The room, the procedures, and numbers you can check
This is a formulation and research facility. It exists so that formulations can be developed, tested and validated in an environment that controls variables rather than adding them. Producing finished vials is one output of that work, not its purpose.
Everything below is published as a figure rather than an adjective. “Clean room” means nothing; 126 air changes an hour through H-14 filtration means something, and you can hold us to it.
Every ISO figure here is a design and engineering classification — derived from the build specification, the filter grade, the measured airflow and the occupancy. That is true of the ISO 7 room, the ISO 5 work zones and the estimated ISO 6, equally. None of them is a certified ISO 14644-1 classification, because that is assigned from counted particles at defined sample locations and we have not published counts. We label them all the same way rather than presenting some as harder than others, and when counts are run they go on this page.
One cleanroom — ISO 7 by design, ISO 6 estimated
A single cleanroom built to ISO 7 design parameters, with an adjoining staging and shipping sub-cleanroom, connected in a stepped chain so the pressure and gravity gradients both run in the same direction — outward and downward. On airflow, filtration and occupancy the room calculates to ISO 6; see below for the working.
| Parameter | Value | Note |
|---|---|---|
| Cleanroom dimensions | 12 × 16 × 7 ft | 1,344 ft³ / 38.1 m³ |
| Staging sub-cleanroom | 6 × 10 ft | Shipping, staging, gowning |
| Terminal HEPA units | 6 × H-14 | Curtained ceiling downflow |
| Airflow per unit | 350–800 m³/h | Three-speed; run high in production |
| Enclosed ISO 5 workstation | 1 | Plus air-curtained open surfaces |
| Pressure differential | 0.05 in w.c. | Cleanroom → staging |
| Cooling capacity | 36,000 BTU | Two independent systems |
| Temperature | ≤ 65 °F / 18.3 °C | Set for gowned personnel |
| Relative humidity | ≤ 35% | Dedicated dehumidification |
Air change rate and recovery
| Fan setting | Total airflow | Air changes/hr | 100:1 recovery |
|---|---|---|---|
| Low — 0.30 m/s | 2,100 m³/h | 55 ACH | ~5.0 min |
| Medium — 0.45 m/s | 3,323 m³/h | 87 ACH | ~3.2 min |
| High — 0.65 m/s | 4,800 m³/h | 126 ACH | ~2.2 min |
At 126 air changes per hour, this room runs above the entire ISO 7 design band. USP <797> specifies a minimum of 30 ACH for an ISO 7 buffer room; general cleanroom design guidance puts the ISO 7 range at roughly 60–120. We are above the top of it, and above ISO 7 specification even at the lowest fan setting. Room air passes a filter roughly every 29 seconds, and a disturbance clears to a hundredth of its peak in about 2.2 minutes.
Published ACH bands put ISO 6 at roughly 150–240 and ISO 7 at 60–120. Those are design heuristics rather than physics: they assume typical industrial occupancy and a corresponding particle generation rate. Steady-state cleanliness is really generation rate divided by airflow times filter efficiency, and two of those three terms here are unusually favourable.
The filtration is H-14, rated 99.995% at the most penetrating particle size — a full grade above the H-13 used in most rooms of this class. Recirculated supply air is effectively particle-free. And occupancy is one or two gowned operators running low-generating processes, not a production floor. Apply 126 ACH to a generation term that small and the expected particle count lands at ISO 6 or better, notwithstanding where 126 sits on a table built for busier rooms.
We label that estimated rather than claimed — and by the same standard, the ISO 7 and ISO 5 figures on this page are design classifications too. None of them is a counted certification. Calling one an estimate and the others facts would be a distinction we have not earned, so we do not make it.
Room class is the backdrop. What matters is the air touching an open vial, and that is never room air. Every surface where product is opened sits under ISO 5 unidirectional flow — either inside the enclosed workstation or directly beneath curtained H-14 downflow, with the curtains keeping the column from spreading and slowing before it reaches work height.
There is no bench in this room where a vial is opened outside an air curtain. Not one. So the relevant question is not what the room classifies at — it is what the air directly above the product classifies at, and that answer is ISO 5 everywhere, unidirectional, downward, continuously.
All units run high whenever product is open. The lower speeds exist for overnight and idle periods; they are never the setting during production. And every surface where a vial is opened is protected by ISO 5 airflow — either inside the enclosed workstation or directly beneath curtained downflow. There is no bench in the room where product is opened outside an air curtain.
