Bose-Einstein Condensation — When Quantum Mechanics Becomes Visible
In 1924, Satyendra Nath Bose sent Einstein a paper. He was an unknown Indian physicist, with no position at a major institution, working in Dhaka. His paper derived Planck’s blackbody spectrum by treating photons as indistinguishable particles, without reference to classical electromagnetism.
Einstein immediately recognized its importance. He translated it into German himself and submitted it to Zeitschrift für Physik. Then he extended it: in two papers (1924, 1925), Einstein applied Bose’s statistics to a gas of massive particles and predicted something extraordinary. Below a critical temperature, a macroscopic fraction of the particles would condense into the lowest energy state, the ground state, forming a new state of matter: the Bose-Einstein condensate (BEC).
Einstein himself was uncertain about the physics: “The theory is pretty, but is there some truth in it?”
It took 70 years to find out. In June 1995, Eric Cornell and Carl Wieman created the first dilute-gas BEC in the laboratory, about 2,000 rubidium-87 atoms cooled to roughly 20 nanokelvin. A few months later, Wolfgang Ketterle’s group created a much larger sodium condensate. They received the Nobel Prize in 2001. Einstein’s prediction was correct.
I. Quantum Statistical Mechanics: The Foundation
To understand BEC, we need quantum statistical mechanics.
Identical particles: In quantum mechanics, identical particles are genuinely indistinguishable, not merely similar. For \(N\) identical particles, the many-body wavefunction must be either symmetric (bosons),
\[\Psi(..., x_i, ..., x_j, ...) = +\Psi(..., x_j, ..., x_i, ...)\]or antisymmetric (fermions),
\[\Psi(..., x_i, ..., x_j, ...) = -\Psi(..., x_j, ..., x_i, ...)\]The grand canonical ensemble: For a system in contact with a heat bath (temperature \(T\)) and particle reservoir (chemical potential \(\mu\)), the mean occupation number of single-particle state \(\epsilon_i\) follows one of three distributions. The Bose-Einstein distribution:
\[\langle n_i \rangle_{BE} = \frac{1}{e^{(\epsilon_i - \mu)/k_BT} - 1}\]the Fermi-Dirac distribution:
\[\langle n_i \rangle_{FD} = \frac{1}{e^{(\epsilon_i - \mu)/k_BT} + 1}\]and the Maxwell-Boltzmann distribution (classical limit):
\[\langle n_i \rangle_{MB} = e^{-(\epsilon_i - \mu)/k_BT}\]For bosons, the chemical potential must satisfy \(\mu < \epsilon_0\) (ground state energy) to keep all occupation numbers positive. As \(\mu \to \epsilon_0^-\),
\[\langle n_0 \rangle_{BE} = \frac{1}{e^{(\epsilon_0 - \mu)/k_BT} - 1} \to \infty\]The ground state occupation diverges: this is BEC.
II. The Critical Temperature and Condensate Fraction
The ideal Bose gas in 3D: For a uniform, non-interacting Bose gas in volume \(V\) with \(N\) particles, the density of states in 3D is
\[g(\epsilon) = \frac{V}{4\pi^2}\left(\frac{2m}{\hbar^2}\right)^{3/2}\epsilon^{1/2}\]At the condensation threshold, the maximum number of particles that can occupy excited states is
\[N_{exc}^{\max} = \int_0^\infty g(\epsilon)\langle n(\epsilon)\rangle d\epsilon = \frac{V}{\lambda_{dB}^3}\zeta(3/2)\]where \(\lambda_{dB} = \sqrt{2\pi\hbar^2/mk_BT}\) is the thermal de Broglie wavelength and \(\zeta(3/2) \approx 2.612\) is the Riemann zeta function.
