Semiconductor & Display Materials — Technical Specification Overview

Trace Metal Profiles, Dielectric Constants and Process Chemical Compatibility — What the COA Doesn’t Show You #

Semiconductor and display materials span a wide functional range — etch chemicals, CMP slurries, dielectric precursors, organic semiconductors, ITO sputtering targets, liquid crystal intermediates. Across all of them, the COA you receive from a Chinese supplier will almost always show bulk purity and, if you’re lucky, a particle count. What it almost never shows is the trace metal profile broken down by ionic species, which is the parameter your fab process actually responds to.

The distinction matters because Chinese national standards governing these materials — published through SAC China Standards (GB/T) — define purity in terms that are aggregated differently from SEMI International Standards. A material certified at 99.999% (5N) purity under GB/T may contain Fe²⁺ at 50 ppb and Na⁺ at 80 ppb and still be fully compliant. SEMI C12 (for bulk process chemicals) sets Fe²⁺ limits at 10 ppb and Na⁺ at 20 ppb for Grade 3. At Grade 4, both drop to 2 ppb. These are not interchangeable grades, and suppliers rarely volunteer the difference.

For display panel materials — particularly organic transport layer intermediates and liquid crystal alignment agents — the relevant parameter shifts from ionic contamination to thermal decomposition onset. Materials arriving with the correct sublimation temperature on the datasheet but a decomposition onset 8–12°C lower than specified will degrade in the evaporation chamber before they deposit correctly. We flag this as a Category B risk in our incoming material assessment protocol (IMA-B), and it surfaces in roughly one in four organic display material lots from first-time suppliers.

The Parameter Most Teams Misdiagnose: Ionic Species Distribution vs. Aggregate Purity #

This is where misqualification happens most often, and the mechanism is straightforward once you see it.

Aggregate purity — expressed as 5N, 6N, or in some datasheets as a weight percentage — is a subtraction value. It tells you what fraction of the material is not the target compound. What it does not tell you is what the remaining 10–100 ppm consists of. A 5N hydrogen peroxide for semiconductor use could have its 10 ppm impurity budget distributed across dozens of species in small concentrations, or it could carry 9 ppm of a single ionic contaminant that happens to have a disproportionate effect on your gate oxide or copper interconnect.

The key insight from process chemistry is that contaminant impact is not linear with concentration. Fe²⁺ in a wet etch bath at 15 ppb will not produce 50% of the damage it causes at 30 ppb — it can produce dramatically more, depending on the redox environment, the substrate metallization layer, and the bath temperature. Na⁺ in a dielectric deposition precursor does not cause proportional threshold voltage shift; it accumulates at the oxide-silicon interface and triggers instability well below the concentration level that bulk purity measurements would flag.

Chinese suppliers operating under GB/T frameworks are not doing anything fraudulent. They are certifying to a different specification. The gap is structural: GB/T standards for electronic-grade chemicals were written to support domestic fab processes running at 90nm nodes and above, where ionic contamination tolerances are genuinely wider. As Chinese fabs have migrated to 28nm and below, some domestic standards have been updated, but the update cycle lags by 2–4 years and tier-2 suppliers typically certify to the older revision.

To confirm whether a supplier’s product actually meets the ionic distribution you need, you need ICP-MS (inductively coupled plasma mass spectrometry) results broken down by individual species — not a summary purity number. The confirmation threshold we use in qualification: any single metallic ion species must be below 10 ppb for SEMI Grade 3 equivalence, and below 2 ppb for Grade 4. If a supplier cannot provide species-level ICP-MS data for three consecutive production lots, the material does not proceed past our first-sample review.

This holds for wet process chemicals and for solid-phase materials like sputtering targets. For ITO targets, the critical species is Sn-to-In ratio uniformity across the target body — not the nominal composition, which is almost always correct, but the spatial distribution, which determines sheet resistance uniformity across the deposited film.

Semiconductor and Display Material Grade Comparison: Key Parameters by Application #

The table below consolidates specification parameters across three functional material categories commonly sourced from Chinese suppliers. Values reflect SEMI standard requirements and our own incoming inspection thresholds.

