How to Read a High-Purity TiO₂ COA for Electronic Ceramic Production

A high-purity TiO₂ certificate of analysis can look reassuring at first glance: a high assay, several low impurity numbers, and a particle-size result. But for electronic ceramics, the useful question is not simply whether the numbers are small. It is whether the certificate identifies the correct lot, uses relevant test methods, reports results on a comparable basis, and controls the variables that matter to your formulation and firing process.

1. Start with certificate identity, not the chemistry table

A certificate of analysis is intended to connect test results to a specific delivered lot. Before discussing purity, verify that the document can be traced to the material you are receiving. At minimum, check the supplier name, product grade, lot or batch number, manufacturing or release date, COA issue date, quantity or shipment reference where applicable, and authorization or quality approval.

The grade name matters. “High-purity titanium dioxide” is a material category, not a complete identity. Different grades can have different crystal phases, particle-size distributions, surface treatments, impurity profiles, and intended applications. For example, a pigmentary rutile TiO₂ optimized for coatings should not be assumed to behave like a high-purity rutile titanium source for MLCC, PTC thermistor, microwave dielectric, or optical-glass production.

Also check whether the certificate references controlled analytical procedures. A result such as “Fe: 0.002%” is much more useful when the supplier can state the sample-preparation route, instrument or method, reporting basis, and relevant detection or quantification limit. If a test method is only identified by an internal code, buyers can request a method summary during supplier qualification.

2. Lot result, specification limit, and typical value are not interchangeable

This distinction is one of the most important parts of COA review:

A TDS may publish a representative TiO₂ content, moisture level, impurity profile, and D50. A lot COA should report the values required for batch release. If the document repeats the same “typical” numbers on every shipment, ask which properties were actually tested for that lot and which were reported by periodic or skip-lot testing.

3. How to read TiO₂ assay and crystal phase

TiO₂ content: ask “on what basis?”

TiO₂ assay is a logical starting point, but it does not describe the entire powder. A reported value may be expressed on an as-received basis or a dry basis. That difference becomes relevant when moisture or other volatile loss is present. Two results should not be compared until the reporting basis is the same.

A high TiO₂ assay also does not reveal which impurities make up the balance, how those impurities are distributed, or whether the powder has the particle contact and reactivity needed by the ceramic process. In barium titanate and other solid-state synthesis routes, for example, composition control, mixing, particle contact, calcination, and residual phases must be considered together.

Rutile content: phase identity has process consequences

Rutile and anatase are both TiO₂, but they are not automatically interchangeable as titanium precursors. Phase identity can affect density, surface characteristics, reaction kinetics, thermal evolution, and process conditions. A rutile-content result should therefore be read as a phase-control parameter, not as another way of stating chemical purity.

For XC-258 electronic-grade titanium dioxide, the published product data identifies a high-rutile material intended for electronic and functional ceramic applications. Whether that phase is appropriate still depends on the customer’s synthesis route and validated firing profile.

4. How to interpret Fe, Si, Na, K, Ca, and Mg

Trace impurities deserve individual review because a total-purity number can hide an application-relevant element. The same nominal TiO₂ assay can be accompanied by different impurity profiles, and different ceramic systems have different sensitivities.

Convert percent to ppm carefully

Electronic-material specifications are often discussed in ppm, while a supplier document may use percent. For mass fraction, the conversion is:

The arithmetic is simple, but interpretation still requires care. A published value of “K + Na: 0.006%” corresponds to 60 ppm combined; it does not prove that K is 30 ppm and Na is 30 ppm. One element could account for most of the combined result. If your formulation has an individual sodium limit, request individual Na and K data.

“Not detected” does not mean zero

An ND result means the analyte was not detected under the stated analytical conditions. It does not establish that the element is absent. Ask for the method detection limit or reporting limit, because “Si: not detected” could mean below 5 ppm under one method and below 50 ppm under another. Those statements are not equivalent for a demanding electronic ceramic process.

The analytical technique also matters. ICP-OES, ICP-MS, and XRF have different strengths, practical detection ranges, matrix effects, and sample-preparation requirements. Even with a sensitive instrument, incomplete digestion can bias a result low. Method comparability is essential when supplier COA data and customer incoming-QC data do not agree.

For a deeper application discussion, see why low iron matters in electronic ceramics and optical glass.

5. Moisture and loss on ignition measure different things

Moisture at 105°C generally targets mass lost during controlled low-temperature drying. It can affect weighing accuracy, dry-basis calculations, slurry solids, powder flow, storage stability, and formulation consistency. The result depends on test temperature, time, sample mass, container geometry, and endpoint definition.

Loss on ignition is measured after heating under a defined, usually higher-temperature procedure. Depending on the material and method, LOI may include residual moisture, adsorbed species, volatile residues, structural water, or other thermally removed components. It is therefore incorrect to use moisture and LOI as synonyms or to add them automatically.

When a COA reports both values, ask:

• What temperature and hold time were used for each test?
• Was the sample pre-dried before the LOI test?
• Are results reported on an as-received or dry basis?
• Does the customer’s incoming test use the same endpoint?

These details are particularly important when a small numerical difference can change a batching calculation or when the material will be used in a moisture-sensitive mixing step.

6. Particle size: D50 is useful, but it is not a complete distribution

D50 is the median particle size: under the stated measurement convention, half of the measured distribution lies below that size and half above it. It is useful for screening, but two powders with the same D50 can have different fine fractions, coarse tails, spans, agglomeration states, slurry behavior, packing, and reaction performance.

