Why this poison is interesting

Because it is not one. A true simple asphyxiant — nitrogen, argon, helium, methane — does nothing to the body whatsoever. It binds no receptor, inhibits no enzyme, is not metabolised and leaves no metabolite. Its entire toxicology is that it occupies volume in the atmosphere that oxygen would otherwise have occupied. This page is in the library because the boundary is genuinely useful: every other gas in Band D poisons something specific, and knowing which category you are in changes what you look for and what protects you.

The second reason is that carbon dioxide does not belong in the category it is usually filed under. It displaces oxygen like any other gas, but unlike nitrogen or argon it is also an acid, a vasodilator, a narcotic at high concentration, and — decisively — the substance the respiratory centre actually measures. The UKHSA information page describes the resulting syndrome as "headache, dizziness, sweating, muscle twitches, increased breathing rate, shortness of breath, drowsiness, fast heart rate and loss of consciousness", with very high concentrations causing "fitting, coma and death".¹ That is not the description of a gas doing nothing.

The third reason is the one that kills people, and it follows from the second. The drive to breathe senses carbon dioxide, not oxygen. So an atmosphere of carbon dioxide produces immediate, unbearable air hunger — a warning proportionate to the danger. An atmosphere of nitrogen or argon produces little or none: the person goes on breathing comfortably, washes their carbon dioxide out entirely normally, and may reach collapse with no warning proportionate to the danger. Peripheral chemoreceptors do sense oxygen, but the hypoxic ventilatory response is weak and partly self-cancelling — the hyperventilation it produces lowers carbon dioxide, which suppresses air hunger again. The gas that feels worse is the one that gives you a chance.

A poison is a drug whose kinetics have escaped its pharmacology.

The toxic principle

For a simple asphyxiant the toxic principle is arithmetic. Air is about 21% oxygen. Any gas introduced into a closed space dilutes that fraction, and the inspired partial pressure of oxygen falls in proportion. Everything downstream — falling arterial oxygen tension, falling saturation, tissue hypoxia, loss of consciousness — is ordinary hypoxia, of exactly the kind produced by altitude or by a blocked airway. There is no toxicological step anywhere in the sequence.

That is why these gases have no antidote in the usual sense and need none: the deficiency is of oxygen, and the treatment is oxygen. On every other page in this band oxygen is either a supportive measure (cyanide, hydrogen sulphide, methaemoglobin inducers) or an active hazard (paraquat). Here it is the correction itself, which makes this the only page in the band where the obvious treatment is also the right one.

Why carbon dioxide is not a simple asphyxiant, in four separate ways

One — it is an acid. Carbon dioxide dissolves to form carbonic acid, so a high inspired concentration produces a respiratory acidosis directly, independent of any oxygen deficit. Two — it is the respiratory stimulus. Central chemoreceptors respond principally to carbon dioxide and pH rather than to oxygen, which is why hypercapnia produces air hunger and hypoxia alone largely does not. Three — it is a narcotic at high concentration, depressing consciousness in its own right. Four — it is a vasodilator, including cerebrally, which contributes to the headache the source lists first among its symptoms.¹ A gas with four pharmacological actions is not inert, and calling it a simple asphyxiant obscures the only feature that ever saves anybody: that it can be felt. Inferred

The silence of an inert atmosphere

This is the single most important idea on the page. Dyspnoea is not a hypoxia alarm — it is largely a carbon dioxide alarm. In an atmosphere of nitrogen or argon, a person continues to ventilate normally and therefore continues to clear carbon dioxide entirely normally: arterial carbon dioxide is normal or low, the chemoreceptors are quiet, and there is often no air hunger at all, and never a warning proportionate to the danger. There is no dyspnoea to prompt escape, and collapse can be the first sign — to the person, and to anyone watching. In a carbon dioxide atmosphere the opposite happens — the gas that is displacing the oxygen is the same gas the body is monitoring, so the warning is loud and immediate. The protective mechanism and the poison are the same molecule. Inferred

One physical property completes the picture and explains the geography of these incidents. Carbon dioxide is denser than air, so it collects in low-lying and enclosed spaces — pits, silos, tanks, cellars, ship holds, fermentation vessels and sumps — and can sit there as a layer while the air above it is entirely breathable. A space can be safe at head height and lethal at floor level, which is why an atmospheric test taken at the top of a shaft says nothing about the bottom of it.

