# Adenosine triphosphate
**Adenosine triphosphate** (ATP) is the [[Molecule|molecule]] that living cells use to move [[Chemical_energy|chemical energy]] from the reactions that release it to the reactions that need it. It is a nucleotide: the base adenine joined to the sugar ribose, which carries a chain of three [[Phosphorus|phosphate]] groups. Cutting the last phosphate off the chain, ATP + H₂O → ADP + Pᵢ, releases free energy, and the cell couples that cut to work it could not otherwise do — pumping [[Ion|ions]] uphill, building [[Polymer|polymers]], contracting a fibre. The quantity that decides whether the coupling works is not the heat of the reaction but the [[Gibbs_free_energy|Gibbs energy]] change, and the textbook figure of −30.5 kJ/mol is only the value under standard conditions.
In the microsim below the reader sets the cell's [ATP]/[ADP] ratio on a logarithmic slider and the number of ATP made per [[Combustion|oxidized]] glucose, and watches two numbers move. The first is the actual free-energy yield of one hydrolysis, `dG = dG0' + R T ln([ADP][Pi]/[ATP])`, which slides from the standard −30.5 kJ/mol to roughly −50 to −60 kJ/mol as the ratio climbs toward the value a living cell holds. The second is the thermodynamic efficiency of respiration, `eta = n |dG_ATP| / 2,870 kJ/mol`, where 2,870 kJ/mol is the free energy available from burning one mole of glucose and n is the ATP yield. A third readout converts the answer into a day: at a steady human metabolic rate the mass of ATP cycled through the body in twenty-four hours is comparable to the body's own mass.
On the [[Energy]] flagship this article is the child of Part IV — Scientific use, section *Cell metabolism* (row E25), where ATP is the unit in which biology's energy books are kept, the biochemical counterpart of the joule ledger that Part I builds for engines and fuels.[^murphy-ch5-forms] It is also the point at which [[Chemical_thermodynamics|chemical thermodynamics]] stops being an abstraction: an [[Endergonic_reaction|endergonic]] step becomes possible only when it is wired to an [[Exergonic_reaction|exergonic]] one, and ATP is the wire.
## Structure
ATP is built in three parts. Adenine, a purine base, is bonded through one of its nitrogens to carbon 1′ of ribose, a five-carbon sugar; the pair is the nucleoside adenosine. A chain of three phosphate groups, labelled α, β and γ outward from the sugar, hangs from the 5′ carbon. The molecular formula is C₁₀H₁₆N₅O₁₃P₃ and the molar mass 507.18 g/mol. The α phosphate is joined to the sugar by a phosphate ester bond; the α–β and β–γ joins are phosphoanhydride bonds, and those are the two that matter energetically.
At the near-neutral [[PH|pH]] of the cytosol the triphosphate chain is almost fully deprotonated. The terminal phosphate's last [[Acid_dissociation_constant|acid dissociation]] has a pKₐ near 6.5, so at pH 7.4 the dominant species is ATP⁴⁻, a small molecule carrying four negative charges within a few tenths of a nanometre of one another. The electrostatic repulsion between them is part of why the chain is a poor place to store charge and a good place to store [[Thermodynamic_free_energy|free energy]]: the products of hydrolysis separate, relax, and are better solvated by [[Water|water]] than the reactant was.
### Metal cation binding
The four-minus charge also means that ATP is never encountered bare. [[Magnesium|Magnesium]] ions bind the β and γ phosphates with a dissociation constant of the order of 0.1 mM, and since free cytosolic Mg²⁺ sits near 0.5–1 mM, most cellular ATP exists as the complex MgATP²⁻. That complex, not ATP⁴⁻, is the true substrate of nearly every kinase and ATPase, and the [[Enzyme|enzyme]] active site is shaped to recognise the metal as well as the nucleotide. The bound cation screens part of the interphosphate repulsion, which lowers the free energy released on hydrolysis by several kJ/mol and, more importantly, holds the flexible chain in the geometry the enzyme needs for in-line attack on the γ phosphorus. Calcium and manganese substitute in vitro; magnesium is what biology uses.
## Chemical properties
In pure aqueous solution ATP is kinetically stable and thermodynamically doomed. The hydrolysis of the terminal phosphoanhydride is strongly favourable, yet the uncatalysed reaction is slow enough that a solution kept cold and near neutral pH survives for days. That gap between what is favoured and what is fast is the whole reason the molecule is useful: it is a [[Spontaneous_process|spontaneous]] reaction with a high activation barrier, so nothing happens until an enzyme lowers the barrier at a place and time the cell chooses. [[Catalysis|Catalysis]] supplies the timing; thermodynamics supplies the push.
