Atmospheric pressure, a one-way seal and a piston.
An airless bottle has no dip tube. The pump group is engineered as a one-way seal: each stroke expels product and blocks air from flowing back, so the bottle interior drops into low pressure. Outside atmospheric pressure then pushes a piston up from the base to follow the product, keeping the formula out of contact with air until the last dose.
The friction between the large piston and the inner wall is the core of the design. It governs both air-tightness and whether the piston travels smoothly — too little friction and the seal leaks, too much and the piston stalls. That is why piston fit is the first parameter a partner factory qualifies, and the first thing we check in QC.
Dose is fixed by pump diameter, stroke and spring force; once set, output is repeatable to about 0.05 ml, and it can be tuned by swapping pump components. Common capacities run 10–100 ml — the format relies on atmospheric pressure, so it suits smaller volumes. Cylindrical bodies are far more common than square. Exteriors take hot-stamp foil, vacuum metallising, spray coating or coloured resin (see the decoration process comparison). Airless is dearer per unit than a standard lotion bottle, but minimum order quantities are not demanding.
| Structure | How it works | Typical use |
|---|---|---|
| Standard airless | Single body, piston at base, pump on top | Serums, essences, lightweight lotions |
| Single-bottle composite | Inner bottle carries the product; decorative outer sleeve; piston in the inner | Prestige skincare where the outer needs metallising or spray |
| Twin-bottle composite | Two independent inner containers with paired pump | Two-phase or dual-active formulas |
| Non-piston (self-sealing valve) | All-plastic valve; a carrier disc rises as product is used until it reaches the top of the body | Mono-material and recyclability projects |
The airless bottle test sheet, with pass thresholds.
These are the numbers a partner factory signs against for an airless programme. They sit under the four inspection gates described in our QC standards.
| Test | Method | Pass threshold |
|---|---|---|
| Vacuum leak | Filled, lying and inverted on white paper in a vacuum chamber | -0.05 MPa for 5 min (quick check) or -0.08 MPa for 15 min, loss ≤0.5%; standard emulsions at -0.06 MPa, volatile formulas at -0.08 MPa; no trace on the paper |
| Inner–outer bottle pull | Tensile tester on the collar-to-outer assembly | ≥50 N without separation; snap fit seated fully; no rubbing between inner and outer wall |
| Inner plug retention | Axial pull on the plug | ≥30 N without pull-out; gasket ≥1 N |
| Anodised aluminium trim bond | Glued trim, cured 24 h, then pulled | ≥50 N |
| Thread clearance | Outer cap to body thread | 0.1–0.8 mm; no thread strip at rated torque on a torque meter |
| Piston seal (static) | Cool 4 h after moulding, assemble piston and body, fill with water, stand 4 h | Resistance felt on the piston; no water passes |
| Piston travel (dynamic) | Stand 4 h, fit pump, dispense to empty | Piston rises cleanly to the top; contents fully expelled |
| Priming | Filled 3/4 with coloured water or lotion, even strokes | First output within ≤8 strokes (coloured water) or ≤10 strokes (lotion) |
| Output per stroke | Weighed over 50 strokes ×3 (method below) | Single stroke ±15%; run average within 5–10% of specified dose |
| Dimensions | Callipers / gauges on mating and overall dimensions | Fit-critical dimensions ≤0.5 mm; overall height/width/depth ≤1.0 mm |
| Visual inspection | 100 W cool-white fluorescent lamp at 50–55 cm | 500–550 lux, 30–35 cm eye distance, 45° ± 15° viewing angle, ≤12 s per piece |
Two things are worth noting. The quick check (-0.05 MPa, 5 minutes) is what a line operator runs; the full test (-0.08 MPa, 15 minutes, ≤0.5% loss) is what goes on the inspection report. And the visual-inspection environment is specified precisely because "looks fine" under warehouse lighting is not a standard.
Lotion pump sizes, outputs and the QC sheet.
