How chemiluminescence detection works
The chemiluminescence detector (CLD) is the standard instrument for measuring nitrogen oxides in emission monitoring, engine testing, and process gas analysis. Its measurement principle relies on a specific photochemical reaction:
NO + O₃ → NO₂* + O₂
Nitric oxide reacts with ozone to produce electronically excited nitrogen dioxide (NO₂*), which emits infrared photons as it returns to ground state. The light intensity is directly proportional to the NO concentration — a clean, linear, and highly sensitive measurement.
To measure total NOx, most CLD instruments include a converter — typically a heated molybdenum or stainless steel catalyst — that reduces NO₂ back to NO before it enters the reaction chamber. The instrument alternates between direct (NO only) and converter (total NOx) modes to report both values.
Where ammonia creates the error
The converter is the weak point. Molybdenum and carbon converters operate at 300–400 °C — conditions under which ammonia is catalytically oxidised to NO:
4 NH₃ + 5 O₂ → 4 NO + 6 H₂O
The NO produced from ammonia is indistinguishable from the NO already present in the sample. When the instrument measures in NOx mode (converter active), it reports the sum of genuine NO and ammonia-derived NO as total NOx. The direct NO channel is unaffected — but since regulatory reporting typically requires total NOx, the converter channel is the one that matters.
Quantifying the error
The conversion efficiency of ammonia to NO in the CLD converter depends on catalyst type, temperature, and age. Typical values:
| Converter Type | NH₃ → NO Conversion | Comment |
|---|---|---|
| Molybdenum (new) | 60–80% | Most common in CEMS applications |
| Molybdenum (aged) | 40–70% | Conversion drops with catalyst deactivation |
| Carbon / graphite | 50–90% | Higher variability, temperature-sensitive |
| Photolytic (UV) | < 5% | Minimal ammonia interference, but lower NO₂ conversion |
Practical impact
A coal-fired power plant with SCR measuring 180 mg/m³ NOx at the stack, with 10 ppm ammonia slip and a molybdenum converter at 70% NH₃-to-NO efficiency, reports approximately 195 mg/m³ NOx — an 8% overstatement. If the emission limit is 200 mg/m³, the plant appears to have only 5 mg/m³ margin when it actually has 20 mg/m³.
Why software corrections are unreliable
Some analyser manufacturers offer software-based ammonia compensation. These corrections require either a separate ammonia measurement (adding another instrument) or an assumed conversion factor. Both approaches have significant limitations:
- Conversion efficiency changes as the converter ages — a fixed correction factor drifts over time
- A separate NH₃ analyser adds cost, complexity, and its own maintenance requirements
- Software corrections are not always accepted by regulatory authorities for compliance reporting
The more robust solution is to remove ammonia from the sample before it reaches the converter — eliminating the error at its source rather than compensating for it after the fact.
Removing ammonia with acid wash filtration
Acid wash filter elements use phosphoric acid (H₃PO₄) impregnated into borosilicate glass microfibre to chemically neutralise ammonia:
NH₃ + H₃PO₄ → (NH₄)H₂PO₄
The reaction is selective: phosphoric acid reacts specifically with basic gases (ammonia, amines) without absorbing NO, NO₂, SO₂, CO, or CO₂. The gas matrix remains intact — only the interferent is removed.
Key Takeaway
Where to position the acid wash element
The acid wash element must be installed upstream of the converter and ideally upstream of the sample cooler. In hot, humid flue gas, ammonia can react with SO₃ to form ammonium bisulphate — a sticky aerosol that deposits in cool sections of the sample line. By removing ammonia while the gas is still hot, the acid wash element also prevents these deposits.
For high-temperature extraction (above 150 °C), use S-type (silica binder) acid wash elements rated to 480 °C. For cold-side extraction or conditioned sample lines below 150 °C, K-type (PVDF binder) elements are the standard choice.
Verifying ammonia removal
The simplest verification is to compare the CLD's direct NO reading with its total NOx reading when the acid wash element is installed. In the absence of ammonia, the difference (NOx minus NO) represents only true NO₂. If this difference matches what you expect from the combustion process (typically 5–10% of total NOx), the ammonia removal is working correctly.
For quantitative verification, challenge the element with a known ammonia concentration and measure downstream with a tuneable diode laser (TDLAS) or FTIR analyser. R+F FilterElements provides application support for these validation tests on request.
Eliminate CLD ammonia interference
RF-AW acid wash elements remove ammonia from sample gas before it reaches the CLD converter — available in 3 sizes, 2 temperature variants, delivery within 2 weeks.
