Why CVD Coatings Fail After Thermal Cycling: Fatigue Analysis

CVD coatings are applied to a variety of high-heat tools due to their ability to withstand harsh environments. However, as time passes, changes do occur after multiple heating and cooling cycles. The coating or the bonding layer may begin to crack. Initially, they are difficult to notice. Then they develop and cause peeling or flaking. This is frequently associated with thermal fatigue; the material is gradually weakened by alternate heating and cooling. In the real production lines, this is manifested by shorter service life of parts and sudden coating failure, although the outside surface is still good.

Thermal Fatigue Mechanism in CVD SiC Coating and TaC Coatings

If a coated part is heated and cooled repeatedly, then it will not act the same each time. Cvd coating SiC and TaC coatings are temperature resistant, yet still are susceptible to stress from repeated thermal cycles. The issue has to do with the various expansion and contraction rates of the various layers. When there are any differences in the coating and the base material (graphite), internal stress will build up over time. Typically, the structure of CVD SiC coatings is extremely hard and stiff. This not only increases the wear resistance of the coating, but also reduces its ability to withstand high temperature and low temperature impacts. A fast heating will cause the outer to expand and the inner to expand slower. This ongoing tension and stress may lead to minute cracks over the years. They are initially hidden, then they start to link after many cycles and cause deterioration of the coating. The stress pattern is similar in case of TaC coating; however, small defects in the coating deposition may result in a more irregular stress pattern. Some users have reported on edge chipping or flaking of parts coated with TaC when they operate the epitaxy tool for a long time.For real production some users reported edge chipping or flaking of TaC coated parts after long production times using the epitaxy tool.In the actual production, it is reported that after using the TaC coated graphite parts for a long time, edge chipping or flaking phenomenon was observed in the parts. This is often seen close to the gas flow paths or heating edges, where there are greater temperature fluctuations.

Crack Initiation and Propagation Under Repeated Heating Cycles

In general, there are no cracks in both the CVD SiC and CVD TaC coatings at the same time. Almost imperceptible and start in early heating cycles, small. The top layer will heat up first and the bottom layer will take longer to heat up. This is an unprecedented one, which will exert pressure on the coating. As the separations begin at the weak points (grain boundaries), they grow and spread as the temperature drops. As the temperature cools, the small separations begin to increase at weak points (grain boundaries) and spread. These are the first signs of the beginning of crack initiation. Initially, the microcracks are short in length and are few in number. Most of them are stable when operating normally and are difficult to detect in the visual inspection. There is an issue if the part is repeatedly heated and cooled. Stress is applied to the weak areas in successive cycles – one after the other. Over time, the microcracks start to grow in length. Once the crack reaches a certain critical length the crack no longer is contained. Starting to link to adjoining cracks. This is the point at which propagation is increased. For the Cvd sic coating , the material is stiff, and the cracks will be straight. The small defects occurred in the coating process and/or surface contamination during the coating prevented the crack path from being as regular as in other coatings, in the case of TaC coatings. One of such applications is the HTO employing a high-temperature reactor with a graphitic susceptor. The operators get an unexpected increase in the amount of particles over time when operating again. When the coating is then examined, the primary cracks typically are found to be present over a wide area however. This is because the cracks gradually grew to a critical extent, causing the sudden change. The key to the matter is that cracks do not grow in any direction. Is able to withstand thermal cycling stresses. If cracks join and become large enough, the strength of the coating decreases rapidly. That's when flaking or delamination will happen, usually due to a slight temperature change and will be noticeable when using.

