An annular denuder–filter pack can separate reactive gases such as ammonia and nitric acid from the related fine-particle ammonium and nitrate that coexist in outdoor air. The result is not a single ‘ammonia level’: it is a set of measurements for particular gases and particle-phase ions collected on different parts of a controlled sampling train. Correct interpretation starts by identifying which component collected which species and whether gas–particle conversion occurred during sampling.
To translate an EPA IO-4.2-type report into a reliable air-quality explanation, check the size-selective inlet, sampling flow, chemical coatings, filter positions, blank corrections, laboratory extract volumes, analysis units and quality checks. Then use explicit denominators for concentration and explicit chemical forms for ammonium, ammonia, nitrate and nitric acid. A method designed to distinguish phases can be misunderstood completely if those species are combined under a vague heading.
This guide covers diffusion to coated walls, particle transmission, sample volume calculations, volatilisation and absorption artefacts, nitrogen-equivalent accounting, data-quality diagnosis and independent worked cases. Numerical examples are invented educational exercises unless linked to a named method. EPA’s 1999 IO-4.2 document described itself as a tentative compendium method requiring thoughtful validation; it should not be misrepresented as a universally binding contemporary compliance standard.
The invisible mixture that ordinary filters can confuse
Outdoor air contains gases and suspended particles at the same time. Ammonia, NH3, is a gas under typical ambient conditions. Ammonium, NH4+, is an ion often found within particles as part of salts. Nitric acid, HNO3, can exist in the gas phase, while nitrate, NO3−, can be part of particulate salts. They are chemically related but cannot be reported as though they were identical substances in the same phase.
Ammonia and nitric acid may react to form ammonium nitrate. The partitioning depends on temperature, relative humidity, chemical composition and available gases. Colder conditions can often favour particulate ammonium nitrate, while warmer or drier sampling conditions may encourage volatilisation. The sampling apparatus can disturb this balance. One reported high nitrate filter mass therefore does not necessarily represent every nitrate molecule that was present in airborne particle form before entering the apparatus.
A plain front filter may trap particles but allow gases to pass or adsorb them on its surface. Particle nitrate may evaporate from the collected deposit as gas-phase nitric acid; incoming gas may also react with collected material. Either process shifts where the chemistry appears in the analysed sample. A filter result can consequently differ from the original gas-particle division even when the pump volume calculation is flawless.
The denuder approach addresses a particular part of this problem. Before air reaches its particle filter, it passes along chemically coated walls designed to capture selected reactive gases. Fine particles, which diffuse to surfaces differently from gas molecules under the operating regime, can continue toward the filter. The method thereby attempts to measure the gas components separately and reduce their interference with particle analysis.
This is not a perfect membrane with a magical chemical sorting function. Particle deposition, incomplete gas uptake, chemical breakthrough, surface reactions and later filter volatilisation are possible. A trustworthy interpretation acknowledges these efficiency and artefact pathways. Saying ‘denuder equals gas and filter equals particles’ is a useful starting model, but not the complete quality-control argument.
Follow the physical sampling train
EPA’s IO-4.2 describes an inlet with an impactor or cyclone-type preseparator, coated annular denuders and a filter pack. The preseparator defines the fine-particle fraction entering the train. The denuders absorb selected gases on treated internal surfaces. The downstream membranes collect particulate components, while additional filter stages can help identify material redistributed during sampling.
A sampler may include alkaline-coated and acid-coated denuders serving different chemical targets. For example, a carbonate coating can collect acidic gases, while an acidic coating can retain basic ammonia. The exact chemicals, concentration, number and order of denuders must match the approved procedure. One cannot simply swap their positions or copy an older field setup from another investigation and assume analyte recovery is unchanged.
An annular denuder is useful because its geometry provides a large coated wall area along a defined flow path. Diffusion helps reactive gas molecules reach those walls, while many fine particles remain entrained in the flowing air and pass onward. This separation depends on residence time, flow regime, coating condition and particle behaviour. Exceeding a design flow can reduce the residence time for gas uptake and alter the intended collection.
The particle filter is not the final stage of interpretation. A downstream backup or chemically active filter may capture certain gases released from material collected on an upstream filter. The presence of nitrate or ammonium in a backup filter can therefore reflect a sampling artefact rather than material that was a separate ambient particle population. The meaning depends on the actual filter chemistry and the method’s correction scheme.
When reading the laboratory report, draw a simple diagram listing components in airflow order. Write ‘what enters’, ‘what this stage should retain’, ‘what may escape’ and ‘what could be generated during collection’ beside each element. This one-page flow map often resolves misunderstandings that a large table of laboratory numbers cannot.
