

• In-plane CTE below Tg (alpha1), the primary acceptance value for substrate materials, typically required between 10 and 17 ppm/°C depending on application.
• In-plane CTE above Tg (alpha2), which characterizes post-transition behavior and stress risk during reflow.
• Glass transition temperature (Tg) determined from the inflection point of the dimensional change curve.
• Percentage of total expansion across a specified temperature window, such as 25 °C to 260 °C for lead-free reflow compatibility.
• Directional comparison between X and Y specimens to quantify weave-induced anisotropy.

Testing Methods

The dominant technique is Thermomechanical Analysis (TMA). A rectangular specimen is placed in a quartz or fused-silica fixture with an extension or expansion probe applying a minimal static load. The furnace ramps temperature at a controlled rate of 5 °C/min or 10 °C/min under nitrogen purge, commonly from ambient to 250–280 °C. A first heating cycle is often run to relieve residual thermal stress from lamination and processing, and the second heating run is used for reporting.

The in-plane CTE is calculated from the slope of the dimensional change curve:

CTE = (ΔL / L0) / ΔT, expressed in ppm/°C, where ΔL is the length change over the temperature interval ΔT and L0 is the original specimen length. Regression is performed over defined segments below and above Tg to obtain alpha1 and alpha2 separately.

Alternative and complementary methods include:

• Optical non-contact dilatometry and digital image correlation (DIC) for full-field in-plane strain mapping on whole substrate panels.
• Moiré interferometry and shadow moiré for high-resolution thermal deformation analysis used in warpage studies.
• Strain gauge methods in which bonded gauges track in-plane strain during controlled thermal cycling.
• Laser speckle and video extensometry for fragile ultra-thin substrates that cannot sustain probe loading.

A typical test plan uses at least three to five specimens per direction to account for material variability, with results reported as averages together with standard deviation.

Applications

XY-direction expansion testing supports multiple stages of the IC substrate and packaging workflow:

• Material qualification and incoming inspection for laminate, build-up film, and substrate suppliers.
• Input data generation for finite element warpage simulation of BGA, FCCSP, FCBGA, and 2.5D/3D interposers.
• Solder joint and micro-bump reliability assessment based on CTE mismatch stress calculations.
• Process control for lamination, curing, and desmear operations where resin state changes alter expansion behavior.
• Comparison and benchmarking of competing substrate materials for high-density, high-frequency, and chiplet architectures.
• Root cause analysis of layer-to-layer registration loss, via cracking, and delamination failures in production.

FAQ

Why is XY-direction expansion measured separately from Z-direction expansion? Glass fabric reinforcement constrains the in-plane directions while resin expansion dominates through the thickness, so the two values can differ by an order of magnitude. XY CTE controls lateral mismatch stress with the die and board, whereas Z CTE affects via and plated hole reliability.

What temperature range is used for testing? Most specifications cover 25 °C to 260 °C to match lead-free reflow conditions, though testing may be segmented at Tg to report alpha1 and alpha2 independently.

Why is copper removed before testing laminate samples? Copper has a CTE of about 17 ppm/°C and a very high modulus, so it mechanically suppresses resin expansion and distorts the true material value if left in place.

Why do X and Y results sometimes differ significantly? Woven glass fabric is not perfectly balanced, and resin content and cure shrinkage vary with flow direction, producing measurable in-plane anisotropy that must be captured by testing both orientations.

How many specimens are required? Industry practice calls for a minimum of three valid specimens per direction per material lot, with additional samples recommended for new material qualifications or statistical process control studies.