Impact of Process Conditions in LPE (ASM) and AIXTRON Systems

The lifetime of CVD SiC and TaC coatings depends strongly on the process conditions in the reactors, such as those in AIXTRON reactors or ASM International LPE systems. Slower changes in temperature, gas flow or pressure can have a slight impact on how stress is developing within the coating. In a number of production runs, it is one of the largest factors in the heating and cooling speed. If the system gets too hot, the outer layer of the coating expands before the inner layer. This causes stress in the interface area. This can begin to cause small cracks to appear much sooner than anticipated when it occurs several times. Typically this stress is minimized when cycle times are slower, but during high throughput operation, operators might try to squeeze out more cycles, increasing the stress on the coating. The gas flows are also very important. In tools such as AIXTRON epitaxy systems, uneven gas distribution can lead to local hot spots on parts being coated or susceptors. These hot spots grow differently from other cooler hot spots nearby. Over time this unequal heating results in a concentration of stress in certain zones. Hence, there are repeating patterns that tend to crack rather than random areas. Another secret factor is pressure stability. ASM International's LPE systems can alter the density of the coating with small pressure changes during deposition. If there is a slight porosity and/or non-uniformity of the coating, it has little thermal cycling property. The cracks can creep faster through the weaker, porous parts than through the dense, well-formed parts. One example of this is long production runs whereby the edge damage occurs to a coated graphite susceptor before any other defect. In many instances, the central section is perfectly good, while the outer sections have a tendency to fail sooner due to increased temperature changes and exposure to gas. Sometimes the operators believe it is a coating quality problem, but often the actual process conditions are the primary cause, rather than the coating quality.

Coating life is not only a matter of materials usage in real use. It will also depend on the stability and balance of the process conditions over hundreds of cycles. Over time, the tiny drift accumulates and slowly takes the coating to the brink.

Role of Coating Thickness and Residual Stress Distribution

The thickness of the CVD SiC and TaC layers has a significant influence on their ability to pass thermal cycling. If it's thin, it will wear off easily. Does not distribute the stress very well, so cracks progress towards the base material sooner. A very thick coating can also be a problem, however. It can hold in more internal stress when it is being deposited; the internal stress does not disappear as soon as it is cooled. Each layer of the coating will react slightly differently at high temperature use. The upper layers heat up and cool down quicker than the lower layers. This time delay increases when the thickness is not even. In some areas, stress increases rather than uniformly across the area. It is often the first zone to fail after a number of cycles. One of the other factors that often goes overlooked is the residual stress. Is not formed during the operation but during coating. SiC and TaC cannot be deposited on the graphite substrate and do not relax perfectly as they cool down. Some parts are stretched, others compressed. An uneven stress remains trapped in the coating. With time, thermal cycling continually increases the amount of stress. Cracks may begin with a relatively small additional load where there is already some tension in the area. When the crack appears it will want to follow the stress path, not the straight visual line. That's why failures are sometimes seemingly random, but actually have an internal pattern of stress. In the real world, such as coated graphite when used for multiple runs at high temperatures. Occasionally the parts that have a little thicker coating are not going to last as long as other parts because the internal stress was already greater from the beginning. Even if it appears to be uniform when first seen, operators might observe peeling and patchy flaking at the edges.

In reality, controlling thickness and minimising residual stress during deposition is much more important to achieving long coating life than changing the type of coating. It's a compromise between strength, flexibility and settling after every thermal cycle.

Extending Lifetime of Susceptors and Coated Components

It's not just one big fix, it's about habits that will help you deal with stress in the long run, to make the life of your susceptors and coated parts longer. For coated parts ( Cvd sic and TaC), most failures occur after a large number of heating cycles with resulting slow damage; the goal is to slow the damage process. The easiest one is controlling the temperature ramp. Rapid heating and cooling can be beneficial to the production process, but can also be very stressful for the coating. Coating expands and contracts on a more gradual ramp which enables a smoother process. Many users will find that they don't need to change materials in order to have less edge cracking and flaking when they have a ramp speed change. Cleaning cycles are not just that simple either. If there is residue on the surface, it is not evenly distributed. This forms hot spots that grow in various places than the rest of the part. These uneven areas over the years turn into weak areas where cracks begin. Frequent and gentle cleaning – stabilizes heat flow. Another major factor is a balance of gas flow in tools. If the gas is not equally distributed between the two sides of the susceptor, then one side may heat up more rapidly than the other. This minor change is replicated in every cycle and will eventually accumulate in the same place creating stress. This is a typical early edge wear or a reoccurrence of cracks in the same direction.