Helium backside cooling removes process heat from a clamped wafer. Helium gas is fed into the microscopic gap between the wafer and the ESC. The gas conducts heat from the wafer to the temperature-controlled ESC body. This cooling is the reason an ESC clamps at all. Without close, even cooling, the wafer would overheat within seconds of the plasma starting. This article explains why plasma processes need wafer cooling, and why solid contact alone is not enough. It also explains how helium is fed and pressurised, and why clamping force and helium pressure are always specified together.
For the full scope of ESC design, materials and applications, see our complete guide to ESCs.
Why plasma processes need wafer cooling
Plasma etch, PVD, CVD and ion implantation all put significant heat into the wafer. Ion bombardment, exothermic surface reactions and radiated plasma energy add together. They push the wafer temperature up quickly within a single process step. That heat must be removed, because temperature directly controls the process result. Etch rate, etch profile, film stress and deposition chemistry all depend on temperature. So a few degrees of drift across the wafer causes critical-dimension variation and yield loss.
So the goal is not only to keep the wafer cool. The goal is to hold it at a set temperature that is uniform from centre to edge, and stable for the whole step. The ESC body is held at a controlled temperature by a circulating coolant. The cooling task is then to move heat efficiently from the wafer into that body. This is why clamping and cooling are two parts of the same job.
Why solid contact alone is poor
A wafer pressed against an ESC looks like flat ceramic on flat silicon. At the microscopic scale, neither surface is flat. The clamping surface is finished with mesas or a textured pattern. The wafer touches only at scattered high points. The real contact area is a small fraction of the apparent area. The wafer rests on a sparse field of microscopic contact points, with vacuum in between. Heat can only cross at those points. A vacuum gap is an excellent thermal insulator. So conduction through solid contact alone is poor and uneven.
Increasing the clamping force does not solve this. Higher force flattens the wafer and improves contact slightly. But the contact area is set by surface texture and flatness, not by how hard the wafer is pulled down. The mesa pattern that limits solid contact is deliberate. It defines a controlled volume behind the wafer that can be filled with gas. The texture and flatness that control this contact are part of the ESC’s internal construction. They are held to single-digit micrometres so the backside gap is consistent across the wafer.
How helium fills the backside gap
Helium enters the gap behind the clamped wafer through holes or channels in the clamping surface. The backside pressure is typically 5–40 Torr. The gas fills the space between the mesas. It bridges the vacuum that previously blocked heat flow. Solid contact carried heat across only a few percent of the area. Helium now conducts heat across nearly the whole backside. This multiplies the effective thermal contact and makes it even. Heat flows from the wafer, through the helium, into the temperature-controlled ESC body, and out through the coolant.
Helium is chosen for clear reasons. It has high thermal conductivity for a gas. It is inert, so it will not react in the chamber. Its small molecule conducts heat efficiently even at the low pressures involved. At a few tens of Torr, conduction across the gap depends on pressure. So helium pressure becomes a control variable. Raising it improves heat transfer, and the recipe sets it to hold the target wafer temperature. ESC body temperature and helium pressure together set the wafer temperature.
Force must exceed helium pressure
The helium under the wafer pushes outward in every direction, including straight up against the clamp. This sets a hard rule: the electrostatic clamping force must be greater than the helium backside pressure. If it is not, the gas pushes the wafer off. If the helium pressure gets close to the clamp force, the wafer lifts at the edge. Helium escapes into the chamber, and cooling collapses just as the wafer is taking full process heat. Both quantities are expressed as a pressure, so the comparison is direct. An ESC holding 20–80 Torr-equivalent must run helium below that figure, with a safety margin.
This is the reason clamping force and helium pressure are never quoted alone. The clamp force sets the upper limit. The helium pressure sets the cooling. The working margin between them defines the process window. An ESC whose clamping force has dropped with age may still pass a basic clamp test. Yet it can fail in production, because it no longer holds enough margin over the helium pressure the recipe needs.
Helium leak rate as a health indicator
Some helium always escapes past the wafer edge. The tool measures this as a leak rate. This is usually the flow needed to hold the backside at its set pressure. A healthy ESC shows a low, stable leak rate. This means the wafer is sealing well against the surface, and the helium is staying in the gap. A rising or unstable leak rate is one of the most useful diagnostic signals a process tool produces. It reports directly on the condition of the clamping surface and the seal.
When the leak rate climbs, the causes are usually mechanical. Common causes are a worn or eroded clamping surface, or particle contamination holding the wafer off the mesas. A damaged sealing band at the wafer edge, or weakened clamping force, can also be the cause. Persistent high helium leak is a common trigger for refurbishment. It is one of the functional parameters we verify when reconditioning a worn ESC to as-new clamping and particle performance.
Multi-zone temperature control
A single helium pressure across the whole wafer cools it uniformly only if the heat load is uniform. It rarely is. Plasma density, gas flow and edge effects give the centre and edge of the wafer different thermal conditions. So a flat backside pressure leaves a radial temperature gradient. Multi-zone ESCs solve this by dividing the backside into concentric zones. A common layout is an inner zone and an outer zone, sometimes more. Each zone is fed with independently controlled helium pressure.
