Gap Filler vs Gap Pad vs Phase Change Material

Three materials, one job: fill the microscopic air gap between a hot component and its heatsink so heat can actually cross the joint. A thermal gap filler, a thermal gap pad, and a phase change material (PCM) all do that — but they behave so differently in your line, your reliability lab, and your BOM that picking the wrong one shows up later as slow takt time, cracked BGAs, or field failures nobody caught in qualification.

The confusing part is that datasheets make them look interchangeable. They quote thermal conductivity in W/m·K, list a thickness, and stop. In reality the choice turns on how each material deforms, how it goes onto the part, and how it ages — not on which one has the biggest number. This guide compares all three across the criteria that actually decide a design-in, gives you a decision framework, and maps each material to real end-markets. It’s the head-to-head companion to our complete thermal interface material selection guide.

The 30-second answer

• Choose a gap filler when the gap is large, uneven, or varies part-to-part, when the component is fragile (thin PCB, BGA), or when you want to automate a dispensing line. It conforms to almost anything at low mounting stress — the joint still has to be held closed, but the force that takes is small next to a compressed pad.
• Choose a gap pad when you want the simplest, cleanest, most reworkable option for a defined, repeatable gap — pick it up, place it, clamp it, done. No dispensing equipment, no cure.
• Choose a phase change material when the interface is flat, thin, and high-power (a CPU/GPU lid or IGBT baseplate), and you want low steady-state thermal resistance with far less pump-out exposure than grease — a trend to confirm on the grade and the clamp you will actually use.

Here is the same decision as a side-by-side matrix.

Comparison table: gap filler vs gap pad vs phase change

These are ZIITEK’s own catalog ranges, shown to indicate roughly what each format* can span — not an industry survey, and deliberately not a cross-vendor comparison. Read them as “what shape of number to expect,” nothing more. As the next section argues, a bulk k figure is not comparable to another bulk k figure unless both are stated with the test pressure and bond-line thickness they were measured at — and on a pad the bond line is whatever your clamp compresses the sheet to, not the figure on the label. The test method and the conditions behind each range sit on the individual datasheet, which is where a design review should be reading them. That is exactly why this table does not rank the three by conductivity.

Now the reasoning behind each row — because the table tells you what, and a design review needs why.

The three materials, in one line each

A thermal gap filler is a dispensable, thermally conductive gel or two-part liquid (also called a form-in-place TIM). You dispense it as a paste; it flows to fit the surface, then either cures in place or stays as a soft, compliant putty. It is the shape-shifter of the group. ZIITEK’s TIF® gel line sits here at roughly 1.5–9.0 W/m·K.

A thermal gap pad is a pre-cured, compressible silicone (or silicone-free) sheet, die-cut to shape. You pick it up and place it. It’s the plug-and-play option, and the one most people mean by “thermal pad.” If you want the fundamentals of pad construction and how thickness is specified, see what is a thermal pad.

Photo by Steve Jurvetson via Wikimedia Commons.jpg>), CC BY 2.0

A phase change material is solid and easy to handle at room temperature, then melts around 45–65 °C — during the product’s first power-up — and wets the interface like a grease that can’t run away. That transition is the whole point, and it’s the part every competing article glosses over; we unpack it in what is phase change material (PCM) TIM. ZIITEK’s TIC®800 PCM transitions at 50–60 °C.

Thermal performance: stop comparing raw W/m·K

The single most common mistake in TIM selection is ranking materials by datasheet conductivity. Bulk conductivity (k) is a material property; what your junction actually feels is thermal resistance across the whole joint:

R″ ≈ (bond-line thickness ÷ k) + contact resistance

A pad rated 6 W/m·K sitting at 1.5 mm can present more resistance than a 3 W/m·K filler that squeezes down to 0.2 mm and wets both faces. Thickness and surface wetting frequently beat raw k. This is also why PCM — despite a middling k — often wins on a flat CPU lid: once it melts it can wet down to a bond line of tens of microns, which a pad at its nominal thickness cannot match. That advantage is conditional — lid flatness, a clamp that holds the joint closed, and enough first-power-up energy to complete the melt all have to hold. If the impedance-versus-conductivity distinction is fuzzy, our explainer on thermal resistance vs thermal impedance walks through the math, and understanding thermal conductivity (W/mK) covers where the k number comes from.

