{
  "id": "stern-gerlach-experiment",
  "title": "Stern–Gerlach Experiment",
  "lang": "en",
  "source_lang": "en",
  "summary": "The Stern–Gerlach experiment demonstrates the quantization of angular-momentum components by splitting an atomic beam in a nonuniform magnetic field.",
  "category": "science",
  "subcategory": "physics",
  "type": "concept",
  "aliases": [
    "Stern–Gerlach experiment",
    "Stern–Gerlach measurement"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/stern-gerlach-experiment/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/stern-gerlach-experiment/",
      "json": "https://aiwiki.page/zh/stern-gerlach-experiment.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The Stern–Gerlach experiment is an experiment in which a beam of neutral atoms passes through a nonuniform magnetic field and separates into distinct components according to the atoms’ magnetic properties. First performed by Otto Stern and Walther Gerlach in Frankfurt in 1922 using silver atoms, it demonstrated that the projection of atomic angular momentum along a measurement axis takes discrete values rather than a continuous range. In modern quantum mechanics, the experiment is a central illustration of spin, quantum measurement, and the dependence of measurement outcomes on the chosen axis."
    },
    {
      "heading": "Historical background",
      "text": "Stern proposed the experiment in 1921 to test space quantization: the prediction, arising from the old quantum theory, that an atom’s angular momentum could have only certain projections along an external field. This contrasted with a classical description in which an ensemble of magnetic dipoles could have a continuous distribution of orientations.\n\nThe experiment was technically demanding. Stern and Gerlach needed a sufficiently narrow atomic beam, a strong magnetic-field gradient, and a means of detecting a very small deposit of silver. Their 1922 result showed a split beam trace rather than the continuous broadening expected from a classical ensemble of freely oriented dipoles.\n\nThe initial interpretation preceded the modern concept of electron spin. George Uhlenbeck and Samuel Goudsmit proposed electron spin in 1925, and the development of quantum mechanics subsequently supplied the appropriate explanation of the silver-atom result.\n\nThis distinction matters historically: the experiment was designed to investigate quantized orbital orientations, but the splitting of ground-state silver atoms is now understood primarily through the spin of their unpaired electron. It was not originally conceived as a direct test of an already established spin theory."
    },
    {
      "heading": "Experimental arrangement",
      "text": "The essential apparatus has four parts:\n\n- Atomic source. An oven produces a beam of neutral silver atoms.\n- Collimation. Narrow apertures select atoms traveling within a small range of directions.\n- Magnetic analyzer. Specially shaped magnet poles produce a strongly nonuniform field.\n- Detector. A plate records the distribution of arriving atoms through the silver deposited on its surface.\n\nNeutral atoms are important because a charged beam would also undergo the much larger deflection associated with the Lorentz force. Neutrality does not prevent magnetic deflection: an atom can possess a magnetic dipole moment even though its total electric charge is zero.\n\nA uniform magnetic field can exert a torque on a magnetic dipole, but it does not provide the spatial gradient required for the beam separation. The decisive feature is the variation of the field across the beam.\n\nThe observed signal is not necessarily a pair of ideal points. Aperture widths, beam geometry, atomic speeds, and the shape of the magnetic field determine the detailed form and width of the separated traces."
