Conductive Ink & Functional Paste — Troubleshooting & Failure Guide

Resistance Drift, Delamination, and Adhesion Failure: Measurable Thresholds and What They Indicate #

Cured sheet resistance is the first electrical parameter most buyers specify, and it is also the first one to drift in service. For screen-printed silver conductive paste on PET substrates, a properly cured trace at 20 µm dry film thickness should measure between 10 and 25 mΩ/sq — anything above 35 mΩ/sq after a standard cure schedule (typically 130°C / 30 minutes for most thermoplastic-compatible silver pastes) warrants process investigation before further qualification. The threshold is not arbitrary: at 40+ mΩ/sq, a 50 mm trace on a membrane switch begins to generate measurable voltage drop under 20 mA operating current, which is the point where switch contact reliability becomes inconsistent.

The data that procurement teams rarely request — but that separates stable suppliers from marginal ones — is resistance stability after humidity aging. The relevant test is 85°C / 85% RH for 500 hours, per IEC 60068-2-78, with a pass threshold of less than 20% resistance increase from initial. In our evaluation of silver paste suppliers across three sourcing events between 2022 and 2024, fewer than half could provide this data from actual test runs rather than formulation estimates.

The adhesion figure matters as much as resistance, particularly for any application involving substrate flexion. A cross-hatch rating of 2 under ISO 2409 — meaning some squares peel at grid intersections — is a reliable early indicator of adhesion failure under repeated bending. For membrane switches rated to 1 million actuations, an initial cross-hatch rating above 1 is disqualifying in our QC-07 material risk procedure.

What Actually Causes Resistance Drift — Process Variables vs. Material Variables #

Resistance drift after deployment is the failure mode most commonly misattributed to the ink itself. In practice, a well-formulated silver paste from a qualified Chinese supplier will hold its electrical properties under stable process conditions. The failure almost always enters through the print and cure process.

Undertreated cure is the primary mechanism. Silver flake-based pastes require the binder resin to fully cross-link or solidify in order to establish particle-to-particle contact across the film. When peak cure temperature is reached but dwell time is cut short — a common outcome when production throughput is prioritized and oven belts are run faster than qualified — residual solvent remains trapped in the film. Residual solvent above roughly 2.5% weight loss (measurable by thermogravimetric analysis against a fully cured reference) creates internal film stress and expands the particle-to-particle gap under thermal cycling. Resistance climbs. The climb is not linear: it accelerates above 60°C operating temperature, which is why the problem often appears acceptable during bench testing and surfaces only after the product has been in service for several months.

Silver migration is the second mechanism, and it operates on a longer timescale. Under DC bias in humid conditions, silver ions migrate from anode to cathode across the binder matrix, forming conductive dendrites that can cause inter-trace shorts. The critical variable is the binder’s moisture vapor transmission rate, not the silver loading. Chinese suppliers who formulate with lower-cost acrylic binders rather than polyester or epoxy-modified urethanes tend to produce pastes with higher MVTR values — and consistently worse long-term isolation resistance under humidity bias testing. Isolation resistance below 10⁸ Ω after 96 hours of humidity bias (50 VDC, 85% RH, per ASTM D257) is a failure condition for any PCB-adjacent conductive ink application.

This is where we see the most significant quality gap in the Chinese supplier base. In our Category B incident tracker, silver migration-related field failures from Chinese-sourced paste account for roughly two-thirds of reported cases — but only about a third of those were attributable to paste formulation. The rest traced back to post-print processing: wash steps that removed the adhesion primer layer, improper topcoat selection that trapped humidity against the trace, or assembly processes that subjected the cured ink to IPA cleaning incompatible with the binder chemistry.

Carbon paste delamination follows a different failure pathway. Carbon-based conductive pastes applied over silver bus bars — a standard construction in heated automotive glass and large-format membrane switches — delaminate when the thermal expansion mismatch between the carbon layer and silver layer exceeds the peel strength of the interface. This is not a material substitution problem or a cure problem. It is a specification problem: the carbon paste must be selected with a coefficient of thermal expansion within roughly 15% of the underlying silver paste’s CTE, and that data is almost never on the COA. Buyers who specify carbon paste by resistivity alone and ignore CTE compatibility will see edge delamination start at approximately 500 cycles of thermal shock between -40°C and +85°C.

What I’d check first in any delamination case: confirm that the carbon paste supplier and silver paste supplier are either the same company or have issued a formal compatibility validation. A one-page compatibility note is not sufficient — request actual peel strength data at the interface after 500 hours of thermal aging.