Two disciplines make the comparison honest:

1. Demand ASTM D5470 data. The ASTM D5470 through-plane method is how credible suppliers measure TIM impedance. A k value is only comparable if it’s stated with the test pressure and bond-line thickness it was measured at.
2. For pads, the datasheet k is a compressed-state number. A pad’s conductivity is measured on a specimen squeezed to a stated deflection under a stated pressure, and your joint only sees that value if the clamp reproduces the same deflection. Compression does not switch conductivity on or off — it changes how much of each face is wetted, which sets the contact-resistance term, and how thin the bond line becomes, which sets the bulk term, so both terms in the equation above move. Under-clamp the pad and the bond line stays thick and the interface resistance climbs, whatever the k says. Fillers and PCM follow the same physics with different levers: on a filler, dispense volume and the dispenser’s stop height set the bond line; on PCM, the melt-and-wet step plus the retained clamp set it. In all three cases the number to compare is resistance at your bond line and pressure, not k on its own.

Mechanical stress and component safety

This is where gap pads lose points and fillers shine. A pad only performs when it’s compressed — typically 10–50% of its thickness — and that deflection generates reaction force on whatever is clamping it. On a stiff assembly that’s fine. On a thin PCB, a large module, or a fragile BGA, that force warps boards and stresses solder joints. Harder, higher-k pads are the worst offenders.

Dispensable gap fillers need far less force: the paste flows into place, and the pressure required is only what it takes to reach the target bond line and keep the parts in contact. PCM behaves the same way once it has melted and wetted, provided the clamp still holds the joint closed while the material is liquid. So when the mechanical budget is tight — sensor modules, large flat lids over delicate silicon, thin-FR4 assemblies — the low-stress materials (filler, PCM) are the safer structural choice, even if a pad looks simpler. Compression behavior is a topic in its own right; see the thermal pad thickness & compression guide for how deflection targets are set.

Gap size and tolerance stack-up

Before anything else, check which regime your gap is in — because the three materials barely overlap on the one axis you can measure with a feeler gauge.

Real assemblies don’t hold a perfect, uniform gap — tolerance stack-up across the housing, PCB, and heatsink means the “1 mm gap” is really 0.6–1.4 mm across a batch. Each material handles that differently:

• Gap filler absorbs variation for free. Dispense a controlled volume and it fills whatever gap the parts present, part to part. This is why fillers dominate where gaps are large or inconsistent.
• Gap pad comes in fixed nominal thicknesses. To cover a variable gap you either over-compress (more stress) or leave air (worse resistance). Stacking two pads to reach a big gap is a common anti-pattern — it adds two more contact interfaces and usually raises resistance.
• PCM is for the opposite regime: thin, flat, well-controlled interfaces measured in tens of microns, not millimeters.

Assembly, automation, and throughput

Form factor decides how the material rides down your line:

• Gap pads are the throughput champion for manual or pick-and-place assembly: no equipment, no cure, no cleanup. Tape-and-reel and peel-and-place kits drop straight into existing flows.
• Gap fillers shine at high-volume automation : a robotic dispenser lays down a precise bead in seconds, no die-cut tooling, easy to change the pattern when the design revs. The trade is capital (a dispenser) plus nozzle maintenance and void inspection.
• PCM is often supplied with sheet format, which is extremely clean for the assembler — but it needs a burn-in / first-power-up step to melt and wet before it hits rated performance. If your takt time can’t absorb that, factor it in.

Photo by Ashley Pomeroy via Wikimedia Commons, CC BY 4.0

Reworkability and field serviceability

If a unit will be opened for service, rework, or upgrade, this row can override everything else. Gap pads win: peel the old one off, drop a new one in. Cured gap fillers lose: they’re effectively permanent, and removing one is messy. PCM is in between — it softens when warm and can be replaced, but it isn’t as clean as a pad. High-mix, repairable, or field-serviceable products lean toward pads for exactly this reason.