    },
    {
      "heading": "Physical principle",
      "text": "Magnetic force in a field gradient\n\nFor a magnetic moment \\(\\boldsymbol{\\mu}\\) in an external field \\(\\mathbf B\\), the interaction potential energy is\n\n\\[U=-\\boldsymbol{\\mu}\\cdot\\mathbf B.\\]\n\nIn a semiclassical description, the corresponding force is\n\n\\[\\mathbf F=\\boldsymbol{\\nabla} \\left(\\boldsymbol{\\mu}\\cdot\\mathbf B\\right).\\]\n\nIf the field near the beam points predominantly along the \\(z\\)-axis and its relevant gradient is also along that axis, this becomes approximately\n\n\\[F_z\\simeq \\mu_z\\frac{\\partial B_z}{\\partial z}.\\]\n\nThus atoms with different allowed magnetic-moment projections experience different transverse forces. The magnet converts an internal magnetic property into a spatial separation that can be detected.\n\nMore generally, the force on a particular internal state is obtained from the position dependence of that state’s magnetic interaction energy. The simple expression above assumes that the state follows the local field appropriately and that transitions between magnetic states can be neglected.\n\nWhy silver gives two principal components\n\nA silver atom in its electronic ground state has the electron configuration\n\n\\[[\\mathrm{Kr}]\\,4d^{10}5s^1.\\]\n\nThe filled shells contribute no net electronic angular momentum. The remaining \\(5s\\) electron has orbital angular momentum \\(L=0\\), but spin \\(S=\\tfrac12\\). Consequently, the atom’s total electronic angular momentum is \\(J=\\tfrac12\\), with two allowed projections:\n\n\\[J_z=m_J\\hbar, \\qquad m_J=+\\frac12,\\;-\\frac12.\\]\n\nHere \\(\\hbar=h/(2\\pi)\\), where \\(h\\) is the Planck constant. The electronic magnetic-moment projection is approximately\n\n\\[\\mu_z=-g_J\\mu_{\\mathrm B}m_J,\\]\n\nwhere \\(g_J\\) is close to 2 and \\(\\mu_{\\mathrm B}\\) is the Bohr magneton. The two projections therefore produce opposite deflections.\n\nThe minus sign expresses the fact that an electron’s magnetic moment is opposite to its spin angular momentum. Labels such as “spin up” should therefore not be identified automatically with an upward deflection on a diagram.\n\nThis two-state account neglects nuclear-spin effects and other fine details of the atomic energy levels. It captures the two principal components resolved in the original experiment.\n\nContrast with a classical prediction\n\nIn a classical model of magnetic dipoles with unrestricted orientations, the projection \\(\\mu_z\\) can vary continuously. An initially unoriented beam would therefore be expected to broaden into a continuous distribution of deflections.\n\nThe Stern–Gerlach result instead reveals discrete components. For the silver ground-state electronic system, there are two possible projections along any selected axis.\n\nThis does not mean that the spin vector is simply a classical arrow restricted to two fixed directions. Quantum theory assigns a definite value to a measured component, while components along different axes are incompatible observables."
    },
    {
      "heading": "Quantum-mechanical description",
      "text": "For an ideal spin-\\(\\tfrac12\\) system, the spin operators can be written as\n\n\\[S_i=\\frac{\\hbar}{2}\\sigma_i,\\]\n\nwhere \\(\\sigma_i\\) are the Pauli matrices. A Stern–Gerlach analyzer oriented along \\(z\\) distinguishes the eigenstates \\(|+z\\rangle\\) and \\(|-z\\rangle\\), whose eigenvalues are \\(+\\hbar/2\\) and \\(-\\hbar/2\\).\n\nAn incoming spin state may be a quantum superposition:\n\n\\[|\\chi\\rangle=a|+z\\rangle+b|-z\\rangle, \\qquad |a|^2+|b|^2=1.\\]\n\nAccording to the Born rule, the probabilities of the two outcomes are \\(|a|^2\\) and \\(|b|^2\\).\n\nThe magnet does more than exert two possible classical forces. Its interaction correlates spin with the atom’s spatial wave function. Schematically,\n\n\\[\\left(a|+z\\rangle+b|-z\\rangle\\right)|\\phi_0\\rangle \\longrightarrow a|+z\\rangle|\\phi_+\\rangle+ b|-z\\rangle|\\phi_-\\rangle,\\]\n\nwhere \\(|\\phi_+\\rangle\\) and \\(|\\phi_-\\rangle\\) describe the separated paths. This is entanglement between an internal degree of freedom and the atom’s motion.\n\nWhen the two spatial packets are sufficiently separated, detecting the path provides information about the spin projection. Spatial separation alone, however, is not identical to irreversible detection: in principle, a coherent apparatus could recombine the paths and reveal interference. Detection and environmental interactions generally introduce decoherence, suppressing observable coherence between them."