Sample volume is the denominator for every channel
The first arithmetic conversion is sampled air volume. In a hypothetical uninterrupted one-hour run at 10 litres per minute, nominal volume equals 10 multiplied by 60, or 600 litres, equivalent to 0.600 cubic metres. If a validated and blank-corrected denuder extract corresponds to 1.2 micrograms of ammonia, its illustrative gas-phase concentration is 1.2/0.600, or 2 micrograms per cubic metre, provided all relevant recovery and flow conditions are met.
A second sample at 8 litres per minute for 90 minutes also collects 720 litres, or 0.720 cubic metres. Suppose it retains 0.72 micrograms of the target compound. The concentration is 1 microgram per cubic metre. Comparing collected masses alone would be misleading: 0.72 micrograms is smaller than 1.2, but the volume also differs. Normalised concentration always needs the actual accepted volume for the relevant interval.
A nominal pump setting is not proof that this volume entered the sample train. Flow can drift, leaks can admit extra air, and blocked filters can create changes in pressure. Use the method’s prescribed flow records and corrections. If measured flow fell during the run, a simple nominal-flow calculation may be biased. Reconstruct volume from acceptable readings or follow the method’s invalidation guidance rather than assigning an arbitrary average.
Some methods express sampled volume at stated standard temperature and pressure conditions. Others describe ambient volume. These are distinct volumetric reference bases. A microgram-per-cubic-metre result at standard conditions should not be compared silently with one at ambient conditions when differences matter. The EPA method specifies how standard volume is calculated; preserve that convention when applying its concentration equations.
The measured analyte mass may come from a solution extract. If the laboratory reports 0.20 micrograms per millilitre and the valid extract volume is 5 millilitres, that is 1 microgram of analyte recovered, before any necessary blank or recovery adjustments. Divide by 0.600 cubic metres and the corresponding illustrative concentration is approximately 1.67 micrograms per cubic metre. Dividing the liquid concentration directly by air volume without multiplying by extract volume would leave the wrong units.
Why ammonia is not the same as ammonium
A gas channel reported as NH3 and a particulate channel reported as NH4+ have different chemical formula masses. The molar mass of ammonia is approximately 17 grams per mole and ammonium is approximately 18 grams per mole. The nitrogen atom accounts for approximately 14 grams in each mole. If the aim is to calculate total reduced nitrogen in a common nitrogen-mass basis, each channel needs its own conversion factor.
For example, 5 micrograms per cubic metre of gaseous NH3 contains 5 times 14/17, approximately 4.12 micrograms of nitrogen per cubic metre. A particulate NH4+ concentration of 3.6 micrograms per cubic metre contains 3.6 times 14/18, or 2.8 micrograms of nitrogen per cubic metre. The hypothetical combined reduced-nitrogen concentration is about 6.92 micrograms N per cubic metre, assuming both channels are valid and represent the same air interval.
It would be inappropriate to add 5 and 3.6 and label the result ‘8.6 micrograms NH3 per cubic metre’. The arithmetic sum of the two species masses exists, but it is not a single chemical-species concentration. Expressing components on a nitrogen basis makes clear what was added. Alternatively, report each species separately and avoid a total unless the intended chemical and environmental question requires one.
The same caution applies to nitric acid and nitrate. Nitric acid has a formula mass of about 63 grams per mole; nitrate ion is about 62. Converting both to nitrogen basis uses the factor 14/63 for HNO3 and 14/62 for NO3−. If a data product labels every nitrate-related result as ‘NO3’, verify whether it actually means nitrate ion, nitric acid equivalent or total oxidised nitrogen expressed as N.
A useful mole calculation: ammonium nitrate, NH4NO3, has formula mass near 80 grams per mole. One mole contains an ammonium portion of approximately 18 grams and a nitrate portion of 62 grams. Therefore a hypothetical 10-microgram mass of pure ammonium nitrate comprises about 2.25 micrograms ammonium and 7.75 micrograms nitrate. The sum returns 10 micrograms only because these are parts of the same salt on a mass basis.
Volatilisation and collection artefacts
Suppose ambient particles initially contain ammonium nitrate. During passage or collection, some of that salt volatilises into ammonia and nitric acid. A downstream filter or denuder stage may collect those released gases in a location that did not contain them as independent ambient gases before sampling. If the analyst assigns every downstream signal to ambient gas without considering filter loss, the gas-particle balance becomes distorted.
