Toxic Gas Adsorbers: A Compound-by-Compound Guide to Choosing the Right Dry Adsorption Media

Not every toxic or corrosive gas is captured the same way, and this is the single most common mistake in specifying a toxic gas adsorber: treating “activated carbon” as a single, interchangeable answer for hydrogen sulfide, ammonia, chlorine, sulfur dioxide, mercaptans, and VOCs alike. It isn’t. Plain activated carbon captures many organic compounds well through physical adsorption, but several of the most dangerous industrial gases barely adsorb onto untreated carbon at all — they need a specific chemical impregnation engineered for that molecule. Get the media wrong and a toxic gas adsorber can look correctly sized on paper — right airflow, right vessel, right contact time — and still fail to protect the people, equipment, and compliance status it was installed for.

This guide goes through the major toxic and corrosive gas classes one at a time: what each one is, how dangerous it actually is in quantified terms, and which adsorption chemistry actually captures it. It’s a companion to our broader guide on gas phase filtration, which covers the standards and compliance side of this problem; this piece goes deeper into the chemistry and media-selection side.

Physisorption Alone Is Not Enough

Activated carbon captures gas molecules two ways, and the difference matters enormously for toxic and corrosive gas duty.

Physical adsorption (physisorption) holds molecules on the carbon’s internal surface through weak van der Waals forces. It works well for larger, non-polar organic molecules — most VOCs, solvent vapours, many odour compounds — and it’s reversible with heat, which is why physisorption-based systems can be regenerated.

Chemical adsorption (chemisorption) requires the carbon to carry a reactive chemical impregnation that the target gas reacts with, converting it into a stable compound that stays on the media rather than simply clinging to it. Small, polar, reactive molecules — hydrogen sulfide, ammonia, chlorine, sulfur dioxide — pass through untreated activated carbon with disappointingly low capacity, because physisorption alone can’t hold onto them well. They need the right impregnation, and different gases need different impregnations. This is the central fact this whole guide is built around.

Hydrogen Sulfide (H₂S)

Where it comes from: wastewater and sewage treatment, anaerobic digestion, pulp and paper, oil and gas processing, and any process involving decomposing organic matter.

How dangerous it actually is: NIOSH sets the Immediately Dangerous to Life or Health (IDLH) concentration for hydrogen sulfide at 100 ppm — the level at which a worker can no longer be assumed able to escape without irreversible harm. What makes H₂S unusually hazardous operationally is olfactory fatigue: at high concentrations it paralyses the sense of smell, so a worker can lose the warning odour cue at exactly the concentration where it matters most.

The right media: potassium hydroxide (KOH) impregnated activated carbon is the standard chemisorption media for H₂S. The alkaline impregnation reacts with H₂S to form potassium sulfide and potassium sulfate, both stable, non-volatile compounds that stay bound to the carbon rather than being released. This is a consumptive, non-reversible reaction — the media has a genuine, calculable capacity, not an indefinite one.

Ammonia (NH₃)

Where it comes from: fertiliser production, refrigeration (ammonia is still widely used as an industrial refrigerant), livestock and poultry facilities, and wastewater treatment.

How dangerous it actually is: NIOSH’s current IDLH for ammonia is 300 ppm. Ammonia is also flammable within a fairly narrow range at high concentrations, which is a separate hazard consideration from its toxicity.

The right media: ammonia is a base, so the effective impregnation chemistry is the mirror image of what works for H₂S — an acid, not an alkali. Phosphoric-acid-impregnated activated carbon is the established chemisorption media for ammonia, documented across multiple patents and engineering literature, including modelling work developed for NASA life-support applications and commercially available products such as Calgon Carbon’s Ammonasorb line. The acid reacts with ammonia gas to form a stable ammonium salt bound to the carbon surface. Using a KOH-impregnated carbon bed (correct for H₂S) on an ammonia stream would be actively counterproductive — an alkaline media has essentially no chemical affinity for an alkaline gas.

Chlorine and Chlorine Compounds

Where it comes from: water and wastewater disinfection, pulp bleaching, chemical manufacturing, and any process using chlorine gas or hypochlorite chemistry.

How dangerous it actually is: NIOSH’s current IDLH for chlorine is 10 ppm — among the lowest IDLH values of any commonly encountered industrial gas, reflecting how severely and quickly it damages lung tissue even at brief exposure.

The right media: chlorine and related oxidising gases need a specialty impregnation engineered specifically for oxidiser chemistry, distinct from both the KOH (alkaline) chemistry used for H₂S and the acid impregnation used for ammonia. This is a case where generic “odour-control carbon” bought off the shelf without specifying the target gas can be a genuinely dangerous substitution — the wrong impregnation family doesn’t just underperform on chlorine, it can react unfavourably with it.

Sulfur Dioxide (SO₂)

Where it comes from: combustion of sulfur-containing fuels, metal smelting and refining, and various chemical processes.

