How to Compare API Purity Levels: Assay, Impurities and Grade Explained for Buyers

Introduction: Two Suppliers, Two "Purity" Figures, One Problem

A formulator receives two quotations for the same mineral API. Supplier A says its material is "98.0% pure." Supplier B says "99.0% pure." The obvious conclusion is that Supplier B is better. That conclusion may be wrong.

The two numbers may have been measured by different methods, calculated on different bases, and reported against different pharmacopoeial editions. One may include moisture in the result and the other may not. One figure may be an assay and the other a calculated "total purity." Without that context, the two figures cannot be compared at all.

"Purity" is the most common word in API sales and one of the least precisely used. In regulatory and quality work it is not one number. It is a set of results: an assay, an impurity profile, limits for specific contaminants, and a pharmacopoeial standard that defines how each is measured and reported.

This guide explains how to compare API purity levels properly: what assay and impurities each tell you, why the calculation basis matters, how pharmacopoeial grade fits in, how to read the purity lines of a Certificate of Analysis, and how to avoid the most common comparison mistakes. It draws on ICH guidelines, FDA and EMA publications and the Indian Pharmacopoeia Commission, with every source listed at the end.

1. What "Purity" Actually Means: Assay, Impurities and Specifications

When a buyer asks "how pure is it?", three different questions are hiding inside one.

Assay: How Much of the Right Substance Is There?

An assay measures the amount of the named substance in the material, using a defined analytical method. It answers the question: how much of the compound itself is present? A monograph normally states an acceptable range, not a single value.

For example, the USP-NF monograph for Ferrous Fumarate (listed under the Dietary Supplement monographs) states that the material "contains NLT 97.0% and NMT 101.0% of ferrous fumarate (C4H2FeO4), calculated on the dried basis." Two features matter here. First, the standard is a range with a lower and an upper limit. Second, the result is "calculated on the dried basis," which means moisture is removed from the calculation. Both are explained below.

Impurities: What Else Is in There?

Impurities are everything that is not the named substance. ICH Q3A(R2), Impurities in New Drug Substances, groups them into three categories:

• Organic impurities: starting materials, by-products, intermediates, degradation products, and reagents, ligands and catalysts.
• Inorganic impurities: reagents, ligands and catalysts, heavy or other residual metals, inorganic salts, and other materials such as filter aids and charcoal.
• Residual solvents: liquids used as vehicles during synthesis. Because their toxicity is generally known, they are controlled under ICH Q3C.

For mineral and metal-salt APIs, inorganic and elemental impurities are usually the most relevant category.

Specification: Who Decides What Is Acceptable?

A specification is the formal list of tests, methods and acceptance limits that a material must meet. ICH Q6A, Test procedures and acceptance criteria for new drug substances and new drug products: Chemical substances (ICH Step 5, effective 1 May 2000), is the reference framework for setting and justifying specifications. In practice, a drug-substance specification typically covers description, identification, assay and impurities, plus any specific tests the substance needs.

Why the Single Number Misleads

A single "purity" percentage compresses all three questions into one figure. It cannot tell you which impurities are present, how they were measured, or whether the number includes moisture. For that reason, regulators and quality teams compare materials on the full set of specification results, not on one headline percentage.

2. The Tests That Make Up a Purity Comparison

When comparing two suppliers, line up the same tests side by side. The list below covers the results that matter for most mineral and nutraceutical-grade APIs.

The sections below take each test in turn.

1. Assay and Its Limits

Compare the stated range (lower and upper limits), the method used and the basis of calculation. An assay of 99.0% against a lower limit of 97.0% is a different margin of safety than 99.0% against a lower limit of 98.5%.

2. Water Content or Loss on Drying

Moisture changes the weight of material you actually receive. A product with higher moisture contains less of the active substance per kilogram. Loss on drying or water content is therefore part of any fair comparison, and it explains why monographs state the assay "on the dried basis."

3. Organic and Inorganic Impurities

Ask which impurities are tested for, which limits apply, and how unspecified impurities are controlled. ICH Q3A describes a threshold concept: a reporting threshold above which an impurity should be reported, an identification threshold above which it should be identified, and a qualification threshold above which its biological safety should be established. These thresholds were written for new drug substances, so for established mineral ingredients buyers more often rely on the limits in the pharmacopoeial monograph and on their own risk assessment.

4. Elemental Impurities

ICH Q3D, Guideline for Elemental Impurities, is a risk-based framework for controlling elemental impurities. It is now in Revision 2 (Step 5), with a legal effective date of 24 September 2022 in the EU (EMA). EMA notes that the implementation strategy is intended for applicants, drug product, drug substance and excipient manufacturers, as well as regulators. For a formulator, the practical need is data from the supplier that supports the formulator's own elemental-impurity risk assessment.

5. Residual Solvents

ICH Q3C(R9), Impurities: Guideline for Residual Solvents, has been legally effective since 29 April 2024 (EMA). It gives toxicologically acceptable levels for residual solvents and encourages the use of less toxic solvents. Residual solvents matter most for APIs made through organic synthesis. For simple mineral salts, the supplier should still state whether solvents are used in the process and, if so, which.

