Introduction
Hermetic seal failure in metallized ceramics is a common reliability issue in electronic packaging, directly affecting device lifespan and stability. This paper systematically analyzes the causes of hermetic seal failures and, based on microstructural images and process data, provides preventive measures and control suggestions, offering a reference for technicians engaged in ceramic packaging.
What is a Hermetic Seal?
A ceramic hermetic seal refers to a sealing structure created by reliably bonding ceramics to metals using processes such as brazing, resulting in a high hermetic seal capable of stable operation in long-term vacuum or pressure environments. Hermetic seals offer excellent electrical insulation, high-temperature resistance, high mechanical strength, and good corrosion resistance, making them suitable for high-vacuum and harsh operating conditions. Typical applications include electronic packaging, sensor packaging, and vacuum devices.
Common Structure of Ceramic-to-Metal Seals
A typical structure of a ceramic-to-metal seal consists of the following layers:
Metal (e.g., Kovar) → Brazing layer → Nickel plating layer → Metallization layer → Ceramic body
This is a multilayer structure composed of different materials, with physical and chemical interfaces between each layer. The effectiveness of a hermetic seal depends on the integrity and reliability of each interface. The photo below shows the microstructure of the ceramic-metal sealing interface. A good microstructure should have clear, well-defined, intact, free of overlap, and of appropriate thickness.
Root Causes of Hermetic Seal Failures—Three Interfaces
In cases of hermetic seal failure, the root cause can often be traced back to the following three interfaces, which deserve our close attention.
Interface Between Ceramic Body and Metallization Layer
The interface between the ceramic body and the metallization layer is the weakest point in the entire sealing structure, and most hermetic seal failures originate here. The sealing process requires adhering these two entirely different materials together, but achieving a perfect seal is not a simple task. The following situations on this interface can lead to hermetic seal failure:
1. Defects in the Metallization Layer Structure
During the sealingprocess between the ceramic body and the metallization layer, if the metallization temperature or formula is not properly controlled, or the thickness of the metallization layer is uneven, it can easily lead to local stress concentration at the interface, resulting in microcracks or even local detachment.
2. Mismatch in Thermal Expansion Coefficients
The thermal expansion coefficients of alumina ceramic and Mo particles in the metallization layer differ, generating tensile stress during cooling, which may lead to overall delamination of the interface.
3. Abnormal Distribution of the Glass Phase at the Interface
In sealing systems containing glass phases, the glass phase plays an important role in interfacial bonding and stress buffering. When improper sintering process control leads to uneven or discontinuous distribution of the glass phase, it weakens the interfacial bonding ability and causes brittle failure under thermal stress.
Transition Zone Between Metallization and Nickel Plating
After the metallization layer is sintered, a nickel plating layer (3-5 μm) is applied to its surface. This interface is the boundary between the metallization layer (Mo + glass phase) and the nickel plating layer. The nickel plating layer has two functions: first, to protect the metallation layer from oxidation; and second, to improve weldability, as the wetting properties of brazing alloys on nickel are far superior to those on Mo. If problems arise at this interface, it can lead to severe hermetic seal failure or electrical disconnection. The following conditions in the transition zone can lead to hermetic seal failure:
1. Insufficient Cleaning or Activation Before Electroplating
After sintering, the metallization layer may have a very thin oxide film or residual contaminants on its surface. If the cleaning or activation treatment is not thorough, it will affect the initial wetting and bonding of the nickel plating layer, resulting in a state of “surface coverage but insufficient bonding” at the interface. Under subsequent thermal cycling or mechanical stress, peeling or localized detachment is likely to occur.
2. Uncontrolled Electroplating Stress
When the temperature, pH value or current density is unstable during the electroplating process, high internal stress may be generated inside the plating layer. This stress is released during subsequent thermal cycling or brazing, which may induce microcracks or weaken the interfacial bonding, and in severe cases, increase the risk of hermetic seal failure.
3. Uneven Metalization Layer Composition
If the distribution of Mo particles or active components in the metallization layer is uneven, it will lead to different degrees of local metallization reaction, resulting in differences in interfacial bonding strength. During subsequent service, these weakly bonded regions are more likely to become failure initiation points, affecting overall reliability.
4. Pinholes in Electroplating
The presence of microscopic pinholes in the plating layer is usually related to impurities in the plating solution, micro-defects on the ceramic body surface, or improper control of electroplating parameters. During subsequent brazing, these pinholes may become channels for localized penetration of brazing filler metal, or they may lead to poor localized wetting, thus becoming the starting point for interface failure or incompatibility, thereby affecting the reliability of the seal.
Interface Between Nickel Plating and Brazing Layer
This is the final interface formed during the sealing process—the area where the brazing filler metal, after melting, flowing, and solidifying, tightly bonds with the nickel plating layer. It is the last link in the entire sealing chain, responsible for completing the final sealing task: the brazing filler metal and the plating layer must fully react to form a continuous, dense, and void-free weld.
The problem with this interface is directly reflected in the helium test failure. However, a more insidious issue is the “factory qualified, but leaking during use” problem—the defect hasn’t yet penetrated the system, but it gradually worsens during subsequent thermal cycling. Below are common problems and their root causes:
1. Temperature Control Issues
If the peak temperature is too low, the brazing filler metal will have poor fluidity, leading to poor wetting. Conversely, if the peak temperature is too high or the holding time is too long, the brazing filler metal will flow excessively, causing excessive dissolution of the Mo metallization layer and forming an excessively thick brittle intermetallic compound.