    },
    {
      "heading": "Successive Stern–Gerlach measurements",
      "text": "Sequences of idealized analyzers make the experiment especially useful for explaining quantum measurement.\n\nRepeating a measurement along the same axis\n\nSuppose a \\(z\\)-oriented analyzer separates an initially unpolarized beam, and only the \\(+z\\) output is retained. Sending this selected beam through a second ideal \\(z\\)-oriented analyzer gives only the \\(+z\\) result.\n\nThe first analyzer and subsequent selection have prepared a state with a definite \\(S_z\\). The second measurement confirms that value, provided no intervening interaction changes the spin.\n\nMeasuring along a different axis\n\nIf the selected \\(|+z\\rangle\\) beam enters an analyzer oriented along \\(x\\), it splits into equal \\(+x\\) and \\(-x\\) components. If only the \\(+x\\) component is retained and then measured along \\(z\\), both \\(z\\) outcomes occur with equal probability.\n\nThe intermediate measurement does not merely reveal a previously unknown \\(x\\)-component while preserving the original definite \\(z\\)-component. The spin operators obey\n\n\\[[S_x,S_z]=-i\\hbar S_y,\\]\n\nso they do not possess a complete common set of eigenstates. This provides a concrete illustration of incompatible observables and the uncertainty principle.\n\nMore generally, if a spin is prepared with projection \\(+\\hbar/2\\) along an axis \\(\\mathbf n\\), and the next analyzer’s axis \\(\\mathbf m\\) makes an angle \\(\\theta\\) with it, then\n\n\\[P(+\\mathbf m\\mid+\\mathbf n) =\\cos^2\\left(\\frac{\\theta}{2}\\right),\\]\n\n\\[P(-\\mathbf m\\mid+\\mathbf n) =\\sin^2\\left(\\frac{\\theta}{2}\\right).\\]\n\nThese analyzer sequences are standard idealizations used in teaching; they should not be confused with the single-analyzer arrangement of the original 1922 experiment."
    },
    {
      "heading": "Significance and limitations",
      "text": "The experiment established a direct connection between quantized angular-momentum projections and observable beam trajectories. Alongside atomic spectroscopy and the Zeeman effect, it helped show why classical descriptions of atomic magnetism were inadequate.\n\nIts broader importance includes:\n\n- State preparation: selecting one output prepares a beam with a specified spin projection.\n- Quantum measurement: an internal observable becomes correlated with a distinguishable detector outcome.\n- Atomic-beam methods: magnetic deflection provides a way to analyze and select internal atomic states.\n- Foundational studies: the apparatus illustrates the difference between a coherent superposition and a statistical mixture, described formally by a density matrix.\n\nSeveral qualifications limit what can be inferred from a simple beam-splitting diagram. The number of resolved outputs depends on the particle’s angular momentum, magnetic structure, field strength, and experimental resolution; not every atomic species gives two beams. Atomic velocity distributions and finite beam widths can blur the components, while changing field directions can cause transitions between states.\n\nThe standard neutral-atom arrangement also cannot be transferred straightforwardly to free electrons. Their electric charge introduces strong orbital deflection and additional beam-dynamical complications.\n\nFinally, the experiment does not by itself select one interpretation of quantum mechanics. The measured splitting and its probabilities are distinct from interpretive questions about state reduction and the measurement problem. Nor should electron spin be understood as the literal rotation of a small classical sphere: spin is an intrinsic quantum angular momentum."
    }
  ],
  "links": [
    "atom",
    "magnetic-field",
    "quantum-mechanics",
    "spin-angular-momentum",
    "lorentz-force",
    "magnetic-dipole-moment",
    "electric-charge",
    "torque",
    "potential-energy",
    "electron-configuration",
    "planck-constant",
    "quantum-superposition",
    "born-rule",
    "wave-function",
    "quantum-entanglement",
    "interference",
    "quantum-decoherence",
    "uncertainty-principle",
    "spectroscopy",
    "density-matrix",
    "electron",
    "measurement-problem"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}