Proton-Proton Chain

What the chain does

The proton-proton chain is the dominant nuclear reaction in the Sun. Its job, like every other hydrogen-burning sequence, is to convert four hydrogen-1 nuclei into one helium-4 nucleus. The chain accomplishes this through a small handful of nuclear reactions whose details depend on which of three branches the helium-3 intermediate takes. The end products are the same in every case: one ⁴He, two positrons (immediately annihilated), two electron neutrinos, and 26.73 MeV of energy released as gamma rays and kinetic energy.

That 26.73 MeV figure is the textbook gross energy. In practice, the two electron neutrinos carry away an average of about 0.6 MeV per cycle (varying by branch), and that energy is irretrievably lost to space because neutrinos do not interact with the rest of the Sun. The thermalized energy that actually drives the Sun's luminosity is closer to 26.1 MeV per ⁴He produced.

The chain was first worked out by Hans Bethe and Charles Critchfield in 1938, the same year Bethe published his Nobel-winning paper on the CNO cycle. Bethe and Critchfield realized that two-proton fusion was possible if a weak-interaction beta-plus decay could occur during the brief moment of nuclear contact, converting one proton into a neutron and producing a deuteron. The cross-section was small enough to be consistent with the Sun's apparent age and luminosity — exactly what was needed.

The three branches

The chain splits into three terminations after the helium-3 stage. Each ends at ⁴He but takes a different path with different neutrino signatures.

Branch	Reaction sequence	Solar fraction	Net energy (MeV)	Key neutrino
pp-I	p+p→²H, ²H+p→³He, ³He+³He→⁴He+2p	~84%	26.20	pp-ν, ≤0.42 MeV continuum
pp-II	p+p→²H, ²H+p→³He, ³He+⁴He→⁷Be, ⁷Be+e⁻→⁷Li+ν, ⁷Li+p→2 ⁴He	~16%	25.66	⁷Be-ν, 0.86 MeV line
pp-III	p+p→²H, ²H+p→³He, ³He+⁴He→⁷Be, ⁷Be+p→⁸B, ⁸B→⁸Be*+e⁺+ν, ⁸Be*→2 ⁴He	~0.02%	19.17	⁸B-ν, ≤15 MeV continuum
pp-IV (Hep)	³He+p→⁴He+e⁺+ν	~3 × 10⁻⁵	19.79	hep-ν, ≤18.8 MeV (highest)

The fractions are temperature-sensitive. Below T ≈ 10⁷ K the pp-I terminator dominates almost completely. As temperature rises through the solar range, alpha capture on ³He becomes competitive and pp-II grows. pp-III stays subdominant in the Sun because ⁸B is unstable and proton capture on ⁷Be has a small cross-section, but it is the dominant source of high-energy neutrinos and was the first to be detected by terrestrial neutrino experiments.

The pp-IV branch (also called hep) is essentially a curiosity: ³He directly captures a proton via the weak interaction to make ⁴He plus a positron and a neutrino. Its cross-section is so small that it contributes negligibly to solar luminosity but produces neutrinos at the highest energies of any solar process. It was first detected by Super-Kamiokande in the early 2000s.

The weak-interaction bottleneck

The first reaction in every branch is

p + p → ²H + e⁺ + ν_e + 0.42 MeV

This step is uniquely slow. Two protons brought into nuclear contact have a few times 10⁻²² seconds to interact via the strong force before they fly apart. During that window, one of the two protons must convert to a neutron via beta-plus decay — a weak-interaction process. The probability of weak emission during the contact time is roughly the ratio of the weak coupling to the strong coupling, squared, times some kinematic factors. The result is a cross-section of order 10⁻²³ barn at solar core temperatures, compared to ~1 barn for typical strong-interaction reactions.

Combined with the additional Coulomb-barrier suppression between two positively charged protons, the reaction rate per proton in the Sun's core is roughly

R_pp ≈ 5 × 10⁻¹⁸ s⁻¹ per proton

That works out to a mean lifetime of about 6 × 10⁹ years for any given proton in the core before it reacts. Out of the Sun's roughly 10⁵⁷ protons, only a small fraction actually inhabit the high-temperature core; of those, only a small fraction find a reaction partner per second; and of those collisions, only one in 10²⁰ or so produces a deuteron. The arithmetic gives the Sun's luminosity: ~3.85 × 10³³ erg/s, or about 4 × 10³³ erg/s.

