Underfill Epoxy Heat Deflection Temperature (HDT) Evaluation: Standards, Test Methods, and Applicati

Heat deflection temperature (HDT) is one of the most critical thermal-mechanical performance indicators for underfill epoxies used in flip-chip, BGA, and wafer-level semiconductor packaging. This article provides a comprehensive technical guide to underfill HDT evaluation, covering specimen preparation, test conditions, applicable standards such as ASTM D648, ISO 75-2, and GB/T 1634.2, data interpretation, and typical acceptance criteria for electronics reliability qualification. Engineers, materials scientists, and quality professionals can use this reference to design robust test plans and compare underfill materials with confidence.

Technical Overview

Underfill epoxy is a thermosetting polymer adhesive dispensed into the gap between a silicon die and its organic or ceramic substrate, or applied as a film at the wafer level. Its primary function is to mechanically couple the die and substrate so that thermo-mechanical stress generated by the mismatch in coefficient of thermal expansion (CTE) is redistributed away from the solder interconnects. This dramatically improves resistance to thermal cycling, mechanical shock, and board-level bending, which is why underfill is indispensable in flip-chip packages, chip-scale packages (CSP), ball grid arrays (BGA), and stacked-die architectures.

Heat deflection temperature (HDT), also known as deflection temperature under load (DTUL), is defined as the temperature at which a standard cured specimen deflects by a specified amount while subjected to a constant flexural stress and heated at a controlled rate. For underfill materials, HDT serves as a practical engineering measure of the upper service temperature limit, reflecting both the glass transition behavior of the cured network and the retention of flexural modulus at elevated temperature. Because solder reflow, burn-in, and high-temperature operation expose the underfill layer to sustained load and heat, a sufficiently high HDT ensures that the material will not soften, creep, or lose stress-transfer capability during processing and field use.

In cured epoxy systems, HDT is closely correlated with glass transition temperature (Tg), crosslink density, and filler loading. Highly filled, fast-cure capillary underfills typically exhibit HDT values in the range of 130–170 °C under 1.8 MPa stress, while high-Tg no-flow and film underfill formulations can exceed 180 °C. Accurate HDT evaluation therefore requires strict control of specimen curing, conditioning, and test parameters in accordance with recognized international standards.

Testing Samples

Standard Cured Flexural Bars

HDT testing is performed on molded or cast rectangular bars rather than on actual electronic packages. Two specimen geometries are widely used, depending on the selected standard:

• ISO 75-2 / GB/T 1634.2 specimens: preferred dimensions of 80 mm × 10 mm × 4 mm, tested flatwise over a 64 mm support span. Bars may be cut from ISO 3167 multipurpose test plates.
• ASTM D648 specimens: dimensions of 127 mm × 12.7 mm × 3.2 mm to 12.7 mm thickness, tested edgewise over a 101.6 mm support span, with 12.7 mm × 3.2 mm being the most common cross-section.

Specimen Preparation and Curing

Liquid capillary underfills are cast into clean silicone or PTFE molds, degassed under vacuum to eliminate entrapped bubbles, and cured following the manufacturer's recommended schedule, for example 150 °C for 1–2 hours. No-flow underfills are similarly cast and cured through a simulated reflow profile. Wafer-level film underfills are laminated between release liners and press-cured under controlled pressure and temperature. An optional post-cure step, such as 165 °C for 1 hour, is applied when it forms part of the qualified production process, since underfill materials are latent-cure systems whose HDT is highly sensitive to degree of cure.

Conditioning and Sample Quantity

Prior to testing, specimens are conditioned at 23 ± 2 °C and 50 ± 5% relative humidity for at least 40 hours in accordance with ISO 291 or ASTM D618, unless dry conditioning is specified to isolate moisture effects. A minimum of three valid specimens, and preferably five, should be tested for each material condition. Specimens are inspected for voids, warpage, flash, and surface defects, and rejected samples must be replaced to ensure statistical validity.

Testing Items

A complete underfill thermal-mechanical characterization typically includes HDT together with supporting measurements that confirm the state of the cured network. Core and optional test items are listed below:

• Heat deflection temperature under 1.80 MPa (ISO 75-2 Method A / GB/T 1634.2) or 1.82 MPa (ASTM D648 Method A) fiber stress, the standard condition for rigid thermosets.
• Heat deflection temperature under 0.45 MPa (ISO 75-2 Method B) or 0.455 MPa (ASTM D648 Method B) fiber stress, used for comparison with lower-modulus or partially cured materials.
• Glass transition temperature by DSC (ISO 11357-2) or DMA (ISO 6721), to verify cure completeness and correlate with HDT.
• Coefficient of thermal expansion below and above Tg by TMA (ISO 11359-2 or ASTM E831), for package stress modeling.
• Flexural strength and modulus (ISO 178 or ASTM D790), confirming the modulus basis of the HDT result.
• Degree of cure verification, such as residual exotherm by DSC, when cure sensitivity is a qualification concern.

