CO₂ is the carbon your plant is built from, fixed by a single slow, flawed enzyme — and the gas only pays inside a triangle with light and temperature, and only if it can reach the enzyme that uses it.
The hub makes the case that CO₂ is the raw material of growth and an accelerator, not a shortcut. This page is the mechanism: the one enzyme that fixes carbon and why its design makes enrichment so valuable, where the returns level off, and why the concentration your sensor reports is not the concentration the enzyme sees. The through-line is the framework's own — the room's number is not the plant's.
What it actually is: the one enzyme, and its ancient flaw
Almost all the carbon in a plant is fixed by Rubisco, the enzyme that grabs a CO₂ molecule and attaches it to a five-carbon acceptor (RuBP) to start the Calvin cycle. Rubisco is the most abundant protein on Earth, because it is, by enzyme standards, painfully slow — a few reactions per second — so plants build it in bulk (25–50% of a leaf's protein). But its decisive trait isn't speed; it's a lack of specificity. Rubisco evolved 3.5 billion years ago in an atmosphere with almost no oxygen, and it never learned to tell CO₂ and O₂ apart. When it binds CO₂ (carboxylation), it does productive work — carbon enters the Calvin cycle and becomes sugar. When it binds O₂ (oxygenation), it triggers photorespiration: a wasteful salvage process that loses carbon the plant had already fixed, burns energy it had already captured, and produces nothing. At today's ambient 420 ppm, photorespiration consumes 25–30% of the carbon Rubisco fixes, and it gets worse as temperature rises — which is why hot C3 plants lose net photosynthesis even under bright light: Rubisco is fixing carbon and then throwing it away. Every CEA crop is a C3 plant and suffers this; the C4 crops (corn, sugarcane) evolved a workaround none of our crops use. The key consequence: raising CO₂ doesn't just add substrate — it tips the balance toward carboxylation and suppresses the wasteful oxygenation that was destroying value.
What it gates: the response curve, and where it levels off
Photosynthesis responds to CO₂ in two phases. From ambient up to roughly 600–800 ppm, the rate is carboxylation-limited — there isn't enough CO₂ to keep all that Rubisco busy — so adding CO₂ produces a strong, nearly linear gain (420→800 often lifts photosynthesis 30–40%). Above roughly 800–1,000 ppm, Rubisco is saturated and the limit shifts to how fast the rest of the Calvin cycle can regenerate RuBP, which depends on light — so further CO₂ produces diminishing returns. Past about 1,200–1,400 ppm, most crops show little additional response under normal light: the gas is consumed and paid for, but the marginal carbon per ppm falls steeply. This curve defines the economic sweet spot at 800–1,200 ppm, with very high light (above 1,000 PPFD) justifying the upper end because abundant light pushes the RuBP-regeneration ceiling higher. At the other extreme sits the CO₂ compensation point — the concentration where fixation exactly equals respiratory release and net growth is zero — which a depleting sealed room drifts toward when its supply fails.
The heart: the CO₂–light–temperature triangle
CO₂'s return depends entirely on whether the other conditions for photosynthesis are present, and this is where most CO₂ advice falls short. Three variables form a triangle, and the gas only pays when all three are matched.
CO₂ × light — the matched pair. CO₂ and light are the two substrates of photosynthesis: carbon and the energy to fix it. At low light, the rate is limited by energy, so enrichment buys little — the plant can't drive the Calvin cycle faster regardless of substrate. At high light (800–1,000+ PPFD), light is abundant and the rate is limited by CO₂, so enrichment to 800–1,200 ppm unleashes the full value of the lighting investment. High light with ambient CO₂ wastes electricity; high CO₂ with low light wastes gas. This is the single most economically important interaction in CEA — the grower who sees it allocates capital fundamentally better than one who treats lights and CO₂ as separate line items.
CO₂ × temperature — the shifting optimum. Because enrichment suppresses photorespiration, the optimal temperature for net photosynthesis moves up with CO₂: at ambient, photorespiration overwhelms carboxylation above ~25–28 °C; at 1,200 ppm, efficient photosynthesis continues to 30–32 °C. A grower who enriches to 1,200 but holds the same daytime temperature they ran at ambient is leaving the benefit on the table — the enriched room can tolerate, and wants, warmer days (within the ceiling protein denaturation still imposes). The triangle is the apex where light's saturation point, temperature's optimum, and CO₂'s response all meet, and they have to move together.