Both anhydride bonds can be cut. Removing the γ phosphate gives ADP and inorganic phosphate; cutting between α and β gives AMP and pyrophosphate, PPᵢ. The second route makes a reaction irreversible, because cells carry a pyrophosphatase that immediately hydrolyses PPᵢ to two phosphates, so the equilibrium is pulled forward twice. ATP is also a [[Chemical_bond|bonding]] partner rather than merely a fuel: it donates phosphate to hydroxyl and carboxyl groups, adenylyl groups to amino acids, and the whole molecule as a ligand to receptors outside the cell.
## Reactive aspects
The reaction the microsim computes is ATP + H₂O → ADP + Pᵢ. Its standard free-energy change under biochemical standard conditions — 25 °C, pH 7, one molar reactants and products, excess Mg²⁺ — is ΔG°′ = −30.5 kJ/mol, about −7.3 kcal/mol. No cell is anywhere near those conditions, and the difference is not a rounding error. The general expression is
`dG = dG0' + R T ln([ADP][Pi]/[ATP])`
with R = 8.3145 J/(mol·K) and T in kelvin. A cell holds ATP at roughly 1–10 mM while keeping ADP an order of magnitude or two lower and Pᵢ in the low millimolar range, so the logarithm's argument is far below one and the correction is large and negative. Working the arithmetic for [ATP] = 3 mM, [ADP] = 0.1 mM and [Pᵢ] = 3 mM at 310 K gives R·T·ln(0.1 × 3 / 3 × 10⁻³) ≈ −2.58 kJ/mol × ln(0.1) ≈ −25 kJ/mol, so ΔG ≈ −55 kJ/mol (derived). That is the number the reader watches move: the slider does not change the molecule, it changes how far the cell holds the reaction from [[Chemical_equilibrium|equilibrium]], and the free energy available is a measure of exactly that distance.
The same picture disposes of the phrase "high-energy phosphate bond". Breaking a bond always costs energy; the release comes from the whole balance sheet — charge separation, better solvation of the products, resonance stabilisation of free phosphate, and the entropy of splitting one particle into two. Fritz Lipmann, who introduced the squiggle notation ~P in 1941, meant a group-transfer potential rather than a special bond, and the group-transfer reading is the one that survives.[^lipmann1941]
Efficiency follows from the same arithmetic. Complete oxidation of one mole of glucose to carbon dioxide and water has ΔG°′ ≈ −2,870 kJ/mol, and modern accounting gives about 30–32 ATP per glucose rather than the older 36–38.[^rich2003] At the standard value that is 32 × 30.5 / 2,870 ≈ 34 % (derived); at the cellular value it is 32 × 50 / 2,870 ≈ 56 % (derived), which is why the sim's second readout climbs as the first one falls. The remainder leaves as [[Heat|heat]], and the comparison is worth making against the [[Heat_engine|heat engine]] treated elsewhere on this flagship: a cell is not bounded by a reservoir-temperature ratio, because it converts chemical potential directly rather than through a thermal intermediate.[^murphy-ch5-forms] ILLUSTRATIVE: the sim holds ΔG°′, temperature and [Pᵢ] fixed and varies only the ATP/ADP ratio and n, so its efficiency curve is a display fit to two dials, not a measured yield.
## Production from AMP and ADP
Cells do not manufacture ATP from scratch; they recharge it. ADP and AMP are phosphorylated back to ATP by three routes — substrate-level phosphorylation, oxidative phosphorylation and photophosphorylation — and adenylate kinase interconverts the pool by the reaction 2 ADP ⇌ ATP + AMP, which lets a cell salvage a second useful phosphate from two spent ones. The total adenine nucleotide pool is nearly constant; what changes, second by second, is how much of it is charged.
