What we destroyed. And why.
A vendor that never fails a batch is either testing nothing or hiding what they find. We expect to fail roughly 3% of batches over time. Here is every one we have failed, the reason, and the action we took.
Failure is part of QC. Hiding it is the failure.
Most research peptide vendors don't talk about their failures. We think that's the deepest signal in the category — every supplier has failures; the difference is whether they're visible or buried. We made our log public because the alternative is opting into the same trust deficit that defines the rest of the market.
This page lists every batch we have started and not shipped: the SKU, the batch ID, the date the batch was destroyed, the failure mode, and the corrective action (if any). Every entry is permanent. Nothing leaves the log.
Destroyed batches, since launch.
Each row shows what failed, when, why, and what changed afterward. Stamp-Red rows are recent failures (last 90 days).
| Batch ID | SKU | Destroyed | Failure mode | Corrective action |
|---|---|---|---|---|
| B-2604-001 | BPC-157 | 2026-04-08 | Purity ≥97.2% — below acceptance threshold | Synthesis run repeated; B-2604-007 passed at ≥98.0% |
| B-2603-009 | TB-500 | 2026-03-12 | Mass spec parent-ion deviation outside tolerance | Investigated upstream; identity confirmed in repeat run B-2603-021 |
| B-2602-016 | GHK-Cu | 2026-02-22 | Visual inspection — color irregularity (uneven copper coordination) | Reconstitution protocol updated; B-2603-011 unchanged result |
| B-2602-002 | Semax | 2026-02-04 | HPLC trailing peak at 6.4 min · purity below threshold | Cleaved-byproduct identified; column wash protocol updated for next run |
| B-2601-019 | MOTS-c | 2026-01-28 | Lyophilization failure — product caking | Freeze-dryer cycle re-validated before next run |
| B-2601-005 | BPC-157 | 2026-01-11 | Mass spec parent-ion match within tolerance, but secondary peak above noise floor | Conservative call: destroyed and re-run rather than shipped at 97.6% |
The four classes of failure mode we log.
Every failure mode falls into one of four categories. We name them on the page so it's easy to track which classes we hit most often — and so you can see what we'd describe as a "false-failure" (conservatively destroyed) versus a structural problem with synthesis.
1. Purity below threshold
HPLC purity below the published acceptance threshold (≥98.0% for the catalog). The single most common failure class. A repeated synthesis run almost always lands above threshold.
2. Identity confirmation failure
Mass spectrometry parent-ion mass outside tolerance — meaning the molecule that came out of synthesis is not the molecule we ordered. Rare but the most serious class. We fully investigate the synthesis SOP before another run.
3. Stability / form failure
Lyophilization caking, color irregularity in copper-coordinated peptides, or any visible defect of form that doesn't show up on chromatography but compromises the product the buyer would receive. Fixed at the lyo or packaging step.
4. Conservative destruction
The batch passed quantitative thresholds but had a marginal result we chose not to ship. A trailing peak just above the noise floor; a single retention-time anomaly the lab couldn't fully explain. These are batches we could have shipped to most vendors' standards. We destroyed them anyway.
What we don't track here, and why.
The transparency log records batch failures — events where a synthesized batch did not become a published COA. It does not record:
- Returns or replacements after delivery (separate process — handled per order; available on request).
- Customer complaints or third-party retest disputes (separate process — see FAQ for the workflow).
- Synthesis partner internal failures (those are theirs to log; we only see the batches that arrive at our QC step).
If you ever third-party retest a Ledger Bioscience batch and find a material disagreement with our published COA, email research@ledgerbioscience.com with the report. We will refund the order and investigate publicly. The disputed batch joins this log if confirmed.