Sterility, Endotoxin and Purity Are Three Different Claims

What a Purity Figure Actually Measures

A purity result on a certificate of analysis almost always comes from reverse-phase HPLC with UV detection, typically at 214 nm or 220 nm where peptide bonds absorb. The instrument separates the sample into peaks by retention time and reports the main peak's area as a percentage of total peak area. A result of 98.5% means the main peak accounted for 98.5% of everything the detector saw elute and absorb at that wavelength, within that run.

That number says nothing about anything that does not absorb UV light at the chosen wavelength, and nothing about whether the vial contains living organisms. A compound can be chemically homogeneous and microbiologically contaminated at the same time, because homogeneity and viability are unrelated properties measured by unrelated instruments. Purity is a claim about the chemical species present relative to each other. It is not a safety claim and was never designed to be one.

It is also worth noting what wavelength choice does to the picture. A compound with a strong chromophore will show a clean, well-resolved peak at 220 nm even if trace non-absorbing material sits alongside it in the vial; a compound that absorbs weakly at the chosen wavelength can show an artificially tidy chromatogram simply because minor species are underrepresented in the signal. Purity, in other words, is always purity relative to what the detector was set up to see, not purity in some absolute sense.

What a Sterility Test Establishes

A sterility test, run to a method such as USP <71>, either filters a defined sample volume through a 0.45 micron membrane and places the membrane into growth media, or inoculates the sample directly into media such as fluid thioglycollate medium and soybean casein digest medium. The media are incubated for fourteen days and checked periodically for turbidity or visible growth, which would indicate a viable organism was present and able to multiply.

A pass result establishes one specific fact: under the conditions of that test, using that sample volume, no viable microorganism grew out within the incubation period. That is a real and useful result. It is also a narrower claim than the word "sterile" suggests to most readers, and the narrowness is worth understanding before relying on it.

What a Sterility Test Does Not Establish

Sterility testing is destructive and statistical by necessity. A lab cannot test every vial in a batch and still have a batch left to sell, so a sample is drawn and the result is extrapolated to the lot under an assumption of uniform processing. If contamination was introduced intermittently during filling rather than uniformly across the batch, a sample that misses the affected vials will pass even though other vials in the same lot are not clean.

• The test detects organisms capable of growing in that specific media under those specific conditions, not every organism that could theoretically be present.
• It does not detect dead organisms or fragments of organisms, which can still carry biological material even though they will not grow.
• It does not measure endotoxin, which is a separate class of contaminant entirely.
• A negative result below the test's detection limit is reported as a pass, not as proof of zero organisms.

None of this makes the test meaningless. It means a sterility pass answers a specific, bounded question, and a careful reader treats it that way rather than as a blanket guarantee.

Because of these limits, a sterility pass is best read as strong supporting evidence rather than a closed case. A buyer relying on it should also want to know the sample size tested against the batch size, and whether the supplier's process includes in-process bioburden checks earlier in manufacturing, which catch contamination trends that a single endpoint test on a small sample can miss entirely.

Endotoxin Is a Separate Measurement Entirely

Endotoxin is lipopolysaccharide, a structural component of the outer membrane of gram-negative bacteria. It is measured with a Limulus Amebocyte Lysate assay, run as a kinetic chromogenic, kinetic turbidimetric, or gel-clot method, and reported in endotoxin units per milligram, EU/mg. This test has nothing to do with whether an organism is alive. Lipopolysaccharide persists in the material after the bacterium that produced it has been killed by filtration, heat, or any other step in a process line.

That is why a bacterium can be present at some earlier stage of manufacturing, get filtered or otherwise eliminated before the final fill, and still leave measurable endotoxin behind in the finished powder. A batch can pass sterility testing with no viable organisms detected and still carry an elevated endotoxin result, because the two tests are looking for entirely different things using entirely different chemistry. A complete certificate of analysis lists sterility (or bioburden, if sterility itself was not tested) and endotoxin as two separate line items with two separate methods, not one combined claim.

Why Lyophilized Powder Is Not Sterile by Default

Lyophilization removes water from a frozen solution under vacuum. It does not sterilize anything. If the bulk solution loaded into the lyophilizer was not sterile-filtered beforehand, freeze-drying simply converts whatever bioburden was present into a dry cake carrying the same contamination, now stable and easy to overlook because dry powder looks clean on inspection.

Peptides generally cannot tolerate terminal heat sterilization, the kind used for many small-molecule solutions in an autoclave, because heat degrades the amide backbone and side chains. So when sterility is claimed for a lyophilized peptide, it has to come from aseptic processing upstream: sterile filtration of the bulk solution through a validated filter, an aseptic fill into vials under controlled conditions, and a lyophilization cycle that does not reintroduce contamination through a compromised chamber or stopper seal. Any break in that chain, and the finished vial is not sterile no matter how stable and inert the dry powder appears on a shelf.

What Filtration Removes and What It Misses

A 0.2 or 0.22 micron sterilizing-grade filter removes bacteria and most fungi by size exclusion, which is the mechanism behind most sterile filtration claims in peptide manufacturing. It is an effective and validated step for that purpose. It is not a complete barrier against every category of contaminant.

Filtration of this pore size does not remove viruses, which are smaller than the filter's rating. It does not remove endotoxin, because lipopolysaccharide molecules and their aggregates are small enough to pass through a 0.2 micron membrane along with the liquid. And the filter itself has to be integrity tested, typically with a bubble point test before and after use, because a single flaw in the membrane defeats the entire step. Filtration is one control in a chain, not a standalone guarantee, and it has to be paired with validated aseptic handling on both sides of it.

Beyond the filter itself, the environment the fill happens in matters just as much. Aseptic fill under a laminar flow hood or in a classified cleanroom is what keeps a sterile-filtered solution sterile on its way into the vial; a filtered solution handled afterward in an uncontrolled environment can pick up contamination downstream of the very step meant to remove it. The filter is one link in a chain, and the chain is only as reliable as its weakest point.

Reading These Three Claims on a Certificate of Analysis

A certificate built to a real standard shows three distinct entries: a purity result with its method, column, and wavelength; a sterility result with its method and pass or fail outcome, or a bioburden count expressed as colony forming units if sterility itself was not run; and an endotoxin result in EU/mg with the assay type named. Each line should be traceable to a specific test, on a specific lot, run on a specific date.

A document that lists a single number labeled something like "purity and sterility" combined is a paperwork shortcut, not a laboratory result, because no single instrument produces both figures at once. A buyer reading a certificate critically should be able to point to which line answers which question, and should ask a supplier directly when that separation is missing.

Why Two Vials From the Same Batch Can Answer Differently

Because sterility testing samples a fraction of a batch, its result is a statistical statement about the lot, not a guarantee about every individual vial in it. If a filling line had an intermittent lapse, isolated to a subset of vials rather than the whole run, a sample drawn from the unaffected majority will pass cleanly while a handful of vials from the same lot would not, if they happened to be tested instead.

Endotoxin can vary vial to vial for similar reasons, particularly if the water or raw material source carried a transient spike partway through a production run. This is the practical reason serious buyers ask for lot-specific documentation rather than a generic spec sheet that gets reused across every batch a supplier ships. A supplier's track record is a reasonable starting point for trust, but it is a different claim from a current, lot-matched sterility and endotoxin result sitting next to the vial in hand.

None of this is a reason to distrust every certificate. It is a reason to read one the way an analytical chemist would: as a record of specific tests, on a specific sample, with specific limits, rather than as a single word that closes the question.