The right media: SO₂ chemisorption uses alkaline-impregnated activated carbon, broadly similar in principle to H₂S media but formulated for SO₂’s specific reaction chemistry. Because SO₂ and H₂S are both acidic gases that respond to alkaline impregnation, some multi-gas media formulations target both simultaneously — worth specifying explicitly if your stream carries both, rather than assuming one media handles both equally well by default.

Mercaptans and Organosulfur Odour Compounds

Where it comes from: the same wastewater, landfill, and organic-decomposition sources that generate H₂S, plus natural gas odorant (mercaptans are deliberately added to naturally odourless natural gas as a safety warning agent).

The right media: mercaptans respond to oxidising impregnation chemistry, and their extraordinarily low odour detection thresholds — often in the low parts-per-trillion range — mean even a well-performing adsorber leaves some residual detectable odour unless the media and contact time are matched specifically to the compound in question, not just to a generic “odour control” specification.

Volatile Organic Compounds (VOCs) as a Class

Where it comes from: paint and coating operations, printing, solvent-based processes, degreasing, and chemical manufacturing.

The right media: most VOCs — aromatic and aliphatic solvents, ketones, esters — are captured well by standard, non-impregnated activated carbon through physisorption alone, which is why VOC adsorption is the one contaminant class in this guide where “activated carbon” as a generic answer is often correct. The real engineering choice for VOCs isn’t usually which impregnation to use, but whether the system should be regenerative (steam or hot-air stripped for solvent recovery, economical at higher, continuous loading) or non-regenerative (replaced when saturated, more economical at lower or intermittent loading). See our dry VOC adsorber guide for that specific decision.

Ethylene: A Different Chemistry Entirely

Ethylene deserves a special note because it breaks the pattern of this whole guide. It’s a small, highly reactive molecule that carbon-based adsorption — physisorption or chemisorption — captures poorly regardless of impregnation. Controlled-atmosphere fruit and produce storage instead uses potassium permanganate (KMnO₄) oxidation, which chemically breaks down the ethylene molecule rather than adsorbing it onto a carbon surface. If ethylene is part of your gas stream — common in food storage and ripening-adjacent facilities — that’s a truly different product, not a variant of a carbon adsorber. See our ethylene scrubber guide.

The Humidity Variable: Why Moisture Changes Adsorption Capacity

This is a factor that’s easy to overlook and directly relevant given how often gas-phase filtration and moisture control show up in the same facility. Water vapour competes with target gas molecules for the same adsorption sites on activated carbon, and the research on this is specific and, for some compounds, surprising.

A study on VOC removal from humidified gas streams using activated carbon cloth found that water vapour had little effect on adsorption capacity for acetone even up to 90% relative humidity, but for benzene, adsorption capacity held steady only until around 65% RH, beyond which it dropped rapidly as capillary condensation of water inside the carbon’s pores began physically excluding benzene molecules from the same sites (Cal, Rood & Larson, *Gas Separation and Purification*, 1996). A separate study on trichloroethylene found its adsorption capacity on hydrophobic activated carbon was largely unaffected by humidity up to 80% RH (Lee et al., *Adsorption*, 2005).

The practical takeaway: humidity’s effect on adsorption capacity is compound-specific, not a fixed derating factor you can assume applies equally everywhere. A media bed sized correctly for a dry process stream can underperform significantly if the same gas arrives in humid air — and by how much depends on which compound you’re capturing, not just how humid the air is. If your gas stream runs consistently above roughly 60–65% RH, this is worth flagging explicitly when specifying media, not discovered after the bed underperforms. Where a facility is fighting both a humidity problem and a toxic-gas problem in the same airstream, see our moisture removal and dehumidification guide — treating both together is usually more effective than treating either in isolation.

Multi-Stage and Multi-Media Systems

Real industrial exhaust streams rarely carry just one contaminant. A wastewater treatment headworks stream might carry H₂S and mercaptans together; a chemical process vent might carry VOCs alongside an acid gas. Where more than one contaminant class is present, layering different media in sequence — for example, a KOH-impregnated stage followed by a general-purpose activated carbon stage — usually outperforms trying to find one universal media that does an adequate job on everything, because each impregnation chemistry is meaningfully optimised for one class of reaction, not several at once.

Signs Your Media Is the Wrong Match, Not Just Exhausted

When a toxic gas adsorber underperforms, the instinct is usually to assume the media has simply run out and needs replacing sooner. Sometimes that’s correct. But a mismatch between the media and the actual gas produces a different pattern, worth recognising because the fix is different — changing media chemistry, not changing the replacement schedule.