The critical condition: BEC occurs when \(N_{exc} = N\), when the excited states can no longer accommodate all the particles. The critical temperature is
\[T_c = \frac{2\pi\hbar^2}{mk_B}\left(\frac{n}{\zeta(3/2)}\right)^{2/3}\]where \(n = N/V\) is the number density. Equivalently, BEC occurs when the thermal de Broglie wavelength becomes comparable to the interparticle spacing:
\[n\lambda_{dB}^3 \geq \zeta(3/2) \approx 2.612\]Physical interpretation: At high temperature, \(\lambda_{dB} \ll n^{-1/3}\) (de Broglie wavelength much smaller than interparticle spacing), exchange effects become negligible, and classical behavior emerges. At \(T \sim T_c\), \(\lambda_{dB} \sim n^{-1/3}\): wavefunctions overlap, quantum degeneracy sets in. Below \(T_c\), there is macroscopic occupation of the ground state.
The condensate fraction: Below \(T_c\),
\[\frac{N_0}{N} = 1 - \left(\frac{T}{T_c}\right)^{3/2}\]At \(T = 0\), all particles are in the ground state; at \(T = T_c\), the condensate fraction vanishes.
Typical experimental values: For rubidium-87 at density \(n \sim 10^{14}\) cm\(^{-3}\), the uniform ideal-gas estimate gives \(T_c \approx 400\) nK. Real trapped gases have geometry- and interaction-dependent transition temperatures, often in the 100 nK range. These temperatures are unimaginably cold, far colder than interstellar space (\(\sim 3\) K).
III. The Order Parameter and Spontaneous Symmetry Breaking
A BEC is characterized by an order parameter, the condensate wavefunction or macroscopic wavefunction:
\[\Psi(\mathbf{r}, t) = \sqrt{n_0(\mathbf{r}, t)} e^{i\phi(\mathbf{r}, t)}\]where \(n_0\) is the condensate density and \(\phi\) is the phase. This is the expectation value of the bosonic field operator:
\[\Psi(\mathbf{r}, t) = \langle\hat{\psi}(\mathbf{r}, t)\rangle\]In a normal Bose gas, \(\langle\hat{\psi}\rangle = 0\); in a BEC, \(\langle\hat{\psi}\rangle \neq 0\).
Spontaneous symmetry breaking: The Hamiltonian is invariant under global \(U(1)\) phase rotation, \(\hat{\psi} \to e^{i\alpha}\hat{\psi}\). Below \(T_c\), the condensate spontaneously selects a phase \(\phi_0\), breaking the \(U(1)\) symmetry. The broken symmetry is associated with conservation of particle number. In the thermodynamic limit, different phase choices give inequivalent ground states: there are infinitely many degenerate condensates, all related by \(U(1)\) rotations.
Off-diagonal long-range order (ODLRO): The signature of BEC in the density matrix formalism (Penrose & Onsager, 1956) is the one-body density matrix:
\[\rho^{(1)}(\mathbf{r}, \mathbf{r}') = \langle\hat{\psi}^\dagger(\mathbf{r})\hat{\psi}(\mathbf{r}')\rangle\]For a normal system, \(\rho^{(1)}(\mathbf{r}, \mathbf{r}') \to 0\) as \(\lvert\mathbf{r} - \mathbf{r}'\rvert \to \infty\). For a BEC,
\[\rho^{(1)}(\mathbf{r}, \mathbf{r}') \to \Psi^*(\mathbf{r})\Psi(\mathbf{r}')\]as \(\lvert\mathbf{r} - \mathbf{r}'\rvert \to \infty\): the off-diagonal elements remain finite at large separations, off-diagonal long-range order. ODLRO is the defining property of a BEC; it holds even in interacting systems where other definitions break down.
IV. The Gross-Pitaevskii Equation
For a dilute, weakly interacting BEC at zero temperature, the condensate wavefunction obeys the Gross-Pitaevskii equation (GPE):
\[i\hbar\frac{\partial\Psi}{\partial t} = \left[-\frac{\hbar^2\nabla^2}{2m} + V_{ext}(\mathbf{r}) + g\lvert\Psi\rvert^2\right]\Psi\]This is a nonlinear Schrödinger equation. The interaction term \(g\lvert\Psi\rvert^2\) encodes the mean-field effect of all other condensate atoms on a single atom. The coupling constant is
\[g = \frac{4\pi\hbar^2 a_s}{m}\]where \(a_s\) is the s-wave scattering length, the key parameter characterizing two-body interactions at ultracold temperatures.
