The ESC components are stacked ceramic and metal layers. Together they clamp a wafer, cool it and release it inside a plasma process chamber. Working from the wafer down, an ESC has a ceramic dielectric clamping surface and one or more buried electrodes. Below these sit a bond layer and a cooled metal base plate carrying helium and coolant channels. Lift-pin holes and edge seals complete the build. This article walks through that cross-section layer by layer. It explains what each part does, and shows how the dielectric material choice drives clamping and thermal performance.
For the full scope of ESC design, materials and applications, see our complete guide to ESCs.
The ESC as a layered stack
An ESC is best understood as a vertical stack of functional layers, not a single block. The wafer sits on top. Directly beneath it is the ceramic dielectric that both insulates and holds. Then comes the buried electrode that carries the clamp voltage, then a bond layer. At the bottom is the cooled metal base plate. It removes heat and provides the mechanical and vacuum interface to the process tool. Each layer has one main job. The layers are matched to each other so that thermal expansion, voltage standoff and flatness stay within specification across the operating temperature range.
The ceramic dielectric clamping surface
The top layer is a thin ceramic dielectric, and it is the most critical part of the ESC. It separates the buried electrode from the wafer. It has to withstand the full chuck voltage of 600–2,000 V DC without breaking down. At the same time, it must be thin enough that the field still produces a usable clamp force. Its bulk resistivity sets the clamping regime. A near-perfect insulator gives Coulombic behaviour. A slightly conductive ceramic gives Johnsen–Rahbek behaviour, with higher force per volt. The surface is held flat to single-digit micrometres, so the wafer sits flat and contacts uniformly across its whole area.
The clamping face is not a plain flat plate. It is finished with a pattern of raised mesas or texture, and this pattern defines two things at once. It defines the real wafer contact area, which sets how much of the wafer actually touches ceramic. It also defines the sealed volume between the mesas, through which helium flows for backside cooling. Reducing the contact area lowers particle generation and makes dechucking easier. But it also reduces conductive heat transfer. So mesa height, diameter and spacing are deliberate engineering trade-offs. The same surface usually carries the sealing band at the wafer edge and around the lift-pin holes. That band keeps the helium contained.
Buried electrode or electrodes
The electrode that carries the high-voltage DC sits a fraction of a millimetre below the clamping surface. The electrode is a thin conductive layer, typically a refractory metal film or mesh sintered into the ceramic. Its job is to hold the charge that induces the opposite charge in the wafer above. The number of electrodes determines whether the ESC is monopolar or bipolar. A monopolar ESC uses a single electrode and relies on the surrounding plasma to complete the circuit. So it only clamps while plasma is present. A bipolar ESC uses two electrodes held at opposite polarity. The circuit closes through the wafer itself. So a bipolar ESC clamps with or without plasma, and it keeps the wafer near ground potential.
In a bipolar design the two electrodes are usually arranged as interdigitated zones, or as inner and outer regions. They are sized so the net clamp force is balanced across the wafer. Three things together fix the clamp force at a given voltage: the electrode geometry, its depth below the surface, and the dielectric thickness above it. The electrode is fully encapsulated in ceramic. So it cannot be inspected or repaired without rebuilding the ESC.
Bond layer and cooled metal base plate
The ceramic puck is joined to a metal base plate through a bond layer. This is commonly an adhesive or brazed joint. It is engineered to survive thermal cycling and to conduct heat efficiently across the interface. The bond has to absorb the difference in thermal expansion between ceramic and metal. An uncontrolled mismatch would crack the puck or delaminate the joint over thousands of process cycles. A degraded bond layer shows up as hot spots and poor temperature uniformity. It is a common reason an ESC is sent for rebuild rather than a simple surface recondition.
Beneath the bond is the base plate, usually aluminium. It provides the mechanical mount, the vacuum and electrical feedthroughs, and the thermal mass. It is machined with internal channels. Coolant passages circulate a temperature-controlled fluid to set the base temperature. Helium passages route backside gas up to the wafer interface. Multi-zone ESCs split the helium and coolant routing into several radial zones, so temperature can be tuned across the wafer radius. The clamp voltage must always exceed the helium backside pressure of 5–40 Torr. If it does not, the gas lifts the wafer and cooling collapses. This is why these channels are designed around the clamp-force budget. The cooling mechanism is covered in full under helium backside cooling.