The GB/T column is not included to criticize Chinese standards — it is included because procurement teams sourcing from China need to know exactly where the gap sits before writing their purchase specification. A supplier offering “SEMI equivalent” without specifying which grade and which revision is not lying, necessarily — but the phrase carries no enforceable meaning until you attach a species-level ICP-MS threshold and a lot-acceptance criterion.

Corrective Actions When Incoming Lots Fail Ionic Spec #

When a material lot arrives and fails at incoming inspection on ionic species distribution, there are four realistic paths. They are listed in order of increasing cost and time commitment.

1. Return and re-source from an alternative supplier. The fastest option if you have a qualified backup supplier on your AVL. Adds 2–4 weeks lead time depending on whether the backup supplier holds safety stock. Appropriate when the failure is a one-off deviation and the primary supplier has a clean track record on prior lots.

2. Request re-purification from the same supplier and re-submit. Some Chinese suppliers — particularly those with in-house distillation or zone refining capability — can re-process a lot and retest. This works for wet chemicals more reliably than for solid-phase materials. Expect a 3–6 week turnaround. Request a new ICP-MS certificate on the re-processed lot; do not accept a repeat of the original COA with an updated date.

3. Adjust your incoming inspection protocol to add species-level ICP-MS as a standard gate. This fixes the detection problem but not the supplier problem. Adds cost of roughly USD 150–300 per lot for external ICP-MS testing. I’d prioritize this as a parallel action — not a substitute for supplier corrective action — because it gives you the data history to escalate or exit the supplier relationship with evidence.

4. Initiate a formal SCAR (Supplier Corrective Action Request) and conduct a raw material traceability audit. This is the thorough option. Most incoming failures on ionic spec trace back to the compounder or raw material source, not to the finishing or packaging step. A traceability audit asks the supplier to document their raw material inputs, their upstream supplier qualifications, and their in-process QC data at the synthesis stage. In our qualification program, we use this path for any supplier where two consecutive lots have failed the same ionic parameter — because at that point it’s a process issue, not a sampling issue.

Options 1 and 3 together fix 80% of cases in the short term. Options 2 and 4 are required if you intend to stay with the supplier long-term.

What to Specify Upfront — Procurement Language That Closes the Gap #

The most common procurement error for semiconductor and display materials from China is specifying purity grade without specifying the standard revision and the ionic species limits individually. “5N purity, SEMI Grade 3” on a PO is insufficient if it is not accompanied by a species-specific acceptance table.

The purchase specification should explicitly state: which SEMI standard revision governs acceptance, the maximum concentration for each ionic species you care about (Fe²⁺, Na⁺, K⁺, Cu²⁺, and Cr³⁺ at minimum for wet process chemicals), the particle count method and threshold, and the lot-by-lot COA requirement including ICP-MS data. For sputtering targets, add composition uniformity tolerance (±0.5 wt% Sn:In ratio across the target face) and density requirement.

Before first volume shipment, request three consecutive lot COAs from production batches, not from qualification samples. The document to request is the supplier’s internal QC specification sheet (not the marketing datasheet) alongside the ICP-MS raw data file from an accredited third-party lab.

For related sourcing considerations on conductive and functional materials for electronic applications, the same ionic specification discipline applies to conductive pastes and transparent electrode materials.

Practical Guidance for Buyers #

When sourcing semiconductor and display materials from Chinese suppliers, the first specification to request is not the purity certificate — it is the ICP-MS breakdown by ionic species from the most recent three production lots. Bulk purity is the easier number to produce and the less informative one. Ionic species distribution is what your process actually responds to, and it is where GB/T and SEMI frameworks diverge most significantly.

The specific risk scenario to anticipate: a supplier qualifies cleanly on first-sample submission, which is often prepared with extra care or sourced from a higher-grade raw material batch. Production volume lots revert to the standard raw material input. If your incoming inspection relies on the supplier’s COA rather than your own third-party ICP-MS gate, you will not catch the drift until it shows up as yield loss or device instability — typically 4–8 weeks into production.