Testing Methods

ASTM D648 Procedure

The specimen is placed edgewise on supports spaced 101.6 mm apart in a silicone oil bath or forced-air oven. A load is applied to produce a maximum fiber stress of 1.82 MPa (264 psi) for Method A or 0.455 MPa (66 psi) for Method B. After a five-minute preload stabilization at 23 °C, the temperature is raised at 2 ± 0.2 °C per minute. The heat deflection temperature is recorded as the bath temperature at which the specimen deflects by 0.25 mm (0.010 in).

ISO 75-2 / GB/T 1634.2 Procedure

The specimen is placed flatwise on supports spaced 64 mm apart, and a bending stress of 1.80 MPa (Method A) or 0.45 MPa (Method B) is applied to the 10 mm face. Following a five-minute stabilization at 23 °C, the heating medium is raised at a uniform rate of 120 K/h, equivalent to 2 °C per minute. The reported HDT is the temperature at which the specimen reaches the specified deflection limit of 0.32 mm for the standard 4 mm thick flatwise bar. Because edgewise and flatwise orientations impose different stress distributions, results from ASTM D648 and ISO 75-2 are similar but not strictly interchangeable, and the standard used must always be stated.

Data Interpretation and Acceptance Criteria

The median value of at least three valid specimens is reported in degrees Celsius, together with the stress level, specimen orientation, dimensions, curing schedule, and conditioning state. For underfill qualification, HDT should exceed the maximum processing or operating temperature encountered by the package with a defined safety margin. Typical acceptance targets are HDT-A ≥ 150 °C for automotive electronics operating at −40 °C to 125 °C or 150 °C, and HDT-A ≥ 180 °C for high-temperature or power packaging. A measured HDT substantially below the DSC-based Tg often indicates incomplete cure, moisture plasticization, or improper specimen preparation and warrants investigation before results are released. All testing should be performed in accredited third-party laboratories or internally audited facilities with calibrated HDT instruments and traceable temperature measurement.

Applications

HDT evaluation of underfill epoxies supports material selection and reliability qualification across the entire advanced packaging supply chain:

• Flip-chip in package (FCIP) and flip-chip on board (FCOB), where underfill must withstand solder reflow and sustained high-temperature operation without softening.
• Ball grid array and chip-scale package corner and edge bonding, where HDT confirms that reinforcement adhesives retain stiffness at elevated ambient temperature.
• Wafer-level chip-scale packaging (WLCSP) and fan-out packaging, where film underfill HDT is matched with molding compound behavior to manage warpage.
• Stacked-die memory modules and 2.5D interposer structures, where multiple cure exposures require verified thermal margins.
• Automotive engine control units, ADAS modules, and power electronics, where qualification to AEC-Q100 Grade 1 or Grade 0 thermal requirements demands rigorously documented HDT data.
• Telecom and 5G infrastructure hardware, in which high power density drives continuous heat exposure of underfilled flip-chip devices.

FAQ

What is a typical HDT value for underfill epoxy?

Most commercial underfill epoxies exhibit HDT between 130 °C and 220 °C under 1.8 MPa stress. Highly filled fast-cure capillary underfills generally fall between 130 °C and 170 °C, while no-flow, molded, and high-Tg film underfills can exceed 180 °C. Exact values depend strongly on resin chemistry, filler loading, and the curing schedule applied to the test specimens.

How does HDT differ from glass transition temperature (Tg)?

Tg is the temperature at which the amorphous polymer network transitions from a glassy to a rubbery state, measured instrumentally by DSC, TMA, or DMA. HDT is an engineering measurement that combines the glass transition with modulus retention under a defined load. HDT usually tracks Tg, but two underfills with identical Tg can show different HDT values because crosslink density, filler content, and post-cure history also affect deflection behavior.

Which stress level should be selected for underfill testing?

Method A at 1.80 MPa (ISO) or 1.82 MPa (ASTM) is the standard choice for rigid thermosets such as cured underfills and provides the best discrimination between formulations. Method B at 0.45 MPa or 0.455 MPa is reserved for comparative testing of lower-modulus materials or for correlation with legacy data sets.

Why must the curing schedule be reported with HDT results?

Underfill epoxies are latent-cure systems, and incomplete crosslinking measurably lowers HDT. Specimens must be cured exactly as in the qualified production process, including peak temperature, dwell time, and any post-cure. Reporting the full cure history is essential for data comparability between material suppliers, accredited third-party laboratories, and end users.

How many specimens are required, and how are results reported?

A minimum of three valid specimens per condition is required by both ASTM D648 and ISO 75-2, although five specimens improve confidence. The median, and preferably the individual values and standard deviation, are reported in degrees Celsius along with the stress level, orientation, specimen geometry, and conditioning state.

Can HDT be measured directly on finished packages?

No. HDT is defined only for standardized bar specimens under controlled flexural stress. The thermal capability of actual packages is instead assessed through TMA, DMA, die shear and adhesion testing, and reliability stress testing such as temperature cycling and high-temperature storage life per relevant JEDEC, IPC, and IEC standards.