Particle-size results are also method-dependent. Laser diffraction data can change with dispersing medium, dispersant, solids concentration, sonication energy and time, optical model, refractive index, and circulation conditions. A number reported after aggressive ultrasonic dispersion may not describe the agglomerates that actually enter your mixer.

If particle size is critical, request D10, D50, D90, distribution span, test method, and sample-preparation conditions. Ask whether the result describes dispersed agglomerates or another operational particle state. Then correlate incoming data with milling energy, slurry viscosity, mixing uniformity, calcination behavior, fired microstructure, and defects.

Method names alone do not guarantee comparable results

A COA should ideally identify the analytical standard or internal method used for each key parameter. Yet even laboratories following the same general technique may obtain different results because of sampling, calibration, digestion, dispersion, blank correction, and rounding. For supplier-customer correlation, exchange method details and test split samples from the same homogenized lot.

7. Worked example: how to review published XC-258 data

The table below uses publicly available XC-258 product data as a reading exercise. These figures are published product reference values; they should not be presented as a lot-specific COA result or contractual acceptance limit unless the applicable sales specification and batch certificate state that explicitly.

This first-pass review suggests useful follow-up questions rather than an automatic approval or rejection. R&D should compare the material with the actual ceramic route. Purchasing should translate validated technical needs into measurable limits, methods, and documentation requirements.

XC-258 is positioned for uses including MLCC ceramic formulations, PTC thermistor ceramics, optical glass, microwave dielectric components, and other electronic applications. Suitability for a specific formulation must still be confirmed through laboratory evaluation and pilot-scale validation.

8. A practical COA checklist for purchasing and R&D

Purchasing review

1. Match the shipment. Verify supplier, exact grade, lot number, dates, and quantity.
2. Check every contractual item. A missing result should not be silently replaced by a TDS typical value.
3. Separate results from limits. Confirm which column is the batch result and which is the specification.
4. Standardize units. Convert percent to ppm only after confirming the reporting basis.
5. Review exceptions. Identify results near a limit, ND statements without thresholds, unusual rounding, and unexplained method changes.
6. Control revisions. Retain the approved specification, COA, supplier change notices, and incoming inspection record.

R&D and quality review

1. Link each parameter to risk. Define which impurity, phase, moisture, or particle-size variable affects the product or process.
2. Confirm test comparability. Align sampling, preparation, method, units, and dry/as-received basis.
3. Look beyond D50 and total purity. Request the impurity profile and full particle-size distribution when needed.
4. Correlate with performance. Compare COA and incoming-QC data with slurry, calcination, sintering, electrical, microstructural, and yield results.
5. Evaluate more than one lot. A sample lot can demonstrate feasibility, but multiple commercial-scale lots are needed to assess consistency.

Common COA red flags

• The certificate has no lot number or does not match the package label.
• Every lot shows exactly the same result to every reported decimal place.
• The document lists only typical values and no acceptance limits.
• “Pass” replaces a numerical result for a parameter that requires trend monitoring.
• ND is reported without a method detection or reporting limit.
• Combined impurity values are used when the customer limits individual elements.
• Moisture, LOI, or assay lacks a reporting basis or test condition.
• D50 is listed without a dispersion and measurement method.
• The analytical method changes without notification or correlation data.

A red flag is a reason to investigate, not always proof that the material is unsuitable. The best response is a documented technical clarification, followed by aligned testing where necessary.

Frequently asked questions

What should a high-purity TiO₂ COA include?

It should identify the supplier, exact grade, lot, relevant dates, and authorization, then report the batch-release parameters required by the specification. For electronic ceramics, these often include TiO₂ assay, phase, trace metals, moisture or LOI, particle size, units, limits, and test methods.

Is 99.6% TiO₂ pure enough for electronic ceramics?

The assay alone cannot answer that question. Suitability depends on the specific ceramic system, impurity profile, crystal phase, particle characteristics, process route, and validated performance limits.

What is the difference between a TiO₂ COA and a TDS?

A COA connects reported test results to a specific lot. A TDS describes the product and often lists specifications or typical properties. A typical TDS value should not be treated as a guaranteed batch result unless it is part of the agreed specification.

How do I convert impurity values from percent to ppm?

For mass fraction, multiply percent by 10,000. For example, 0.002% equals 20 ppm. Confirm that both values use the same wet, dry, or as-received basis before comparing them.

Does “not detected” mean the impurity is absent?

No. It means the analyte was below the detection or reporting threshold of the stated method under the test conditions. Request the numerical threshold if the impurity is critical.

Are moisture and LOI the same?

No. Moisture is usually measured by controlled drying at a lower temperature, while LOI uses a defined higher-temperature procedure and can include other volatile or thermally removed matter. Review both test conditions.

Is D50 enough to qualify electronic-grade titanium dioxide?

No. D50 describes the median of a measured distribution, but it does not reveal the fine fraction, coarse tail, span, agglomeration state, or test-preparation conditions. Request D10, D90, distribution information, and the method when particle size is critical.

Why might supplier and incoming-QC results disagree?

Differences can come from sampling, sample heterogeneity, digestion, dispersion, calibration, instruments, reporting basis, detection limits, or rounding. Testing a properly split, homogenized sample with exchanged method details is a practical way to investigate.

Can one good COA qualify a new TiO₂ supplier?

No. One lot can support initial screening, but supplier qualification should also examine multiple lots, method correlation, change control, production-scale trials, quality systems, capacity, and ongoing consistency.