The Thomas-Fermi approximation: In the limit of large condensates (\(Na_s/a_{HO} \gg 1\), where \(a_{HO}\) is the harmonic oscillator length), the kinetic energy is negligible:
\[0 = \left[V_{ext}(\mathbf{r}) + g\lvert\Psi\rvert^2 - \mu\right]\Psi\]For a harmonic trap \(V_{ext} = \frac{1}{2}m\omega^2 r^2\),
\[n_0(\mathbf{r}) = \lvert\Psi\rvert^2 = \frac{\mu - \frac{1}{2}m\omega^2 r^2}{g}\]for \(r < R_{TF}\), and \(0\) otherwise. The Thomas-Fermi radius is
\[R_{TF} = \sqrt{\frac{2\mu}{m\omega^2}} = a_{HO}\left(\frac{15Na_s}{a_{HO}}\right)^{1/5}\]The inverted parabola density profile, characteristic of the Thomas-Fermi condensate, is observed experimentally and is a signature of BEC.
Collective modes: Linearizing the GPE around the equilibrium,
\[\Psi = (\sqrt{n_0} + \delta\psi)e^{-i\mu t/\hbar}\]gives the Bogoliubov-de Gennes equations:
\[\hbar\omega u = \left[-\frac{\hbar^2\nabla^2}{2m} + V_{ext} - \mu + 2gn_0\right]u + gn_0 v\] \[-\hbar\omega v = \left[-\frac{\hbar^2\nabla^2}{2m} + V_{ext} - \mu + 2gn_0\right]v + gn_0 u\]These describe the collective excitations, the density oscillations of the condensate. For a uniform condensate, the Bogoliubov spectrum is
\[\epsilon_k = \sqrt{\frac{\hbar^2k^2}{2m}\left(\frac{\hbar^2k^2}{2m} + 2gn\right)}\]which has two limits. At low \(k\) (phonon regime), \(\epsilon_k \approx \hbar c_s k\): a linear dispersion, sound waves, with speed of sound \(c_s = \sqrt{gn/m}\). At high \(k\) (free particle regime), \(\epsilon_k \approx \frac{\hbar^2k^2}{2m} + gn\): a quadratic dispersion, free particles with a mean-field energy shift. The crossover occurs at the healing length:
\[\xi = \frac{1}{\sqrt{8\pi n a_s}}\]The healing length is the length scale over which the condensate density recovers from a local perturbation. It sets the size of vortex cores.
V. Superfluidity and the Two-Fluid Model
BEC is intimately connected to superfluidity, but they are not identical.
Landau’s criterion for superfluidity: A superfluid can flow without viscosity if the velocity is below the Landau critical velocity:
\[v_c = \min_k \frac{\epsilon_k}{\hbar k}\]For a BEC with Bogoliubov spectrum, \(v_c = c_s = \sqrt{gn/m}\). Above \(v_c\), the fluid can create excitations and dissipate energy, normal fluid behavior; below \(v_c\), no excitations can be created, superfluid behavior. This is Landau’s criterion, and for a weakly interacting BEC it explains the onset of superfluid behavior.
The two-fluid model (Tisza, Landau): Below \(T_\lambda\), the superfluid consists of two interpenetrating components. The normal fluid, with density \(\rho_n(T)\), carries entropy and has viscosity. The superfluid, with density \(\rho_s(T) = \rho - \rho_n\), carries no entropy and has zero viscosity. At \(T = 0\), \(\rho_s = \rho\) (all superfluid); at \(T = T_\lambda\), \(\rho_s = 0\) (all normal).
The superfluid density is not the condensate fraction, a subtle but important distinction. In helium-4, the condensate fraction at \(T = 0\) is only about \(10\%\) (due to strong interactions), but the superfluid fraction is \(100\%\). Interactions deplete the condensate without destroying superfluidity.