Toxicokinetics

This table is mostly empty and its emptiness is the content. A simple asphyxiant has no toxicokinetics, because kinetics describe what the body does to a substance and the body does nothing to these substances at all. The rows that are populated belong almost entirely to carbon dioxide.

The library's emptiest kinetics table — and why that is the finding
Parameter	Therapeutic	In overdose	Why it changes
Absorption	Inert gases dissolve minimally in blood and do nothing when they do	Irrelevant. The harm is in the atmosphere, not in the patient	The row that separates this page from every other in the library. For nitrogen or argon there is no absorbed dose, no concentration and nothing to measure. The exposure is a property of the room.
Carbon dioxide absorption	Freely diffusible; normally excreted by the lung down a gradient	The gradient reverses. A high inspired concentration means carbon dioxide moves into the blood rather than out of it	This is what makes carbon dioxide an actual poison rather than a diluent. The lung's normal function is turned around, so hypercapnia develops even with perfectly preserved ventilation.
Distribution	—	Carbon dioxide equilibrates rapidly across all tissue and across the blood–brain barrier	Rapid central equilibration is why the narcosis and the headache appear early, and why the syndrome the source describes is dominated by neurological symptoms.¹
Metabolism	None for any gas on this page	None	No bioactivation, no metabolite, no enzyme. Shared with nitrous oxide and paraquat — but not with hydrogen sulphide, which is rapidly metabolised to thiosulphate and sulphate — and unlike those three, here there is no target either.
Elimination	Exhaled	Immediate on removal to air. Carbon dioxide is cleared by the first few normal breaths	Elimination is instantaneous and complete, which is why a rescued patient either recovers quickly or is suffering hypoxic-ischaemic injury sustained before rescue. There is no residual poisoning to treat.
The exposure axis	—	Atmospheric concentration and duration. "The higher the carbon dioxide concentration, the worse symptoms may get"¹	There is no level, no nomogram and no clearance. The history is the atmosphere and the geometry of the space, and for carbon dioxide the fact that it is denser than air means the concentration varies with height.
The warning signal	—	Present for carbon dioxide, absent for inert gases	The only variable on this page that changes survival, and it is not a property of the patient. It is whether the displacing gas happens to be the one the chemoreceptors measure. Inferred
Dialysability	—	Never assessed and incoherent. EXTRIP has published no recommendation covering these gases²	Recorded only for consistency with the rest of the library. There is no xenobiotic in the patient to remove — the abnormality is a partial pressure in a room.

Metabolism and the metabolites

There are none, for any gas on this page, and unlike nitrous oxide — where the same sentence opens the section and is followed by the metabolism the gas destroys — here there is genuinely nothing to describe. What follows instead is the physiology that decides whether the person is warned.

Two atmospheres, one falling oxygen fraction, and a warning that depends on which gas replaced it

Gas accumulates in an enclosed spaceCarbon dioxide is denser than air and pools low. A space can be breathable at head height and lethal at floor level
Inspired oxygen fraction falls"Higher levels in an enclosed space can reduce the amount of oxygen available in the air. This can lead to asphyxiation"¹
If the displacing gas is inert — nitrogen, argon, helium, methaneCarbon dioxide cleared normally; chemoreceptors quietNo air hunger. No distress. No warning. The single most dangerous branch on this page
If the displacing gas is carbon dioxideArterial carbon dioxide rises; chemoreceptors fireAir hunger, tachypnoea, panic — "increased breathing rate, shortness of breath"¹. The warning arrives
Carbon dioxide only — direct pharmacological actionsRespiratory acidosis, cerebral vasodilatation, narcosis"Headache, dizziness, sweating, muscle twitches... drowsiness, fast heart rate"¹ — effects no inert gas produces
Both — falling arterial oxygenTissue hypoxiaOrdinary hypoxia, of the kind produced by altitude or a blocked airway
Loss of consciousness, then hypoxic-ischaemic injury"Fitting, coma and death" at very high concentration¹. In an inert atmosphere this arrives without any preceding distress
On removal to air — immediate reversal of the exposureNothing persists. What remains is whatever hypoxic injury was sustained before rescue

Elimination and accumulation

Nothing accumulates in the patient. What accumulates is the gas in the space, and that is an engineering fact rather than a physiological one — but it is the one that determines the incident. Carbon dioxide's density means it collects at the bottom of enclosed volumes and can persist there for a long time, undisturbed, in a space that ventilates poorly by design.

