Thermal Interface Material — Supplier Qualification Guide

What the COA Doesn’t Tell You — and What It Should #

Every TIM supplier will hand you a COA with thermal conductivity front and center. It’s the number that gets quoted in sales conversations, the number procurement teams put on their approved vendor list criteria, and the number that, frankly, tells you the least about what will happen in your assembly line.

Thermal conductivity is a bulk material property measured under idealized lab conditions — typically per ASTM D5470 at a fixed pressure and controlled surface finish. Your actual assembly doesn’t look like that. Your heatsink surface has a Ra of 0.8–1.6 µm, your clamping torque varies by ±15% across operators, and your bond line thickness shifts with every component height tolerance stack-up.

The parameters that matter — BLT consistency, compression deflection, outgassing rate, and long-term thermal impedance drift — appear on fewer than one in three COAs we review from Chinese suppliers. That gap is not accidental. It reflects what’s easy to test and what suppliers prefer not to commit to.

Head-to-Head Comparison — TIM Formats and What They Actually Deliver #

The format decision drives everything downstream: which COA fields are relevant, which incoming tests are feasible, and where your qualification risk lives.

Graphite sheet is the only format where through-plane conductivity — the number that matters for vertical heat transfer — is a fraction of the headline in-plane number. We flag this specifically because Chinese supplier datasheets frequently lead with the in-plane value (150–400 W/m·K) without making the direction explicit. For a buyer specifying a CPU-to-heatspreader interface, that number is essentially irrelevant. Through-plane for graphite is typically 5–10 W/m·K, which is competitive with a good silicone pad but not the order-of-magnitude advantage the headline suggests.

For the most common industrial and EV battery applications we qualify — where assembly pressure is controlled and the interface geometry is well-defined — phase change pads represent the lowest qualification risk from Chinese suppliers. The pre-formed geometry means BLT is measurable before installation, and lot-to-lot consistency is trackable with basic incoming tools. Thermal grease is the opposite: application variability at the line introduces a degree of freedom that makes incoming inspection nearly meaningless without process audit data.

The Overlooked Variable — Lot-to-Lot Filler Loading Consistency #

COA values reflect a tested sample from a given lot. What they don’t capture is whether the next lot used the same filler loading, the same particle size distribution, or the same silicone base viscosity as the approved sample.

Chinese TIM compounders — particularly second- and third-tier suppliers operating in Dongguan and Shenzhen — frequently source aluminum nitride, boron nitride, or alumina filler from multiple spot-market vendors depending on quarterly price. A 10% shift in filler particle size distribution can reduce effective thermal conductivity by 0.5–1.2 W/m·K without any change to the COA headline number, because the measurement conditions don’t capture the assembly-level interface resistance contribution.

In our supplier risk-tiering process (what we internally call the M-3 Consistency Gate), we require six consecutive production lot COAs before recommending any Chinese TIM supplier for volume qualification. Of the 19 suppliers we evaluated in 2023–2024, only seven could produce this data without gaps or substituted test dates. Three of those seven showed conductivity variation exceeding ±0.8 W/m·K across lots — enough to shift thermal performance meaningfully in a tight thermal budget design.

The scenario where this becomes a real problem: a buyer qualifies a TIM supplier on an ISIR sample, approves it, and places monthly releases. Twelve months later, field returns start showing elevated junction temperatures. The root cause, traced back through our M-3 gap analysis, is a filler source change at the compounder that happened in month four and was never disclosed. The COA showed values within tolerance on every shipment. Incoming hardness and compression curve testing would have caught the drift by month five.

This is the argument for functional incoming inspection — not documentation review.

Implementation Notes — After You’ve Chosen a Supplier #

The qualification decision is the beginning of the risk management process, not the end. Once a supplier is conditionally approved, the incoming protocol needs to be specific and non-negotiable.

