This article presents a complete inspection specification for the Martens heat resistance test of underfill adhesives used in flip-chip, BGA, and CSP electronics packaging. It covers specimen preparation, apparatus, test conditions, step-by-step procedure, acceptance criteria, and reporting requirements in accordance with recognized standards such as GB/T 1699, DIN 53462, ISO 75, and ASTM D648. The content is intended for quality engineers, materials scientists, and reliability specialists responsible for incoming material control and supplier qualification.
Technical Overview
Underfill adhesives are low-viscosity, typically epoxy-based encapsulants that are dispensed along the edge of flip-chip, ball grid array (BGA), and chip-scale package (CSP) assemblies. Capillary action draws the liquid adhesive into the gap between the silicon die and the substrate, and thermal curing converts it into a rigid, highly filled polymer that mechanically couples the die to the board. The cured underfill redistributes thermo-mechanical stress away from the solder interconnects, suppresses solder joint fatigue during thermal cycling, and protects the joints from moisture, contamination, and mechanical shock. Because the cured adhesive must retain stiffness and dimensional stability at elevated operating, burn-in, and rework temperatures, its heat resistance is a critical acceptance property.
The Martens heat resistance temperature, also called the Martens temperature, is the temperature at which a rigid cured specimen reaches a defined deflection under a specified bending stress while being heated at a constant rate. For cured underfill materials, the test is generally performed according to the Martens method described in GB/T 1699 and the equivalent DIN 53462 procedure. Related but distinct thermal indicators include the heat deflection temperature (HDT) determined per ISO 75 and ASTM D648, the Vicat softening temperature per ISO 306 and GB/T 1633, and the glass transition temperature (Tg) measured by DSC or DMA per ISO 11357. These properties are correlated because they all reflect the onset of softening in the glass transition region, yet they are not interchangeable and must be evaluated against their own specifications.
- Purpose of the specification: to define a uniform method for measuring and judging the Martens heat resistance temperature of cured underfill adhesives.
- Scope: incoming material inspection, supplier release testing, qualification of new grades, and validation after formulation or process changes.
- Governing methods: GB/T 1699 and DIN 53462 for the Martens test, with ISO 75 / ASTM D648 and ISO 306 / GB/T 1633 as supplementary thermal performance methods.
Testing Samples
Standard Martens Specimen Bars
The Martens test is performed on cast or machined cured bars with nominal dimensions of 120 mm in length, 15 mm in width, and 10 mm in thickness, with dimensional tolerances as specified in the applicable standard. The bars must be produced from the same underfill formulation, mixture ratio, and cure profile as the production material, including any required post-cure. Specimen surfaces shall be free of bubbles, voids, cracks, and unmixed streaks, and edges shall be smooth and free of chips. Specimens containing visible defects or thickness non-uniformity exceeding the permitted tolerance shall be rejected before testing.
Sampling Plan
At least three valid specimens shall be tested from each sample lot, and five specimens are recommended whenever the result is used for formal acceptance decisions. Samples shall be taken from representative containers of the lot in accordance with the agreed incoming inspection plan, and retention samples shall be stored under the conditions specified by the material supplier. For two-part underfill systems, the hardener and resin shall be mixed at the specified ratio, degassed if required, and cast within the stated pot life.
Conditioning Before Test
Cured specimens shall be conditioned in a standard laboratory atmosphere of 23 ± 2 °C and 50 ± 5 % relative humidity per GB/T 2918 and ISO 291 for a minimum of 24 hours, preferably 48 hours, prior to testing. Specimens shall be protected from contamination, ultraviolet exposure, and thermal history changes between cure and test. Any deviation from the standard conditioning shall be documented in the test report.
Testing Items
The primary inspection item is the Martens heat resistance temperature of the cured underfill adhesive, expressed in degrees Celsius. The following checks and supplementary items constitute the full inspection content:
- Specimen appearance verification: absence of voids, cracks, delamination, and cure irregularities.
- Dimensional verification: length, width, and thickness measured with a caliper or micrometer accurate to 0.02 mm.
- Martens heat resistance temperature (°C): primary acceptance item.
- Cure verification (when specified): degree of cure or residual exotherm by DSC per ISO 11357 to confirm the cure profile was achieved.
- Supplementary thermal data (optional or by agreement): Tg by DSC or DMA, coefficient of thermal expansion by TMA, HDT per ISO 75 / GB/T 1634, and Vicat softening point per ISO 306 / GB/T 1633.
Testing Methods
Apparatus
A Martens heat resistance tester shall be used, comprising a stirred and temperature-controlled oil bath with a suitable heat transfer medium, vertical specimen holders, a calibrated lever and weight loading system, and dial indicators or electronic displacement sensors for reading deflection. Temperature shall be measured with calibrated thermometers or thermocouples accurate to within ± 0.5 °C, and bath temperature uniformity shall be maintained so that all specimens experience essentially the same thermal environment. The applied load shall be calculated for each specimen so that the nominal surface bending stress equals 5 MPa, using the measured specimen width and thickness and the lever geometry defined by the apparatus standard.
Test Conditions
- Initial bath temperature: at least 50 °C below the anticipated Martens temperature of the material.
- Heating rate: 50 °C/h, uniform and continuous, with the tolerance stated in the referenced standard.
- Applied bending stress: nominal surface stress of 5 MPa at the outer fiber of the specimen.