If the weak interaction were a million times stronger, the Sun would burn through its hydrogen in a million years instead of ten billion. The slowness of the pp first step is therefore directly responsible for the long stellar lifetimes that allow planets the time to form, biology to start, and intelligence to evolve. There is a serious anthropic argument that the universe must be one whose weak coupling is roughly the value we observe, since stars with much faster pp would not have time to support life on their planets, and stars with much slower pp would not produce enough luminosity to be detectable habitats at all.

Subsequent steps and the deuteron's fate

Once a deuteron forms, it survives only briefly. The next reaction

²H + p → ³He + γ + 5.49 MeV

has a strong-interaction cross-section and proceeds within a few seconds in the dense solar core. The deuterium abundance in the Sun is therefore vanishingly small at any moment — about 10⁻¹⁷ by mass — even though deuterium is being created at the rate set by the pp first step. The deuterium we measure on Earth and in molecular clouds is primordial, made during Big Bang nucleosynthesis; stellar interiors destroy deuterium efficiently rather than make it.

The next reaction is the branching point. Helium-3 has two main fates: fusing with another ³He (pp-I) or capturing an alpha particle to make ⁷Be (pp-II/III). The cross-sections for both are well-measured at solar energies, and the temperature dependence is what selects which branch dominates in a given star.

Conditions inside the Sun

The Sun's core has the following standard-solar-model values:

• Central temperature: T_c ≈ 1.567 × 10⁷ K (about 15.7 million K)
• Central density: ρ_c ≈ 153 g/cm³
• Central pressure: P_c ≈ 2.34 × 10¹⁷ dyne/cm² (~230 billion atm)
• Hydrogen mass fraction at the centre: X_c ≈ 0.34 (down from primordial ~0.71)
• Helium mass fraction at the centre: Y_c ≈ 0.64 (built up over 4.6 Gyr)
• Energy production rate at the centre: ε ≈ 10 erg/g/s

The total power produced is the volume integral of ε over the energy-generating region, which is roughly the inner 25% by radius (about 1% by volume but ~30% by mass). The Sun's burning is concentrated almost entirely in the inner core.

The pp-chain rate scales as ε ∝ T⁴ at solar core temperatures, so a 10% temperature change shifts the luminosity by about 50%. This is much weaker temperature sensitivity than the CNO cycle (T¹⁵–²⁰), and it is what makes pp-burning stars stable: small thermal perturbations lead to small luminosity perturbations, which the star can buffer with a tiny structural adjustment. Stars on the CNO regime are far more thermally sensitive and develop convective cores to handle the steep gradient.

Worked numerical example: cycles per second in the Sun

The Sun's luminosity is L_⊙ = 3.846 × 10³³ erg/s. Each pp-chain cycle deposits roughly 26.1 MeV of thermal energy (after subtracting neutrino losses). Convert MeV to erg:

1 MeV = 1.602 × 10⁻⁶ erg
26.1 MeV = 4.18 × 10⁻⁵ erg per cycle

Number of cycles per second:

N_cycles/s = L_⊙ / E_cycle
           = 3.846 × 10³³ / 4.18 × 10⁻⁵
           ≈ 9.2 × 10³⁷ cycles per second

Each cycle consumes 4 protons, so the Sun is converting hydrogen to helium at

dN_p/dt ≈ 4 × 9.2 × 10³⁷ = 3.7 × 10³⁸ protons/s
        = 6.2 × 10¹⁴ g/s of hydrogen consumed

which is about 620 million tonnes per second. About 0.7% of that mass is converted to energy via E = mc²:

dM/dt = L_⊙ / c² = 3.846 × 10³³ / (3 × 10¹⁰)² ≈ 4.3 × 10¹² g/s
      = 4.3 × 10⁹ kg/s ≈ 4.3 million tonnes per second of mass-loss-to-energy

So 620 Mt/s of hydrogen is fused, of which 4.3 Mt/s is converted to photons and neutrinos, and 615.7 Mt/s of helium is added to the core. The Sun has roughly 1.989 × 10³³ g of mass, of which currently ~10⁵² g (about 7 × 10⁻⁸ of the total per second of operation) is being processed per second. At this rate the Sun will exhaust core hydrogen in about 5 billion more years, consistent with stellar evolution models.

