Atmospheric carbon monoxide (CO) measurement is a cornerstone of ambient air quality monitoring, industrial emission control, and indoor air safety assessment, yet its accuracy is frequently compromised by a wide range of interference factors. This technical article systematically analyzes the spectral, chemical, physical, and sampling-related interferences that affect CO determination, and reviews the standard test methods, instruments, and quality-control measures used by accredited third-party laboratories to eliminate or correct these interferences.
Technical Overview
Carbon monoxide is a colorless, odorless, and highly toxic gas produced primarily by incomplete combustion of fossil fuels, biomass, and other carbonaceous materials. In the atmosphere, background CO concentrations typically range from 50 to 200 ppb in clean rural environments, while urban roadside locations can exceed 10 ppm during peak traffic hours. Because regulatory limits and health-based thresholds are often expressed at low concentration levels, the measurement of atmospheric CO demands high sensitivity, high selectivity, and rigorous control of interference factors.
Interference refers to any physical or chemical effect that causes an analytical instrument to produce a biased response attributable to a substance or condition other than the target analyte. For CO, the most significant interferences arise from the spectral overlap with water vapor, carbon dioxide, methane, and nitrous oxide in infrared-based analyzers, as well as cross-sensitivity to hydrogen and reducing gases in electrochemical sensors. Additional systematic errors originate from temperature and pressure fluctuations, sampling line contamination, zero and span drift, and matrix differences between calibration gases and real air samples.
International and national standards, including ISO 12039, ISO 8186, ISO 21258, EN 14626, ASTM D3162, ASTM D546, and the Chinese national standard GB 9801, specify both the reference measurement methods and the performance criteria for interference rejection. Understanding and quantifying these interference factors is essential for laboratories, air quality monitoring agencies, and industrial facilities that must demonstrate the metrological traceability and defensibility of their CO monitoring data.
Testing Samples
Ambient Air Samples
Ambient air is collected at urban monitoring stations, suburban background sites, traffic-adjacent locations, and regional background observatories. Continuous in-situ monitoring systems draw air directly into the analyzer, while grab samples can be collected in passivated stainless-steel canisters or Tedlar-type bags for subsequent laboratory analysis. Ambient CO levels are generally low, making the signal-to-interference ratio a critical concern.
Stationary Source Emission Samples
Flue gas samples from combustion sources such as boilers, furnaces, incinerators, and engines contain CO at concentrations ranging from a few ppm to several thousand ppm. These matrices are characterized by high moisture content, elevated CO2 levels, and the presence of nitrogen oxides, sulfur dioxide, and hydrocarbons, all of which act as potential interference sources. Samples are typically extracted through heated probe lines and conditioned before analysis.
Indoor Air Samples
Indoor environments, including residential buildings, parking garages, tunnels, and industrial workshops, are sampled to assess occupational and public exposure. Indoor CO originates from cooking appliances, unvented heaters, tobacco smoke, and infiltration of vehicle exhaust. The wide variation in humidity and volatile organic compound (VOC) content in indoor air introduces interference challenges distinct from those of outdoor monitoring.
Calibration and Quality Control Gases
Certified CO standard gas mixtures in synthetic air or nitrogen, balance-gas blanks, zero air, and interference-test gases containing known concentrations of CO2, H2O, CH4, N2O, H2, SO2, and NO are used to characterize and verify the interference rejection performance of measuring systems. The matrix compatibility between calibration gases and actual samples is itself a significant interference-related variable.