For phase change pads and silicone pads, our standard incoming checklist covers:

• Thickness measurement: 10 samples per lot, caliper or micrometer, acceptance criterion ±0.05 mm from nominal (tighter than most supplier datasheets specify)
• Shore OO hardness (for silicone pads): 5 samples per lot, accept if within ±5 points of ISIR baseline — deviations signal filler or base polymer changes
• Compression deflection curve: 1 sample per lot at 10 psi, 30 psi, and 50 psi; compare to approved baseline curve; flag if BLT at 50 psi deviates >8% from baseline
• Visual/dimensional: 100% for cut accuracy, 5% AQL for surface contamination per ASTM E2234 sampling plan

For thermal grease, incoming inspection shifts to rheology — viscosity at 25°C (typical acceptance window: ±15% of approved baseline) and non-volatile content. Thermal testing on grease lots is expensive and slow; the practical proxy is viscosity plus an oil bleed test per ASTM D1742.

On qualification timeline: allow a minimum of 90 days from first sample receipt to volume approval. That timeline accommodates three production lots, allows for incoming data accumulation, and includes one thermal cycling validation run — typically 500 cycles, -40°C to 125°C per IEC 60068-2-14 — on assembled test vehicles. Compressing this timeline is the single most common cause of costly mid-ramp disqualifications.

One specific red flag in early shipments: if Shore hardness on silicone pads increases by more than 8 points between your ISIR sample and the first production lot, the supplier has changed either the crosslink density or the filler ratio. Neither change will necessarily show on a conductivity COA. Catch it at incoming or catch it in the field.

Practical Guidance for Buyers #

When sourcing TIM from China, don’t open the conversation with thermal conductivity. Open it with bond line thickness tolerance and lot consistency data. Those two parameters require more manufacturing discipline than conductivity, and a supplier’s willingness to commit to them in writing tells you more about their process control capability than any single datasheet value.

The specific risk to guard against: a supplier who performs well on ISIR but ships inconsistent production lots. The mechanism, as described above, is usually filler source variation — something that happens at the raw material level, below the supplier’s finished-goods QC radar. Incoming compression curve testing at 50 psi, compared against your approved baseline, will catch this before it reaches your assembly line. A deviation of more than 8% in BLT at 50 psi should trigger hold-and-investigate, not just documentation follow-up.

Before committing to volume, insist on three consecutive production lot COAs plus one thermal impedance measurement per ASTM D5470 at your specified assembly pressure — not at the supplier’s standard test condition. The delta between those two measurements is the number your thermal engineers actually need. If a supplier can’t or won’t provide assembly-pressure-specific data before volume commitment, that resistance is itself a qualification signal.

For related sealing and interface consumable qualification protocols, see our guides on pump and valve seals and o-rings and static seals.

Does thermal conductivity on the COA accurately reflect assembly performance?

Not reliably. COA conductivity is measured at a standardized pressure and surface condition that rarely matches your assembly. The more predictive value is thermal impedance measured at your actual clamping pressure — typically 10–50 psi depending on format — which can differ from the nominal conductivity-derived value by 20–35% in real assemblies.

What’s the minimum sample size for incoming inspection on TIM pads?

For thickness and Shore hardness, 10 samples per lot is our standard for lots under 5,000 pieces. For compression deflection, 1 sample per lot is sufficient if you’re comparing against a locked baseline curve rather than absolute values.

How do I detect a filler source change without sending material to a lab?

Shore OO hardness and compression deflection at 50 psi are your two fastest proxies. A filler ratio or particle size change will almost always shift at least one of these before it shows up in thermal performance. Full XRF or TGA analysis is the confirmatory tool, but the incoming proxies catch roughly 80% of meaningful changes before you need the lab.

Should I qualify a TIM supplier against GB/T standards or ASTM?

It depends on your downstream compliance requirements. GB/T 10294 and ASTM D5470 measure thermal conductivity differently — GB/T uses a guarded hot plate method, ASTM D5470 uses a pressure-dependent steady-state column. For EV battery applications shipped to Europe or North America, ASTM D5470 results are more directly comparable to your design data. For domestic Chinese OEM supply chains, GB/T is acceptable. Specify which standard you require in the purchase order, not just in the RFQ.

What’s a realistic qualification timeline for a new Chinese TIM supplier?

90 days minimum from first sample to volume approval, assuming no failures requiring root cause and resubmission. We’ve seen buyers compress this to 30 days under schedule pressure and spend six months unwinding the consequences. The 90-day window isn’t conservative — it’s the minimum needed to accumulate three production lots and run one thermal cycling validation.

Published by sinoraw.com Technical Team | Request a sourcing consultation