What Is a Thermal Pad? How They Work, Types & W/mK
A thermal pad fills the gap between a chip and heatsink. How pads work, W/mK vs impedance, thickness from measured gap plus compression, materials and reuse.

A thermal pad is a soft, pre-formed sheet of thermally conductive material placed between a heat-generating component and a heatsink or chassis. It fills the microscopic air gaps between the two surfaces so heat can flow across the joint instead of being trapped by insulating air. It is one type of thermal interface material (TIM), alongside thermal paste, gels, and phase change materials. On datasheets and in catalogs the same part is also called a thermally conductive pad, a thermal interface pad, or a gap pad — the terms are interchangeable.
Unlike paste, a thermal pad arrives at a fixed thickness and hardness. That makes it clean, repeatable, and ideal for bridging large or uneven gaps — but it also means the pad only performs when it is specified and compressed correctly. This guide explains what thermal pads are, how they work, what each material type is good for, and what the numbers on a datasheet — W/mK, impedance, Shore hardness — actually mean once the pad is in your stack-up. It is written by the engineering team at ZIITEK, a thermal-management-materials manufacturer that has supplied pads, greases, and gap fillers to EV, telecom, server, and LED customers since 2006.
What does a thermal pad do?
Two metal surfaces that look flat are, at a microscopic level, rough. Press a chip against a heatsink and they only touch at a few high points; the rest of the interface is air. Air conducts heat poorly — about 0.026 W/mK at room temperature — so those gaps act as insulation and drive the component’s junction temperature up. Even a modest 1–3 W/mK pad conducts roughly 40–100× better than the air it displaces.
A thermal pad solves this by deforming to fill those voids. Because it is soft and compressible, it “wets out” against both surfaces, displacing the air and creating a continuous conductive path from the hot component to the cooler heatsink. The result is lower thermal resistance across the joint and a cooler, more reliable part.
Most thermal pads also do a second job: electrical insulation. Silicone-based pads are dielectric, so they let heat pass while keeping a component electrically isolated from a grounded heatsink — useful around MOSFETs, VRMs, and power ICs. But breakdown voltage belongs to the grade at a thickness, not to the category: a 1.0 mm and a 0.5 mm sheet of the same compound do not withstand the same voltage. Graphite and carbon-fiber pads conduct, so they cannot isolate at all.
How does a thermal pad work?
The mechanism is conduction across a filled gap:
- Contact & conformance — under mounting pressure, the pad compresses and conforms to both mating surfaces, eliminating trapped air.
- Conduction — heat moves from the die, through the pad’s conductive filler network, into the heatsink.
- Steady state — the pad stays where it was placed instead of migrating the way a low-viscosity paste can; compression set, oil bleed and outgassing differ by compound, so read the specific datasheet, not the category.
The pad body is usually a silicone (or silicone-free) elastomer loaded with conductive fillers — aluminum oxide, zinc oxide, boron nitride, graphite, or ceramic particles — that carry the heat. The more (and better-connected) the filler, the higher the conductivity, but also the stiffer and denser the pad. Every pad is a balance between conductivity, softness, and manufacturability.
What is a thermal pad made of?
| Base / filler | Typical role | Trade-off |
|---|---|---|
| Silicone elastomer | Most common carrier; flexible, wide temperature range, dielectric | Can outgas low-molecular-weight siloxanes in sensitive optics |
| Silicone-free (e.g. ZIITEK Z-Paster®) | For optics/relays where siloxane outgassing is a risk | Often narrower temp range |
| Aluminum oxide / zinc oxide filler | Inexpensive conductivity boost | Higher loading = stiffer pad |
| Boron nitride | High conductivity with low dielectric constant | Higher cost |
| Graphite / carbon fiber | Highest conductivity, thin pads (ZIITEK’s TIR®700: 9–35 W/mK) | Electrically conductive — not an insulator |
This matters for selection: a “silicone thermal pad” and a graphite pad can both be called thermal pads, yet one insulates and one conducts electrically. Always check the datasheet, not just the W/mK number.