Testing Items
A comprehensive interference analysis for atmospheric CO measurement typically covers the following test items:
- CO concentration determination in ambient air, indoor air, and source emissions, reported in mg/m3, ppmv, or ppbv
- Cross-sensitivity testing against known interferents including water vapor, carbon dioxide, methane, nitrous oxide, hydrogen, acetylene, and other hydrocarbons
- Zero drift and span drift evaluation over defined time intervals
- Temperature coefficient and pressure coefficient determination of the analyzer response
- Linearity verification across the calibrated measurement range
- Detection limit and quantification limit assessment under realistic sample matrices
- Response time and lag time measurement, including sampling system residence effects
- Sample loss and memory effect evaluation for canisters, bags, and transfer lines
- Uncertainty budget compilation incorporating all identified interference contributions
Testing Methods
Non-Dispersive Infrared Spectrometry (NDIR)
NDIR is the most widely deployed technique for continuous CO monitoring, as specified in ISO 12039 for stationary sources, EN 14626 for automated ambient air systems, ASTM D3162, and GB 9801. The method relies on absorption of infrared radiation by CO in the 4.6 to 4.7 micrometer wavelength region. Principal interference factors include:
- Water vapor: broadband IR absorption and light scattering can generate positive bias; most instruments incorporate gas-by-gas conditioning, optical filtering, or algorithmic compensation
- Carbon dioxide: absorption bands centered near 4.3 micrometers partially overlap the CO filter bandpass, producing cross-response in high-CO2 matrices such as flue gas
- Nitrous oxide and methane: weak but measurable absorption overlap requiring reference-cell or filter correlation correction
- Optical contamination: dust or condensate on cell windows attenuates the beam and mimics concentration changes
- Source and detector aging: gradual intensity drift that must be corrected through frequent zero and span checks
Gas Chromatography with Flame Ionization Detection (GC-FID) and Methanizer
The reference method described in ISO 8186 and ISO 21258 separates CO chromatographically, then catalytically hydrogenates it to methane over a heated nickel catalyst before FID detection. Because separation precedes detection, spectral interference is effectively eliminated; however, interferences shift to the chemical domain:
- Catalyst poisoning by sulfur compounds, chlorinated species, or high CO2 loads reduces methanization efficiency and biases results low
- Co-eluting hydrocarbons and oxygenates can mask or distort the CO-derived methane peak
- Incomplete separation from ambient methane peaks requires careful column selection and temperature programming
- Carrier gas purity and detector flame stability influence baseline noise and detection limits
Electrochemical Sensor Analysis
Potentiostatic electrochemical cells are common in portable monitors and personal dosimeters. Their compactness is offset by pronounced cross-sensitivity:
- Hydrogen: typically the dominant interferent, producing response equivalent to tens of percent of the same CO concentration
- Sulfur dioxide, nitrogen oxides, hydrogen sulfide, alcohols, and silanes: alter the working electrode reaction and cause either positive or negative bias
- Temperature and relative humidity: change electrolyte conductivity and gas diffusion rates, requiring environmental compensation algorithms
- Sensor aging: electrolyte depletion and electrode passivation cause sensitivity loss and baseline drift over months of service
Fourier Transform Infrared Spectroscopy (FTIR)
FTIR provides multi-component capability with classical least-squares or multivariate spectral deconvolution. It resolves overlapping H2O, CO2, and CO features quantitatively, but residual interference arises from inadequate spectral libraries, pressure-broadening mismatches between reference and sample spectra, and detector non-linearity at high absorbance.
Laser Spectroscopic Methods (TDLAS, QCLAS, CRDS)
Tunable diode laser and quantum cascade laser absorption spectroscopy, along with cavity ring-down techniques, target isolated rotational-vibrational CO lines with extremely high selectivity and sub-ppb sensitivity. Residual interference factors include spectral line overlap with H2O or CO2 lines at elevated pressure, etalon fringing, and path-length calibration errors.
Hopcalite Oxidation and Gas Filter Correlation Methods
Classical hopcalite (manganese-copper oxide) catalyst methods convert CO to CO2 and infer CO from the heat of reaction or CO2 formation. Interferences include catalyst deactivation by humidity and sulfur compounds, and the CO2 background of the sample itself. Gas filter correlation NDIR instruments improve selectivity by comparing sample spectra against pure CO reference paths, substantially suppressing broadband H2O and CO2 interference.
