
In brief: Passivation documentation is the written trail showing a surface was cleaned and treated, by whom, on what date and with what result. US GMP wants those entries dated, signed and kept in chronological order, which is why a record written up afterwards is a problem even when the job was done properly. Records, or copies of them, also have to be producible at the site where the work happened, and that is where outsourced passivation most often comes unstuck.
A passivation finding is often not about the chemistry at all. The work was done, the surface was fine, and the batch was good. The record is what failed.
Passivation documentation is the written trail behind the job. It shows that a surface was cleaned and treated, by whom, on what date, and with what result. In a regulated plant that trail is not an extra. An inspector cannot see a passive film by looking at it, so the record is what gets examined instead.
This is the gap that catches good plants. The work is real and the paperwork is thin.
Anything written down that shows the treatment happened and worked. For most plants that means three things: the procedure you followed, the log entry for the job, and the test result that closed it out.
US GMP names the log explicitly. Records of major equipment cleaning and maintenance belong in equipment logs, though routine work such as lubrication and adjustment is left out. Those logs must "show the date, time, product, and lot number of each batch processed" (21 CFR 211.182). Where a vessel is dedicated to one product, the same clause puts those records in the batch record instead.
The rule also covers signatures. The people who did the work and double-checked it "shall date and sign or initial the log indicating that the work was performed".
The procedure behind the log matters just as much. It should name the chemical, the concentration, the contact time and the rinse, so that two operators working from it would do the same thing. Our guide to choosing a passivation chemical covers how that choice gets made in the first place.
Notice what those clauses do not ask for. None of them asks for a certificate, a letterhead or a bound report. A purchase order can ask for one, and some standards do.
Because of when they were written rather than what they say. Which rule bites depends on where the work sits in your process.
Passivation run as a step inside a batch is production and process control. Those procedures "shall be documented at the time of performance" (21 CFR 211.100(b)). Passivation done as equipment maintenance between campaigns lands in the equipment log instead, where entries "shall be in chronological order" (21 CFR 211.182).
Take a job done on Tuesday and written up on Friday. It may be accurate. It may describe the job perfectly. If the passivation was a batch step, it has still missed documentation at the time of performance. If it was a maintenance job, that Tuesday date has to go somewhere. Appended after Friday's entries it breaks chronological order, and a late entry invites a look at the ones around it.
Gaps matter in the other direction too. A log with no gaps at all, across a plant that shut down for two weeks in the monsoon, invites the same scrutiny. Real records have holidays in them.
This is why a tidy record can be worse than a rough one. Neat handwriting in a single ink across three weeks of entries suggests one sitting rather than twenty. Failures that begin on the metal rather than on paper are a separate subject, and we cover those in common passivation failures.
Messy, and written as the work happens. The mess is the point. An entry is worth something precisely because nobody knew how the job would end when the first line was written.
Four habits make it obvious that a record is real:
None of those habits costs money. What they cost is the discipline of carrying the log to the job instead of rebuilding it afterwards.
Corrections deserve their own note. A crossed-out line with an initial beside it reads as an honest record of a real morning. A line scribbled over, or covered with correction fluid, reads as something being hidden, whether or not anything was.
The equipment cleaning log, and then whatever it connects to. The cleaning procedure behind it usually traces to ASTM A380. Batch production records are a separate document. They name the people "performing and directly supervising or checking" each step that matters (21 CFR 211.188). The log is where a passivation job usually lands. The batch record is where the batch it touched is written up.
| Record | What it covers | What it has to carry |
|---|---|---|
| Equipment cleaning log (211.182) | One vessel, job by job | Date, time, product and lot number, signed by whoever did and double-checked the work, or just the verifier where the cleaning was automated |
| Batch production record (211.188) | One batch of product | Who performed and who supervised or checked each step that matters |
| Test result | One check on the surface | The method by name, and what was seen |
| Vendor certificate | The supplier's own statement | Nothing these rules ask for |
So the questions arrive in a settled order. What was done, when, to which equipment, tied to which batch, and who signed. A gap in any one of those turns a routine look into a deeper one.
Each answer opens the next, which is why the order holds. Without a date you cannot tie the job to a batch. Without a signature you cannot say who stands behind the date.
This lands hardest on plants that export. India has around 3,000 drug companies and roughly 10,500 manufacturing units (IBEF). The ones supplying regulated markets answer to inspectors who look at records like these for a living. What to record for each individual test is set out in passivation testing methods.