Vortices: In a superfluid, the velocity field is irrotational,
\[\mathbf{v}_s = \frac{\hbar}{m}\nabla\phi\]where \(\phi\) is the phase of the condensate. For a simply connected region, \(\nabla \times \mathbf{v}_s = 0\). Circulation is quantized:
\[\oint \mathbf{v}_s \cdot d\mathbf{l} = \frac{2\pi\hbar}{m} n, \quad n \in \mathbb{Z}\]Rotation is accommodated by quantized vortices, topological defects where the density vanishes and the phase winds by \(2\pi n\). The vortex core size is on the order of \(\xi\) (the healing length). In rotating BECs, vortices form an Abrikosov lattice, a triangular array of vortices, with density
\[n_v = \frac{m\Omega}{\pi\hbar}\]where \(\Omega\) is the rotation frequency. This has been observed directly in rotating BECs (Cornell, Ketterle groups, 2001-2003).
VI. Feshbach Resonances: Tunable Interactions
One of the most powerful tools in ultracold physics is the Feshbach resonance. Near a Feshbach resonance, the s-wave scattering length \(a_s\) can be tuned continuously by an external magnetic field:
\[a_s(B) = a_{bg}\left(1 - \frac{\Delta B}{B - B_0}\right)\]where \(a_{bg}\) is the background scattering length, \(B_0\) is the resonance field, and \(\Delta B\) is the resonance width. Near \(B_0\), \(a_s \to \pm\infty\) (the unitary limit); far from \(B_0\), \(a_s \to a_{bg}\).
What this enables: By tuning through \(B_0\), you can change the interaction from repulsive (\(a_s > 0\)) to attractive (\(a_s < 0\)).
The BEC-BCS crossover: For two-component fermions (two spin states of \(^6\)Li or \(^{40}\)K), when \(a_s > 0\), pairs of fermions form molecules, bosonic dimers that BEC: this is the BEC side. When \(a_s < 0\), Cooper pairing gives BCS superconductivity: this is the BCS side. At the unitary point (\(a_s \to \infty\)), the system is a strongly interacting quantum fluid with universal behavior, its properties determined only by \(n\) and \(T\), independent of the microscopic details.
The BEC-BCS crossover was realized experimentally in 2003-2004 (Ketterle, Grimm, Jin groups). It connects BEC (dilute bosons), BCS superconductivity, and, qualitatively, high-\(T_c\) superconductivity in a single experimental system.
Efimov states: Near a Feshbach resonance, three-body physics exhibits Efimov states (Efimov, 1970), an infinite tower of three-body bound states with geometric spectrum:
\[E_n^{(3)} = e^{-2\pi n/s_0} E_0^{(3)}\]where \(s_0 \approx 1.00624\) for identical bosons. The ratio of successive Efimov length scales is \(e^{\pi/s_0} \approx 22.7\), and the corresponding ratio of binding-energy scales is \(e^{2\pi/s_0} \approx 515\). This log-periodic scaling is a discrete scale invariance: the three-body system repeats itself under a change of length scale by about \(22.7\). Efimov states were first observed in ultracold cesium (2006) and are now a rich field of study.
VII. Optical Lattices and Quantum Simulation
One of the most powerful applications of BECs is quantum simulation. By shining counterpropagating laser beams, you create an optical lattice, a periodic potential from the AC Stark effect:
\[V_{latt}(\mathbf{r}) = V_0 \sin^2(kx)\]Atoms tunnel between sites, and the system is described by the Bose-Hubbard model:
\[\hat{H} = -J\sum_{\langle i,j\rangle}\left(\hat{a}_i^\dagger\hat{a}_j + \hat{a}_j^\dagger\hat{a}_i\right) + \frac{U}{2}\sum_i \hat{n}_i(\hat{n}_i - 1) - \mu\sum_i \hat{n}_i\]Two competing terms decide the physics: the kinetic energy (tunneling \(J\)) lets atoms delocalize, favoring the superfluid; the interaction energy (\(U\)) makes atoms localize, favoring the Mott insulator.
The superfluid-Mott insulator transition: At \(U/J \ll (U/J)_c\), the system is a superfluid with delocalized atoms. At \(U/J > (U/J)_c\), it becomes a Mott insulator with integer occupation per site (one atom per site at unit filling). This quantum phase transition was realized by Greiner et al. (2002), a landmark experiment in ultracold physics.
