An ESC works by applying high-voltage DC to one or more electrodes buried beneath a thin ceramic dielectric. The electric field induces an opposite charge in the wafer resting on the surface, and the resulting electrostatic attraction clamps the wafer flat against the ESC. Because the force is generated by an electric field rather than vacuum or a mechanical ring, an ESC can hold a wafer inside the vacuum of a plasma chamber while leaving the wafer edge fully exposed to the process. This article explains the working principle step by step.
For the full scope of ESC design, materials and applications, see our complete guide to ESCs.
The basic working principle
Every ESC is built around the same physics. A conductive electrode sits a fraction of a millimetre below the clamping surface, separated from the wafer by a ceramic dielectric. When the ESC’s power supply applies a DC voltage (typically between 300 and 3000 V) to the electrode, charge accumulates on it. That charge induces an equal and opposite charge in the wafer above. Opposite charges attract, so the wafer is pulled down onto the ESC. The attraction is distributed across the whole wafer area rather than concentrated at the edge, which is what gives the ESC its flat, uniform hold.
The magnitude of that hold (the clamping force) depends on the applied voltage, the thickness and properties of the dielectric, and the size of the gap between wafer and surface. In practice the force is reported as a pressure, and a production ESC typically delivers 20–80 Torr-equivalent. How that figure is specified and measured is covered in our guide to ESC clamping force.
Two clamping regimes: Coulombic and Johnsen–Rahbek
The way the charge behaves inside the dielectric splits ESCs into two families, set by the bulk resistivity of the ceramic:
- Coulombic: the dielectric is a near-perfect insulator (above 10¹⁵ Ω·cm). Charge stays on the electrode and the wafer-to-electrode separation equals the full dielectric thickness. Release is fast; voltages are higher.
- Johnsen–Rahbek (JR): the ceramic is slightly conductive (roughly 10¹⁰–10¹³ Ω·cm). Charge migrates to the ESC surface, the effective gap shrinks to microscopic contact points, and the ESC produces much higher force at the same voltage, with slower release and more temperature sensitivity.
The difference is the single most important behavioural distinction between ESCs. We cover it in detail, including the resistivity thresholds and temperature dependence, in Coulombic vs Johnsen–Rahbek ESCs.
One electrode or two: monopolar and bipolar
The working principle also depends on how many electrodes the ESC uses. A monopolar ESC has a single electrode and relies on the surrounding plasma to complete the electrical 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 it clamps with or without plasma and keeps the wafer near ground potential. The trade-offs are explained in monopolar vs bipolar ESCs.
Why clamping enables cooling
Clamping is not an end in itself. It exists to make wafer cooling possible. Plasma processes pour heat into the wafer, and that heat must be removed to hold critical dimensions. Because the wafer only touches the ESC at microscopic high points, helium is fed into the gap behind it at 5–40 Torr to conduct heat to the temperature-controlled ESC body. The electrostatic force has to exceed that helium pressure, or the gas simply lifts the wafer off the surface. This is why an ESC’s clamping force and helium pressure are always quoted together, and it is the mechanism behind helium backside cooling.
Releasing the wafer: the dechuck step
When the process finishes, the voltage is switched off and the wafer should release. In practice some charge remains trapped in the dielectric, so the ESC runs a controlled dechuck sequence (often a brief reverse-polarity pulse and plasma-assisted discharge) to bleed that charge before the lift pins move. Johnsen–Rahbek ESCs need more careful dechucking because they store charge in the ceramic by design. Persistent sticking points to an ageing dielectric, a topic covered under dechucking and residual clamping.
An ESC clamp cycle, step by step
The physics above describes a static hold, but in a production tool the ESC runs through a defined sequence for every wafer. Understanding that sequence is what lets an engineer place a fault: most ESC problems show up at one specific step, and the symptom usually identifies the cause.
- Load onto the ESC. The robot end-effector places the wafer on the lift pins, which lower it onto the clamping surface. The wafer must seat flat against the mesas; a particle trapped under the wafer here tilts it, opens the helium gap, and is a common source of backside damage and uneven clamping.
- Clamp voltage ramp. The power supply ramps the electrode to the set voltage (600–2,000 V DC). Ramping rather than stepping limits inrush and lets the charge distribute evenly. If the voltage is too low for the regime, or the dielectric is leaking charge, the wafer never fully seats and the next steps fail.
- Helium backside flow on. Once clamp is confirmed, helium is admitted behind the wafer at 5–40 Torr. The clamping force must exceed this pressure; if it does not, the gas lifts the wafer edge, helium leaks into the chamber, and backside cooling collapses.
- Process runs. Etch, PVD, CVD or implant proceeds while the ESC holds the wafer flat and conducts heat away through the helium. The clamp voltage and helium flow are held steady and monitored; a mid-process clamp loss or helium spike usually aborts the wafer.
