Substrate Glass Thermal Stability Testing: Methods, Standards, and Quality Evaluation

Substrate glass thermal stability testing evaluates the ability of display, semiconductor, and photovoltaic glass substrates to withstand thermal shock, thermal cycling, and sustained high-temperature exposure without cracking, warping, or irreversible dimensional change. This guide explains the key thermal parameters, applicable ISO, IEC, ASTM, GB, and JIS test methods, typical sample requirements, and acceptance criteria used by manufacturers and accredited third-party laboratories. It serves as a practical reference for quality, process, and reliability engineers working with glass substrates.

Technical Overview

Substrate glass is the foundational platform material for flat-panel displays, OLED devices, photovoltaic modules, touch panels, and an expanding range of semiconductor packaging applications such as glass carriers and through-glass-via (TGV) interposers. In every one of these applications, the glass must survive demanding thermal environments, including thin-film deposition at 300°C to 500°C, annealing steps approaching the strain point, laser or frit sealing at 400°C to 600°C, solder reflow above 250°C, and in-service temperature cycling from -40°C to +150°C in automotive and industrial electronics.

Thermal stability testing quantifies how a glass substrate responds to these stresses. It measures resistance to rapid temperature transients (thermal shock), endurance under repeated heating and cooling (thermal cycling), expansion behavior (coefficient of thermal expansion), structural relaxation temperatures (strain point, annealing point, softening point), and permanent dimensional change after heat exposure (thermal shrinkage or compaction). Because thermally induced stress is one of the dominant causes of substrate breakage and alignment error in high-volume manufacturing, thermal stability verification is a mandatory element of incoming material qualification, new product introduction, and failure analysis.

Key Thermal Performance Parameters

• Thermal shock resistance: the maximum temperature difference (ΔT) a sample withstands without fracture when rapidly transferred from a hot to a cold medium.
• Thermal cycling endurance: survivability and dimensional stability after repeated excursions between low and high temperature extremes.
• Coefficient of thermal expansion (CTE): typically reported over 20°C to 300°C in ppm/K; governs interfacial stress with silicon, metals, and functional coatings.
• Strain point and annealing point: temperatures bounding the annealing range, which control stress relaxation and the allowable process thermal budget.
• Softening point: the temperature at which glass begins to deform under its own weight, defining the upper processing limit.
• Thermal shrinkage (compaction): irreversible dimensional contraction after heat treatment near the strain point, a critical yield factor for large-generation display panels.

Testing Samples

Testing samples are usually prepared as coupons cut from production sheets, full-size panels, or polished wafers, depending on the substrate format and the customer specification. Edge grinding and polishing are essential before testing, because raw cut edges introduce flaws that cause premature fracture and non-representative results.

Alkali-Free Display Glass Substrates

Alkali-free aluminoborosilicate glasses are the standard platform for TFT-LCD and OLED manufacturing because mobile alkali ions degrade thin-film transistor performance. Typical thermal characteristics include a CTE of 3.0 to 3.8 ppm/K (20°C to 300°C), a strain point of 620°C to 690°C, an annealing point of 670°C to 730°C, and a softening point of 830°C to 970°C. These substrates are tested extensively for thermal shrinkage because compaction directly affects mask alignment accuracy during repeated TFT process steps.

Borosilicate Glass Substrates

Low-expansion borosilicate glasses, with a CTE of approximately 3.3 ppm/K, are used in optical components, sensor packaging, and analytical instruments. Their high strain point of roughly 500°C to 530°C and excellent thermal shock resistance make them suitable for applications involving repeated sterilization or aggressive thermal cycling.

Soda-Lime Glass Substrates

Soda-lime glass provides a lower-cost option for applications with moderate thermal demands, such as photovoltaic cover glass, mirrors, and entry-level display products. With a CTE of 8.5 to 9.5 ppm/K and lower characteristic temperatures (strain point near 510°C, annealing point near 545°C, softening point near 730°C), its high-temperature processing window is limited, but thermal shock and cycling qualification remain necessary reliability requirements.

Glass Wafers and Semiconductor Carriers

Polished glass wafers from 100 mm to 300 mm in diameter are used for wafer-level packaging, temporary-bond carriers, TGV interposers, and MEMS capping. These samples are evaluated for thermal cycling compatibility with silicon, where matching the CTE to roughly 3 ppm/K is essential, and for warp stability after simulated deposition, bonding, and debonding exposures.

Chemically Strengthened Cover Glass

Ion-exchanged cover glass for touch displays must retain its engineered compressive stress profile after thermal events such as lamination curing and solder reflow. Thermal stability programs for strengthened glass combine thermal exposure with post-test surface stress and fracture strength measurement to verify that the ion-exchanged layer has not been excessively relaxed by heat.

Testing Items

A comprehensive substrate glass thermal stability evaluation typically includes the following items, selected according to the end application and the applicable customer or industry specification:

• Thermal shock resistance (critical temperature difference determination, water quench or air quench)
• Thermal cycling endurance (-40°C to +85°C, -40°C to +125°C, or -40°C to +150°C, 50 to 1,000 cycles)
• High-temperature storage (150°C to 500°C, 24 to 1,000 hours)
• Coefficient of thermal expansion measurement (20°C to 300°C standard range, extended ranges on request)
• Strain point, annealing point, and softening point determination
• Thermal shrinkage (compaction) after simulated process heat treatment
• Warp, sag, and surface flatness retention after thermal exposure
• Post-thermal strength retention testing (ring-on-ring, four-point bending) and edge quality inspection

Testing Methods

Thermal Shock Resistance Testing

In the classical quench method, samples are heated in a forced-convection oven to a set temperature, held for 10 to 30 minutes to reach uniform temperature, then transferred within a few seconds to a quench medium, typically a water bath at ambient temperature or 0°C. Samples are subsequently inspected under strong grazing illumination or with dye penetrant to detect cracks and edge chips. The temperature difference is increased stepwise, often in 10 K increments, until the failure threshold is reached, and the median failure ΔT is reported as the thermal shock endurance. Reference procedures include ISO 718 for thermal shock and thermal shock endurance of glass, ASTM C149 for the transfer-to-water principle, and GB/T 4547 for thermal shock testing of glass. For substrates that must not contact water, an air-to-air shock method using a forced cold-air chamber is applied instead.