Reliability: pump-out, compression set, and outgassing

Steady-state numbers hide the failure modes that show up 1,000 thermal cycles later. The three materials fail in different ways:

• Pump-out is the classic risk for one-part fillers (and greases): repeated expansion/contraction slowly walks the material out of the joint, leaving voids. Two-part cured formulations and PCM grades are usually reported as less prone to it, but that is a trend across formulations, not a property of the material class. Crosslink density, filler loading, bond-line thickness, and the power/temperature cycle you actually run all move the result, so the question is what pump-out or cycling data exists for the grade in your joint — not whether the family is immune. A fixed-temperature dwell does not settle it either: that test holds the joint still, while pump-out is driven by cycling.
• Compression set is the pad-specific ager: a pad that’s held under load for years can take a permanent set and lose contact pressure, raising resistance.
• Outgassing matters for optics, sensors, and sealed enclosures. Silicone materials can release low-molecular-weight siloxanes; if that’s a concern, look at a silicone-free construction — ZIITEK’s 6060-11 non-silicone pad is one option — and screen the actual build against ASTM E595 (TML ≤ 1.0%, CVCM ≤ 0.1%). Silicone-free removes siloxanes as a source; it does not make the part volatile-free, because binders, plasticizers, and process residues can still outgas, so the screen has to run on the construction you will install. The same supplier's non-silicone range is a case in point: Z-Paster 100 15-02F is quoted at 1.5 W/m·K over −20 to 125 °C with TML 0.30%, and that TML belongs to that model — it does not transfer to 6060-11, and neither figure is a zero-outgassing claim. NASA’s public outgassing database is a useful reference for how materials are rated.

Qualification should cover thermal cycling (e.g., −40 → 125 °C), damp heat (85 °C / 85% RH), dielectric strength if the TIM is also an insulator, and flammability — UL 94 V-0 is standard for most electronics. What has actually been run, and what has not: ZIITEK specifies the TIF800TU and TIF700UU pads against 150 °C × 1000 h and 85 °C / 85% RH × 1000 h validation conditions. Those are test conditions, not a pass verdict or a retention figure — the cold-shock entry for these grades is filed as −40 °C ↔ 125 °C over 1000 h, which is a duration, not 1000 cycles. If your qualification plan needs a cycle number or a before/after value, ask for the current report and the fixture pressure it was run at instead of carrying a number across from another grade. For a plain-language primer on the material family, Wikipedia’s thermal interface material entry is a reasonable starting point. One warning if you go reading further: Wikipedia’s phase-change material entry describes the same physics put to a completely different job — storing thermal energy in buildings and packaging, not thinning a bond line. It’s useful background on the phase transition itself, but it is not about PCM as a thermal interface material, and conflating the two is the single most common source of confusion on this topic.

Cost and total cost of ownership

Per-gram or per-sheet pricing is misleading. Count the whole picture:

• Gap pads : low material and labor cost at moderate volume; inventory cost if you carry many thicknesses; waste from die-cut skeletons.
• Gap fillers : material can be cheaper per joint at scale and there’s near-zero geometric waste, but you pay for the dispenser (capital), maintenance, and cure/cycle time.
• PCM : material cost is moderate; the hidden cost is the burn-in step and tighter surface-flatness requirements.

At low-to-medium volume, pads usually win on TCO. At high volume with automation already in place, dispensed fillers often pull ahead. PCM earns its cost only where its thermal edge on a hot, flat interface is worth it.

Match the material to your application

These are starting points, not verdicts — your gap, your stress budget, and your rework requirement still govern, in that order. But notice how often the answer is not “the highest W/m·K product.” Our broader walk-through, how to choose a thermal interface material, extends the same logic to greases and tapes.

When NOT to use each

Vendor product pages rarely tell you this, so here it is plainly:

• Don’t use a gap pad to bridge a large or highly variable gap by stacking or heavy over-compression — you’ll add interfaces or over-stress parts.
• Don’t use a cured gap filler where the unit must be opened and reworked — you’re building in a permanent joint.
• Don’t use PCM where the interface isn’t flat and thin, or where production can’t accommodate a burn-in step.
• Don’t use any silicone TIM near optics/sensors without checking outgassing — evaluate a silicone-free construction and verify outgassing on that build too.

How to spec it right

Before you commit, put every candidate on equal footing:

• Get ASTM D5470 impedance curves, each point tagged with test pressure and bond-line thickness — not a lone k number.
• For pads, get the conductivity at your actual deflection , and the clamp force that deflection implies.
• Ask for reliability data at your conditions: thermal cycling, 85/85, and pump-out/compression-set results — with the conditions attached, since hours at temperature and number of cycles are not interchangeable, and neither means much without the fixture pressure.
• If outgassing matters, require ASTM E595 (or better) data and consider silicone-free chemistry.
• Confirm UL 94 V-0 and any dielectric requirements.