Why rescuers die here for a different reason than they do with hydrogen sulphide

Two pages in this band end with people dying while trying to help, and the mechanisms are not the same. With hydrogen sulphide the rescuer is poisoned: a few breaths cause knockdown, and the warning smell has already failed. With an asphyxiant the rescuer is not poisoned at all — they simply enter an atmosphere without enough oxygen in it, and with an inert gas they get no warning that this is happening. The practical consequence is identical and the monitoring is different: an oxygen meter detects an asphyxiant atmosphere and says nothing at all about hydrogen sulphide, while a gas-specific detector does the reverse. Neither instrument covers the other's hazard, and a space cleared on one reading is not cleared.

The clinical corollary is more hopeful than most of this band. The exposure ends completely at the moment of extrication. Carbon dioxide is cleared by the first few normal breaths and inert gases were never doing anything to clear. A patient who is rescued early recovers fully, and everything that does not recover is hypoxic-ischaemic injury sustained before rescue — which places the whole of the prognosis in the hands of how quickly, and how safely, somebody got them out.

Target organs — and why those

There is no organ selectivity here, and saying so plainly is more useful than manufacturing an argument. The organs that fail are the organs that always fail in hypoxia, in the usual order. The cards below record that, and one of them is not an organ at all.

Brain

TargetNeurones, by ordinary hypoxia

Why hereNo selectivity mechanism, and that is the finding. The brain fails first because it has the highest oxygen demand and the least anaerobic reserve — the same reason it fails first in cyanide and hydrogen sulphide poisoning, and in cardiac arrest, and at altitude. Here nothing is being poisoned; the oxygen is simply not arriving. For carbon dioxide there is an additional and separate central action — narcosis and cerebral vasodilatation — that an inert gas does not have. Established

At the bedside"Headache, dizziness, sweating, muscle twitches, increased breathing rate, shortness of breath, drowsiness, fast heart rate and loss of consciousness", with "fitting, coma and death" at very high concentration.¹ With an inert gas, the early half of that list is largely absent and loss of consciousness can be the first sign.

Heart

TargetMyocardium, by ordinary hypoxia

Why hereFails after the brain and for the same unremarkable reason. Carbon dioxide adds a respiratory acidosis and a sympathetic response, which is why tachycardia appears in the source's symptom list.¹ There is no direct cardiac toxicity from any gas on this page, which distinguishes it from every other entry in the band. Inferred

At the bedsideTachycardia, then bradycardia, hypotension and arrest as hypoxia deepens. From that point it is an ordinary hypoxic arrest and is managed as one.

The respiratory centre — as an alarm rather than a target

TargetCentral chemoreceptors

Why hereThe most important entry on this page and it is not an injury. Central chemoreceptors respond principally to carbon dioxide and pH rather than to oxygen. A carbon dioxide atmosphere therefore triggers the alarm loudly; an inert atmosphere leaves it silent while the oxygen disappears. The whole difference in survivability between the two exposures sits in this card. This is standard respiratory physiology and no source cited on this page states it — it is badged accordingly and flagged in the appraisal. Inferred

At the bedsideCarbon dioxide: severe air hunger, tachypnoea, agitation and an overwhelming urge to leave. Inert gas: comfortable breathing, no dyspnoea, and no reason to leave. A history of collapse without any preceding breathlessness points towards the second.

Everything else

TargetWhatever the hypoxic-ischaemic insult reached

Why hereIncluded to close the section honestly. Post-hypoxic organ injury here is not toxicological — it is the consequence of a period of inadequate oxygen delivery, identical in kind to that following any arrest, and it is not attributable to a property of the gas. Established

At the bedsideAcute kidney injury, transaminitis, rhabdomyolysis and post-hypoxic encephalopathy in survivors of prolonged exposure — determined by the duration of the insult, not by which gas caused it.

