Atmospheric carbon monoxide (CO) sample pre-treatment is the decisive step that determines whether whole-air samples collected in canisters, bags, or continuous monitoring lines yield accurate and defensible concentration data. This technical article explains container preparation, moisture management, cryogenic preconcentration, and dilution procedures aligned with ISO 4224, ISO 8186, EN 14626, ASTM D3162, JIS B 7953, and GB 9801. It is written for environmental engineers, laboratory analysts, and air quality professionals who need a complete overview of the pre-treatment workflow, quality control requirements, and practical applications.
Technical Overview
Carbon monoxide is a colorless, odorless, and highly toxic gas produced primarily by the incomplete combustion of fossil fuels and biomass. Its concentration in the atmosphere spans a very wide range, from tens of parts per billion in clean background air to several parts per million near roadways, tunnels, and combustion sources. Because of this wide dynamic range, and because common optical detectors are subject to interference from water vapor and carbon dioxide, careful pre-treatment of air samples is essential before instrumental determination.
Although CO is chemically stable and adsorbs weakly to most inert surfaces, pre-treatment remains critical for four reasons: it preserves the representativeness of the collected sample, it removes particulate matter and moisture that interfere with detection, it concentrates trace-level CO to quantities detectable by gas chromatography, and it dilutes high-concentration samples into the linear calibration range of the analyzer. Recognized reference procedures are established in ISO 4224 and ISO 8186, EN 14626, ASTM D3162, JIS B 7953, and the Chinese national standard GB 9801, while ambient limit values are specified in GB 3095 and occupational exposure limits in GBZ 2.1.
A typical pre-treatment experiment is therefore an end-to-end evaluation covering container cleaning, controlled field sampling, laboratory drying and preconcentration, instrumental measurement, and statistical verification of blanks, recoveries, and hold times. Accredited third-party laboratories apply this workflow to produce results that are traceable to national metrology standards and acceptable for regulatory reporting.
Testing Samples
Whole-Air Canister Samples
Electropolished or passivated stainless-steel canisters, typically 3.2 L or 6 L in volume with optional inert silica lining, are the reference containers for atmospheric CO. They may be filled instantaneously as grab samples or integrated over 1 to 24 hours using calibrated critical orifices or electronic mass-flow controllers. After collection, samples are commonly pressurized with ultrapure nitrogen or helium to 1.5 to 2 atmospheres, which stabilizes the sample and permits multiple analytical aliquots from a single container.
Flexible Bag Samples
Polyvinyl fluoride film bags and multilayer aluminum-foil laminate bags are lightweight alternatives for short-term indoor air and workplace screening. Their hold time is limited, generally 24 to 48 hours, because of permeation effects and possible background CO off-gassing from polymer materials. Bag samples must always be accompanied by field blanks to demonstrate container cleanliness.
Continuous Monitoring Line Samples
For fixed monitoring stations and online analyzers, the pre-treatment train consists of an inert intake line, a 2 micrometer particulate filter, a water removal device such as a perfluorinated membrane dryer, and a constant-flow, constant-pressure delivery system feeding the detector. This configuration supports continuous NDIR analyzers operated under EN 14626 and JIS B 7953.
Calibration Standards, Blanks, and Spiked Samples
Every analytical batch includes certified CO-in-air mixtures traceable to a national metrology institute, zero air for blank verification, field blanks, and matrix-spiked canisters. These materials quantify cleaning efficiency, calibration linearity, and analyte recovery across the full pre-treatment sequence.
Testing Items
The core deliverables of the atmospheric CO sample pre-treatment experiment include both concentration data and performance statistics:
- CO mass concentration in mg/m³ or volume fraction in ppm and ppb for ambient, indoor, and workplace air.
- Time-weighted average concentrations derived from flow-controlled integrated samples.
- Container blank levels and hold-time stability data for canisters and bags.
- Recovery, breakthrough, and dilution-linearity verification results.
- Method detection limits, duplicate precision expressed as relative percent difference, and spike recovery statistics.
- Pressure-, temperature-, and humidity-corrected sample volume calculations for accurate mass reporting.
Testing Methods
Container Cleaning and Preparation
Canisters undergo repeated pressurization with humidified zero air to approximately 200 kPa followed by evacuation below 0.05 torr, typically for at least three cycles. A small amount of ultrapure water vapor is left in the container to passivate active wall sites. Each canister is leak-tested, finally evacuated, and blank-verified by analysis; acceptance requires blank CO below 10 percent of the method detection limit.