The reverse kind of artefact can occur when a gas adsorbs to a collection substrate or reacts with material already there. A particle filter may show an apparent gain in an ion that did not arrive as particle mass. In a humid environment, droplets and reactive particles can create additional pathways. Chemical composition and sampler configuration control the effect, so a correction valid in one campaign should not be assumed universal.
EPA’s IO-4.2 specifically discusses chemical pathways involving dissociation of ammonium nitrate from filters and interactions involving ammonium bisulfate. The method was developed to support estimation of atmospheric acidic and basic species while recognising these interfering reactions. Readers should consult the correction and filter-handling details rather than taking a chromatogram peak at face value as a direct ambient-phase count.
If a filter-based system and a denuder-filter system disagree, the first hypothesis should not be that one laboratory is careless. Different sampling artefacts may explain parts of the difference. Field comparison literature documents that volatilisation and gas retention can change apparent gas and particle fractions. A paired experiment should examine temperature, flow, filter chemistry and the specific compounds rather than applying a blanket judgement across all species.
A critical distinction is between total collected amount and partitioned amount. Some gas–particle exchanges inside the apparatus shift material between collection stages while leaving a related total closer to conservation. Other losses may escape the train or change chemistry in ways that corrupt even the total. Summing stages can therefore help diagnose redistribution but is not automatically a perfect reconstruction of ambient composition.
Coatings, breakthrough and blank corrections
A denuder’s coated wall has a finite capacity and chemical specificity. If concentration or exposure time produces loading beyond the validated range, gas can pass through without complete uptake. A second denuder or downstream check may reveal such breakthrough. A low apparent gas concentration in the first denuder does not prove clean air if the coating was exhausted or the flow exceeded specifications.
Blank measurements identify contamination from reagents, surfaces, filters and extraction steps. In a hypothetical example, a sample extract contains 1.0 microgram ammonium but an applicable procedural blank contains 0.1 microgram. Whether and how to subtract that background must follow the validated protocol. Unapproved blank subtraction can conceal contamination or artificially make a weak result appear precise.
Reagent purity matters because trace inorganic ions are common laboratory contaminants. Clean gloves, properly conditioned equipment, reliable water, correct storage and protected transport are central to the method. The analyst should record which batch and coating solution was used for each denuder. A sequence of elevated blanks from the same reagent batch can be more revealing than a single anomalous atmospheric concentration.
Ion chromatography separates dissolved ions within an extract, usually allowing measured nitrate, sulfate and ammonium quantities to be related to collected material. The laboratory must document calibration standards, peak identity, interferences and final extraction volume. A reported concentration in milligrams per litre of extract is not yet an ambient micrograms-per-cubic-metre result until extract volume, blank treatment and sampled air volume have been incorporated.
The working sample label must distinguish gas denuder, front particle filter and backup filter. If two vials are accidentally switched in the electronic result table, a scientifically plausible-looking concentration can be assigned to the wrong phase. Independent review of the chain-of-custody sheet, extraction notebook and stage-specific chemical expectations helps detect such transcription errors.
What an apparent nitrate shift might mean
Imagine a winter monitoring day where the particle filter indicates substantial nitrate and the ammonia-gas denuder records a moderate NH3 concentration. A warmer afternoon produces more gas-phase material and less apparent particle nitrate. This pattern is compatible with temperature-dependent partitioning, but other chemical drivers and sampling artefacts must be considered before claiming that a specific reaction caused the full difference.
The field record should include ambient temperature, relative humidity, pressure and flow alongside the sampling times. Where measurements integrate over hours, a single meteorological observation at the start is inadequate to describe all conditions. Wind and source influence also change over time, so a measured gas concentration need not correspond to a single source activity.
Consider a front-filter nitrate value that appears unexpectedly low while a backup filter has a large signal. An evaporation pathway may be plausible. Yet a backup-filter response can also arise from other causes, depending on materials and conditions. Review the method’s prescribed recovery calculation and performance criteria. A backup stage is a diagnostic part of a chemical sampling system, not an automatic additive ‘missing nitrate’ column.
If a reported particulate ammonium concentration is high but measured particulate sulfate and nitrate are low, do not instantly infer a stoichiometric impossibility. Other anions or measurement artefacts may matter. A charge-balance calculation can reveal whether measured major ions are mutually consistent, but it cannot prove every unmeasured ion absent. Convert measured concentrations to equivalents before discussing neutrality.
