Why ≥99% and not 98%: a practical comparison

The choice of ≥99% as a working purity threshold can read as arbitrary — why not 98%, why not 99.5% — until you compare what the chromatograms actually look like at each level. This article does the side-by-side: the math of impurity load, the synthesis byproducts that appear at 2% versus 1%, and the research-application sensitivity that drives where the threshold gets drawn. The mechanics of how purity is measured and why ≥99% emerged as the dominant standard live in the cornerstone on what ≥99% purity actually means.

How does impurity load scale across the purity range?

Purity is reported as the target peak's percentage of the total integrated chromatogram area. The impurity load is the complement — 100% minus the reported purity. Small changes in the headline percentage are large changes in the impurity load, because the math is multiplicative across the small numerator.

The step from 99% to 98% looks small in headline percentage but doubles the absolute impurity load. The step from 99.8% to 98% is a 10× change. Whether 10× of a small number is large enough to matter depends entirely on the research context — which is what makes the question of "how pure is pure enough" application-dependent rather than purely analytical.

Which synthesis byproducts appear consistently below 99%?

A non-target peak on an HPLC chromatogram is not just "an impurity" — it usually has a specific synthesis origin, and the byproducts that appear at 2% impurity load are not the same ones that appear at 0.2%. The deeper coverage of every byproduct category is in common HPLC artifacts and what impurity peaks mean; the focused list below is what reliably shows up when batches drop below the 99% threshold.

• Truncated sequences — peptides missing one or more residues from a failed coupling step. At 99.8% purity these are typically <0.1% area each; at 98% they aggregate to several tenths of a percent.
• Oxidation products — methionine, tryptophan, or cysteine residues oxidized during synthesis. At 99% purity, oxidation products are usually contained; below that, they accumulate as resolvable peaks.
• Deletion sequences — peptides missing a single residue from the middle of the chain. Often co-elute close to the target; their accumulation is one driver of the 99→98 step-change.
• Residual protecting groups — fragments from the protecting groups used during solid-phase synthesis that didn't fully cleave. Distinctive retention times; appear consistently below 99%.
• Counter-ion artifacts — peaks from the trifluoroacetic acid (TFA) used in synthesis or HPLC purification. Usually identifiable; their visibility on the chromatogram is part of what distinguishes batches that have been carefully purified.

A practiced eye reading a chromatogram can often estimate the synthesis quality from the impurity peak pattern alone — even before reading the integrated purity number. The visual skill is covered in reading an HPLC chromatogram.

Where does research-application sensitivity sit on the threshold?

The same impurity load is consequential or inconsequential depending on what the material is being used for. An immunogenicity study is more sensitive to impurity content than a structural binding study; a primary-cell assay is more sensitive than an isolated-enzyme reaction; a reference-standard preparation has tighter requirements than a routine in-vitro screen.

• Immunogenicity / immune-cell assays. Trace impurities can produce off-target signaling responses that confound the experiment. The case for ≥99% (or higher) is strongest here, in the same envelope where endotoxin sensitivity tightens.
• Reference-standard preparation. When the peptide will be used as a reference for downstream quantitation, every percent of impurity affects the calibration accuracy. Higher purity is non-optional.
• Structural / binding studies. Most isolated-protein binding assays are not very impurity-sensitive — a 98% pure peptide may produce indistinguishable results from a 99.8% pure preparation. The threshold here is application-driven rather than dictated.
• Method development and screening. Routine in-vitro screening on isolated systems can often tolerate higher impurity loads, particularly when the impurity profile is well-characterized and consistent across batches.

The practical consequence is that vendor purity standards function as a default for the most sensitive applications. A supplier holding ≥99% is producing material suitable across the breadth of typical research applications without forcing the buyer to evaluate sensitivity case-by-case. Below 95%, the impurity profile begins to dominate; reputable suppliers don't ship there.

Why is detection-limit math the upper bound, not 100%?

A purity number above 99.95% sounds better than 99.5%, but in practice it usually isn't. Modern HPLC methods reliably resolve peaks down to roughly 0.1% area. Below that threshold, the integration software can't reliably distinguish a real low-level impurity from baseline noise or a tail of the target peak. A reported purity of 99.95% implies the lab is integrating peaks below 0.05%, which is below most methods' validated sensitivity range.

Past 99.5%, the chromatogram's shape and the methodology validity matter more than the marginal percentage. A 99.4% reading paired with a clean chromatogram and a documented method per USP <621> carries more interpretive weight than a 99.95% reading with no chromatogram and unspecified integration parameters. The number alone is undertwined from the substance without the methodology.

Sources and further reading

• USP <621> Chromatography — system suitability requirements and reporting standards for HPLC purity assays.
• USP <1086> Impurities in Drug Substances and Drug Products — guidance on classifying and reporting impurity profiles.
• FDA Guidance for Industry: Q3A(R2) Impurities in New Drug Substances — the ICH-aligned framework for impurity identification thresholds.