Field Collection Procedures
- Install an inert intake line with an in-line particulate filter positioned away from direct exhaust, pump motors, and operator breathing zones.
- For integrated sampling, attach a calibrated critical orifice or flow controller sized to fill approximately 80 percent of the canister volume over the target period, for example about 2.2 mL/min for a 3.2 L canister over 24 hours.
- Record initial vacuum, final pressure, ambient temperature and pressure, and sampling flow verification before and after collection.
- Complete chain-of-custody documentation and transport samples to the laboratory within the validated holding period.
Laboratory Pre-Treatment Train
In the laboratory, samples first pass through a particulate filter and a perfluorinated membrane dryer, which transmits CO quantitatively with recoveries above 99 percent while removing water. High-concentration samples are dynamically diluted with zero air to fall within the calibration range. For trace-level gas chromatographic analysis, CO is preconcentrated cryogenically on molecular sieve 5A at liquid nitrogen temperature, with a preceding trap removing water and carbon dioxide. The focused CO is then thermally desorbed onto a porous-layer open tubular column for separation.
Instrumental Determination
- Non-dispersive infrared spectrometry with gas filter correlation, per ISO 4224, EN 14626, ASTM D3162, GB 9801, and GB/T 18204 series methods, suitable for concentrations above approximately 0.05 ppm.
- Gas chromatography with a heated nickel methanizer coupled to a flame ionization detector, or a reduction gas detector, per ISO 8186, achieving detection limits of roughly 1 to 5 ppb with cryofocusing.
- Cavity-enhanced optical techniques with dry-gas correction or spectroscopic water compensation for sub-ppb research-grade measurements.
Quality Assurance and Quality Control
Each batch requires a multipoint calibration curve prepared at humidity levels matching the samples, continuing calibration verification within 10 percent, duplicate samples with relative percent difference below 10 percent, matrix spike recoveries between 90 and 110 percent, and blank confirmation after every cleaning cycle. Hold-time studies must demonstrate stability, which for humidified passivated canisters typically extends to 30 days, whereas bag samples must be analyzed within 24 to 48 hours.
Safety Considerations
CO is acutely toxic, so sampling near known sources requires personal CO monitors and adequate ventilation. Cryogenic preconcentration with liquid nitrogen demands insulated gloves, face protection, and oxygen-deficiency monitoring, and compressed calibration cylinders must be secured and leak-checked on a scheduled basis.
Applications
Pre-treated atmospheric CO samples support a broad range of regulatory, occupational, and scientific objectives:
- Ambient air quality compliance monitoring against limit values in GB 3095 and equivalent regional standards.
- Roadside, tunnel, and parking garage ventilation studies and traffic emission assessments.
- Indoor air quality investigations in offices, schools, arenas, and enclosed recreational spaces.
- Occupational exposure assessments producing 8-hour time-weighted averages for comparison with GBZ 2.1 limits.
- Atmospheric chemistry research, where CO serves as a tracer for combustion plumes and biomass burning.
- Verification of fixed CO alarm systems, combustion appliance diagnostics, and post-incident fire investigations.
FAQ
Why is pre-treatment necessary when carbon monoxide is chemically stable?
Stability alone does not guarantee accuracy. Water vapor and carbon dioxide broaden and overlap infrared absorption bands in NDIR detection, trace ppb-level concentrations require cryogenic preconcentration to exceed detector thresholds, and time-weighted averages demand flow-controlled capture that only a validated pre-treatment sequence can provide.
How long can collected samples be stored before analysis?
Humidified, passivated canisters generally preserve CO for up to 30 days with recoveries above 95 percent. Film and foil bags must be analyzed within 24 to 48 hours, and every project should rely on laboratory-validated hold-time data rather than default assumptions.
Does drying the sample remove carbon monoxide?
No. Perfluorinated membrane dryers selectively permeate water while transmitting CO quantitatively, with typical recoveries above 99 percent. Chemically reactive scrubbers and unsvalidated sorbent traps should be avoided because they may oxidize or fractionate CO.
What detection limits can be achieved after pre-treatment?
NDIR with gas filter correlation typically reaches 0.05 to 0.1 ppm, gas chromatography with methanizer or reduction gas detection combined with cryofocusing reaches 1 to 5 ppb, and cavity-enhanced optical methods can achieve sub-ppb performance on dried gas streams.
Can one canister sample be used for multiple analytes?
Yes. Because CO, methane, carbon dioxide, and many non-methane hydrocarbons are stable in passivated canisters, a single pressurized whole-air sample can support multi-analyte analysis, provided the pre-treatment train and holding times are validated for every target compound.