- Helium off. Backside gas is shut off and pumped out before declamping, so there is no trapped pressure pushing the wafer up as the clamp releases. Skipping or shortening this step can pop the wafer off its position.
- Controlled dechuck. The supply switches off and typically applies a brief reverse-polarity pulse, often combined with a plasma-assisted discharge, to bleed charge trapped in the dielectric. Done too fast, residual clamping remains; done poorly, the wafer can jump when the pins move.
- Lift pins raise the wafer. With charge cleared, the pins lift the wafer clear for the robot. A wafer that sticks, shifts, or “pops” at this point indicates incomplete dechuck or surface adhesion, both signs the dielectric is holding residual charge it should have released.
The dechuck step is the one most often blamed for yield loss and tool downtime, because residual clamping worsens as an ESC ages. The mechanisms and the recovery routines are covered in dechucking and residual clamping.
What fab engineers monitor on an ESC
An ESC is not a passive fixture; it is an instrumented part of the process. A handful of parameters, logged per wafer or per lot, tell an engineer whether the ESC is healthy long before it fails outright. The most useful ones, and what a drift in each is telling you, are set out below.
| Parameter | What it tells you | Trouble sign |
|---|---|---|
| Clamp voltage | The voltage needed to reach the target clamping force for the regime | Voltage creeping up to hold the same force suggests a degrading dielectric or contaminated surface |
| Leakage current | Current flowing through the dielectric at clamp voltage | Rising leakage current is the clearest early indicator of dielectric degradation and a path to electrical breakdown |
| Helium flow and leak-back rate | How much backside helium is needed, and how fast it bleeds when supply is closed | High flow or a fast leak-back rate means the wafer is not sealing: weak clamp or surface wear |
| Dechuck time | How long the wafer takes to release after power-off (normally < 1–5 s) | Lengthening or inconsistent dechuck time points to trapped charge and residual clamping |
| Particle performance | Added particles on the wafer backside or in the chamber per pass | A particle excursion traced to the ESC indicates surface erosion or breakdown of the clamping mesas |
None of these readings is meaningful in isolation. The value comes from trending each one across thousands of wafers: an ESC rarely fails suddenly, and a slow, correlated drift across clamp voltage, leakage and dechuck time is the signature of an ESC approaching end of life.
Signs an ESC is degrading
ESCs wear. Thermal cycling, ion bombardment and millions of clamp cycles gradually change the dielectric and the clamping surface, and the monitored parameters above translate into a recognisable set of symptoms on the tool.
- Weak clamp and helium leaks. The ESC no longer holds the wafer tightly enough to contain backside helium, so flow climbs and cooling becomes uneven. This shows up as temperature non-uniformity and drifting process results.
- Rising leakage current. Increasing current through the dielectric at the same voltage signals that the ceramic is losing its insulating quality, the precursor to dielectric breakdown and an electrical fault.
- Slow or failed dechuck. Trapped charge accumulates and the wafer is slow to release, sticks, or shifts on the pins. Residual clamping is one of the most disruptive ageing symptoms because it directly threatens the wafer.
- Particle excursions. An ESC that suddenly adds particles is shedding material from an eroded surface, and every affected wafer is a yield risk until the source is removed.
- Surface erosion. The mesas and texture that define real contact area and helium volume wear down, flattening the surface and degrading both clamping and cooling at once.
Once these symptoms appear, the ESC needs either refurbishment or replacement rather than continued tuning on the tool. We restore worn ESCs through L1 surface reconditioning and L2 full rebuild to as-new clamping and particle performance, with full functional test, which returns a degraded ESC to service at a fraction of the cost and lead time of a new OEM unit. The decision between repair and replacement is set out in the complete guide to ESCs.
Frequently asked questions
What is the working principle of an ESC?
DC voltage on a buried electrode induces an opposite charge in the wafer above it; the two attract, clamping the wafer flat against the ceramic surface so it can be held and cooled inside a vacuum chamber.
How much voltage does an ESC use?
Typically 600–2,000 V DC, depending on the clamping regime and the force required. Johnsen–Rahbek ESCs reach a given force at lower voltage than Coulombic ones.
Does an ESC need plasma to work?
A monopolar ESC does, because the plasma completes its circuit. A bipolar ESC does not, because its two opposite-polarity electrodes close the circuit through the wafer itself.
Why does a wafer sometimes stick to the ESC after the power is off?
Switching off the voltage does not always release the wafer, because charge can stay trapped in the dielectric, known as residual clamping. The ESC clears it with a controlled dechuck, usually a reverse-polarity pulse and a plasma-assisted discharge, before the lift pins move. Sticking that gets worse over time points to an ageing dielectric and an ESC that needs reconditioning or replacement.