Sampling System Interference Control
Regardless of the detector technology, sampling and conditioning subsystems must be evaluated:
- Heated sample lines prevent condensation that both removes analyte and deposits corrosive interferents
- Inert materials such as PTFE, Sulfinert-treated stainless steel, or passivated canisters minimize adsorptive losses and CO off-gassing from polymer surfaces
- Particle filters prevent optical and sensor fouling without introducing reactive media
- Leak testing guards against dilution with ambient air, which itself contains background CO
- Drying strategies must be validated because permeation dryers and some desiccants can alter the sample matrix or introduce contaminants
Applications
- Urban and regional ambient air quality networks fulfilling national monitoring obligations under GB 3095 and comparable international frameworks
- Continuous emission monitoring systems (CEMS) at power plants, waste incinerators, cement kilns, and industrial boilers under ISO 12039 and ISO 21258
- Occupational health and workplace exposure assessments in tunnels, mines, warehouses, and enclosed parking facilities
- Indoor air quality investigations following cooking, heating, and combustion incidents
- Engine and vehicle exhaust characterization for emissions certification and research
- Instrument qualification and type-approval testing of automated measuring systems against EN 14626 performance criteria
- Laboratory method validation, interference studies, and measurement uncertainty evaluation for accreditation purposes
- Atmospheric research including biomass burning plume tracking and long-range transport studies of trace CO
FAQ
Why is water vapor the most common interference in CO measurement?
Water vapor absorbs strongly and broadly across the infrared spectrum, including the 4.6 micrometer region used for CO detection, and it is present in ambient air at concentrations typically thousands of times higher than CO. Even a small fractional overlap therefore produces a significant apparent CO signal. Condensed water additionally scatters light and damages sensors, which is why heated lines, drying, and algorithmic humidity compensation are standard countermeasures.
How does CO2 interfere with NDIR CO analyzers?
The intense CO2 absorption band centered near 4.3 micrometers lies adjacent to the CO bandpass filter window. In high-CO2 matrices such as combustion flue gas, the wings of the CO2 band and filter leakage allow CO2 radiation absorption to be registered as CO, creating a positive bias. Gas filter correlation, narrowband interference filters, and matrix-matched calibration are used to suppress this effect, as required by ISO 12039 performance testing.
Do electrochemical CO sensors respond to hydrogen?
Yes. Hydrogen is electrochemically active at the same working electrode potential as CO in most sensor designs, and cross-sensitivity of 10 to 60 percent is not uncommon. Applications near fuel cells, battery charging areas, or hydrogen-containing processes require either hydrogen-compensated sensor designs, selective filtration membranes, or parallel hydrogen measurement with mathematical correction.
Can drying a gas sample remove CO along with the water?
CO is a stable, non-polar molecule with negligible solubility in dryers and condensate under normal conditions, so physical drying generally does not remove CO. However, aggressive chemical desiccants, permeation dryers with degraded membranes, or very cold traps can alter matrix composition or introduce contamination, so any conditioning step must be validated by recovery experiments using certified CO standards.
How are interference effects quantified during method validation?
Accredited third-party laboratories introduce certified interferent gases at concentrations representing worst-case sample matrices, record the analyzer response, and calculate an interference factor as apparent CO concentration divided by interferent concentration. The results feed into the measurement uncertainty budget in accordance with ISO methodology and the type-approval test procedures of EN 14626 and ISO 12039.
What practical measures minimize CO measurement interference in the field?
Effective countermeasures include routine zero and span calibration with certified gases, periodic multi-point linearity checks, temperature-controlled installation, heated and leak-tested sample lines, scheduled filter and catalyst replacement, automatic data validation flags, and regular interference verification using known interferent concentrations. Combined, these controls keep total interference-induced bias well within the tolerance required by air quality and emission standards.