Packaging technologies for NTC thermistors have been continuously iterated with the consistent goal of enhancing sealing performance, temperature resistance and long-term reliability. Early temperature sensors mostly adopted epoxy-packaged components. While such solutions feature low cost and simple processes, they exhibit two inherent defects under the wide-temperature-range and long-service-life operating conditions of new energy systems.

First, poor thermal expansion matching: Epoxy resin has a significant mismatch in the coefficient of thermal expansion with ceramic chips and metal leads. Under prolonged thermal cycling, chip cracking and silver electrode delamination tend to occur, causing resistance to rise continuously until complete failure.

Second, weak hermetic protection: The epoxy material itself lacks sufficient densification, allowing moisture to easily penetrate into components. Under DC bias, this triggers silver ion migration, resulting in abnormal resistance drop and insulation failure. These two fatal defects cannot meet the long-term reliability requirement of 8–10 years.

To address the core pain points of epoxy packaging, the industry has gradually shifted to glass packaging solutions. Inorganic glass delivers better matching of thermal expansion coefficients with ceramic chips and Dumet leads, generating low internal interfacial stress after high-temperature sealing and mitigating cracking induced by temperature variation. Meanwhile, glass boasts excellent airtightness, which effectively blocks moisture, oil contaminants and corrosive gases. Its upper temperature resistance limit and anti-aging capability are improved by orders of magnitude compared with epoxy packaging, resolving the two chronic drawbacks of epoxy solutions from the packaging root.

Glass-packaged components fall into two mainstream structures. The first is the axial lead structure, with two leads extending reversely from both ends of the glass. During assembly, one lead must be bent by 180°, which not only easily causes mechanical damage at the joint between glass encapsulation and leads, but also prevents the chip from closely fitting the sensor housing, greatly reducing the temperature response speed.

The second structure is the radial single-end lead design (also known as single-end glass-sealed structure). The leads require no bending; under identical specifications, the chip can be positioned directly at the front end of the sensor housing, delivering the shortest heat conduction path and fastest response speed. Besides, neatly arranged parallel leads are compatible with automated welding production lines, making this structure gradually become the mainstream core component for new energy temperature sensors.

Nevertheless, this radial single-end glass-sealed thermistor, which solves the pain points of epoxy packaging, still carries unavoidable inherent structural weaknesses. Two leads extend in parallel from the same side of the glass tail, with spacing equivalent only to the thickness of the thermistor chip. For assembly of temperature sensors, the leads must be forcibly stretched to widen the gap — this is the core inducement for microcracks forming at the glass-lead joint. Whether glass-packaged products adopt axial or radial leads, once the glass encapsulation breaks, moisture may ingress. Under the effect of an electric field, metal ion migration may occur, driving resistance to decline progressively until full failure.

Targeting this pain point, two mature improvement routes have been developed domestically to support large-scale industrial application. Fundamentally, however, both approaches operate on the premise that "glass will crack, followed by remedial measures", and cannot eliminate failure risks at the source.

Most NTC temperature sensors deployed in new energy vehicles and energy storage equipment in China adopt conventional radial single-end glass-sealed NTC thermistors as core components. The industry generally improves finished product reliability via peripheral measures including optimized secondary packaging processes, multi-layer resin potting and metal housing installation.

This approach can slow moisture ingress and buffer external vibration to satisfy operating requirements under conventional working conditions, yet it belongs entirely to "patch-style peripheral optimization" with three insurmountable limitations: ◾ Internal microcracks generated during assembly cannot be eliminated; defects are formed prior to packaging. ◾ Organic potting materials continuously age under long-term high temperature, with moisture permeability rising year by year, making them incapable of providing 8–10 years of long-lasting protection. ◾ The intrinsic structure featuring stress concentration at the sealing interface between leads and glass, and at glass-lead joints remains unchanged, as does the physical property enabling easy water molecule penetration. Such measures can only delay failure time instead of eradicating risks.

To tackle the two pain points of "cracking upon stretching narrow-spaced leads and subsequent ion migration", the industry has developed an improved solution of "structural load reduction + material backup": cross bonding replaces parallel bonding to pre-widen lead spacing before delivery and reduce tensile force during assembly. Gold electrodes and gold-plated leads are also adopted to suppress silver ion migration even if glass cracking occurs.

Although this scheme lowers the probability of crack formation during assembly, it still follows the logic of passive remediation with clear underlying deficiencies: ◾ It merely reduces crack probability instead of arresting crack propagation and solder joint failure. Wider spacing only lessens assembly tensile stress; welding thermal shock and long-term vibration & temperature cycling can still initiate microcracks. Glass is inherently hard and brittle with no toughness. Initial cracks keep expanding under cyclic stress, spread to sealed lead joints and cause interface separation, poor contact, fluctuating temperature measurement signals, and eventually product failure. ◾ It prevents ion migration but cannot block moisture erosion of chips. Gold electrodes and gold-plated leads deliver no hermetic capability. Once cracks penetrate the encapsulation, continuous moisture intrusion corrodes the thermistor ceramic chip over time and triggers gradual resistance drift — a latent fault difficult to screen before delivery and progressively deteriorating during service. ◾ Precious metal processes substantially increase costs. Raw gold materials plus multiple special procedures including electroplating, cleaning and inspection significantly raise unit manufacturing costs, creating prominent cost pressure for large-scale matching in vehicle and energy storage scenarios.

Solutions including peripheral packaging reinforcement, wide-spacing design with gold electrodes and gold-plated leads are all passive remedial optimizations. They fail to eliminate stress concentration at the structural level and cannot fundamentally resolve the core issue of glass cracking. For new energy equipment designed with an 8–10 year service life, major hidden risks persist regarding reliability margin during long-term operation.

In the global high-end automotive and industrial energy storage temperature measurement sectors, glass-sealed NTC thermistors with ceramic bases are recognized as high-reliability core components, and represent a fundamental structural solution to mitigate cracking at the glass-lead joint of glass-sealed components.

This structure incorporates a ceramic base at the glass-lead joint of radial glass-sealed components encapsulating thermistor chips. At high temperatures, the glass phase wets the ceramic substrate and induces interfacial elemental interdiffusion to form a dense transition layer with chemical bonding. This upgrades the original line-type force bearing and line sealing to annular surface force bearing and surface sealing, dispersing stress from the source and extending the hermetic sealing path. It can mitigate thermal shock and mechanical tensile stress generated during sensor fabrication and welding to reduce thermistor microcracks at the source; additionally, it resists long-term equipment vibration and temperature cycling, greatly boosting airtight performance and significantly increasing the difficulty of moisture penetration. NTC temperature sensors assembled with such thermistors achieve reliability and service life improved by orders of magnitude compared with ordinary glass-sealed components.

Nevertheless, mass production of this structure entails high technical thresholds. Core challenges including precise matching of thermal expansion coefficients between glass and ceramic, control of high-temperature integrated sealing processes, and consistency assurance for mass production have long remained unbroken for domestic enterprises.

During the period when domestic manufacturers cannot independently mass-produce glass-sealed thermistors with ceramic bases, two alternative solutions have emerged domestically to meet the demand of long-lifespan equipment. The first relies on external multi-layer resin encapsulation, potting and metal housing sealing of temperature sensors for peripheral reinforcement. The second adopts structural and material improvements featuring cross-bonded wide-spacing leads paired with gold electrodes and gold-plated leads. In some scenarios, peripheral packaging reinforcement is applied simultaneously. Both solutions can satisfy the usage requirements of medium-to-long-life equipment to a certain extent, but neither addresses the essential structural problem of stress concentration at the glass-lead joint of conventional single-end glass-sealed components. A do