Testing CVD TaC Coating Adhesion Strength and Peeling Risks

For special coatings on the graphite, CVD TaC and Cvd sic coating parts of high-temperature chip making machines, the bonding performance of the coating to the part is most important on the life of the parts. Weak bond causes small crack and peel in heat and cooling processes, generating the contaminants or uneven protection and the failure of the process. Many parts look well at beginning in factory and the edge of it peel off after using several times. Therefore, the bond strength should be tested before the parts used in factory to avoid the unexpected failure in the machines.

Standard Adhesion Tests for CVD TaC Coating on Graphite Substrates

To find out the relative strength, several qualitative tests are carried out in factories; first is the scratch test. Here, a diamond tip is drawn over the coating under increasing force, if the film cracks, flakes off or debonds from the graphite layer prematurely it shows poor adhesion. This is a simple test but does show clear results prior to parts entering service.

Another quick test is the tape test where adhesive tape is applied to the film and then pulled off; if flakes come away with the tape the bond with the graphite base may be weak. This is a quick preliminary test used in laboratories prior to more rigorous testing.

Repeated heating and cooling is also extremely popular, because the process mimics that of a real application perfectly. Tough coatings show no ill effects, while weaker coatings begin to flake or peel at the edge. In one factory situation, the coatings failed on the second check after being repeatedly heated (peeling at the edges) though this was due to a contamination of the substrate prior to application.

Pull-off testing (where a stud is glued to the coating and is then pulled until the substrate pulls away from it) is the other test, which gives a value on a continuum, is good for comparison of batches but can take a considerable time to perform. Indentation testing is used as a visual and quantifiable test, where a hard point is driven into the coating and the extent of crack formation is measured. If performed correctly together, these tests ensure good coating integrity.

Early Warning Signs of TaC Coating Peeling in Epitaxy Applications

These types of machines have very few instances where the coating delamination happens unexpectedly. Most of the time a little bit of change occurs in the part, that goes unnoticed. Color change of the coating appears fairly early. A new coating should appear even and smooth but under stress some portion appears dull in color, particularly along edges, corners and areas with high heat.

There is often also very minor edge lifting that appears like a standing layer of coating at the boundaries if viewed in the light at the proper angle and or at higher magnification. In one factory line a clean up worker noticed fine dust particles next to the edge of the part, and found flaking was already starting at the rim.

Hairline cracks also indicate trouble. The initial cracks are fine and superficial and appear in zones of heat stress, eventually connecting to form larger regions of spalling if the crack is not repaired. Other subtle indicators include increased levels of particulate or dust inside the chamber.

Even though no apparent defect might be visible on the coating, particles may spall off the coating during operation and can be observed on monitors prior to a catastrophic failure. Roughness can increase, as a once smooth coating develops a grainy appearance due to degradation of the top layer of the coating.

Because a small flaw on the edge can develop into a larger failure that compromises quality and necessitates machine downtime, identifying these early indicators is important.

Thermal Cycling Effects on CVD TaC Coated Epitaxial Parts

Because coated components are repeatedly heated up and cooled down during the day there are very strong stresses developed on the coating in time. When the components are heated up, the coating and the graphite expand, but at different rates. So, each heat cycle develops stressing between the two materials at interface. When the component cools down, the stress turns the other way round. One heating and cooling cycle will not cause any damaging, however during hundreds of such cycle small internal stress develops.

The coating would be relatively smooth and stable, but after repeated cycling of temperature changes small stress areas would develop at the sharper edges or uneven heating. The middle of component would remain for long, the outside of the component would be wearing out in actual tools. Small crack on the edge of the components is a common example, after numerous usage it would develop at edge and would start growing as stress forces it apart after each heat cycle and small flakes would peel off.

With repeated heating and cooling the internal matrix of the coating wears out, so that few components would pass the initial test, however after a prolonged usage (a number of months) would still be prone to damage. In one case, parts have started to peel off after few hundred cycles due to rapid heating creating extremely large stress as opposed to the slow heating.

Surface Pretreatment Methods to Improve Coating Bonding Strength

There is significant preparation that must be carried out on graphite parts before applying any type of coating so the layer bonds. Poor preparation on a good coating will simply result in a coating that fails too soon. Deep cleaning of a graphite part is one aspect of this preparation. Graphite easily traps a very fine dust, residual oils, or machining particles within its pores. These items physically prevent the coating from contacting directly with the base.

Even a microscopic film of oil will create weak zones in the bond, which will become sites for the peeling to initiate. Dedicated cleaning solutions are typically employed or sometimes the part is baked in a furnace to eliminate any deep seated residues.

Next is to ensure the surface is slightly roughened to increase the effective grip the coating will have than if the surface were perfectly smooth. This is generally achieved by sandblasting or manipulating machine tools to achieve a rough but relatively stable gripping surface for the coating to mechanically "hook" into, or by the use of plasma treatments which momentarily bombard the graphite surface with gas to activate it and chemically predispose it to be compatible with coating materials while having no adverse effects on the substrate shape.

Heat treatments of graphite parts are used prior to coating, since high-temperature baking of a part can drive out trapped gases within the material. These gases could potentially escape while a coating is being deposited, creating the ideal site for a weak spot or a blister under the coating. These processes are generally combined.

Failure Analysis of TaC-Coated Rings Used in LPE (ASM) Systems

Coated graphite rings work under extreme conditions including high temperature, high gas flow, and thermal cycles.Failure mechanisms of coated graphite rings are most frequently associated with coating stress, surface quality and thermal fatigue.

Edge delamination where coating tends to delaminate from the edges of sharp edges and thin section as they are subjected to accelerated cooling and heating rates compared to bulk parts. After the initiation of the lift at the edge, this phenomenon grows similar to an opening zipper along the edge.

Tiny flaking can occur on the inside surface where the gas flow is more intensive and generates hotspots locally. A batch of fine black dust was discovered in one system that has been associated with the fine cracks on the inner wall of the ring. Interface bonding failure is another fundamental mechanism; weak adhesion below the layer is attributed to sudden, large area peel off usually from surface that was not properly cleaned or gases trapped at the interface that expand.

Rapid thermal cycling, on the other hand, creates the thermal shock in which thermal stresses can rapidly build up at the coating/base interface. In one production process a ring only lasted a few months. Although seemingly normal initially, the particulate level steadily increased and upon removal a large section of the coating had delaminated in the vicinity of the attachment points.

This failure resulted from non-uniform heating and poor edge cleaning prior to coating. While the failures usually grow from small defects with time, regular inspection and stable heating temperature may be good preventative measures against failures of this type.