Timeline of effects

Two atmospheres, and only one of them tells the person what is happening

Time

What you seeWhat is happening

Immediate — carbon dioxideThe warning fires
What you seeHeadache, dizziness, sweating, muscle twitches, "increased breathing rate, shortness of breath".¹
What is happeningArterial carbon dioxide rises because the lung's normal excretory gradient has reversed. Chemoreceptors respond to exactly the gas that is displacing the oxygen, so the alarm is proportionate to the hazard. Inferred
Immediate — inert gasThe silence
What you seeNothing. Breathing feels normal. No dyspnoea, no distress, no prompt to leave.
What is happeningVentilation is unimpaired, so carbon dioxide is cleared normally and the chemoreceptors have nothing to report. The oxygen fraction is falling and no physiological system is monitoring it in a way that produces symptoms. Inferred
Seconds–minutesLoss of consciousness
What you see"Drowsiness, fast heart rate and loss of consciousness."¹ With an inert gas this may be the first sign of anything at all.
What is happeningArterial oxygen falls below what cerebral metabolism requires. For carbon dioxide, narcosis contributes independently of the hypoxia; for an inert gas, hypoxia is the whole of it.
MinutesThe second casualty
What you seeA colleague enters the space to help and collapses in turn.
What is happeningThe atmosphere has not changed and the rescuer has no protection against it. With an inert gas they receive no warning either. This is a scene-safety mechanism rather than a clinical one, and it is why these incidents produce multiple casualties.
MinutesCardiovascular collapse
What you seeBradycardia, hypotension, arrest.
What is happeningOrdinary hypoxic arrest. From this point the gas is irrelevant and the patient is managed as any other hypoxic cardiac arrest.
On removal to airImmediate and complete reversal of the exposure
What you seeRapid recovery if rescue was early; otherwise the picture of hypoxic-ischaemic injury.
What is happeningCarbon dioxide is exhaled within a few breaths and inert gases were never doing anything. Nothing persists, so the outcome is set entirely by how long the exposure lasted.

What the mechanism predicts at the bedside

Do not enter the space. The atmosphere that dropped the patient is unchanged, and with an inert gas it will give you no warning either.
An oxygen meter and a gas-specific detector answer different questions. An oxygen reading detects displacement and says nothing about hydrogen sulphide; a sulphide detector says nothing about nitrogen. A space cleared on one instrument is not cleared.
Carbon dioxide is denser than air and pools low. A test at the top of a shaft does not describe the bottom of it.
A collapse with no preceding breathlessness suggests an inert gas rather than carbon dioxide — because air hunger is a carbon dioxide response, not a hypoxia response.
A patient found distressed, tachypnoeic and headachey suggests carbon dioxide, which has pharmacological actions an inert gas does not.¹
Extrication is the treatment and oxygen is the antidote — the only page in this band where oxygen corrects the actual lesion rather than working around it.
Expect full recovery if rescue was quick. Nothing persists in the patient; there is no metabolite, no target and no residual poisoning.
Everything that does not recover is hypoxic-ischaemic injury sustained before extrication, and is managed as post-arrest care rather than as poisoning.
Ask about the space, not about a dose. Geometry, ventilation, what was stored or fermenting in it, and how long the patient was in it are the whole history.
Expect more than one casualty. These incidents characteristically involve a rescuer, and the second patient may be the more salvageable one.
There is no role for dialysis and the question is incoherent² — there is no substance in the patient.

The antidote, from the poison's side

This is the one page in the band where the antidote is obvious, available, and actually corrects the lesion. It is worth pausing on how unusual that is: across Band D, oxygen has been a supportive measure that does not touch the mechanism (cyanide, hydrogen sulphide, methaemoglobin inducers) or an active accomplice to it (paraquat). Here the deficiency is of oxygen and the treatment is oxygen.