• Odour or gas breakthrough happens almost immediately after a media change, well before any reasonable capacity calculation would predict exhaustion. This points to the media having little real affinity for the actual target gas — a classic sign of, for example, standard non-impregnated carbon installed on an ammonia or H₂S stream, where physisorption alone provides only weak, short-lived capture.
• Spent media looks and smells essentially unchanged from fresh media. Chemisorption media that has actually reacted with its target gas typically shows visible colour change or a distinct residual odour tied to the reaction products (KOH-impregnated carbon spent on H₂S, for instance, characteristically darkens and takes on a sulfurous smell from the sulfide/sulfate reaction products). Media that looks untouched after months in service, on a stream that should have loaded it, is a sign it was never really capturing the target gas.
• Performance is fine on dry days and degrades on humid ones, tracking weather rather than runtime. This points to a humidity-sensitive compound (see above) running near or above the RH threshold where competitive adsorption starts cutting into working capacity, rather than a media-chemistry mismatch as such.
• One contaminant is well controlled while another, present in the same stream, isn’t, despite both nominally being “handled” by the installed media. This is the multi-contaminant-stream problem described above: one media chemistry rarely captures chemically dissimilar gases equally well, and a system specified against only the more obvious or more concentrated contaminant can leave a second one essentially unaddressed.

If any of these patterns show up, the productive question isn’t “how often should we replace this,” it’s “was this the right media chemistry for what’s actually in our airstream” — and that’s a specification question, not a maintenance one.

Worked Example: Estimating Media Service Life

Media replacement frequency is a real operating cost, and it’s calculable from the same numbers used to size the system in the first place.

Example. A wastewater treatment exhaust stream runs at 8,000 CMH carrying 8 ppm of H₂S. At standard conditions, 1 ppm of H₂S corresponds to about 1.394 mg/m³, so the mass loading is:

8,000 m³/h × (8 × 1.394) mg/m³ × 24 h ÷ 1,000,000 ≈ 2.14 kg of H₂S per day

A KOH-impregnated carbon bed containing 800 kg of media, with an illustrative working capacity of 15% by weight (real working capacity varies by product and operating conditions — this figure should always be confirmed against the specific media’s datasheet, not assumed), holds about 120 kg of H₂S capacity before it’s exhausted:

800 kg × 15% ÷ 2.14 kg/day ≈ 56 days, roughly two months, before replacement is due.

That figure is sensitive to two things worth stress-testing before you commit to a change-out schedule: the actual working capacity of the specific media product (10% capacity instead of 15% cuts service life to about 37 days; 20% extends it to about 75 days), and humidity. If the working capacity above were derated by 40% because the stream runs consistently humid, service life falls to roughly 34 days — a full three weeks earlier than the dry-air estimate. Build a margin into the change-out schedule rather than running to the exact calculated date, and verify actual saturation with outlet monitoring rather than relying on the calculation alone.

What to Specify

Getting the right media starts with specifying the actual gas, not a generic category:

1. Name the specific compound(s), not just “toxic gas” or “odour” — H₂S, NH₃, Cl₂, SO₂, mercaptans, and VOCs each need different media, sometimes in the same vessel as separate stages.
2. Concentration and mass loading, not just airflow — two streams with identical airflow can need very different media volumes if their concentrations differ.
3. Humidity of the gas stream, especially if it runs above roughly 60% RH.
4. Whether multiple contaminant classes are present, so a multi-stage bed can be designed rather than discovered as necessary after underperformance.
5. Required outlet concentration or removal efficiency, tied to whichever standard applies — an occupational exposure limit, an equipment-protection target, or an emissions permit.

Where NextAir’s Dry Adsorbers Fit

NextAir Systems manufactures dry adsorption systems matched to the specific gas, not sold as one generic “activated carbon filter” regardless of application:

• Odor Control Adsorbers and Odor Removal Units — engineered for H₂S, mercaptans, and ammonia, with media chemistry selected for the actual target compounds.
• Dry VOC Adsorbers — for solvent-laden streams, in regenerative and non-regenerative configurations.
• Ethylene Scrubbers — potassium permanganate oxidation for controlled-atmosphere storage.
• Centralized Air Purification Systems — layering gas-phase adsorption with particulate and pathogen control for buildings and control rooms.

Contact us with the specific gas, its concentration, and your airflow, and we’ll recommend media chemistry matched to what’s actually in your stream, not a generic answer.

Talk to NextAir Systems About Your Specific Gas

Send us the compound, concentration, and airflow you’re dealing with, and we’ll match the adsorption media to what’s actually in your gas stream. Contact us.

References and Further Reading

• CDC/NIOSH — NIOSH Pocket Guide to Chemical Hazards: Chlorine (IDLH 10 ppm)
• CDC/NIOSH — Immediately Dangerous to Life or Health (IDLH) Values database
• CDC/NIOSH — IDLH documentation: Chlorine, May 1994 (revised IDLH derivation)
• Cal, M.P., Rood, M.J. & Larson, S.M. — Removal of VOCs from humidified gas streams using activated carbon cloth, *Gas Separation and Purification* (1996), via USGS
• Lee, J.W. et al. — Adsorption dynamics of water vapor on activated carbon, *Adsorption* (2005)
• Roohi, S. et al. (NASA/HX5/Jacobs Technology) — Dynamic Modeling of Ammonia Removal with Phosphoric-Acid-Treated Activated Carbon, 48th International Conference on Environmental Systems (2018)
• USPTO — Acid-impregnated activated carbon and methods of forming and using the same (ammonia chemisorption patent)

Frequently Asked Questions