The test report shows every fitting at 2.0×10⁻¹⁰ atm·cc/s. Twenty fittings. Identical values to two significant figures. Real test data doesn't look like that.
Why Helium — The Actual Reasons
Atmospheric background concentration is negligible. Helium constitutes approximately 5.2 ppm of the atmosphere. Mass spectrometer detection works on signal-to-noise ratio — the lower the background concentration of the tracer gas, the more sensitively the instrument can detect a small leak signal above that background. Argon comprises 0.93% of the atmosphere — a background level roughly 180 times higher. A leak that produces a detectable helium signal would be completely masked by argon background noise.
Mass number 4 is nearly interference-free. At mass number 4, a mass spectrometer encounters almost no interfering species from normal atmospheric constituents. The only potential interferent is molecular deuterium (D₂), present at negligible concentrations. Nitrogen at mass 28 shares that position with carbon monoxide — using nitrogen as a tracer gas requires distinguishing two species at identical mass numbers.
Complete chemical inertness. Helium does not react with system materials, vacuum system components, or any other gas present during testing.
Higher leak rate than process gases. Most specialty process gases are several times heavier than helium — HCl at 36.5 g/mol versus helium at 4 g/mol. Through molecular-flow leak paths, helium flows approximately three times faster than HCl at equivalent pressure. Helium leak testing is conservative: a fitting that passes helium testing will have a lower actual leak rate for the process gas in service.
These combined properties allow modern helium mass spectrometer leak detectors to achieve sensitivity down to 10⁻¹² atm·cc/s.
What 1×10⁻⁹ atm·cc/s Actually Means
This acceptance threshold reflects the accumulated requirements of semiconductor equipment manufacturers — Applied Materials, Tokyo Electron, Lam Research — and industry specifications including SEMI F1, converging on a common per-fitting acceptance criterion for UHP gas delivery systems.
The physical meaning: 1×10⁻⁹ atm·cc/s represents one billionth of a cubic centimeter of gas at atmospheric pressure passing through the leak point per second.
Annualized: 10⁻⁹ cc/s × 3.15×10⁷ seconds/year ≈ 0.03 cc/year — approximately two-thirds of a single water drop over an entire year of continuous service.
Why not 10⁻⁸? For acutely toxic gases — HF, AsH₃, PH₃, WF₆ — a sustained leak rate of 10⁻⁸ atm·cc/s, multiplied across equipment lifetime and the number of fittings in a system, produces cumulative contamination that is not acceptable in a cleanroom environment.
Why not 10⁻¹⁰? Test time increases substantially. More importantly, 10⁻⁹ atm·cc/s is approximately the practical limit that correctly installed VCR metal-seal fittings can consistently achieve in field conditions.
Testing Methods
Vacuum Chamber Method Component placed inside a sealed vacuum chamber, pressurized internally with helium, chamber evacuated externally. Sensitivity: down to 10⁻¹² atm·cc/s. Application: component-level acceptance testing at the manufacturer.
Spray Method Assembled system evacuated internally, helium spray gun applied to each fitting and weld sequentially. Identifies precisely which joint is leaking. Sensitivity: 10⁻¹⁰ to 10⁻¹² atm·cc/s. Application: standard acceptance method for prefabricated gas sticks and panels — every VCR fitting and orbital weld tested individually, documented separately.
Sniffer/Probe Method System pressurized internally with helium, probe moved across exterior of fittings and welds. Sensitivity: approximately 10⁻⁷ atm·cc/s. Application: periodic maintenance inspection of installed systems that cannot be disconnected. Not appropriate for installation acceptance — a fitting that passes sniffer testing has not been tested to UHP acceptance criteria.
How Leak Test Reports Are Falsified or Invalidated
Missing calibration record Before any measurement, the instrument must be verified against a calibrated standard leak with a known, certified leak rate. A report without a calibration record cannot be verified.
Insufficient dwell time The spray method requires the helium gun to dwell at each fitting long enough for helium concentration to build to a detectable level. Rushing through — one to two seconds per fitting — does not allow enough time for signal development. Correct practice is a minimum of three to five seconds per fitting.
Test pressure below service pressure Leak rate scales with pressure differential. A fitting tested at atmospheric pressure may pass, then leak at the actual operating pressure of 30, 50, or 150 bar. Test conditions should reflect actual service conditions.
Elevated helium background Residual helium from adjacent testing or purging operations elevates the noise floor and masks genuine leak signals. A valid test requires establishing a stable, low background baseline before measurements begin.
Fabricated results A legitimate spray-method test of twenty VCR fittings will produce twenty different readings — varying by fitting geometry, spray gun distance, background fluctuation, and instrument response time. If twenty fittings all report identical values, that data was not measured. Real instruments and real test conditions produce variation. Identical values to multiple significant figures is the signature of copy-paste data entry.
Instrument connection failure The mass spectrometer's vacuum-side connection must be correctly made to the system under test. An incorrectly connected instrument measures its own internal background — every fitting appears to pass because the system was never in the measurement loop. On-site witness of the test setup is the only reliable prevention.
What a Valid Test Report Must Contain
- Instrument model and serial number
- Calibration record: standard leak value, calibration date, instrument response
- Test method and test pressure
- Background helium level at test start
- Individual measurement for each fitting and weld, with fitting identification
- Acceptance threshold applied
- Technician name and test date
A report that shows only a summary pass/fail without individual fitting data is not a compliant leak test record for UHP component qualification.
Previous: Single-Stage vs. Two-Stage Regulators
Next: Bellows Valves vs. Diaphragm Valves — Valve Selection for Corrosive Gas Service
Related: MTR Traceability — What to Verify When Qualifying UHP Component Suppliers
Browse related products: Valves