The range is listed by grade: silicone and carbon-fiber and non-silicone. Ask for samples against your gap and compression target.
Thermal pad vs thermal paste (vs grease vs putty)
This is the comparison every reader wants, so here it is up front. For the full breakdown see thermal pad vs thermal paste.
| Thermal pad | Thermal paste | |
|---|---|---|
| Form | Solid, pre-cut sheet | Viscous compound |
| Best for | Larger/uneven gaps, multiple components | Tight, flat CPU/GPU lids |
| Bulk conductivity | ~1–18 W/mK (carbon-fiber pads reach ~35) | ~1–14 W/mK |
| Bond line achievable | Thicker (0.2–20 mm) | Very thin (tens of µm) |
| Cleanliness | Clean, no mess; whether it can be reused depends on the grade and the compression it saw | Messy, single-use |
| Assembly | Fast drop-in, automatable | Requires dosing skill |
| Long-term | No drying-out mechanism; compression set, oil bleed and outgassing are grade- and assembly-specific | Can dry out / pump out |
Reuse is not a property of thermal pads in general. A pad compressed inside its rated range and lifted without tearing may go back in for a fit check; one that has taken a compression set, picked up contamination, or been over-compressed should be replaced — the second installation will not have the same thickness or contact. Ask the supplier for the rule on that grade.
The short version: use a pad when you need to span a real gap, insulate, or assemble at volume; use paste when the two surfaces are flat and nearly touching and you want the absolute lowest resistance. For flexible, dispensable filling of very large or varying gaps, a gap filler or phase change material may beat both.
Thermal conductivity and W/mK: how to read a thermal pad spec
The headline number on any pad is its thermal conductivity in W/mK (watts per meter-kelvin) — how well the bulk material moves heat. Higher is better, all else equal.
To put real numbers against that, here are our own catalog ranges rather than vague industry averages. ZIITEK’s TIF® silicone pads run 1.25–18 W/mK; our boron-nitride TIF®100N pads 1.5–5.0 W/mK; our silicone-free Z-Paster®100 pads 1.5–10 W/mK; and our TIR®700 carbon-fiber pads 9–35 W/mK. These are ZIITEK grades, not an industry survey — other manufacturers’ ranges will differ, so read the datasheet for the part you are actually buying. The wide spread is the point: “thermal pad” spans more than an order of magnitude in conductivity, and the top of that range comes from carbon fiber, not silicone.
But here is the trap engineers fall into: bulk W/mK does not equal real cooling performance. What actually determines the temperature drop across the joint is thermal impedance — which folds in bond line thickness and the contact resistance at each surface. A 6 W/mK pad that conforms well under low pressure can outperform a stiff 12 W/mK pad that won’t wet out. We unpack this fully in understanding thermal conductivity (W/mK) in TIM — read it before you spec on W/mK alone.
Two more numbers worth knowing:
- Thermal impedance (°C·in²/W or K·cm²/W) — the real-world resistance of the pad at a given thickness and pressure. Lower is better. It is characterized under ASTM D5470 , the standard test method for the thermal transmission properties of thermally conductive electrical insulation materials — if a datasheet quotes impedance without naming a thickness and pressure, the number is not comparable to anyone else’s.
- Dielectric / breakdown voltage — how much voltage the pad withstands before it conducts. The figure means nothing without a thickness: a per-mm value (V/mm) is tied to the thickness it was measured at, and an absolute VAC rating to the sheet tested. Values are grade-specific too — ZIITEK’s TIS® thermally conductive insulating materials, for instance, span 1500–6000 VAC. Check it if the pad sits near live traces; graphite and carbon-fiber pads are not insulators at all.
Thermal pad thickness: how to choose and measure
Thickness is where thermal pads are most often specified wrong. The starting point is arithmetic: the pad’s nominal (uncompressed) thickness must exceed the measured gap by enough that assembly compresses it to what the grade’s compression–pressure curve allows at your clamp load. Compression is what forces the pad to wet out; an uncompressed pad barely works.