A lotion pump has around eleven parts — actuator, upper stem, lock collar, gasket, closure, pump seal, lower stem, spring, housing, glass ball and dip tube — which is why new tooling is expensive and why stock (common-mould) pumps are the default unless a brand is committed to a proprietary shape.
| Parameter | Standard options / value |
|---|---|
| Neck sizes | Φ18, Φ20, Φ22, Φ24, Φ28, Φ33, Φ38 mm |
| Output per stroke | 0.15/0.2 cc · 0.5/0.7 cc · 1.0/2.0 cc · 3.5 cc · 5.0 cc · 10 cc and above |
| Lock head | Guide-block lock · screw lock · clip lock · no lock |
| Structure | External-spring pump · plastic-spring pump · leak-resistant lotion pump · high-viscosity pump |
| Feed | Airless (piston) or dip-tube |
| Spring position | External spring never touches the formula, so a corroding spring cannot contaminate the product; internal-spring pumps are cheaper but need a compatible spring alloy |
| Test | Pass threshold |
|---|---|
| Dry-pump (air) strokes | <8 strokes for bottles ≤1,000 ml; <10 strokes above 1,000 ml |
| Priming with product | ≤5 strokes aqueous; ≤10 strokes non-aqueous |
| Dose stability | Even and stable; average ≥3 g (ml) per stroke on standard pumps; 1–2 ml on 250 ml styling-gel pumps |
| Dispensing efficiency | ≥95% aqueous · ≥90% non-aqueous of fill recovered |
| Return time | Actuator returns in ≤3 s |
| Service life | ≥1,500 continuous actuations |
| Feel | Springy, no sticking or dead travel |
| Dip tube length | Measured to 0.02 mm; tolerance +2/-1 mm; 5 mm clearance from the base on a 250 ml bottle |
Output measurement method. Fill to nominal volume, prime with 15 strokes, weigh; press 50 continuous strokes, weigh again. (Weight before − weight after) ÷ 50 = single-stroke dose. Repeat three times and average. The result is what the specification is signed against.
What fails in the field. The seal at the glass or steel ball under the spring is critical: if it leaks, product flows back into the bottle on each stroke and suction drops, so the pump starts "losing" doses. A loose actuator-to-valve fit causes back-flow and resistance at the nozzle. Both faults are visible at IQC sampling, which is why we sample pumps before they go near the filling line.
Sprayer sizes, dose and torque.
A fine-mist sprayer works on atmospheric balance. Its parts are the actuator, spray insert, centre stem, lock collar, gasket, piston core, piston, SUS304 stainless spring, housing and dip tube (PP, LDPE or HDPE) with a glass ball. Perfume uses aluminium crimp collars (semi- or full-crimp); screw collars are plastic.
| Parameter | Value |
|---|---|
| Neck sizes | 12.5 mm to 24 mm, chosen to match the bottle finish |
| Output per actuation | 0.1–0.2 ml; typical for perfume, toners and styling water |
| Dosing error | Within 0.02 g per actuation |
| Spring | SUS304 stainless steel |
| Collar types | Crimp (aluminium, semi- or full-crimp) for perfume · screw (plastic) for general use |
| Collar torque | Specified by diameter, 58.8 N to 224 N, verified on a torque meter |
| Cap flip test | Overcap cycled open and closed with no cracking or loosening |
| Service life | n = Vtotal ÷ Vi × 1.2 (1.2 is the safety factor); ≥1,500 continuous actuations |
| Tooling | Many components, so tooling is costly — stock sprayers are the default |
Spray quality is judged on an even mist with no droplets or streams, plus the counted priming strokes and the weighed dose. A sprayer that passes on water and fails on your formula is a viscosity or compatibility issue, which is why sample sprayers are tested with your actual bulk.
Trigger sprayers: buy stock, mind the rubber.
A trigger sprayer has a dozen-plus components — housing, base, spring, glass ball, gasket, thread collar, dip tube, valve, nozzle, trigger, piston and secondary stem — in ABS, PP, PE and EVA, sometimes with an aluminium shroud. It is a home-care and hair-care format more than a skincare one.
Three sourcing rules. One: use stock (common-mould) units wherever possible; a proprietary trigger means high tooling cost and a long lead time. Two: check the body-to-shroud connection — a weak joint is the most frequent transit failure. Three: the cylindrical rubber sealing component has no anti-ageing or acid/alkali resistance and hardens and stops sealing below 0°C; contaminated by product it degrades, shortens the unit's life and coarsens the spray. Specify the elastomer against your formula and your shipping route.
Frequently asked
Specify the pump before you price the bottle.
Tell us the formula, viscosity and target dose. Your quote comes with the pump or sprayer specification and the test sheet above filled in for your components.
Related reading.
- Airless bottles — formats, capacities and decoration options we source for airless programmes.
- QC standards with the numbers — the four inspection gates these tests sit under.
- Decoration process comparison — metallising, spray, hot-stamp and print for pump and airless exteriors.
- Perfume bottle supplier — crimp sprayers, glass and collars for fragrance.
- Dropper bottles — the alternative when a serum does not need a pump.