≥99% vs ≥99.5% Purity: What That 0.5% Gap Actually Means in Formulation

A spec sheet that says "≥99% purity" and one that says "≥99.5% purity" look like they're describing almost the same material — a rounding error apart. In practice, that 0.5 percentage points isn't a single number moving up or down. It's a budget, and different batches — even of the same compound — can spend it very differently: residual solvent, moisture, unreacted precursor, or related synthesis byproducts. Treating purity as one clean quality score, rather than reading what actually fills the remainder, is where a lot of formulation surprises start.

What Actually Fills the Missing 0.5% (or 1%, or 2%)

An assay result of, say, 99.6% on a certificate of analysis (COA) means that 99.6% of the sample, by the analytical method used — almost always HPLC for the actives in this category — is confirmed to be the target molecule. The remainder is not automatically "0.5% of something bad." On a properly built spec table, it is broken out into several distinct line items, each with its own limit:

• Loss on drying / moisture content — residual water or solvent trapped in the crystal lattice or on the powder surface, reported separately from assay.
• Residual process solvent — trace amounts of the extraction, reaction, or crystallization solvent that didn't fully clear during drying.
• Related substances / synthesis byproducts — structurally similar molecules carried over from the reaction or bioconversion pathway: unreacted precursor, reaction intermediates, or over-reduction/over-oxidation products.
• Residue on ignition / sulfated ash — inorganic residue from catalysts, buffers, or salts used during processing.

Two products both labeled "≥98.0% purity" can carry that 2% very differently — one mostly as moisture, another mostly as a single related substance. A single assay percentage doesn't distinguish between the two, but your formulation might care a great deal which one it is.

Why the Same Gap Matters Differently by Application

How much this 0.5-2% actually matters depends on what you're making. In parenteral or other highly sensitive pharmaceutical applications, regulatory pharmacopoeias set tight, individually-named limits on specific related substances and residual solvents, because material entering the body by injection bypasses the metabolic buffering that oral or topical exposure gets — so the composition of that remaining fraction is scrutinized closely, not just its total size. In a food-grade powder blend or a topical formulation, by contrast, a broader loss-on-drying or ash allowance is normal, expected, and compliant. This is a general pattern across regulated industries, not a claim about the regulatory status of any specific product — but it explains why the "same" 0.5% gap can be a non-issue in one application and a disqualifying one in another.

Purity Percentage and Contaminant Safety Are Different Axes Entirely

It's worth separating two things that a headline purity number can make look related but aren't: assay purity (how much target molecule is present) and contaminant safety (heavy metals, residual solvents, microbial limits). A 99.5%-assay powder and a 97%-assay powder can both pass or both fail heavy-metal and microbial testing independently of their purity tier — purity tells you about the chemical composition of the non-target fraction in only a general sense, while heavy metals (lead, mercury, cadmium, arsenic) and microbial limits (total plate count, yeast/mold, E. coli, Salmonella) are measured and controlled as their own, separate specifications. A higher purity number is not a proxy for a cleaner contaminant profile — you have to check both.

Three Products, Three Purity Tiers: What the Full Spec Table Shows

Comparing three real actives at different assay tiers makes the point concrete:

• β-Nicotinamide Mononucleotide (NMN, CAS 1094-61-7) — assay spec ≥99.5% (a recent batch tested at 99.6%), loss on drying ≤1.0% (0.4% actual), residue on ignition ≤0.5% (0.2% actual), lead ≤2 ppm (<1 ppm actual). This is also the one of the three referenced against USP/EP monograph expectations.
• Nicotinamide Riboside Hydrogen Malate (NRHM, CAS 2415659-01-5) — assay spec ≥98.0% (98.6% actual), with a spec line most buyers never think to ask for: "single maximum related substance ≤0.5%" (0.2% actual). That single line matters more than the headline number — it caps how much any one impurity is allowed to dominate the non-assay fraction, rather than just bounding the total.
• Berberine Hydrochloride (CAS 633-65-8) — assay spec ≥97.0% (97.6% actual), loss on drying allowance ≤5.0% (2.8% actual), a materially looser moisture tolerance than NMN's ≤1.0%, consistent with this botanically-extracted alkaloid salt's typical handling, and referenced against the Chinese Pharmacopoeia (CP).

Ranked purely by headline purity, these would read 99.5% > 98.0% > 97.0%, in that order, and nothing more. Reading the full spec tables shows three materials with genuinely different impurity structures, different pharmacopeial references (or none at all), and different moisture tolerances — none of which the single percentage communicates on its own.

Read the Full Spec Table, Not Just the Headline Number

The practical takeaway is not "always buy the highest purity number." It's that the purity percentage is the entry point into a spec table, not a substitute for reading it. Before comparing two quotes on purity alone, pull the full COA and check: what specific categories make up the non-assay fraction, whether any single related substance is separately capped, what the moisture/LOD allowance is, and whether heavy metal and microbial limits are documented independently. A 99.5%-purity material with an undocumented related-substance profile can be a worse fit for your process than a well-characterized 98.0% material — the number alone won't tell you which.

What to Ask a Supplier For

• The full quality spec table, not just the assay line — including loss on drying, residue on ignition, and any related-substance limits.
• Which pharmacopoeia (if any) the specification is referenced against, and whether that's stated on the COA itself, not just in marketing copy.
• Heavy metal and microbial results as their own section, independent of the purity figure.
• Batch-to-batch consistency — ask to see two different COAs for the same product to confirm the numbers aren't cherry-picked.