### Production, aerobic conditions
The dominant route in an oxygen-using cell is [[Cellular_respiration|oxidative phosphorylation]]. Electrons stripped from fuel molecules pass down a chain of membrane complexes to oxygen, and the free energy released is spent pumping [[Proton|protons]] across the inner mitochondrial membrane. The resulting electrochemical gradient — a concentration difference plus a membrane voltage, and therefore describable by the same [[Nernst_equation|Nernst]] arithmetic as an [[Electrochemical_cell|electrochemical cell]] — drives ATP synthase, a rotary motor whose c-ring turns as protons return. Peter Mitchell proposed this indirect coupling in 1961 against considerable resistance, and it is now the standard account.[^mitchell1961][^nobel1978] John Walker's crystal structure of the F₁ head in 1994 showed its three catalytic sites caught in three different states, as Paul Boyer's binding-change mechanism had predicted, and direct observation of the rotating shaft followed in 1997.[^abrahams1994][^noji1997][^nobel1997]
### Production, anaerobic conditions
Without oxygen the chain stops and the cell falls back on substrate-level phosphorylation, in which a phosphate is transferred to ADP directly from a [[Chemical_reaction|reaction]] intermediate whose own group-transfer potential is higher than ATP's. Glycolysis nets two ATP per glucose this way, roughly a fifteenth of the aerobic yield, and the pathway can only keep running if the electron carrier it reduces is reoxidised — by lactate in muscle, by ethanol and carbon dioxide in yeast. The energy left in those products is the reason fermentation is a poor bargain thermodynamically and a good one when speed matters more than yield.
### ATP production during photosynthesis
In [[Photosynthesis|photosynthesis]] the [[Photon|photons]] of [[Sunlight|sunlight]] do what oxidation does in respiration: they drive electrons uphill and a proton gradient across the thylakoid membrane follows. The same rotary enzyme, in chloroplast form, converts that gradient to ATP, and the ATP together with reduced NADP powers the fixation of carbon dioxide into sugar. The chain from sunlight to a molecule of ATP to a molecule of glucose to an animal that eats the plant is the food chain in [[Energy_transformation|energy-transformation]] terms, and each link discards most of what it received.[^murphy-ch14]
### ATP recycling
The pool is small and the flux through it is enormous. A human body contains only a few hundred grams of adenine nucleotide at any instant, yet an adult at rest running on about 2,000 kcal per day — a continuous 97 W, close enough to call it a 100 W bulb — turns over a mass of ATP through the day that is of the same order as body mass, because each molecule is charged and discharged many hundreds of times.[^murphy-ch5-units][^biochem-turnover] That is the sim's third readout, and it is the clearest statement of what ATP is: not a store of energy but a [[Energy_storage|carrier]] in constant circulation, with a turnover time measured in seconds and a standing inventory that would be exhausted in under a minute if resupply stopped.
## Biochemical functions
The uses of ATP divide into three families: a source of free energy, a source of phosphate for covalent modification, and a signal in its own right. Beyond the cases below it activates amino acids for protein synthesis by adenylylation, and at high cytosolic concentration acts as a hydrotrope that keeps proteins soluble — a role owing nothing to its free energy and everything to its charge.
### Cellular energy production
Most ATP in a resting animal is spent on a short list: ion pumping, protein synthesis, and cytoskeletal turnover. The Na⁺/K⁺-ATPase alone, discovered by Jens Skou in 1957, can account for a fifth or more of a cell's [[Basal_metabolic_rate|basal]] budget, exporting three sodium ions and importing two potassium ions per ATP against their gradients.[^skou1957][^nobel1997]
### Intracellular signaling
Kinases transfer ATP's γ phosphate onto serine, threonine or tyrosine side chains, switching target proteins on or off; phosphatases reverse it. ATP is also the precursor of cyclic AMP, the second messenger that carries hormone signals inward, and the substrate for most [[Enzyme_kinetics|enzyme-kinetic]] work on regulation.
### DNA and RNA synthesis
As a nucleoside triphosphate ATP is one of the four monomers of RNA, incorporated whole with loss of pyrophosphate; the deoxy form, made by ribonucleotide reductase, serves DNA. Here ATP is not fuel but building material, and the pyrophosphate release is what makes polymerisation effectively irreversible.
### Muscle contraction
In muscle, ATP binding to myosin releases it from actin and hydrolysis re-cocks the head for the next stroke, so ATP is spent both to move and to let go. Rigor mortis is a fibre whose ATP has run out and whose bridges cannot detach.
### ATP binding cassette transporter
ABC transporters use two ATP-binding domains as a clamp: binding and hydrolysis drive a cycle of conformational change that carries a substrate across a membrane. The family spans bacterial nutrient importers and the multidrug efflux pumps behind chemotherapy resistance.