A well-organised dataset maintains the original measured form for every species as well as any derived common-basis totals. This allows a reader to reproduce both the atmospheric gas/particle interpretation and the nitrogen budget. Hiding the originals behind one final ‘total ammonia’ number makes it hard to distinguish a true chemical conclusion from a bookkeeping convention.
Work a comparison with two sampling intervals
A hypothetical morning sample runs for one hour at a validated 12 litres per minute and yields 1.44 micrograms corrected particulate ammonium. Volume is 720 litres or 0.720 cubic metres, so concentration is 2 micrograms NH4+ per cubic metre. An afternoon sample runs for ninety minutes at the same flow and yields 2.16 micrograms. Its volume is 1.08 cubic metres and its concentration is also 2. Both samples show the same normalised value despite different collected masses.
Now suppose the afternoon gas denuder retains 3.24 micrograms of NH3 over the same 1.08 cubic metres. Its reported gas concentration is 3 micrograms NH3 per cubic metre. On nitrogen basis, this is 3 times 14/17, or about 2.47 micrograms N per cubic metre. The particulate ammonium contributes 2 times 14/18, or about 1.56. The combined reduced-nitrogen concentration is around 4.03 micrograms N per cubic metre for the afternoon, assuming all channels are valid.
A common spreadsheet error converts gas NH3 using 18 instead of 17 because the user has been working with ammonium. At 3 micrograms NH3 per cubic metre, the incorrect conversion gives 3 times 14/18, or 2.33 micrograms N, compared with the correct approximate 2.47. The difference is not huge here, but consistent speciation requires that every formula match its declared molecular species.
There is also a potential mistake in averaging. If the morning and afternoon periods have different accepted durations, do not take a simple mean of their phase concentrations as though they sampled equal time. Weight by accepted interval length for a time-average objective, provided phase measurements and quality controls are comparable. If either period lacks valid sampling volume, no amount of weighting rescues it.
The report should make clear whether its gas/particle ratio compares species masses, moles or nitrogen equivalents. For example, dividing 3 micrograms NH3 by 2 micrograms NH4+ yields 1.5 as a species-mass ratio, not a molar ratio. A molar calculation divides by 17 and 18 grams per mole, producing a different dimensionless comparison. Both can be useful, but they answer different definitions of ‘ratio’.
Common questions answered carefully
Does a denuder measure PM2.5 mass? Its purpose in this application is to separate and quantify selected gases and particle chemical constituents; it is not automatically a total PM2.5 gravimetric mass determination. The stage and method matter. A high ammonium concentration can coexist with a low total particle mass only within the constraints of chemistry and measurement, and the measured species must be interpreted in their shared time and volume context.
Can a front filter reading prove that nitrate was all in particles outdoors? No. Collected nitrate may be affected by gas uptake or losses from salts during sampling. The denuder and backup stages are designed partly to diagnose and address such changes. Their interpretation requires the complete method and validated corrections, not isolated readings.
Can one call an ammonium value ‘ammonia’ for simplicity? That loses the phase distinction and the difference in chemical formula mass. Use ammonia for NH3, ammonium for NH4+, and say explicitly when combining them on a nitrogen basis. A small vocabulary difference can prevent a substantial scientific misinterpretation.
Is a zero gas-channel result proof there is no ammonia? Not necessarily. A below-detection result may have reporting limitations; breakthrough, coating failure, leaks or missing exposure can also affect results. Inspect method sensitivity and sampling controls before turning a zero-looking number into an absolute environmental claim.
What is the transferable habit? Read a sampling train as a sequence of physical and chemical transformations. Mark where particles are separated, where gases are captured, which reactions might occur on a surface, how analyte mass becomes an ambient concentration and where an independent control tests the inference. That reasoning generalises to water filters, analytical cartridges and laboratory separations.
Diagnose whether a collection channel is trustworthy
A denuder that seems to capture no nitric acid may actually be performing badly rather than observing zero ambient gas. To investigate, examine the documented inlet, flow stability, coating application and extraction recovery. A low amount on an upstream coated section alongside an unexpected increase on a downstream section can suggest incomplete capture or breakthrough, but stage chemistry and approved validation determine how far that inference can go.
A significant amount on a blank-coated component is another warning. The blank may reveal reagent contamination, handling problems or trace species introduced during extraction. A field sample and its blank do not necessarily experience exactly the same physical exposure, so the method defines which background measurements and corrections are appropriate. Preserve the blank’s original result rather than erasing it after subtraction.