Before committing to volume, require a minimum of three consecutive production-lot COAs with species-level ICP-MS data, each showing Fe²⁺ below your acceptance threshold (10 ppb for SEMI Grade 3 equivalence, 2 ppb for Grade 4). For OLED organic materials, add a TGA decomposition onset test with a minimum threshold of 290°C. Run your own incoming spot test on the first three volume lots regardless of supplier certification. That incoming gate costs less than one yield excursion.

For teams also evaluating PCB and electronic substrate materials, the trace contamination discipline described here translates directly — ionic contamination in laminate resin systems causes similar reliability failures through different mechanisms.

On the question of whether to requalify annually: some procurement teams do annual requalification as standard practice; others only requalify after a supplier reports a formulation or raw material change. Our practice for semiconductor-grade chemicals is annual requalification for any supplier running below 18 months of tracked lot history, and change-triggered requalification after that. There is no universal answer here — the calculus changes if your supplier has a stable, audited raw material chain and consistent ICP-MS data across two years of production.

FAQ #

Is a Chinese supplier’s GB/T certificate equivalent to SEMI C12 Grade 3 for semiconductor wet process chemicals?
No — and the gap is specific: GB/T tolerances for Fe²⁺ and Na⁺ are typically 3–10× wider than SEMI C12 Grade 3 limits. A GB/T-compliant material may carry Fe²⁺ at 50 ppb where SEMI Grade 3 allows 10 ppb. Treat GB/T certification as a starting point for supplier dialogue, not as acceptance documentation.

What is the minimum ICP-MS data we should require on a COA for semiconductor-grade chemicals?
Species-level data for at least Fe²⁺, Na⁺, K⁺, Cu²⁺, and Cr³⁺, from an accredited third-party lab, for each production lot. A summary purity number without species breakdown does not give you the information your process qualification requires.

Can Chinese suppliers actually achieve SEMI Grade 4 purity levels (2 ppb per ionic species)?
Some can. The capability exists at first- and second-tier suppliers with dedicated zone refining or sub-boiling distillation equipment. The challenge is not whether they can produce it — it is lot-to-lot consistency at production volume. Across our 2023–2024 qualification program, roughly one in three Chinese suppliers claiming Grade 4 capability could demonstrate it across six consecutive lots.

For ITO sputtering targets, what is the most common out-of-spec condition seen at incoming inspection?
Density below the 6.8 g/cm³ threshold, which indicates residual porosity from the sintering process. Low-density targets produce non-uniform film deposition and higher particle generation during sputtering. This is measurable with Archimedes density testing — a faster and cheaper check than full compositional analysis, and a reliable early screen.

Should we specify decomposition onset temperature for OLED organic transport layer materials?
Yes. The datasheet sublimation temperature is not sufficient. You need a TGA (thermogravimetric analysis) result showing decomposition onset above 290°C at a 10°C/min ramp rate under N₂ atmosphere. Materials with onset below that threshold will partially decompose in the evaporation source before they deposit, which shifts your layer thickness and affects device efficiency in ways that are difficult to trace back to the material without this data.

How do we evaluate lot-to-lot consistency before committing to volume orders?
Request three consecutive production-lot COAs — not qualification samples, which are often prepared separately. Look specifically at whether the Fe²⁺ and Na⁺ values are consistent within ±30% across the three lots. Wide variation across lots on the same parameter, even if all three pass individually, signals an unstable raw material input or in-process control issue.

If a material passes our incoming ICP-MS check but we still see yield issues, where do we look next?
The next variable to check is particle size distribution in the liquid, not just particle count. A lot with 40 particles/mL at ≥0.5 µm can still carry problematic contamination if there is a tail of particles at ≥0.2 µm that your count method did not capture. Also verify that the chemical compatibility of your storage and transfer system has not introduced secondary contamination — some ionic species migrate from container materials, particularly at elevated storage temperatures.

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