Does Particle Size Distribution Affect Print Quality More Than Silver Loading? #

Yes, and the effect is measurable. Silver loading — the weight percentage of silver flake in the formulation — is the number most buyers fixate on when comparing pastes. Loading levels between 65% and 80% by weight are common in commercial thermoplastic silver pastes, and the difference in cured resistivity across that range is real but often smaller than the difference caused by particle size distribution (PSD) variation within a single supplier’s production lots.

A paste with 72% silver loading and tight D90 < 8 µm PSD will consistently outperform a 78% loading paste with D90 variability spanning 6 to 14 µm across lots — because irregular large flakes create print defects at fine-line geometries (below 200 µm line width) and reduce film uniformity. In our AVL gate review for membrane switch applications, PSD specification is now a hard requirement alongside resistivity: D50 within ±1.5 µm and D90 within ±2.5 µm across five consecutive production lots.

For applications where line width is above 500 µm and current density is the primary concern, silver loading matters more than PSD. The calculus shifts at fine-geometry applications.

Practical Guidance for Buyers #

When sourcing conductive paste from China, the first specification to request is not silver loading — it is cured sheet resistance at your actual cure schedule, not the supplier’s standard test condition. Chinese suppliers routinely publish resistivity data at an ideal cure condition (135°C / 60 min in a laboratory oven with no substrate heat sink effect). Your production tunnel oven at the same set temperature will deliver 10–15°C less to the ink surface when running at production belt speed, which shifts cured resistance upward by 15–25% in our measured experience.

The specific risk to guard against in initial qualification is lot-to-lot PSD shift. A supplier who passes initial sample approval with excellent resistivity data can deliver production lots with wider PSD that print inconsistently at your line geometry. Request COA data from three consecutive production lots before granting AVL approval — not three samples from the same lot re-tested.

Before volume commitment, insist on a 500-hour humidity aging test (85°C / 85% RH) conducted on your substrate, at your cure schedule, with your print parameters. Supplier-provided aging data on their own test substrate is useful background, but it does not substitute for validation on your stack. Specify a pass threshold of less than 20% resistance increase and provide it in writing before testing begins, so there is no ambiguity about acceptance criteria.

For buyers working with conductive-functional-materials or related pcb-electronic-substrates, incoming inspection protocols need to be calibrated to the specific binder chemistry — not treated as generic film testing.

Frequently Asked Questions #

Can I use the same silver paste qualification protocol for carbon and silver-carbon blend pastes?

No — carbon and silver-carbon blends require separate evaluation for resistivity stability under humidity, because carbon black’s surface chemistry interacts with moisture differently than silver flake. A silver paste protocol that passes at 10⁸ Ω isolation resistance will not catch carbon paste failures that manifest as gradual resistivity increase above 60% RH without any catastrophic short.

My paste passes initial COA but shows 35–50% resistance increase after 200 thermal cycles. What’s the most likely cause?

That delta — appearing after thermal cycling but not immediately after cure — almost always points to either solvent retention above 2.5% or an adhesion primer incompatibility between the paste binder and substrate surface energy. Run a TGA on a freshly cured sample from the production lot against a lab-cured reference. If residual weight loss at 200°C exceeds the reference by more than 1.5 percentage points, the cure cycle is the issue. If TGA is clean, check substrate surface energy with a dyne pen: values below 38 mN/m on PET before printing indicate insufficient corona treatment and will cause adhesion degradation under thermal stress.

What silver loading percentage should I specify for a heated seat element application?

It depends on your line width and the operating current. For 500 µm line widths at currents above 500 mA, 70–75% loading with a polyester binder is a reasonable starting spec. For narrower lines under 300 µm, prioritize PSD tightness (D90 < 8 µm) over loading percentage.

Is GB/T testing equivalent to IEC for conductive ink electrical characterization?

SAC China Standards (GB/T) electrical test methods for conductive inks — including resistivity and adhesion — allow wider test condition tolerances than the equivalent IEC or ASTM methods. A supplier who reports compliance only to GB/T is not lying, but their numbers may not be directly comparable to your IEC-referenced datasheet values. Always request the specific test standard and conditions used for each reported parameter, not just the result.

How many lots should I sample before approving a Chinese conductive paste supplier?

Three is the minimum for our qualification threshold; five is the standard for any supplier entering our preferred vendor list. Fewer than three lots cannot reveal PSD or resistivity drift trends that only appear across production batches made from different raw material deliveries.

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