The Mott insulator shows incompressibility, an energy gap, and no phase coherence. The superfluid shows sharp interference peaks, phase coherence, and compressibility. The transition is driven purely by quantum fluctuations (at \(T = 0\)): a quantum phase transition.
Quantum simulation of lattice models: By engineering optical lattices, you can simulate the Fermi-Hubbard model (high-\(T_c\) superconductivity), Heisenberg spin chains (quantum magnetism), topological band structures (topological insulators), gauge fields (artificial magnetic fields), and disordered systems (Anderson localization). For generic strongly interacting systems, these problems become classically intractable as the Hilbert space grows exponentially. Ultracold atoms in optical lattices are, in this sense, quantum simulators for condensed matter physics.
VIII. Spinor Condensates and Topological Defects
When atoms have internal spin degrees of freedom, the condensate is a spinor BEC. For spin-1 bosons (e.g., \(^{23}\)Na, \(^{87}\)Rb), the order parameter is a three-component vector:
\[\Psi = \begin{pmatrix}\Psi_{+1}\\\Psi_0\\\Psi_{-1}\end{pmatrix}\]The interaction Hamiltonian is
\[\hat{H}_{int} = \frac{c_0}{2}n^2 + \frac{c_2}{2}\lvert\mathbf{F}\rvert^2\]where \(\mathbf{F}\) is the spin density. In the ferromagnetic case (\(c_2 < 0\), e.g., \(^{87}\)Rb), spins align, giving a magnetized condensate. In the antiferromagnetic or polar case (\(c_2 > 0\), e.g., \(^{23}\)Na), spins anti-align, giving spin-nematic order.
Topological defects in spinor BECs: The order parameter manifold determines the possible topological defects. For the polar phase, the order parameter manifold is \(\cong (S^2 \times U(1))/\mathbb{Z}_2\), which allows vortices (\(\pi_1\)), skyrmions (\(\pi_2\)), and monopoles (\(\pi_2\)). Half-quantum vortices, with circulation \(h/2m = \pi\hbar/m\), are topologically stable in the polar phase. More exotic spin-2 condensates, particularly cyclic and biaxial-nematic phases, can support non-Abelian vortices, providing a laboratory for noncommuting topological defects.
IX. Analogue Gravity
One of the most remarkable applications of BECs is analogue gravity (Unruh, 1981; Garay et al., 2000).
The acoustic metric: In a flowing BEC, phonons (Bogoliubov quasi-particles) propagate in an effective curved spacetime, the acoustic metric:
\[ds^2 = \frac{n_0}{mc_s}\left[-(c_s^2 - v^2)dt^2 - 2v_i dx^i dt + \delta_{ij}dx^i dx^j\right]\]where \(\mathbf{v}\) is the superfluid velocity and \(c_s\) is the speed of sound. Where the flow crosses from subsonic to supersonic (\(\lvert\mathbf{v}\rvert = c_s\)), an acoustic horizon can form, the analogue of a black hole event horizon. In the black-hole configuration, phonons cannot escape from the supersonic region, just as light cannot escape a black hole.
Analogue Hawking radiation: Unruh showed (1981) that an acoustic horizon should emit thermal radiation, analogue Hawking radiation, at temperature
\[T_H = \frac{\hbar}{2\pi k_B}\left\lvert\frac{\partial(v - c_s)}{\partial x}\right\rvert_{horizon}\]Steinhauer reported its observation in a BEC in 2016. The result became a landmark claim in analogue-gravity experiments, although its interpretation has been debated.
The significance is one of scale: Hawking radiation from real black holes is unmeasurably tiny (\(T_H \sim 10^{-8}\) K for solar-mass black holes), while the BEC analogue lets us study the physics of horizons in the laboratory.
Analogue cosmology: A BEC with time-varying interactions (via Feshbach resonance) can simulate expanding universes. The phonon spectrum can simulate the expansion of the universe and the generation of cosmological perturbations, and analogue inflation has been simulated in ultracold atom experiments.