That is why generic gap bands — “tight”, “moderate”, “large” — fail: the thickness that suits a gap depends on the grade’s compression range and on the load the component will tolerate. Work from the measured gap and a compression target instead:
nominal thickness ≈ measured gap ÷ (1 − target compression)
At 20% and 30% compression the numbers work out as:
| Measured gap | Nominal thickness at 20% compression | at 30% compression |
|---|---|---|
| 0.3 mm | 0.38 mm | 0.43 mm |
| 0.5 mm | 0.63 mm | 0.71 mm |
| 1.0 mm | 1.25 mm | 1.43 mm |
| 3.0 mm | 3.75 mm | 4.29 mm |
Round up to the nearest stocked thickness, then check both ends of the tolerance stack: at the largest gap the pad must still be compressed enough to conform and wet out, and at the smallest gap the compression must stay inside the grade’s rating and below the component’s load limit. A thickness that is right at nominal can still fail a stack-up that runs ±0.2 mm.
To measure the gap accurately, assemble the heatsink with a soft modeling putty or a thickness gauge in place of the pad, disassemble, and measure the crush height at each component. Choosing thickness, compression ratio, deflection force, and Shore hardness together is its own discipline — the full method (with datasheet reading and mounting-pressure limits) lives in our thermal pad thickness & compression guide.
As the gap falls toward 0.2–0.3 mm the required nominal approaches the thinnest stocked sheet — ZIITEK’s TIF® line bottoms out at 0.25 mm — and the compression window narrows with it. At that point compare a thin pad against paste or a gel.
A note on hardness: pads are rated in Shore OO (softest, e.g. 5–70) or Shore A. Softer pads conform under lower pressure and protect fragile components; harder pads hold shape but need more clamping force. ZIITEK’s TIF® silicone pad line, for example, spans roughly 1.25–18 W/mK, 0.25–12 mm thickness, and 5–70 Shore OO — the range exists precisely because no single pad fits every gap.
Types of thermal pad
- Silicone gap pad — the default. Dielectric, wide temperature range; ZIITEK’s TIF® line covers 1.25–18 W/mK.
- Silicone-free pad — for optics, cameras, and relays where siloxane outgassing must be avoided; ZIITEK’s Z-Paster®100 runs 1.5–10 W/mK. The 6060-11 grade is listed at 6 W/mK, 60 Shore OO, 1–5 mm, −45 to 125 °C and ≥5500 V/mm, with 0.234 °C·in²/W impedance that applies only at 1.0 mm and 30 psi. Silicone-free removes the siloxane source; it does not by itself make an assembly low-outgassing.
- Graphite / carbon-fiber pad — the conductivity ceiling for pads, and thin and light with it: ZIITEK’s TIR®700 carbon-fiber pads reach 9–35 W/mK and work at low clamping pressure. The catch is electrical: they are non-insulating , so they cannot sit across live traces the way a silicone pad can.
- Boron-nitride pad — high conductivity with a low dielectric constant, for RF-sensitive designs; ZIITEK’s TIF®100N covers 1.5–5.0 W/mK.
- Phase change pad — solid at room temperature, softens at operating temperature to flow like paste; see what is phase change material (PCM) .
A “gap pad” is just a thermal pad doing one job
You will see gap pad, thermal gap pad, thermal interface pad and thermally conductive pad used interchangeably with thermal pad. They name the same part from different angles: thermal interface pad and thermally conductive pad describe what it is — a conductive sheet used as a TIM; gap pad and thermal gap pad describe the job it was chosen for. Strictly, a gap pad is a thermal pad selected to fill a defined air gap between a component and an enclosure or heatsink, at low stress. If your design has a genuine, repeatable gap to span at volume, a gap pad is usually the right sourcing term.