## Abiogenic origins
ATP is present in every known organism, which places it before the last universal common ancestor and makes its prebiotic assembly a real problem. The difficulty is phosphate: on the early Earth most phosphorus was locked in sparingly [[Solubility|soluble]] apatite, and condensing phosphates into anhydride chains in bulk water runs uphill. Proposed escapes include evaporite settings, mineral polyphosphates, and reduced phosphorus delivered by meteoritic iron–nickel phosphide, which hydrolyses to reactive phosphite.[^pasek2008] None is settled, and the molecule's universality is stronger evidence of antiquity than any proposed synthesis is of mechanism.
## ATP analogues
Because ATP is consumed by the enzymes that study it, structural biology relies on analogues that bind but resist hydrolysis. Replacing the bridging oxygen between the β and γ phosphates with an imido group gives AMP-PNP, and with a methylene group AMP-PCP; replacing a terminal oxygen with sulfur gives ATP-γ-S, hydrolysed very slowly and transferring a thiophosphate that phosphatases remove reluctantly. Each substitution shifts the metal-binding geometry, so an analogue's structure is evidence about a state, not a photograph of the native complex.
## Medical use
Clinically, the interesting molecule is usually not ATP itself but adenosine, its fully dephosphorylated relative, which acts on cell-surface receptors to slow conduction through the atrioventricular node; a rapid intravenous bolus is used to interrupt paroxysmal supraventricular tachycardia. [citation needed] ATP released from damaged cells is a danger signal at purinergic receptors, and the collapse of cellular ATP is the proximate cause of death in ischaemia: pumps fail, ion gradients run down, and [[Water|water]] follows the ions inward. Measuring ATP by the light output of firefly luciferase is a routine assay of cell viability and of surface hygiene.
## History
Karl Lohmann in Heidelberg and, independently, Cyrus Fiske and Yellapragada Subbarow in Boston isolated a labile phosphate compound from muscle in 1929; it was ATP.[^lohmann1929][^fiske1929] Its structure was established over the following decade, and Alexander Todd's group completed the first chemical synthesis in 1948, work recognised by the 1957 Nobel Prize in Chemistry for research on nucleotides and nucleotide coenzymes.[^nobel1957] Fritz Lipmann's 1941 review set out the idea of a group-transfer potential and of ATP as the common currency between the reactions that release free energy and those that consume it, and shared the 1953 prize with Hans Krebs.[^lipmann1941][^nobel1953]
How the molecule is made took longer to settle than what it does. Through the 1950s the search was for a chemical intermediate linking oxidation to phosphorylation; Mitchell's 1961 proposal that the link is a [[Proton|proton]] gradient across a membrane replaced the missing intermediate with a [[Chemical_potential|potential]] and earned the 1978 prize.[^mitchell1961][^nobel1978] The rotary mechanism that converts that gradient into chemistry was confirmed structurally and then visually in the 1990s, and Boyer, Walker and Skou shared the 1997 prize.[^abrahams1994][^noji1997][^nobel1997]
## See also
- [[Bioenergetics]]
- [[Cellular_respiration]]
- [[Photosynthesis]]
- [[Chemical_energy]]
- [[Gibbs_free_energy]]
- [[Enzyme]]
- [[Energy_storage]]
## References
[^lipmann1941]: Lipmann, Fritz (1941). "Metabolic generation and utilization of phosphate bond energy." *Advances in Enzymology and Related Subjects of Biochemistry* 1: 99–162.
[^lohmann1929]: Lohmann, Karl (1929). "Über die Pyrophosphatfraktion im Muskel." *Die Naturwissenschaften* 17 (31): 624–625.
[^fiske1929]: Fiske, Cyrus H.; Subbarow, Yellapragada (1929). "Phosphorus compounds of muscle and liver." *Science* 70 (1816): 381–382.
[^mitchell1961]: Mitchell, Peter (1961). "Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism." *Nature* 191 (4784): 144–148.
[^abrahams1994]: Abrahams, Jan Pieter; Leslie, Andrew G. W.; Lutter, René; Walker, John E. (1994). "Structure at 2.8 Å resolution of F1-ATPase from bovine heart mitochondria." *Nature* 370 (6491): 621–628.
[^noji1997]: Noji, Hiroyuki; Yasuda, Ryohei; Yoshida, Masasuke; Kinosita, Kazuhiko (1997). "Direct observation of the rotation of F1-ATPase." *Nature* 386 (6622): 299–302.