Cross-checking against the solar neutrino flux: 9.2 × 10³⁷ cycles/s × 2 neutrinos/cycle = 1.84 × 10³⁸ neutrinos/s emitted by the Sun. Spread over a sphere at 1 AU (4π·(1.496 × 10¹³)² = 2.81 × 10²⁷ cm²), that gives a flux at Earth of

Φ_ν = 1.84 × 10³⁸ / 2.81 × 10²⁷ ≈ 6.5 × 10¹⁰ ν/cm²/s

Direct neutrino measurements (Borexino, SNO+, Super-K) give 6.0 × 10¹⁰ ν/cm²/s at Earth, dominated by pp-neutrinos. The agreement is excellent and validates the pp-chain energy budget in the Sun directly.

The solar neutrino problem and its resolution

For three decades after Ray Davis's 1968 Homestake experiment, solar neutrino measurements detected only ~1/3 of the flux predicted by standard solar models — almost entirely the high-energy ⁸B neutrinos from pp-III, the only branch Davis's chlorine detector was sensitive to. The shortfall was called the "solar neutrino problem" and triggered intense theoretical scrutiny: was the Sun's core cooler than predicted? Was the solar model wrong? Were the nuclear cross-sections off?

The actual answer turned out to involve neutrino physics rather than astrophysics. The Sudbury Neutrino Observatory (SNO), a 1000-tonne heavy-water Cherenkov detector commissioned in 1999, could distinguish electron neutrinos from muon and tau neutrinos. SNO's 2001-2002 measurements showed that the total neutrino flux from ⁸B decay matched standard solar model predictions almost exactly — but only one third of the arriving neutrinos were still electron-flavoured. The other two thirds had oscillated into muon and tau flavours en route from the Sun.

This required neutrinos to have nonzero mass, contradicting the original Standard Model assumption. Takaaki Kajita (Super-Kamiokande) and Arthur McDonald (SNO) shared the 2015 Nobel Prize in Physics for the discovery. The resolution simultaneously vindicated the proton-proton chain — the Sun was burning exactly as theory said — and overturned a foundational assumption of particle physics.

Borexino, an Italian liquid-scintillator detector commissioned in 2007, has since measured the pp-I, pp-II, and pp-III neutrino fluxes individually and confirmed the standard solar model prediction for each branch to ~5% precision. In 2020 it announced the first detection of CNO-cycle neutrinos in the Sun at the predicted ~0.8% level, completing the experimental verification of solar nuclear burning.

Variants and extensions

• CNO cycle. The other major hydrogen-burning sequence; uses C, N, O nuclei as catalysts. Dominant in stars heavier than ~1.3 M☉ at solar metallicity. Crossover with pp at the Sun's core would happen at T ≈ 1.8 × 10⁷ K.
• Hot pp chains. In dense, hot environments such as accreting white dwarfs and X-ray bursts, alternative pp-chain branches operate at proton-rich conditions, including ³He(α,γ)⁷Be(p,γ)⁸B at very high rates. These are not relevant to main-sequence stars but appear in classical novae and rp-process burning.
• Big Bang nucleosynthesis. The first six minutes of the universe ran a hot, fast version of the pp chain, producing the primordial 25% ⁴He abundance and trace deuterium and lithium. The slowness of pp first step at lower densities prevented further burning beyond ⁴He.
• Pre-main-sequence deuterium burning. Brown dwarfs and very young stars destroy primordial deuterium via ²H(p,γ)³He at temperatures as low as 10⁶ K. This sets the lower mass limit for deuterium-burning objects at ~13 M_J — the conventional brown dwarf / planet boundary.
• Lithium depletion. Pre-MS stars cooler than the lithium-burning temperature (~2.5 × 10⁶ K) preserve their initial Li abundance; hotter ones destroy it via ⁷Li(p,α)⁴He, the same step as the terminator of pp-II. The lithium dip in the Hyades cluster is a clear observational signature.