A supplier that can hand you all of that quickly is telling you something about their engineering discipline.

Where ZIITEK fits

As a thermal-materials manufacturer, ZIITEK makes all three families, which means the recommendation isn’t captive to one format:

• TIF® gel / gap filler (dispensable, ~1.5–9.0 W/m·K) for large, uneven, low-stress gaps and automated lines.
• TIF® gap pads (~1.25–16 W/m·K) for defined, reworkable gaps — the TIF800TU and TIF700UU pages carry the current datasheet for that route.
• TIC®800 PCM (~0.95–9.6 W/m·K, transition 50–60 °C) for flat, high-power lids that need the lowest resistance.
• Silicone-free route — the 6060-11 pad (6 W/m·K, 60 Shore OO, 1–5 mm) where siloxane outgassing is the constraint; rated impedance for that grade is quoted per thickness and pressure (0.234 °C·in²/W at 1.0 mm and 30 psi), so ask for the point that matches your stack.

Note that the PCM has the lowest headline conductivity of the three — and is still the right answer on a flat, high-power lid, because it wins on bond-line thickness and wetting, not on k. That’s this entire article in one line, and it’s why we publish the number rather than quietly omitting it.

The TIF® gel and TIF® pad lines carry UL 94 V-0 recognition under ZIITEK’s UL Yellow Card (E331100); for a PCM grade, ask for the flammability rating on that specific datasheet rather than assuming the company-level listing covers it. Across the range, materials are RoHS/REACH compliant and produced under IATF 16949 / ISO 9001 processes, and are benchmarked as drop-in alternatives to common incumbents. If you’re speccing for a specific gap, thermal budget, and assembly method, request samples of the grade you are weighing — TIF800TU or TIF700UU on the pad route, 6060-11 where silicone is excluded — with ASTM D5470 curves taken at your bond line and pressure, or ask for engineering support for your design.

FAQ

What’s the difference between a thermal gap filler and a gap pad? A gap filler is dispensed as a liquid/gel that conforms to the gap and then cures or stays soft; a gap pad is a pre-cured solid sheet you place and compress. Fillers handle large, uneven, or variable gaps at low mounting stress; pads are simpler to assemble and easy to rework but need clamp force and come in fixed thicknesses.

Is a phase change material better than a gap pad or gap filler? For a flat, thin, high-power interface (like a CPU/GPU lid), PCM usually gives lower thermal resistance than a pad and a much thinner bond line than either — and supplier data usually reports it as less prone to pump-out than grease, though that depends on the grade and on how the joint is clamped and cycled. For large or uneven gaps, a filler or pad is the right tool — PCM only shines on thin, flat, well-controlled joints.

What is a phase change thermal pad, and does it need burn-in? It’s a PCM supplied in pad form. It’s solid at room temperature and melts on first power-up (around 45–65 °C) to wet the interface — so yes, it reaches rated performance only after that first thermal excursion (the “burn-in”).

What are thermal gap pads made of? Typically a silicone (or silicone-free) polymer matrix loaded with ceramic fillers such as alumina, boron nitride, or aluminum nitride, sometimes with a fiberglass or film carrier. Filler type and loading set the conductivity, hardness, and electrical behavior.

Which is better for high-volume automated assembly, a gap filler or a gap pad? Dispensable gap fillers suit high-volume automation — a robot lays a precise bead with no die-cut tooling and easy pattern changes. Pads are excellent for manual or pick-and-place lines and need no dispensing equipment. The tie-breaker is whether you already have (or want) a dispenser, and whether rework is required (favoring pads).

Can I use thermal putty instead of a pad? Thermal putty is essentially a soft, non-curing gap filler — it conforms like a filler with low stress and no dispenser needed for small runs, but it’s messier to rework and doesn’t offer a pad’s clean handling. It’s a reasonable middle ground for uneven gaps in low-to-medium volume.

How do gap fillers and PCM compare to thermal paste? Thermal paste (grease) gives a very thin bond line but is prone to pump-out and dry-out over time. PCM offers grease-like thinness with far less pump-out exposure than a grease; fillers and pads trade some thinness for the ability to span real gaps. As with the other two, ask what cycling data exists for the specific grade rather than reading the family as immune. See thermal pad vs thermal paste for that head-to-head.