You do, and this is where outsourced passivation goes wrong. US GMP says records, "or copies of such records", must be "readily available for authorized inspection" (21 CFR 211.180). They must be available "at the establishment where the activities described in such records occurred".
Copies count, and so does anything your team can pull up by computer. What does not count is detail you never received.
Read that against a common arrangement. A vendor treats your reactor, keeps the batch numbers and bath readings in its own file, and sends you a one-page summary. You cannot produce what you were never given.
Fix this in the purchase order rather than after the audit. Name the pack you expect, and make the final payment depend on receiving it. That pack is the procedure, the log entry, the bath readings, the test result and the signatures.
Ask for it before the crew leaves site. Chasing a contractor for April paperwork in October is its own small project. That is why we hand the passivation paperwork across on the day.
Longer than the job itself suggests. Where a record ties to a batch, US GMP holds it "for at least 1 year after the expiration date of the batch". A product with a three-year shelf life therefore carries its records four years from manufacture. Certain OTC products exempt from expiry dating run instead to three years after the batch was distributed.
That clock has a practical effect on storage. A log book that lives in a maintenance cupboard for four years needs to survive a monsoon, a shift handover and an office move. The rule accepts true copies such as photocopies and microfilm, so scanning the book is reasonable insurance.
Where you keep that copy matters as much as having it. A shared drive your QA team can open during an inspection meets the test. A vendor portal you have to email for does not.
There is a second reason to keep records close. The rule says written records must be maintained so the data can be used for evaluating product quality standards "at least annually". Passivation history is part of what that review can draw on, and records you do not hold cannot be drawn on at all.
Pick one piece of equipment and follow its paper trail end to end. That single exercise finds more than a checklist will, because it tests whether the pieces join up.
Start at the vessel and work outwards. Find the procedure, then the log entry, then the test result, then the signature on each. Note every point where you had to ask a colleague for something. An auditor would have asked you the same question, and waited for the answer.
The following scenario is fictional. It is an illustrative example, not a specific project or client.
A QA officer at a plant near Ahmedabad pulls the file for one reactor. The site procedure is current, and the vessel is in service without complaint. But the April passivation was done by a contractor, and all the file holds is an invoice with a line item.
No log entry, no bath readings, no test result, nobody's signature. The work may well have been fine. Everything that would show it sits in someone else's filing cabinet, 400 kilometres away. None of that can be fixed on the morning of an inspection.
Where the trail breaks, the practical move is to fix the present rather than argue about last year. An inspection establishes where the surface stands today. The record then starts again from a known point.
The thing to remember is simple. An auditor cannot see your passive film, so the record is the work as far as they are concerned.
On Monday, open the cleaning log for your most critical vessel. Check three things: that entries run in date order, that each is signed, and that the last outsourced job left its pack with you.
If any of those three is missing, that is your week's job. Where the paperwork needs sorting out before a visit rather than during one, our team can help.
On its own, no. A certificate says a job was done; it does not show the date, the equipment, the batch or who signed. Keep it, but keep the procedure, the log entry, the readings and the test result with it. Those are what an inspector reads.
Better not to. Where the passivation is a batch step, GMP wants it documented at the time of performance. Where it is equipment maintenance, the log has to run in chronological order. A week written in one sitting tends to be obvious either way. Filling the log at the equipment takes less time than defending a run of late entries.
Do not reconstruct it. A record created now with an old date is a much bigger problem than a missing one. Note the gap honestly, then re-establish where the surface stands today and document from there. For what to look for on the surface itself, see identifying rouge contamination.
Not directly. They apply when you supply regulated markets or when a customer requires them. The purchase terms we see from Indian buyers increasingly ask for the same evidence, so the habit is worth building either way. The test side of that evidence is covered in our ASTM A967 guide.
Cleaning validation is a study. It proves your process removes what it should, under defined conditions, across enough runs to be convincing. Passivation documentation is different: it shows that a specific treatment happened on a specific day to a specific vessel. One does not substitute for the other.
Sources. US FDA, 21 CFR 211.182, Equipment cleaning and use log, for the required log fields, the signature requirement and the chronological order clause · 21 CFR 211.180, General requirements, for the retention period and for records being readily available at the establishment where the work occurred · 21 CFR 211.188, Batch production and control records · IBEF, Indian Pharmaceutical Industry. Every clause quoted here was pulled from the eCFR on the day of writing. India's own Schedule M was reviewed and deliberately not quoted, because the copy available to us predates the 2023 revision and cannot be relied on as current.