Lift-pin holes, seals and edge features
Several smaller features run through the whole stack. They are easy to overlook but essential to operation. Lift-pin holes are precision bores, typically three of them. Pins rise through these bores to lift the wafer off the surface for robot handling after dechuck. Each hole needs its own seal so it does not leak helium during processing. The wafer-edge seal band contains the helium within the clamped area. Its condition directly affects how much backside pressure the ESC can hold before it leaks. Around the periphery, edge and focus-ring features protect the ESC from direct plasma erosion and shape the process at the wafer edge. Wear or contamination at any of these seal features is a frequent cause of helium leak-rate failures. It is also a typical trigger for refurbishment.
Summary of ESC components
The table below summarises the main components of an ESC. It shows the function each one performs and its typical material or specification. It reads from the wafer down, following the same cross-section described above.
| Component | Function | Typical material / spec |
|---|---|---|
| Dielectric clamping surface | Insulates electrode from wafer; sets clamping regime; holds wafer flat | Al₂O₃, AlN or doped ceramic; flat to single-digit µm |
| Mesas / surface texture | Define real contact area and helium volume; ease dechuck and lower particles | Patterned ceramic; height and spacing engineered |
| Buried electrode(s) | Carry HV DC to generate the clamping field | Refractory metal film/mesh; single (monopolar) or dual (bipolar) |
| Bond layer | Joins ceramic puck to base plate; conducts heat; absorbs expansion mismatch | Adhesive or brazed joint |
| Base plate | Mechanical mount, thermal mass, vacuum/electrical interface | Aluminium, machined |
| Coolant channels | Circulate fluid to set base temperature | Internal passages; single or multi-zone |
| Helium channels | Route backside gas to wafer interface for cooling | 5–40 Torr backside pressure |
| Lift-pin holes | Allow pins to raise wafer for handling after dechuck | Typically 3 sealed bores |
| Edge / seal features | Contain helium; protect ESC from plasma erosion | Seal band; edge/focus-ring |
How dielectric material choice drives performance
Of all the layers, the dielectric ceramic does the most to determine how an ESC behaves. Its bulk resistivity sets the clamping regime, and its thermal conductivity sets how evenly the wafer is cooled. A resistivity above 10¹⁵ Ω·cm gives Coulombic clamping: fast release, higher voltage, low temperature sensitivity. A resistivity of roughly 10¹⁰–10¹³ Ω·cm gives Johnsen–Rahbek clamping, with higher force per volt but slower release and more temperature sensitivity. The full comparison is in our guide to Coulombic vs Johnsen–Rahbek ESCs. Three material families dominate:
- Alumina (Al₂O₃): robust, well understood and cost-effective. It is usually Coulombic, though doping the alumina lowers its resistivity into the Johnsen–Rahbek band.
- Aluminium nitride (AlN): far higher thermal conductivity than alumina, which gives better wafer temperature uniformity. It is common in demanding etch applications and naturally suited to JR operation.
- Doped or composite ceramics: alumina with titania or other additives, engineered to place the bulk resistivity precisely in the JR window.
The dielectric sets both clamp force and thermal uniformity. It is also the layer that wears. Plasma erosion thins the surface, alters the mesa geometry and shifts resistivity. This is why clamping force drifts and dechucking worsens over an ESC’s service life. Where the body remains sound, ESC Technology can recondition the surface (L1) or fully rebuild the ESC (L2). Either route restores as-new clamping and particle performance, with a full functional test.
Frequently asked questions
What are the main components of an ESC?
From the wafer down, the layers are: a ceramic dielectric clamping surface with mesas, one or more buried electrodes, and a bond layer. Below that is a cooled aluminium base plate carrying helium and coolant channels. Lift-pin holes and edge seals run through the stack.
What are the mesas on an ESC surface for?
The raised mesas define the real wafer contact area. They also define the sealed volume that holds helium for backside cooling. Their height and spacing are engineered to balance heat transfer against low particle generation and easy dechucking.
How does dielectric material affect ESC performance?
The dielectric’s resistivity sets the clamping regime. Alumina above 10¹⁵ Ω·cm is Coulombic, while doped or JR-band ceramics give higher force per volt. Its thermal conductivity sets cooling uniformity, which is why aluminium nitride is chosen for demanding etch.