[^skou1957]: Skou, Jens Christian (1957). "The influence of some cations on an adenosine triphosphatase from peripheral nerves." *Biochimica et Biophysica Acta* 23: 394–401.
[^rich2003]: Rich, Peter R. (2003). "The molecular machinery of Keilin's respiratory chain." *Biochemical Society Transactions* 31 (6): 1095–1105. (Source of the modern 30–32 ATP per glucose accounting in place of the older 36–38.)
[^pasek2008]: Pasek, Matthew A. (2008). "Rethinking early Earth phosphorus geochemistry." *Proceedings of the National Academy of Sciences* 105 (3): 853–858.
[^nobel1953]: Nobel Prize Outreach. "The Nobel Prize in Physiology or Medicine 1953 — Hans Krebs and Fritz Lipmann." https://www.nobelprize.org/prizes/medicine/1953/summary/
[^nobel1957]: Nobel Prize Outreach. "The Nobel Prize in Chemistry 1957 — Lord Todd." https://www.nobelprize.org/prizes/chemistry/1957/summary/
[^nobel1978]: Nobel Prize Outreach. "The Nobel Prize in Chemistry 1978 — Peter D. Mitchell." https://www.nobelprize.org/prizes/chemistry/1978/summary/
[^nobel1997]: Nobel Prize Outreach. "The Nobel Prize in Chemistry 1997 — Paul D. Boyer, John E. Walker and Jens C. Skou." https://www.nobelprize.org/prizes/chemistry/1997/summary/
[^murphy-ch5-forms]: Murphy, Thomas W. (2021). *Energy and Human Ambitions on a Finite Planet*. Chapter 5 (energy and power units), Table 5.2, p. 90 (the energy forms, with chemical energy entered as H − T·S). Portal Book 097, https://open.umn.edu/opentextbooks/textbooks/energy-and-human-ambitions-on-a-finite-planet
[^murphy-ch5-units]: Murphy (2021), *Energy and Human Ambitions on a Finite Planet*, Chapter 5, pp. 93–95 (1 cal = 4.184 J; a 2,000 kcal/day diet is 8.368 MJ over 86,400 s = 96.85 W, "very close to 100 W", and the book warns against quoting it to four figures). Portal Book 097.
[^murphy-ch14]: Murphy (2021), *Energy and Human Ambitions on a Finite Planet*, Chapter 14 (the food-chain framing of biological energy flow; page to pin). Portal Book 097.
[^biochem-turnover]: Standard biochemistry texts give the human standing adenine-nucleotide pool as a few hundred grams and the daily ATP turnover as of the order of body mass; no Portal Book on the Energy shelf covers ATP, and a biochemistry text is requested for the index (chapter and page to pin).
## External links
- [Energy and Human Ambitions on a Finite Planet](https://open.umn.edu/opentextbooks/textbooks/energy-and-human-ambitions-on-a-finite-planet), Murphy (2021) — Portal Book 097, the Energy shelf's energy-accounting text
- [Biology 2e](https://openstax.org/details/books/biology-2e) at OpenStax, whose metabolism chapter carries the ATP cycle at introductory level
- The Wikipedia pair's external links list the structural and enzymological databases
<!-- MATTERSIM:BEGIN g24 — Matter & Energy Cluster microsim (framework build, specs/sims/Adenosine_triphosphate.json); do not hand-edit inside -->
**Microsim — three.js (Wikitube framework), pending deploy:** *Adenosine triphosphate* will play here once `https://wikitube-3d-microsims.netlify.app/matter/Adenosine_triphosphate.html` is live.
<!-- pending: <div class="wt-sim" data-src="https://wikitube-3d-microsims.netlify.app/matter/Adenosine_triphosphate.html" data-title="Adenosine triphosphate"></div> -->
<!-- MATTERSIM:END -->
## Wikipedia : Wikitube
**Strict pair:** [Wikipedia](https://en.wikipedia.org/wiki/Adenosine_triphosphate) : [Wikitube](https://en.wikitube.io/wiki/Adenosine_triphosphate) · pinned revision [1373471144](https://en.wikipedia.org/w/index.php?oldid=1373471144) · 2026-09-11
## Previous hub tags
Hubs: `Life_Physics`. Portals: [[PORTAL_Energy]].
---
*Matter & Energy Cluster child articles, wave 1 · 2026-09-11 · drafted · Energy row E25 · sim pending (matter/Adenosine_triphosphate).*