6. Monograph-Specific Limit Tests

Many monographs include named limit tests, such as insoluble matter, specific ions or heavy metals, that apply to that particular substance. These tests are what separate a pharmacopoeial-grade material from one sold only against an in-house limit. Ask for the full list, not a summary.

7. Microbiological and Physical Tests Where the End Use Needs Them

Nutraceutical and food-fortification uses may need microbiological limits. Physical properties such as particle size and bulk density affect processing rather than purity, but they belong in the same comparison because they affect how a material performs once it is in your process.

3. Compendial Grade: USP, BP, EP and IP

"Pharmaceutical grade" is a phrase, not a standard. The real standard is the pharmacopoeia and the specific monograph a material is tested against.

What a Pharmacopoeial Monograph Gives You

A monograph defines the identity tests, the assay method and limits, and the impurity and limit tests for one substance. When two suppliers both state that they test to the same monograph and edition, their results become directly comparable. When one tests to a monograph and the other to an in-house specification, they are not.

Why Compendial Standards Carry Regulatory Weight

In the United States, FDA's Compliance Policy Guide Sec. 420.100 states that a drug recognized in an official compendium that fails to conform to compendial standards for quality, strength or purity is adulterated, "unless the differences from such standards are plainly stated on the drug's label" (issued 10/1/80, revised 5/1/92). This is a finished-drug rule, but it shows why compendial compliance is treated seriously across the supply chain.

The Pharmacopoeias a Buyer Will Meet

The USP-NF is the compendium most often named for the US market, the European Pharmacopoeia and British Pharmacopoeia for Europe and the UK, and the Indian Pharmacopoeia (IP) for India. A buyer should state in the enquiry which pharmacopoeia and which edition applies to the destination market.

The Indian Pharmacopoeia Is Updated Regularly

The Indian Pharmacopoeia Commission (IPC), established by the Ministry of Health and Family Welfare in 2005, publishes the IP. Its ninth edition was published in 2022 and was effective from 1 December 2022. In 2026, the Union Health Minister released the tenth edition, IP 2026, which incorporated 121 new monographs and brought the total to 3,340 monographs (IPC). The IPC pages reviewed did not state the effective date of IP 2026, so buyers should confirm it with the IPC before specifying an edition.

Monographs Change

Because pharmacopoeias are revised, a result reported against an old edition may not meet a newer one. Always record the edition on the CoA and in the purchase specification.

Quick Reference: The Guidelines Behind a Purity Specification

4. Comparing Like With Like: Basis, Method and Edition

Four differences cause most false comparisons.

1. Dried Basis vs "As Is"

A result "on the dried basis" removes moisture from the calculation. A result "as is" includes it. The two can differ noticeably for the same material.

Illustrative example (hypothetical figures):

• Supplier A reports an assay of 98.0% as is, with 2.0% loss on drying.
• Supplier B reports an assay of 99.0% on the dried basis, with 2.0% loss on drying.

At first glance B looks better. Convert both to the same basis:

• Supplier A on the dried basis: 98.0 ÷ (1 − 0.02) = 100.0%
• Supplier B as is: 99.0 × (1 − 0.02) = 97.0%

On a like-for-like basis, A's material contains more of the active substance. The headline figures pointed the wrong way.

2. Different Analytical Methods

Two methods can give slightly different results for the same sample. ICH Q2(R2), Validation of Analytical Procedures (Step 5, Revision 2, legal effective date 14 June 2024 per EMA), covers how methods for assay, purity, impurities and identity are validated, including accuracy, precision, specificity, detection limit, quantitation limit, linearity and range. When two CoAs disagree, check whether the methods are the same and whether they are validated.

3. Different Limits Behind the Same Result

A result of 98.5% means something different against a limit of "not less than 98.0%" than against "not less than 95.0%." Always compare the result and the limit together.

4. Different Editions or Standards

A result against IP 2022 and a result against USP-NF are different statements. A result against an in-house specification is a third. Convert the comparison to one named standard before judging.

A Five-Step Procedure for a Fair Comparison

1. Request the full specification and a recent CoA from each supplier. Do not compare brochures.
2. Name one standard. Pick the pharmacopoeia and edition that applies to your product and market, and ask each supplier to report against it.
3. Convert to one basis. Restate each assay on the dried basis, using the reported loss on drying, so every figure means the same thing.
4. Compare limits as well as results. Note the lower and upper limits and how much margin each batch has.
5. Compare impurity data line by line. Put the impurity results, limits and test methods in one table.

If a supplier cannot provide enough data to complete these steps, that is itself a finding.

5. How to Read the Purity Lines of a Certificate of Analysis

A CoA is the main place where purity is reported, and it is worth reading line by line. (A general guide to reading a CoA is in our blog on choosing between generic and branded API suppliers.) For purity specifically, check these points.

• Is there a stated limit beside every result? A result with no limit cannot be judged.
• Is the basis stated? Look for "dried basis," "anhydrous basis" or "as is."
• Is the method named? The pharmacopoeial method or the in-house method should be identified.
• Is the standard and edition stated? For example, "IP 2022" or "USP-NF."
• Are impurities listed individually or only as "complies"? A bare "complies" is weaker evidence than a numerical result.
• Do the results move from batch to batch? Identical results on every batch can mean the report was copied rather than measured. Natural variation within limits is normal.

