HTOL vs THB vs HAST vs Temperature Cycling vs HTS: Conditions, Purpose and Failure Modes Compared
Most semiconductor reliability tests are accelerated tests: they apply stresses such as temperature, voltage, humidity and temperature changes more strongly than in actual use to speed up failures, checking lifetime and failure rate in a short time. HTOL operates devices at high temperature under voltage and evaluates degradation of the chip itself, such as TDDB and electromigration. The temperature-humidity-bias test (THB) and HAST are moisture-resistance tests that use humidity and voltage to accelerate moisture-driven failures such as corrosion; HAST uses pressurized steam to perform a similar evaluation in less time. Temperature cycling repeats high and low temperatures to evaluate fatigue caused by differences in thermal expansion between materials, and the high-temperature storage test evaluates degradation driven by temperature alone, without voltage. Because each targets different failures, product qualification combines them.
| Aspect | HTOL | Temperature-humidity-bias (THB) | HAST | Temperature cycling | High-temperature storage (HTS) |
|---|---|---|---|---|---|
| Purpose | Check chip lifetime and failure rate | Check package moisture resistance | Check moisture resistance in a short time | Check the resistance of joints and packages to thermal fatigue | Check resistance to degradation driven by temperature alone |
| Stresses applied | High temperature and voltage | Temperature, humidity and voltage | Temperature above 100 °C, humidity, pressure and voltage | Temperature changes (repeated high and low temperatures) | High temperature only |
| Bias applied? | Yes. Devices are operated with signals at or slightly above the maximum operating voltage | Yes. Applied so that power dissipation stays small, to limit self-heating | Usually yes. Unbiased HAST, without bias, also exists | Generally no | No |
| Typical conditions | Around 125 °C. Higher for some temperature grades | 85 °C/85% RH at atmospheric pressure (the 85/85 test) | Pressurized steam at about 110 to 130 °C and 85% RH, e.g. 130 °C/85% RH | e.g. −55 to 125 °C or −65 to 150 °C, holding for a set time at each extreme | e.g. 150 °C. Higher for some products |
| Typical duration | About 1,000 hours, with interim measurements at e.g. 168 and 500 hours | About 1,000 hours | About 96 hours at 130 °C/85% RH | Hundreds to thousands of cycles | About 1,000 hours. Sometimes shortened at higher temperature |
| Main mechanisms accelerated | Degradation and latent defects of the chip itself, such as TDDB, BTI and electromigration | Electrochemical reactions driven by moisture ingress (corrosion, ion migration) | Same as THB, accelerated more strongly by faster moisture ingress | Fatigue from repeated strain due to differences in thermal expansion coefficients | Degradation driven by temperature alone (intermetallic growth, thermal degradation of materials, loss of stored charge) |
| Typical defects found | Electrical parameters out of spec (e.g. threshold voltage shift), functional failures | Corrosion of aluminum pads and interconnects, increased leakage current, shorts between conductors | Same as THB: corrosion, increased leakage current, shorts | Cracks in solder joints and wire bonds, mold compound delamination, die and passivation cracks | Increased resistance or lifting of wire bonds (Kirkendall voids), data retention failures in nonvolatile memory |
| Acceleration model | Arrhenius model (temperature) and a voltage acceleration model | Peck's model (a power of relative humidity times an Arrhenius temperature term) | Peck's model (same as THB) | Coffin-Manson relationship (a power of the temperature swing) | Arrhenius model |
| Sample condition | Packaged parts mounted on boards and operated under voltage | Packaged parts. In qualification, often run after preconditioning that applies moisture soak and reflow heat | Packaged parts. Like THB, often run after preconditioning | Individual components (often after preconditioning); board-level tests with parts mounted also exist | Packaged parts placed in a hot chamber. No electrical connection is needed |
| Role (example standards) | Qualification test for lifetime and failure rate. JEDEC (JESD22 series), AEC-Q100 and others | Qualification test for moisture resistance. JEDEC (JESD22 series), AEC-Q100 and others | Sometimes used in place of THB. JEDEC (JESD22 series), AEC-Q100 and others | Qualification test for packages and joints. JEDEC (JESD22 series), AEC-Q100 and others | Qualification test for temperature-only degradation. JEDEC (JESD22 series), AEC-Q100 and others |
| Cautions | Control temperature so self-heating does not push the junction temperature too high | If the chip heats up, relative humidity at its surface drops and humidity-driven failures become less likely | The high temperature can cause failures that never occur in actual use, such as changes in mold compound properties | A different test from thermal shock (faster temperature changes) and power cycling (self-heating) | Without voltage, it cannot evaluate degradation driven by electric field or current, such as TDDB and electromigration |
Figures are typical orders of magnitude and vary by process generation and product.