Where it gets confusing is the neighboring category: a gap filler is dispensed as a liquid and cures or stays soft in place, rather than arriving pre-cut. Pads win on cleanliness and drop-in assembly; fillers win on wildly varying gaps and near-zero mounting stress. We compare all three formats side by side in gap filler vs gap pad vs phase change.
How to choose a thermal pad: a selection checklist
Work through these in order — it mirrors how a design engineer specs a TIM. For the full material-agnostic framework see how to choose a thermal interface material.
- Measure the gap (and its tolerance across units). This sets thickness.
- Set the required conductivity from your power dissipation and temperature budget — don’t over-buy W/mK you can’t use.
- Check electrical needs — must the pad insulate (silicone) or conduct (graphite)?
- Pick a hardness that your mounting pressure and component fragility allow (Shore OO for delicate parts).
- Confirm reliability — operating temperature range, compression set, outgassing, flammability (UL 94 V-0 for most electronics).
- Match the format — sheet, die-cut, or adhesive-backed for your assembly line.
Applications: where thermal pads are used
CPUs and GPUs, VRAM and memory modules, MOSFETs and VRMs, M.2/NVMe SSDs, LED modules, power-supply and battery packs, telecom and 5G radios, EV and energy-storage electronics, and automotive control units. In practice, wherever a component sits a small, defined distance from a cooling surface — and especially where insulation or clean high-volume assembly matters — a thermal pad is the go-to interface.
Frequently asked questions
Do thermal pads actually work? Yes — especially for low-to-moderate power parts, uneven gaps, and multi-component boards. For a flat, high-power CPU lid where you can achieve a near-zero bond line, paste still gives lower resistance. Match the material to the joint.
Is a thermal pad better than thermal paste? Neither is universally better. Pads win on gap-filling, cleanliness, insulation, and assembly speed; paste wins on minimum thermal resistance for flat, tight interfaces. See the full comparison.
What thickness thermal pad do I need? Measure the gap with the heatsink assembled, divide the measured gap by (1 − compression) for the compression your grade’s curve allows, and round up to a stocked thickness. Then check the tolerance extremes: still in contact at the largest gap, still inside the grade’s compression rating and the component’s load limit at the smallest. Too thin and it won’t contact both surfaces; too thick and it over-stresses the component.
What does W/mK mean on a thermal pad? It is the bulk thermal conductivity of the material. Higher is generally better, but real cooling depends on thermal impedance at your actual thickness and pressure — not W/mK alone.
How long does a thermal pad last? There is no category answer: life depends on the grade, the compression it is held at, the temperature and the cycle profile. TIF800TU and TIF700UU, for example, were run at 150 °C for 1000 h and 85 °C / 85% RH for 1000 h, and the cold-shock entry is filed as −40 °C ↔ 125 °C over 1000 h — a duration, not a cycle count. Those are conditions that were run, not a pass result, a retention figure or a service-life guarantee; ask for the current report. Unlike grease, it holds its position rather than flowing out.
Getting the pad right
A thermal pad is deceptively simple: a soft sheet that fills a gap. But its value only shows up when thickness, conductivity, hardness, and compression are specified together for your joint — which is exactly where generic advice stops and engineering begins. Start from the TIM selection guide for the whole landscape, and if you’re specifying pads into a product at volume, ZIITEK’s engineering team can help match a pad’s conductivity, thickness, and hardness to your design; sample availability is confirmed on the quote.
Sources: Thermally conductive pad — Wikipedia for the naming conventions and materials background; Thermal conductivity and resistivity — Wikipedia for the 0.026 W/mK conductivity of air; ASTM D5470 for the standard test method behind quoted thermal-impedance figures. All product specification ranges (TIF®, TIF®100N, TIR®700, TIS®, Z-Paster®100) are ZIITEK’s own catalog values, not industry averages; the TIF800TU and TIF700UU conditions are as supplied by ZIITEK, with no cycle count stated and no pass/fail or retention figure implied.