The reading that earns its keep: the CO₂ the leaf actually gets
Here is CO₂'s place in the translation-gap family, and it has two faces. The first is the stomatal gate: CO₂ can only reach Rubisco if the stomata are open, and the gas must diffuse a long path — room air, through the boundary layer, through the stoma, through the mesophyll, into the chloroplast — losing concentration at every step. Even under full opening and good airflow, the CO₂ that Rubisco actually sees is only 60–70% of the room concentration, and far less if stomata are closing. When VPD climbs past the closure threshold, the diffusion path narrows, internal CO₂ drops, and the room's enriched gas becomes gas floating in the room, not gas fixing carbon in the plant. Enrich to 1,200 while letting VPD shut the stomata, and you are enriching the room, not the plant. (There's a beneficial flip side: enriched CO₂ lets the plant fix carbon with less stomatal aperture, giving it a margin of water-conservation against moderate VPD stress — but if VPD forces full closure, no room concentration overcomes a shut valve.) The second face is the boundary-layer depletion zone: even at a perfect room 1,200, a leaf deep in still canopy air consumes CO₂ faster than diffusion replaces it, so the leaf surface can sit hundreds of ppm below the room — interior leaves running at 700–800 while the sensor reads 1,200. This is why airflow appears in every chapter: stripping the boundary layer is what makes the room's CO₂ real at the leaf. And there's a cascade to respect — enrichment partially closes stomata, which lowers transpiration, which reduces calcium delivery and slightly warms the leaf, so adding CO₂ can quietly create a calcium problem while solving a carbon one. CO₂ is never a standalone number.
Adjustment, the coupling trap, and the clean answer
How you supply CO₂ is itself a coupling decision. Compressed CO₂ (bottled or bulk) is clean — no heat, no moisture, no byproducts — with precise control, and it's the standard for sealed indoor rooms. Combustion burners make CO₂ cheaply but also produce heat (the cooling system must remove it), water (the dehumidifier must remove it), and, if poorly tuned, trace ethylene — a plant hormone potent enough at 50 parts per billion to trigger premature senescence, leaf curling, and flower drop. An unmaintained burner can damage the crop while enriching it: a coupling problem of the first order. The clean answer is to use clean gas (or a well-tuned, CO-monitored burner), and then to make the enrichment pay by matching it to light and temperature, delivering it with airflow, and keeping stomata open with good VPD — the same clean-intervention logic the whole site runs on. The dynamic side — a controller injecting against a logged setpoint — is real but stays deferred here; the lever on this page is understanding the triangle well enough to set it. The larger coupling that dominates greenhouse practice is the ventilation tradeoff, which the context split takes up next.
The context split: a light-period tool, and the ventilation tradeoff
CO₂ is a light-period tool and nothing else. In the dark there is no photosynthesis, so the plant consumes no CO₂ — it produces it through respiration, and a sealed room's CO₂ slowly rises overnight. Injecting CO₂ in the dark is simply wasted; the correct practice is to shut injection off at lights-off and resume at lights-on (an optional 5–15 minute delay lets the accumulated respiratory CO₂ be used first, though the yield difference is negligible). The harder context is the ventilation–CO₂ tradeoff that defines greenhouse climate management: ventilation is the primary tool for temperature and humidity, but opening the vents brings in ambient air and crashes the enriched CO₂ within minutes. The grower faces a continuous either/or — ventilate (manage heat and humidity, lose CO₂) or seal (hold CO₂, accept the heat and humidity). The solutions run a spectrum: open greenhouses that accept ambient CO₂ for simple low-cost climate control; fully sealed indoor facilities that pay the energy cost for complete control including CO₂; and semi-closed greenhouses that keep vents shut much of the day with partial mechanical cooling. The right point on that spectrum is specific to climate, crop value, and energy cost — which is why generic enrichment advice so often fails. The crop-by-crop targets live in the matrix.
Measurement and instrument discipline
CO₂ is measured with an NDIR sensor (non-dispersive infrared — CO₂ absorbs at 4.26 µm). A unit accurate to ±50 ppm is adequate for control; place it at canopy height, mid-canopy, away from injection points (which read artificially high) and doorways (which read diluted). Calibrate at least annually, since NDIR sensors drift — and be careful with automatic baseline calibration, which works in rooms that regularly return to ambient but introduces error in rooms that are continuously enriched. The single most valuable thing the sensor does is catch depletion: an alarm when a 1,200 ppm room reads below 600 during the light period can save a day's lost growth from an empty tank or a failed regulator. And size injection to the canopy's peak consumption, seal the room against the 10–30% that leaks out, and — non-negotiably — protect workers: CO₂ is an asphyxiation hazard at high concentration, so breathing-height alarms and an interlock that shuts injection off above ~5,000 ppm are required, not optional.
Synthesis: what mature CO₂ management looks like
Pulling it together: CO₂ is an accelerator, not a shortcut. Match it to light and temperature — the triangle — because enrichment only pays when the energy is there to use it and the temperature lets the carbon stick. Deliver it to the leaf with airflow, and keep the stomatal gate open with good VPD, or the room's gas never reaches the enzyme. Run it only in the light period. Seal the room and use clean gas or a maintained burner. And measure it at the canopy, with a depletion alarm and worker-safety interlocks. The same gas returns 30–50% more in a well-run environment than a poorly-run one — the grower who has stopped seeing CO₂ as a number on a controller and started seeing it as the apex of a triangle has turned the highest-potential operational expense in the room into its highest actual return.