Red Flags

• A single "purity: 99%+" with no method, basis or limit.
• No impurity data at all.
• A CoA with no batch number or date.
• Results that sit exactly on a limit on every batch.
• A refusal to share the specification.

None of these proves a problem, but each one is a reason to ask for more information before approving a supplier.

6. How to Verify What a Supplier Tells You

A CoA is a supplier's statement. Buyers who can verify it are in a stronger position.

1. Test incoming batches. Run the identity and assay tests that matter most for your process on a sample of each incoming lot, at least until the supplier has a track record.
2. Use an independent laboratory. For a new supplier, send one or more batches to an accredited third-party laboratory and compare the result with the CoA.
3. Use reference standards. Pharmacopoeial reference standards allow a laboratory to run the monograph method properly.
4. Review validated methods. Ask the supplier for method-validation summaries for the key tests (see ICH Q2(R2) above).
5. Trend results over time. Keep a record of several batches. A supplier whose results are stable and within limits is more reliable than one with a single good batch.
6. Keep retained samples. A retained sample from each batch makes it possible to investigate any later complaint.
7. Audit where possible. A site or remote audit under a GMP framework such as ICH Q7 (FDA's published version is dated August 2001) is the strongest check of a supplier's quality system.

When Your Result and the CoA Disagree

A small difference between a buyer's result and a supplier's CoA is common, because laboratories, instruments and sample handling differ. A larger difference needs investigation. A sensible sequence is to check that both laboratories used the same method and basis, repeat the test on the retained sample, share the raw data with the supplier, and, if the gap remains, send the sample to an independent laboratory. Agree in advance, in the supply agreement, how such disputes will be resolved. A supplier that responds to a discrepancy with data, not defensiveness, is showing you how it will behave over the long term.

7. Higher Purity Is Not Always Better: Fit for Purpose

It is tempting to treat the highest assay as the best choice. In practice, the right purity is the one that meets the specification for the intended use at a sensible cost.

Match the Grade to the End Use

The same mineral salt may be bought for a pharmaceutical formulation, a dietary supplement, food fortification or an animal-nutrition premix. Each has its own standard and its own set of required tests. Buying a tighter specification than the application needs adds cost without adding value. Buying a looser one creates a compliance risk.

This table is general guidance. Requirements vary by market, so confirm them with your regulatory team. For the certifications that matter by end use, see Certifications a Nutraceutical Exporter Should Have.

Consistency Often Matters More Than the Peak Number

A material that always delivers 99.0% inside its limits is easier to formulate with than one that swings between 97.5% and 100.5%. Process teams value predictability, so batch-to-batch consistency is a quality attribute in its own right.

Count the Whole Cost

A small price difference per kilogram can disappear if a batch fails incoming testing or needs rework. Compare suppliers on total cost, including testing, documentation review and the risk of rejection.

Write Your Own Purchase Specification

The strongest protection for a buyer is a written purchase specification that both sides sign. It should name the pharmacopoeia and edition, the assay range and basis, the loss-on-drying or water limit, the impurity and limit tests that matter for your product, the documents that must accompany each batch, and how changes will be notified. A purchase specification turns a vague promise of "high purity" into something that can be checked batch by batch. For the full list of documents to request, see our guides on documents an API exporter should provide and on certifications a nutraceutical exporter should hold.

8. A Side-by-Side Comparison Framework

Use the table below to compare two or more suppliers on the same terms.

9. Shreeji Industries' Approach

A fair comparison starts with the right information. Buyers evaluating Shreeji's materials can request the batch-specific CoA and, for Dried Ferrous Sulphate, the Technical Data Sheet, and then compare them against any other supplier's documents using the framework in this guide. Shreeji's minimum order quantity for Dried Ferrous Sulphate is 500 kg.

Where to Go Next

10. Common Misconceptions

Misconception 1: "99% Purity Means 1% Impurities"

Not necessarily. The balance of an assay can include moisture and other components, and the assay itself has analytical variability. Impurities are measured separately.

Misconception 2: "A Higher Percentage Always Means a Better Product"

Only if the figures were measured by the same method, on the same basis and against the same standard. Otherwise the numbers are not comparable.

Misconception 3: "Pharma Grade Is a Defined Standard"

It is a marketing phrase. The defined standard is the pharmacopoeial monograph the material is tested against.

Misconception 4: "A Certificate of Analysis Guarantees Quality"

A CoA is a supplier's report. It is strong evidence when it is complete, consistent and verifiable, and weak evidence when it is not.

Misconception 5: "An Assay Over 100% Means an Error"

Many monographs allow an upper limit slightly above 100 percent, as the USP-NF Ferrous Fumarate example above shows (NMT 101.0%), because analytical methods have inherent variability.

11. Frequently Asked Questions

12. Sources and References

Data sourced from regulatory authority websites, government publications and, where noted, industry associations.