- Endpoint criterion: the temperature at which the indicating pointer registers a deflection of 6 mm.
- Ambient conditions: standard laboratory atmosphere as defined for conditioning.
Test Procedure
- Measure and record the width and thickness of each specimen at three positions and verify conformity to the dimensional requirements.
- Clamp each specimen vertically in the holder in the orientation defined by the apparatus standard so that bending occurs about the specified axis, and confirm firm, defect-free seating.
- Calculate the required weight for each specimen based on its measured cross-section and mount it on the loading lever to produce the 5 MPa bending stress.
- Set the deflection indicator to zero after the load is applied and the system has stabilized.
- Immerse the loaded specimens, start the bath agitation, and heat at 50 °C/h without interruption.
- Observe the deflection indicator continuously and record the bath temperature at the exact moment the deflection reaches 6 mm; this temperature is the Martens heat resistance temperature for that specimen.
- Repeat the measurement for all valid specimens and inspect each tested bar after removal for anomalies that could invalidate the result.
Calculation, Precision, and Reporting
The Martens heat resistance temperature of the lot is reported as the arithmetic mean of the valid individual results, rounded to 1 °C. Individual values from sound specimens should normally agree within a few degrees; a spread greater than 3 °C shall trigger an investigation into specimen quality, cure state, or apparatus calibration before the result is accepted. The test report shall include material identification and lot number, cure and post-cure schedule, specimen dimensions and number, conditioning details, heating rate, applied stress, individual and mean Martens temperatures, the standard referenced, the testing laboratory, and the test date. Testing may be performed by an in-house laboratory or by accredited third-party laboratories operating under an ISO/IEC 17025 quality system.
Acceptance Criteria and Failure Handling
The mean Martens heat resistance temperature shall meet or exceed the minimum value stated in the product datasheet or purchase agreement. Typical cured capillary and no-flow underfill adhesives exhibit Martens temperatures in the approximate range of 110 °C to 160 °C depending on the resin system, filler loading, and cure profile, with values of 130 °C or higher commonly specified for general-purpose flip-chip grades and higher values for high-Tg automotive and high-temperature applications. If a lot fails the requirement, the batch shall be quarantined and the root cause investigated, including cure profile verification, mixing ratio of two-part systems, storage life and pot life compliance, and specimen casting quality. Retesting shall use freshly cast bars and, where possible, a confirmed reference material to validate apparatus condition.
Applications
The Martens heat resistance inspection specification is applied throughout the electronics packaging supply chain wherever cured underfill adhesives must demonstrate stable mechanical performance at elevated temperature:
- Incoming material inspection of capillary underfills, no-flow underfills, and molded underfill compounds at assembly and packaging sites.
- Supplier qualification, annual requalification, and change notification assessment for underfill material suppliers.
- Validation of new cure profiles, ovens, or reflow conditions that could affect the final crosslink state of the adhesive.
- Reliability engineering and failure analysis supporting investigations of solder joint cracking, underfill delamination, or warpage at high temperature.
- Automotive, industrial, and high-density computing applications where underfills operate near or above conventional glass transition regions and require documented thermal endurance data.
- Research and development of new epoxy and modified resin formulations, where Martens temperature is tracked alongside Tg, modulus, and coefficient of thermal expansion.
FAQ
What exactly is the Martens heat resistance temperature of an underfill adhesive?
It is the temperature at which a cured 120 mm × 15 mm × 10 mm bar of the adhesive, loaded in bending at a nominal surface stress of 5 MPa and heated at 50 °C/h, deflects by 6 mm at the indicating point. It provides a single, reproducible number describing the onset of softening of the cured thermoset under load.
How does the Martens temperature differ from Vicat softening point and heat deflection temperature?
The three methods apply different loads and endpoints. The Martens test uses a 5 MPa bending stress and a 6 mm pointer deflection; the Vicat test per ISO 306 measures flat-ended needle penetration of 1 mm under a 10 N or 50 N load; and HDT per ISO 75 and ASTM D648 measures deflection under edgewise or flatwise bending stresses of 1.80 MPa or 0.45 MPa. All three fall within or near the glass transition region, but their numeric values differ systematically and shall not be compared directly.
Can the test be performed on an actual assembled circuit board?
No. The Martens method requires standardized cast bars with controlled dimensions and loading geometry. Assembled hardware is instead evaluated with complementary techniques such as DSC or DMA on cured coupon samples, thermomechanical analysis, or functional reliability testing of the assembled package.
What is a typical acceptance value for underfill adhesives?
Acceptance values are grade-specific and defined in the supplier datasheet or the purchase specification. General-purpose flip-chip underfills commonly require Martens temperatures of at least 130 °C, while high-Tg automotive or high-temperature grades may require 150 °C or more. The specification limit shall always be agreed contractually rather than assumed.
Who is qualified to perform this inspection?
Any laboratory equipped with a calibrated Martens tester and operating in accordance with GB/T 1699 or DIN 53462 may perform the test, including manufacturer in-house laboratories and accredited third-party laboratories. Calibration of the temperature measurement, heating rate control, and loading system is a prerequisite for valid results.
How should a failed result be handled?
The lot shall be quarantined and the failure verified by retesting freshly cast specimens from the same batch. The investigation shall cover the cure schedule, mixing ratio for two-part systems, storage and pot life compliance, casting defects, and apparatus calibration before a disposition of reject, concession, or supplier corrective action is issued.