Thermal Cycling Testing

Thermal cycling is performed in environmental chambers according to IEC 60068-2-14 (Test Nb two-chamber or Test Nc single-chamber method) and the corresponding GB/T 2423.22 standard. A representative automotive-grade profile is -40°C to +150°C with 15 to 30 minutes dwell at each extreme and 50 to 500 cycles, with ramp rates controlled to avoid unrealistically steep gradients. After cycling, samples undergo visual and microscopic crack inspection, non-contact warp measurement, coating adhesion verification, and comparison of fracture strength against untested controls. This test reveals cumulative degradation mechanisms, such as subcritical crack growth, film delamination, and progressive dimensional drift, that single shock exposure cannot detect.

Coefficient of Thermal Expansion Measurement

CTE is measured by push-rod dilatometry in accordance with ASTM E228 and ISO 7991, with JIS R 3104 as the recognized glass-specific method. Rod or coupon samples, typically 25 mm to 50 mm in length, are heated at 3 K/min to 5 K/min from room temperature to 300°C or to an extended upper limit, and the mean CTE is calculated from the recorded length change. For ultra-low-expansion glasses or shrinkage studies requiring higher precision, laser interferometric dilatometry can be applied. Accurate CTE data are essential for matching glass to silicon, metal frames, seal frits, and encapsulation layers in order to minimize thermomechanical stress.

Strain Point, Annealing Point, and Softening Point Determination

The strain point and annealing point are determined by fiber elongation per ASTM C336 or by beam-bending viscometry per ASTM C598, while the softening point (Littleton point) is determined per ASTM C338. These three characteristic temperatures define the annealing schedule, the stress-relaxation behavior, and the maximum permissible process temperature for each glass composition. A high strain point is particularly valued for display substrates because it permits higher deposition and annealing temperatures without irreversible substrate compaction or sag.

High-Temperature Storage and Thermal Shrinkage Testing

Samples are heat treated on a defined schedule, for example 400°C to 600°C for 1 to 10 hours to simulate TFT fabrication, and then measured for irreversible dimensional change using laser scanning or machine-vision metrology. Compaction is commonly limited to the range of 10 ppm to 20 ppm or below for large-generation panel substrates, because higher values degrade inter-layer overlay accuracy. Warp and sag after storage are quantified by non-contact profilometry against pre-test baselines, and post-anneal CTE shifts are checked for materials intended for repeated thermal processing.

Applications

Thermal stability testing of substrate glass supports qualification and reliability programs across the electronics and energy industries:

• TFT-LCD array and color filter substrate qualification for notebook, monitor, and television panels.
• OLED substrate and encapsulation glass, including frit-sealing thermal compatibility verification.
• Semiconductor packaging applications such as glass carriers, TGV interposers, and wafer-level optics.
• Photovoltaic front cover glass and TCO-coated glass for module durability certification.
• Touch panel and chemically strengthened cover glass for consumer, industrial, and medical devices.
• Automotive display modules, where -40°C to +95°C or wider cycling survival is mandatory.

FAQ

What temperature difference must substrate glass withstand to pass thermal shock testing?

Acceptance levels depend on the application. Display substrate coupons typically demonstrate a critical ΔT of 100 K to 180 K by water quench, strengthened cover glass specifications frequently require survival at 120 K or above, and low-expansion borosilicate can exceed 180 K. Specifications are usually written as zero failures at a defined ΔT for a defined sample count, for example 10 out of 10 samples surviving a 140 K shock.

How is thermal cycling different from thermal shock testing?

Thermal shock applies a single, rapid temperature transient to determine the fracture threshold, while thermal cycling applies repeated, slower transitions between temperature extremes to reveal cumulative damage such as crack propagation, film delamination, and dimensional drift. Cycling is performed to IEC 60068-2-14 and GB/T 2423.22, whereas shock methods follow ISO 718, ASTM C149, and related glass standards.

What sample preparation and quantity are required?

Coupons must have ground and polished edges to remove cutting flaws, and samples should be cleaned, dried, and conditioned at 23 ± 2°C and 50 ± 5% relative humidity before testing. Typical programs use 10 to 30 samples per test condition to ensure statistical validity, with thickness, initial dimensions, and initial warp recorded before exposure.

How are test results judged as pass or fail?

Common acceptance criteria include no cracks, chips, or edge defects after exposure; warp and shrinkage within the specified tolerance; retention of fracture strength within a defined percentage of the initial value; and intact coating adhesion. Statistical criteria, such as median thermal shock ΔT exceeding a specified minimum, are applied when stepwise testing to failure is required.

How long does a typical thermal stability test program take?

Thermal shock screening typically requires 1 to 3 working days, a 100 to 300 cycle thermal cycling program takes 3 to 7 days, and CTE or viscosity point measurement requires approximately 3 to 5 working days. A full thermal qualification package, including post-test dimensional and strength verification plus reporting, is generally completed within 2 to 4 weeks by accredited third-party laboratories.