The room is not simply pressurised
It is designed with a defined direction of travel for air, and for anything heavier than air. Supply is ceiling-mounted H-14 filter units recirculating room air, plus a dedicated makeup-air fan drawing from an interior entryway through an H13 filter box.
Exfiltration is low, by design. The room loses most of its air at floor level rather than through a ceiling return. That does two jobs at once: it continuously sweeps out anything that settles, and it gives dense gas a defined exit. The makeup air filter sits near the argon station so incoming clean air moves directly across the point where argon is used.
The cleanroom is held above the pressure of every space it touches at all times — during production, overnight, and while unoccupied. It is never allowed to equalise and never allowed to go negative.
This matters because pressure is the only contamination control that works when nobody is watching. Filters clean the air being moved; gowning controls what a person brings in; segregation controls what happens while product is open. Positive pressure is what keeps unfiltered air out of the room at 3am on a Sunday — and the moment it lapses, every other control is working against an ingress it cannot see.
Plumbed, not bottled at the bench
| Element | Specification |
|---|---|
| Gas purity | 99.999% argon |
| Regulation | Dual-stage, dedicated to the lab |
| Distribution | Hard-plumbed copper trunk |
| Stations | 6, quick-disconnect |
| Drops | Short direct runs, purged before every use |
This is a real conflict between two correct practices, and worth explaining rather than hiding. H-14 downflow at 0.65 m/s easily overcomes argon's gravity-driven settling — so leaving the unit directly above the argon bench running actively destroys the inert blanket you are trying to establish, dilutes the purge, and creates turbulent mixing that folds room air back into the argon.
So that single unit is switched off during purging. The room's other five maintain classification, and the work zone is protected by the argon blanket itself during that window. It is a temporary, deliberate, single-unit disable — not a shutdown, and not a period when the room is unprotected.
Because the facility routes dense gas deliberately downhill and out, the argon path through the building is known rather than assumed — which is what makes it monitorable. Fixed oxygen monitoring with a non-depleting zirconium oxide sensor, alarmed at the 19.5% OSHA action level, sits at the terminal end of the door chain where that path ends.
Sterile and endotoxin-free are two different problems
They require separate solutions, and every component is processed for both. Vials run USP 70% IPA to strip oils and handling residue, then a spin, then three separate rinses of in-house distilled water with a spin between each, then a final alcohol rinse and spin. Stoppers take the same four-stage sequence, alcohol first, in batches of 500 to 1,000. Bottle caps take the same sequence and are then autoclaved.
Dilution is a recognised depyrogenation route on its own — sequential rinses in endotoxin-free water drive contamination down by orders of magnitude on any surface. So components reach the heat step already at or below the acceptable endotoxin limit.
Dry heat depyrogenation
Three ovens, each with a defined role. Cycles are calculated as FD — the equivalent minutes of exposure at 250 °C — using the industry-standard depyrogenation z-value of approximately 46.5 °C. The acceptance baseline is a 3-log endotoxin reduction, which FDA endotoxin guidance associates with dry heat cycles on the order of 250 °C for 45 minutes.
| Oven | Cycle | FD | vs 3-log | Load |
|---|---|---|---|---|
| Main glassware | 290 °C / 35 min | 255 min | 8.5× | Beakers, bottles, flasks |
| Vial oven | 371 °C / 15 min | 6,082 min | 200×+ | Vials, high throughput |
| Tall-glass oven | 232 °C / 2 hr 15 min | 55 min | 1.8× | Graduated cylinders |
Dwell timing starts when probes confirm the glass has reached temperature — not when the oven setpoint is reached. That distinction is the difference between a validated cycle and a number on a dial, and it is where most informal depyrogenation quietly fails.
The inputs are held to the same standard as the room
Water is generated in-house by distillation with endotoxin testing per batch, and water for injection with a certificate of analysis is used for formulation. USP or injection-grade materials are used throughout. Sterile filtration is through 0.22 µm.
Measurement is where formulation stops being a recipe and becomes a quantity. Analytical balances are verified against certified weights, positive-displacement dispensing is held to a fraction of a percent, and fills are confirmed gravimetrically rather than assumed. Lyophilisation runs under monitored conditions, and the vacuum is broken with argon before any stopper seats.
None of this is exotic. It is what the good drugs get. It is simply absent from this market, because it is slow, it costs money, and nothing forces anyone to do it.
The room, the procedures and the formulas are under continuous improvement — because the standard we are building toward is the one used for drugs that cost a thousand dollars a vial.
Why this lab exists