Detection Limit Testing for Atmospheric Carbon Monoxide

Detection limit testing for atmospheric carbon monoxide (CO) establishes the lowest concentration that can be reliably distinguished from background noise using non-dispersive infrared, gas chromatographic, or electrochemical analytical systems. Accurate determination of the method detection limit (MDL) is essential for validating ambient air quality data, qualifying monitoring instruments, and demonstrating compliance with air quality standards such as GB 3095, ISO 4224, EN 14626, and ASTM D3162. This article outlines the principles, sample types, testing items, statistical procedures, and applications of CO detection limit verification in accredited third-party laboratories.

Technical Overview

Carbon monoxide is a colorless, odorless, and highly toxic gas generated mainly by incomplete combustion of fossil fuels and biomass. Major sources include motor vehicle exhaust, industrial boilers, residential heating, and open burning. Because CO in clean ambient air may exist at sub-ppm concentrations, analytical systems must demonstrate sufficient sensitivity to quantify levels far below regulatory limits such as the 1-hour average of 10 mg/m³ and the 24-hour average of 4 mg/m³ specified in GB 3095-2012.

The detection limit represents the lowest concentration of a target analyte that can be detected, but not necessarily quantified, with a stated level of confidence. Several related quantities are evaluated during detection limit studies:

• Instrument Detection Limit (IDL): the minimum concentration distinguishable from instrument noise, typically estimated from the standard deviation of repeated zero-gas readings.
• Method Detection Limit (MDL): the minimum concentration measurable with 99% confidence that the reported value is greater than zero, determined through replicate analysis of low-concentration samples carried through the complete analytical procedure.
• Limit of Quantification (LOQ): the lowest concentration that can be quantified with acceptable accuracy and precision, commonly estimated as three to four times the MDL.

Detection limit studies are performed in accordance with recognized protocols, including US EPA procedures for MDL determination, HJ 168 for environmental method validation, and the performance-based requirements of ISO 4224, EN 14626, GB 9801, ASTM D3162, and JIS B 7953. Traceable calibration gases, controlled laboratory conditions, and rigorous statistical evaluation ensure that the reported detection limit is defensible and reproducible.

Testing Samples

Zero Gas and Blank Samples

High-purity zero air or nitrogen with CO content below the instrumental blank serves as the baseline reference. Replicate zero readings are used to characterize baseline noise and drift, which feed into the calculation of the instrument detection limit. Zero gas must be certified traceable to a national metrology institute to guarantee the integrity of the blank signal.

Certified Low-Level CO Calibration Standards

Certified reference gas mixtures of CO in synthetic air, gravimetrically prepared and traceable to national measurement standards, are used for detection limit verification. Standards are typically selected at concentrations of one to five times the estimated MDL so that the added signal is statistically distinguishable from the blank. Multi-point calibration series spanning the working range are used to confirm linearity and response slope.

Collected Ambient Air Samples

Real ambient air samples may be collected in inert fluoropolymer bags or passivated stainless-steel canisters, conditioned and leak-checked before sampling. Holding times and stability requirements are observed to prevent loss or contamination. These samples demonstrate the practical detection capability of the full method, including sampling, transport, and analysis steps.

Testing Items

Detection limit evaluations for atmospheric CO typically encompass the following performance characteristics:

• Instrument detection limit derived from baseline noise statistics at the analytical wavelength or detector channel.
• Method detection limit from at least seven replicate determinations of a low-concentration sample processed through the entire method.
• Limit of quantification established from the MDL and verified with a quantifiable low-level standard.
• Baseline noise and zero drift over a defined observation period.
• Linearity and calibration verification across the working concentration range.
• Repeatability precision expressed as relative standard deviation at near-limit concentrations.
• Response time and rise time of the analyzer to a step change in CO concentration.
• Blank contribution from the sampling system, including bags, canisters, and transfer lines.

Testing Methods

Non-Dispersive Infrared Spectrometry (NDIR)

NDIR with gas filter correlation (GFC) is the most widely deployed technique for continuous ambient CO monitoring, as specified in GB 9801, ISO 4224, EN 14626, and ASTM D3162. CO absorbs infrared radiation in a characteristic band near 4.6 μm; correlation filtering removes cross-interference from water vapor, carbon dioxide, and hydrocarbons. Modern GFC-NDIR analyzers achieve detection limits on the order of 0.05 to 0.1 ppm, equivalent to approximately 0.06 to 0.12 mg/m³ at 25 °C and 101.325 kPa.

Gas Chromatography with Methanizer and FID

For laboratory-based determinations requiring enhanced sensitivity, CO is separated chromatographically and catalytically converted to methane over a nickel catalyst before detection by flame ionization detector (FID). This approach offers stable response, excellent selectivity, and method detection limits down to approximately 0.02 mg/m³, making it suitable for background-site and research measurements.

Electrochemical and Optical Sensor Systems

Portable electrochemical sensors and laser-based optical instruments, including tunable diode laser and cavity-enhanced absorption platforms, are used for screening surveys and indoor assessments. Their detection limits are generally higher than those of GFC-NDIR systems but remain adequate for occupational exposure levels and indoor air guidelines when validated through the same MDL protocols.

Statistical Determination of the Detection Limit

The standard MDL protocol requires at least seven replicate analyses of a sample spiked at a concentration within one to five times the estimated detection limit, processed through all steps of the method. The standard deviation of the replicate results is multiplied by the one-sided Student t value at 99% confidence with n−1 degrees of freedom, which equals 3.143 for seven replicates. The resulting MDL is verified by confirming that the spike concentration is within the acceptable range relative to the MDL and that the associated precision meets the method criteria. The limit of quantification is then reported as a defined multiple of the MDL, typically three to four times.

Applications

Detection limit verification for atmospheric CO supports a broad range of regulatory, industrial, and scientific activities:

• Qualification and periodic performance verification of continuous analyzers in national and regional ambient air quality monitoring networks.
• Compliance demonstration against ambient air quality standards and permit conditions for industrial facilities.
• Indoor air quality assessments in parking garages, tunnels, offices, and residential buildings with combustion appliances.
• Instrument acceptance testing, type approval, and calibration services performed by accredited third-party laboratories.
• Atmospheric chemistry research, including background monitoring, emission inventory validation, and urban airshed studies.
• Quality assurance programs requiring documented sensitivity data for data validation and auditing purposes.

FAQ

What is the difference between the detection limit and the limit of quantification? The detection limit is the lowest concentration at which the analyte can be reliably distinguished from the blank with 99% confidence, while the limit of quantification is the lowest concentration that can be measured with acceptable accuracy and precision, typically three to four times higher.

Why are at least seven replicate measurements required? Seven replicates provide a statistically acceptable estimate of the standard deviation with the 99% confidence Student t factor, balancing rigor against practical analytical workload in line with established MDL protocols.

What concentration units are used for atmospheric CO? Results are commonly reported in mg/m³ or ppmv. At 25 °C and 101.325 kPa, 1 ppmv of CO corresponds to approximately 1.15 mg/m³, so detection limits should always be reported with the reference conditions stated.

How often should detection limits be re-evaluated? Detection limits should be established during method validation, re-confirmed annually, and re-evaluated whenever the instrument, calibration gases, sampling system, or analytical procedure is materially changed.

Which standards govern CO detection limit testing? Relevant references include GB 3095 and GB 9801 for ambient air quality and measurement, ISO 4224 and EN 14626 for infrared spectrometric measurement, ASTM D3162 for continuous nondispersive infrared analysis, JIS B 7953 for automatic CO analyzers, and US EPA MDL determination guidance for the statistical procedure.