The analyst should distinguish instrument detection from environmental absence. If the laboratory quantification limit for ammonia is high relative to the sample, a printed zero may represent an undetectable or unquantifiable amount. Some systems use codes, inequalities or blank cells for those situations. Importing them into a spreadsheet as numerical zeros changes the distribution, especially where many observations are near the laboratory’s lower range.
The coating volume and surface treatment should be recorded as part of the method history. Two denuders that look similar may have different gas-collection properties when their coatings differ. A reader should not conclude that all glass tubes are functionally interchangeable. A successful method requires controlled chemistry as well as controlled geometry.
Sample filtration creates pressure differences that may evolve as particulate material accumulates. If the pump does not maintain the method’s accepted flow, the volume used for concentration normalisation may become wrong. This is especially consequential for long sampling periods or dusty locations. Reviewing the logged flow and leak tests is more informative than checking whether the pump was audible during a field visit.
Work through equivalent and charge-balance reasoning
Chemists often examine ionic charge as well as mass. Sulfate carries two negative charges per ion, nitrate one negative charge, and ammonium one positive charge. Comparing their concentrations in micrograms per cubic metre without conversion to moles or charge equivalents can create a false impression of imbalance. The correct procedure divides each mass concentration by its formula mass and multiplies by the ion’s charge magnitude.
For a deliberately simplified particle mixture containing 18 micrograms/m³ NH4+ and 62 micrograms/m³ NO3−, the approximate molar concentrations are one micromole/m³ of each. Their charges balance: one micromole of positive ammonium and one micromole of negative nitrate. This is consistent with the stoichiometry of ammonium nitrate under the specified simplification. It does not prove the real aerosol contains only this salt or has no unmeasured ions.
Consider a separate sample with 36 micrograms/m³ NH4+ and 96 micrograms/m³ sulfate ion, SO4²−. Ammonium equals about two micromoles/m³; sulfate equals about one micromole/m³ using formula mass near 96 grams per mole. Sulfate carries two negative charges, so total positive and negative equivalents can balance. The ordinary mass ratio 36/96 is not the correct charge comparison.
An invalid calculation would add 36 and 96 and infer that the collected air contained 132 micrograms/m³ of ‘neutral ammonium’. It contains measured ionic species under the method definitions. Their association as particular chemical salts requires a model or additional evidence. Charge balance is an independent plausibility test, not a complete molecular identification method.
If the charge equivalents do not balance, examine missing anions or cations, gas contamination, analytical recovery, instrument interference and reporting bases. Poor closure can motivate further investigation but does not automatically identify which ion is wrong. A species may be omitted from the analytical programme altogether. Review the entire sampled composition before imposing a chemical correction that forces artificial neutrality.
Evaluate changing partitioning with a paired example
Imagine valid winter and summer samples from a hypothetical station. Winter gas NH3 is 2 micrograms/m³ and particulate NH4+ is 4; summer gas NH3 is 6 and particulate NH4+ is 2. Both pairs change, but their species mass sums do not by themselves represent one conserved ammonia species. To compare on a common nitrogen basis, convert winter to 2 times 14/17 plus 4 times 14/18, around 4.76 micrograms N/m³.
For the summer values, 6 times 14/17 plus 2 times 14/18 gives approximately 6.50 micrograms N/m³. The overall reduced-nitrogen concentration increased in the hypothetical data rather than simply shifting between phases at constant total. That distinction matters. A report claiming ‘only temperature changed partitioning’ would overlook evidence that sources, transport or total reduced nitrogen also changed.
The gas fraction on nitrogen basis can also be calculated. In winter it is approximately 1.65/4.76, or 34.6 per cent, while in summer it is 4.94/6.50, or 76.0 per cent. The calculation is transparent only when the numerator, denominator and species basis are declared. A gas fraction constructed from raw 2/(2+4) or 6/(6+2) has a slightly different chemical convention.
Such seasonal examples are teaching devices, not observed evidence of any particular agricultural or urban area. Real data could be influenced by emissions, atmospheric acids, meteorology, transport and sampler artefacts. To test the claim that warming shifted the gas-particle equilibrium, add compatible nitrate, nitric-acid, temperature and humidity measurements. Seasonal comparison alone is not mechanistic proof.
Distinguish method limitation from an environmental trend
The EPA IO-4.2 compendium document reflects an established methodological effort but explicitly describes limitations and need for evaluation. This historical status matters when deciding whether a field dataset may be used for a particular regulatory decision. A researcher may use a validated modern adaptation, but should name exactly which protocol and modifications were applied rather than claim every annular-denuder experiment follows one immutable standard.
