Nuclear fusion is a reaction in which two or more atomic nuclei combine to form a larger nucleus. The difference in mass between the reactants and products is manifested as either the release or the absorption of energy. This difference in mass arises as a result of the difference in nuclear binding energy between the atomic nuclei before and after the fusion reaction. Nuclear fusion is the process that powers all active stars, via many reaction pathways.
Fusion processes require an extremely large triple product of temperature, density, and confinement time. These conditions occur only in stellar cores, advanced nuclear weapons, and are approached in fusion power experiments.
A nuclear fusion process that produces atomic nuclei lighter than nickel-62 is generally exothermic, due to the positive gradient of the nuclear binding energy curve. The most fusible nuclei are among the lightest, especially deuterium, tritium, and helium-3. The opposite process, nuclear fission, is most energetic for very heavy nuclei, especially the actinides.
Applications of fusion include fusion power, thermonuclear weapons, boosted fission weapons, neutron sources, and superheavy element production.


== History ==


=== Theory ===

American chemist William Draper Harkins was the first to propose the concept of nuclear fusion in 1915. Francis William Aston's 1919 invention of the mass spectrometer allowed the discovery that four hydrogen atoms are heavier than one helium atom. Thus in 1920, Arthur Eddington correctly predicted fusion of hydrogen into helium could be the primary source of stellar energy. 
Quantum tunneling was discovered by Friedrich Hund in 1927, with relation to electron levels. In 1928, George Gamow was the first to apply tunneling to the nucleus, first to alpha decay, then to fusion as an inverse process. From this, in 1929, Robert Atkinson and Fritz Houtermans made the first estimates for stellar fusion rates.  
In 1938, Hans Bethe worked with Charles Critchfield to enumerate the proton–proton chain that dominates Sun-type stars. In 1939, Bethe published the discovery of the CNO cycle common to higher-mass stars.


=== Early experiments ===

During the 1920s, Patrick Blackett made the first conclusive experiments in artificial nuclear transmutation at the Cavendish Laboratory. There, John Cockcroft and Ernest Walton built their generator on the inspiration of Gamow's paper. In April 1932, they published experiments on the reaction: 

73Li + p →  8X → 2 42He
where the intermediary nuclide was later confirmed to be the extremely short-lived beryllium-8. This has a claim to the first artificial fusion reaction. 
In papers from July and November 1933, Ernest Lawrence et. al. at the University of California Radiation Laboratory, in some of the earliest cyclotron experiments, accidentally produced the first deuterium–deuterium fusion reactions:  

21D + 21D → 31T + p
21D + 21D → 32He + 10n
The Radiation Lab, only detecting the resulting energized protons and neutrons, misinterpreted the source as an exothermic disintegration of the deuterons, now known to be impossible. In May 1934, Mark Oliphant, Paul Harteck, and Ernest Rutherford at the Cavendish Laboratory, published an intentional deuterium fusion experiment, and made the discovery of both tritium and helium-3. This is widely considered the first experimental demonstration of fusion.
In 1938, Arthur Ruhlig at the University of Michigan made the first observation of deuterium–tritium (DT) fusion and its characteristic 14 MeV neutrons, now known as the most favourable reaction: 

21D + 31T →  42He + 10n


=== Weaponization ===

Research into fusion for military purposes began in the early 1940s as part of the Manhattan Project. In 1941, Enrico Fermi and Edward Teller had a conversation about the possibility of a fission bomb creating conditions for thermonuclear fusion. In 1942, Emil Konopinski brought Ruhlig's work on the deuterium–tritium reaction to the project's attention. J. Robert Oppenheimer initially commissioned physicists at Chicago and Cornell to use the Harvard University cyclotron to secretly investigate its cross-section, and that of the lithium reaction (see below). Measurements were obtained at Purdue, Chicago, and Los Alamos from 1942 to 1946. Theoretical assumptions about DT fusion gave it a similar cross-section to DD. However, in 1946 Egon Bretscher discovered a resonance enhancement giving the DT reaction a cross-section ~100 times larger.
From 1945, John von Neumann, Teller, and other Los Alamos scientists used ENIAC, one of the first electronic computers, to simulate thermonuclear weapon detonations.
The first artificial thermonuclear fusion reaction occurred during the 1951 US Greenhouse George nuclear test, using a small amount of deuterium–tritium gas. This produced the largest yield to date, at 225 kt, 15 times that of Little Boy. The first "true" thermonuclear weapon detonation i.e. a two-stage device, was the 1952 Ivy Mike test of a liquid deuterium-fusing device, yielding over 10 Mt. The key to this jump was the full utilization of the fission blast by the Teller–Ulam design.
The Soviet Union had begun their focus on a hydrogen bomb program earlier, and in 1953 carried out the RDS-6s test. This had international impacts as the first air-deliverable bomb using fusion, but yielded 400 kt and was limited by its single-stage design. The first Soviet two-stage test was RDS-37 in 1955 yielding 1.5 Mt, using an independently reached version of the Teller–Ulam design.
Modern devices benefit from the usage of solid lithium deuteride with an enrichment of lithium-6. This is due to the Jetter cycle involving the exothermic reaction:

63Li + 10n → 42He + 31T
During thermonuclear detonations, this provides tritium for the highly energetic DT reaction, and benefits from its neutron production, creating a closed neutron cycle.


=== Fusion energy ===