Removal from the atmosphere: The definitive intervention. The exposure ends completely and immediately, because there is nothing in the patient to clear except carbon dioxide, which the first few breaths handle. Established
Oxygen: Genuinely the antidote, and the only occurrence of that in this band. It restores the deficiency directly rather than compensating for a lesion elsewhere. Contrast paraquat, where oxygen is fuel for the mechanism — the two pages are the opposite ends of Band D's argument about when to give it.
Ventilation: Supports a patient who is not breathing adequately after a hypoxic insult. It is post-arrest care rather than antidotal, because the exposure has already ended.
Why there is nothing to scavenge, block or reverse: There is no toxic species. No receptor is occupied, no enzyme inhibited, no protein modified, no metabolite made. A simple asphyxiant is the only entry in this library against which the entire concept of an antidote is undefined, and carbon dioxide's few pharmacological actions all resolve on exhalation.
Extracorporeal removal: Never assessed and incoherent.² Recorded only for consistency with the rest of the library.

Critical appraisal

The central claim of this page — that dyspnoea is a carbon dioxide response rather than a hypoxia response, and that this is why inert atmospheres kill silently — is not supported by any citation on this page. It is standard respiratory physiology, it is badged inferred in all four places it appears, and it is the first thing that should be tested. If it cannot be sourced, the badge is correct and the emphasis may still be too confident.
The four ways carbon dioxide is not a simple asphyxiant are likewise uncited, badged inferred. Acid formation, chemoreceptor stimulation, narcosis and vasodilatation are textbook physiology; the source lists a symptom set consistent with them¹ without attributing them.
The claim that carbon dioxide is denser than air is elementary physical chemistry and is not cited here. It is load-bearing for the confined-space geography and note that it is asserted rather than sourced.
No concentration threshold appears anywhere on this page, deliberately. The source states only that "the higher the carbon dioxide concentration, the worse symptoms may get"¹ and this page goes no further. Printing the widely quoted percentage bands would be an occupational-hygiene standard repurposed as a clinical threshold, which this library does not do.
No lethal concentration, no time to unconsciousness and no description of apparatus or circumstances appears here. This is a page about an atmosphere that kills without warning, and the editorial boundary is tighter than usual as a result. The physiology of the absent warning is included because it explains rescuer deaths and is necessary for clinicians; nothing operational accompanies it. A forensic case series retrieved during sourcing was deliberately not cited for this reason.
No evidence-tier downgrade appears. Nothing on this page is contested by any source found, and manufacturing a badge from the thinness of the evidence would invert the house rule, which requires a citation for the doubt rather than an absence of citation for the claim.
The organ cards deliberately decline to offer a selectivity mechanism, because there is none. That is an argument rather than an omission, and it is the point of including the page.
The distinction drawn between rescuer deaths here and on the hydrogen sulphide page is this page's own reasoning, not a finding — as is the observation about the two classes of detector, which is included because it has practical consequences and is not sourced.
The 'oxygen is the antidote' framing is rhetorical as much as pharmacological. Oxygen is not an antidote in the sense the rest of this library uses the word; the sentence is making the point that this is the only entry where restoring the deficiency is the treatment. It is flagged so that it is not read as a pharmacological claim.
EXTRIP's silence² is recorded purely for consistency and is the most meaningless instance of it in the library.

References

1
Carbon dioxide: general information. UK Health Security Agency, Compendium of Chemical Hazards. ("Higher levels in an enclosed space can reduce the amount of oxygen available in the air. This can lead to asphyxiation and the higher the carbon dioxide concentration, the worse symptoms may get. This can cause headache, dizziness, sweating, muscle twitches, increased breathing rate, shortness of breath, drowsiness, fast heart rate and loss of consciousness. Exposure to very high concentrations can cause fitting, coma and death." Note: the compendium publishes a full toxicological overview for chlorine, ammonia, hydrogen cyanide, hydrogen sulphide, hydrogen fluoride and nitrobenzene, but for carbon dioxide it publishes general information and incident management only — there is no toxicological overview, which is why this page is the least evidenced in the band.) gov.uk
2
EXTRIP Workgroup — published recommendations index. Extracorporeal Treatments in Poisoning Workgroup. Cited for the absence of any recommendation covering carbon dioxide or the simple asphyxiant gases. extrip-workgroup.org/recommendations

Carbon dioxide and the simple asphyxiants

At a glance