2. Physical Defects
These are usually related to process condition control. For example, an unreasonable temperature profile or uneven clamping pressure may cause slight deformation of the metal lid or affect the flow and venting process of the molten brazing filler metal. During brazing, if gas fails to escape in time and is trapped in the molten brazing filler metal, it will form localized pores or voids after solidification. These defects weaken interfacial continuity and are a significant factor affecting hermeticity.
3. Nickel Plating Oxidation and Improper Filler Control
If metallized ceramics are stored for extended periods or under improper conditions, an oxide layer (NiO) may form on their surface. This reduces the wettability and interfacial bonding between the brazing filler metal and the nickel layer, thus affecting the formation of the metallurgical bond. Furthermore, improper filler metal usage can also affect sealing quality: insufficient filler metal may result in incomplete gap filling, while excessive filler metal may cause uncontrolled flow or localized overflow, and in some cases, even lead to interfacial erosion or abnormal structural stress, thereby affecting overall reliability.
The Internal Connections of the Three Interfaces
These three interfaces are not isolated but interconnected—a problem at any interface breaks the entire sealing chain.
The first interface (ceramic/metallization layer) is the foundation of the entire sealing. The quality of glass phase migration during sintering determines the strength of this layer’s “grip”—it not only supports itself but also determines whether the layer above can stand firmly.
The second interface (metallization layer/nickel plating layer) is a chemical barrier. If the plating is not tight (blistering, pinholes), during brazing at the third interface, the molten brazing filler metal will penetrate downwards through these gaps, causing the previously mentioned excessive dissolution of the Mo metallization layer, fundamentally destroying the seal.
The third interface (plating/brazing layer) is the final sealing layer and the endpoint of stress transmission. Due to the different coefficients of thermal expansion of the three materials, the shrinkage stress during cooling will be transmitted downwards layer by layer. If an excessively thick brittle intermetallic compound forms in the third layer, it will not only fail to buffer stress but will also directly transfer the pressure to the second or even the first layer, leading to plating delamination or ceramic cracking.
These three interfaces are interconnected. This is why some products pass helium testing at the factory but gradually leak after thermal cycling—a single microscopic defect initially does not penetrate, but under repeated thermal stress, the weak points at multiple interfaces gradually expand and connect, eventually forming a through-hole leakage channel.
Key Points for Process Control
Given the failure mechanisms associated with the three interfaces described above, we can implement control measures during production by focusing on the following key stages:
| Control Stage | Control Objective | Target Interface |
| Ceramic processing | Surface roughness Ra 0.8-1.6μm, no contamination | Interface 1 |
| Metallization layer sintering | Uniform glass phase distribution; thickness 15-25μm | Interface 1 |
| Pre-plating activation | Thoroughly removes oxide film and glass phase residue | Interface 2 |
| Plating layer control | Thickness 3-5μm, no pinholes, no peeling. | Interface 2 |
| Brazing temperature | Peak temperature accurate to +/- 5℃, with appropriate holding time | Interface 3 |
| Brazing pressure | Evenly distributed, with sufficient pressure at corners | Interface 3 |
| Brazing filler metal dosage | Calculate precisely to avoid insufficient or excessive dosage | Interface 3 |
| Storage control | Vacuum packaging to prevent oxidation of the plating layer | Interface 3 |
Conclusion
Hermetic seal failure can usually be traced back to microstructural changes at three key interfaces. The failure of each interface is not caused by a single factor. Through systematic failure analysis, identifying the specific interface where the failure occurred and the dominant factors, and optimizing process control, we can continuously improve sealing quality and reliability.
If you have encountered similar hermeticity problems in your daily production where the cause is unclear, please feel free to discuss and exchange ideas with us—sometimes a different perspective might just reveal the answer.
Frequently Asked Questions
Q1: Is hermetic seal failure always due to sealing process issues?
A1: Not necessarily. Poor brazing wettability and brazing temperature are common causes of hermetic seal failure during the sealing process. However, besides this process, multiple factors such as material compatibility, the bonding strength and thickness of the metallization layer, and defects in the ceramic itself all have a direct impact.
Q2: Why does the product still leak even though it passed the factory helium test?
A2: The product may leak even though it passed the factory helium test, possibly due to the following reasons:
1. Stress-induced defect propagation: Microcracks, weak interfacial bonding, or residual stress areas present at the factory gradually develop under subsequent thermal cycling or mechanical stress, eventually forming a through-hole leakage path.
2.Misjudgment of a large leak: The leakage path is too large, and the helium has already leaked out before testing, leading to a misjudgment of passability.
3.Cavity connection: Isolated cavities within the weld gradually connect during temperature cycling, forming a leakage path.
Q3: How can I quickly determine with the naked eye which interface the hermetic seal failure occurred on?
There are several quick methods:
1. If the fracture surface reveals white or the original ceramic color, it usually indicates insufficient bonding between the ceramic and metallization layers (a sintering process issue).
2. If the fracture surface has a gray, powdery metallic appearance, it’s likely due to insufficient internal strength of the metallization layer.
3. If the brazing filler metal and ceramic have separated, but a complete metallization layer remains on the ceramic, it indicates poor wetting of the metallization layer/brazing filler metal.
