How to read a peptide COA

The answer is at the top. Nine parts, and what each one can tell you.

01

The report itself.

ID, version and page count on every page. Dates in order: received, tested, issued.

report ID version 2 page 3 of 9 received YYYY-MM-DD tested YYYY-MM-DD issued YYYY-MM-DD ≤ ≤

02

The sample, as received.

The lab’s own photo of the vial, as it arrived. Anything the client supplied is marked.

lab photo seen by the lab container cap fill lot printed client-supplied product name label claim lot

03

Identity, by mass.

Similar peptides can leave the column together. Their masses differ: look for the error in ppm.

Identity levels
one peak t MS two masses A B m/z error (ppm) = (observed − expected) ÷ expected × 10⁶

04

Purity is a share.

The main peak’s share of all peaks, at one wavelength. Not water, salts, microbes or endotoxin.

λ 214 nm main peak impurities baseline purity = main peak area ÷ all peak area

05

Content is an amount.

mg of peptide in the vial, against a reference standard, with its uncertainty and its basis.

total powder in the vial content, as free base salt, water, other worked example 8 9 10 11 12 mg per vial label claim 10.3 ± 0.4 mg (k = 2) 103 % of label claim

06

Safety tests.

Each answers its own question and needs its own sample.

Endotoxin (LAL)
Bacterial debris that causes fever. It is there even when no bacteria live.
EU per vial · EU per mg
LOQ · spike recovery
Sterility (USP <71>, 14 days)
Do live microbes grow from the sample?
growth · no growth
units tested · controls
Residual solvents
Solvents left from synthesis or purification.
named solvents
each solvent named

A faster microbial screen or a bioburden count is not a sterility test.

07

A pass needs a rule.

A verdict needs a stated specification and a decision rule. Near a limit, the uncertainty decides.

Our decision rules
limit ± U specification → conforms inconclusive does not conform measured value

08

Sign-off.

Who released the report, and when. Usually at the end.

analyst ran the tests reviewer checked the raw data authoriser released the report

09

Check it at the source.

Open it on the lab’s own website. Compare the numbers, the status and the fingerprint.

Verify a report
hkanalytical.com/verify your PDF SHA-256 f3a6 18d3 fccc 2ec4 4fe8 2ca2 = f3a6 18d3 fccc 2ec4 4fe8 2ca2 in your browser recorded at issue all 64 digits
example data · not a test result

Warning signs.

On any COA, from any lab.

  1. 01 Identity from retention time alone.
  2. 02 Purity with many decimals and no uncertainty.
  3. 03 mg with no basis and no reference standard.
  4. 04 “Pass” with no specification.
  5. 05 A test date before the receipt date.
  6. 06 A sterility result sooner than 14 days after receipt.
  7. 07 No way to check the report on the lab’s own website.
  8. 08 One report for many products or sellers.

We read the public reports of five labs buyers use and compared them on these parts, with our own. How we compare