A network may retain historical measurements for long-term analysis while instruments and sample preparation evolve. If coatings, filter media, extraction methods or laboratory detection limits changed, a discontinuity in the data could be methodological. Compare the transition date to the monitoring protocol before asserting a real atmospheric shift. A method-change flag is often as important as the plotted concentration difference.
Meteorological representation deserves scrutiny. A gas-phase ammonia result averaged over twelve hours is not directly comparable to a five-minute humidity observation as though both represent the same period. Time-integrated chemistry and time-resolved weather should be aligned according to the analysis question. An abrupt temperature change can affect equilibrium and sampler conditions, but a single snapshot cannot describe a full shift.
A denuder system used near concentrated agricultural emissions may face different loading and breakthrough challenges from one used at a remote rural site. A design fit for low ambient concentrations may require modified sampling duration or other validated controls at higher concentrations. The correct scientific response is to adapt and validate the method, not assume that a plausible-looking number demonstrates unchanged performance.
Troubleshooting cases with precise answers
Case one: a particle filter reports 0.90 micrograms of nitrate over an accepted 0.30 cubic metres. The concentration is 3 micrograms NO3−/m³, before any method-specific artefact correction. If an analyst writes 3 micrograms HNO3/m³, the formula has changed without a stated conversion. On an equivalent nitrogen basis, nitrate contributes 3 times 14/62, approximately 0.677 micrograms N/m³.
Case two: the laboratory extracts a denuder with 10 millilitres of solution and reports an ammonia-equivalent mass concentration of 0.12 micrograms per millilitre. Recovered amount is 1.2 micrograms. If sampled volume was 0.40 cubic metres, the result is 3 micrograms/m³ on the stated ammonia mass basis. The extract volume must be multiplied before dividing by sampled air volume, and the exact analytical chemical basis must be retained.
Case three: a technician reports that front-filter nitrate decreased while the backup-filter signal increased. That pattern is consistent with possible re-volatilisation but is not uniquely diagnostic. The reader should check temperature, species captured by each substrate, blank behaviour and the method’s recovery calculation. Without those checks, adding the filters may overestimate or misclassify the original particle mass.
Case four: two colocated systems differ by fifteen per cent. A correlation coefficient close to one over a week would show that they track changes similarly, not that they agree without bias. Examine paired relative differences, flow records and chemical-stage handling. Their agreement may vary across concentrations or humidity regimes; a single summary statistic can hide those dependencies.
Case five: a laboratory provides 8 micrograms/m³ for ‘ammonia and ammonium’ but no definition of its chemical mass basis. The proper question is whether this refers to NH3 mass, NH4+ mass, nitrogen equivalents or arithmetic addition of distinct reported species. The number cannot be responsibly compared with a paper reporting ‘total NHx as N’ until the basis is made explicit.
Case six: a reader interprets gas NH3 greater than particulate NH4+ as proof of fresh agricultural emissions. That observation may be consistent with several conditions but is not source-specific. Temperature, acidity, transport, analytical quality and other emissions all influence the concentrations. Source attribution needs additional evidence rather than a gas/particle ratio alone.
Building an auditable environmental explanation
Every published concentration should be traceable to a sample identifier, location, time interval, stage of the sampling train, measured extracted mass, blank handling, recovery checks and accepted air volume. Provide concentration units and chemical formula in the table heading. A separate note should explain whether values have been converted to a common nitrogen basis and whether gas-particle artefacts were corrected.
It is better to leave an invalid sample visibly missing than to fill it with a fabricated value for a smoother plot. Likewise, do not merge distinct phases or ions because the public report has limited column space. A short table with NH3 gas, NH4+ particle, HNO3 gas and NO3− particle can communicate much more honestly than one unexplained column labelled ‘ammonia pollution’.
Finally, an instrument’s analytical sophistication does not erase the need for plain language. Explain that gas molecules diffuse to treated walls, particles continue to a filter, compounds can change phase while being collected, and laboratory measurements require conversion into air concentration. Give the reader a worked denominator and a question they can independently check. This is how a complicated chemical sampling system becomes useful evidence rather than a mysterious set of lab numbers.
Primary and connected reading
US EPA, official inorganic ambient-air methods compendium index.
Effects of filter type and temperature on volatilisation losses of ammonium salts, EPA HERO record.
eduKate guide to TEOM particle-mass readings and semivolatile loss.
eduKate guide distinguishing aerosol number concentration from particle mass.
eduKate approach to mathematical units, diagnosis, worked examples and checks.