The edge zone can run at a different pressure from the centre. This lets the process engineer tune heat removal along the wafer radius. The wafer temperature profile becomes flat, and critical dimensions stay uniform. More demanding ESCs also segment the coolant base or add embedded heaters for finer control. Each added zone increases the plumbing and sealing complexity inside the ESC. This is one reason a multi-zone ESC is harder to manufacture, qualify and repair than a single-zone design.
Summary of key parameters
The table below summarises the parameters that control helium backside cooling. It also shows the role each one plays in holding the wafer at temperature.
| Parameter | Typical value / role |
|---|---|
| Helium backside pressure | 5–40 Torr; sets the rate of heat transfer across the gap |
| Clamping force | 20–80 Torr-equivalent; must exceed helium pressure with margin |
| Coolant gas | Helium — inert, high thermal conductivity, small molecule |
| Backside gap | Defined by surface mesas/texture; held flat to single-digit µm |
| ESC body temperature | Set by circulating coolant; the heat sink for the wafer |
| Helium leak rate | Flow to hold set pressure; low and stable = healthy seal |
| Temperature zones | One or more concentric zones for radial uniformity |
Running a helium leak-up (leak-back) test
A helium leak-up test is also called a leak-back test. It is the standard check a fab runs before product wafers go on the tool. It confirms that the ESC and wafer are sealing well enough to cool reliably. The test does not read the steady-state flow the recipe uses to hold pressure. Instead, it isolates the backside volume and watches how quickly the pressure decays. This reports directly on the integrity of the seal between the wafer and the clamping surface. The test is run during preventive maintenance, and after an ESC is installed or reseated. It is also run whenever helium leak rate or temperature data suggest the seal has degraded.
The procedure is simple. Most tools can run it through a maintenance or diagnostic screen rather than a production recipe.
- Place a known-good test wafer on the ESC. Apply the normal clamping voltage so the wafer is held at full clamping force.
- Open the helium supply. Let the backside gap fill to a defined set pressure within the working 5–40 Torr band, and allow the reading to stabilise.
- Close the isolation valve to seal the backside volume off from the helium supply. This traps the gas behind the clamped wafer.
- Start a timer and record the backside pressure over a fixed interval. Watch the rate at which it rises or decays from the isolated set point.
- Compare the measured pressure change against the tool’s pass/fail limit for that ESC and process. If the result is borderline, reclamp the wafer and repeat.
A low, slow pressure change means the wafer is sealing well and the trapped helium is staying in the gap. A high leak-up rate means gas is escaping faster than it should. That points to a poor seal at the wafer edge, or a worn or contaminated clamping surface. It can also mean the clamping force no longer holds enough margin over the helium pressure. The test removes steady-state supply flow from the measurement, so it isolates the seal itself. That makes it more sensitive to early surface degradation than the production leak rate alone.
Troubleshooting poor backside cooling
Poor backside cooling shows up in the process result before it shows up as an obvious fault. The usual symptoms are localised hot spots on the wafer, and a centre-to-edge temperature gradient that the recipe can no longer flatten. Critical-dimension drift, or etch-rate variation that tracks with thermal non-uniformity, is another symptom. When cooling is suspected, the diagnosis works through the gap, the seal and the surface in turn. Each failure mode leaves a different signature in the helium pressure, leak rate and temperature data.
- Worn mesas and reduced contact: erosion of the surface mesas changes the backside gap and the real contact area. Heat transfer becomes weaker and uneven, even when pressures look normal.
- Insufficient clamp force relative to helium pressure: clamping force that has dropped with age may no longer hold enough margin over the helium pressure the recipe needs. The wafer then lifts slightly at the edge under gas load.
- Leak paths at the wafer edge: a damaged or worn sealing band lets helium escape around the wafer edge. This raises the leak rate and leaves the edge without enough cooling.
- Particle build-up: particles trapped between the wafer and the mesas hold the wafer off the surface locally. This opens a leak path and creates a cold or hot region, depending on the gas distribution.
- Blocked helium holes or channels: a partly blocked feed hole leaves part of the backside without gas. This produces a persistent hot spot that does not respond to a pressure increase.
Particle contamination and blocked feed paths can sometimes be cleared by cleaning. Worn surface features are different. Eroded mesas, a degraded sealing band, and a clamping surface that no longer seats the wafer correctly are mechanical wear. They are a trigger for refurbishment, not a maintenance fix. ESC Technology refurbishment reconditions the worn surface. It restores clamping and particle performance to as-new condition, with a full functional test. The ESC returns to a state where the backside gap, the seal and the leak rate all fall back within specification.
Frequently asked questions
Why is helium used for backside cooling instead of another gas?
Helium has high thermal conductivity for a gas. It is chemically inert, so it will not react in the process chamber. Its small molecule conducts heat efficiently even at the 5–40 Torr pressures used in the backside gap. These properties make it the standard coolant gas for ESCs.
What happens if helium pressure exceeds the clamping force?
The helium pushes the wafer off the ESC. The wafer lifts at the edge and helium leaks into the chamber. The backside gap loses its gas and cooling collapses. This is why clamping force must always exceed helium backside pressure with a working margin, and why the two are specified together.
What does a rising helium leak rate indicate?
A rising or unstable helium leak rate usually means the wafer is no longer sealing well against the clamping surface. Causes include surface wear or erosion, particle contamination, a damaged edge seal, or weakened clamping force. It is a common diagnostic signal and a frequent trigger for ESC refurbishment.