Where the pp chain shows up

• Solar luminosity calibration. The Sun's bolometric luminosity, 3.846 × 10³³ erg/s, is set entirely by the pp-chain rate at the central temperature and density. Any standard solar model assumes pp-chain energy generation; departures are tested against helioseismology and neutrino flux measurements.
• Borexino solar neutrino spectrum. The Borexino detector, in a 280-tonne ultra-pure scintillator buried under 1400 m of rock at Gran Sasso, Italy, measured pp-neutrinos at 6.1 ± 0.5 × 10¹⁰ /cm²/s at Earth in 2014, ⁷Be neutrinos at 4.99 ± 0.13 × 10⁹ /cm²/s in 2008, and ⁸B neutrinos at 5.7 × 10⁶ /cm²/s — all matching standard solar model predictions.
• Stellar age dating via lithium. The pp-II terminator destroys ⁷Li. Stars cooler than the Li-burning temperature retain their primordial ⁷Li, and warmer stars deplete it. The Li abundance vs. T_eff in cluster stars (the "Li dip") provides an independent cluster age estimate, complementary to the main-sequence turnoff method.
• Stellar mass-radius-luminosity relations. The pp-chain's T⁴ rate fixes the luminosity-mass slope L ∝ M⁴ for low-mass main-sequence stars. The exact slope depends on opacity and structure, but the underlying nuclear scaling is from pp.
• Habitable zone calculations. The pp-chain's slow rate gives M dwarfs (0.1–0.6 M☉) main-sequence lifetimes of 100 Gyr – 10 Tyr, far longer than the current age of the universe. This is why M dwarfs are heavily targeted in habitability surveys (TRAPPIST-1, Proxima Centauri b) — they offer extreme stellar lifetime stability that follows directly from pp-chain physics.

Common pitfalls

• Calling the chain "fusion of two protons." The chain begins with two-proton fusion, but four protons participate per cycle. The clean way to state it is: net reaction 4 ¹H → ⁴He.
• Forgetting neutrino energy losses. The 26.73 MeV gross figure includes neutrino-carried energy that escapes the Sun. Only ~26.1 MeV per cycle thermalizes and contributes to luminosity. The exact amount depends on which branch the cycle takes.
• Conflating pp-I, pp-II, pp-III with separate nuclear reactions. All three branches share the same first three steps (p+p→²H, ²H+p→³He, then a branch point at ³He). They differ only in the post-³He pathway.
• Quoting a single temperature exponent. The pp-chain rate scales as roughly T⁴ at the Sun's core temperature, but the exponent shifts with temperature. Below 10⁷ K it is closer to T⁶; above 1.5 × 10⁷ K it falls toward T³ as the cross-section approaches the Gamow peak energy.
• Treating the Sun's core as isothermal and isodense. Standard solar models give T and ρ profiles that vary by ~30% across the inner 0.2 R_⊙. The integrated burning is dominated by the very centre but extends outward. Don't compute luminosity assuming a single core temperature.

Summary

The proton-proton chain is the simplest answer to the question "where does the Sun get its energy." It fuses four hydrogen nuclei into one helium-4 by way of three branches that all start with the same pp-first-step bottleneck, release 26.73 MeV gross per cycle, and emit two neutrinos that escape directly to space. The chain dominates in stars cooler than ~1.3 M☉ where the CNO cycle's steep T¹⁵–²⁰ scaling cannot compete with pp's gentler T⁴. Solar neutrinos measured at Borexino and SNO have confirmed the energy production rate of every pp branch to within ~5%, and the same neutrino measurements demonstrated neutrino oscillations and nonzero neutrino mass. The slowness of the weak-interaction first step is what gives the Sun a 10-billion-year main-sequence lifetime — and, by extension, why biology had time to evolve here.