X. Beyond Mean Field: Strongly Correlated Systems
The GPE is a mean-field theory, valid when \(na_s^3 \ll 1\) (dilute, weakly interacting). Beyond mean field, the leading correction to the ground state energy is given by the Lee-Huang-Yang correction:
\[E/N = \frac{gn}{2}\left[1 + \frac{128}{15\sqrt{\pi}}\sqrt{na_s^3} + ...\right]\]The \(\sqrt{na_s^3}\) term is the Lee-Huang-Yang (LHY) correction, the leading beyond-mean-field effect.
Quantum droplets: In mixtures of two BECs, or in dipolar BECs, the LHY correction can stabilize quantum droplets, self-bound liquid-like states. In a two-component mixture, the attractive mean-field interaction is tuned close to collapse, while the repulsive LHY correction from quantum fluctuations stabilizes it. The result is self-bound droplets with surface tension from quantum fluctuations, a nearly constant-density interior, and liquid-like density. These were observed in 2018 (Cabrera et al., Semeghini et al.). They are a new state of matter, a liquid whose collapse is prevented by quantum fluctuations.
Dipolar condensates: Atoms with large magnetic dipole moments (\(^{164}\)Dy, \(^{168}\)Er, \(^{52}\)Cr) have long-range anisotropic interactions. Dipolar BECs show a roton-maxon dispersion relation, a supersolid phase (simultaneous superfluid and crystalline order), and self-bound droplets.
The supersolid, a state that is simultaneously a superfluid (zero viscosity) and a crystal (periodic density modulation), was observed in dipolar BECs in 2019. It was long thought impossible, how can a system be both superfluid and solid, and yet it is possible: nature found a way.
XI. BEC in Other Systems
BEC is not only about ultracold atoms.
Exciton-polariton condensates: In semiconductor microcavities, photons are strongly coupled to excitons, creating polaritons, half-light, half-matter quasiparticles. Polaritons are bosons with extremely light effective mass (\(\sim 10^{-4} m_e\)), and BEC of polaritons occurs at temperatures \(T \sim 10-300\) K, far above ultracold atom condensates. Room-temperature polariton condensation has been achieved. The condensate shows spontaneous coherence, superfluidity, and vortices, but it is a driven-dissipative condensate, constantly losing photons and gaining from the pump. This non-equilibrium character makes it fundamentally different from equilibrium BECs.
Photon BEC: Photons in a dye-filled microcavity can thermalize via absorption and re-emission. A BEC of photons was realized by Klaers et al. (2010). Photons acquire an effective mass from the cavity, thermalize with the dye molecules, and below a critical temperature, form a BEC.
Magnon BEC: Magnons, spin-wave quasiparticles in magnetic materials, can form BECs at room temperature under microwave pumping. This has been observed in YIG (yttrium iron garnet) films.
Nuclear matter BEC: In dilute nuclear matter, alpha particles (\(^4\)He nuclei) may form condensate-like phases. Neutron superfluidity in neutron stars comes from Cooper pairing of neutrons, and at low density, neutron matter is closely connected to BCS-BEC crossover physics.
XII. Why BEC Matters
Let me state explicitly why Bose-Einstein condensation is important.
As a phenomenon: BEC is the cleanest realization of quantum mechanics at macroscopic scales. A condensate containing \(10^6\) atoms behaves as a single quantum object, described by a single wavefunction with a single phase. It makes quantum interference visible to the naked eye, in absorption images.
As a laboratory: Ultracold atoms are the most controllable quantum systems available. You can tune interactions continuously (Feshbach), engineer arbitrary potentials (optical lattices), measure single atoms (quantum gas microscopes), prepare highly controlled quantum states, and control dimensionality (1D, 2D, 3D). This makes BEC systems ideal for testing many-body theories, simulating condensed matter models, studying quantum phase transitions, and exploring non-equilibrium dynamics.
As a connection: BEC connects quantum statistics (Bose-Einstein), superfluidity (Landau, Tisza), superconductivity (BCS), quantum field theory (spontaneous symmetry breaking, Goldstone modes), high energy physics (the Higgs mechanism, spontaneous symmetry breaking, vacuum condensates), cosmology (condensate descriptions in some inflationary models), and quantum information (topological defects and non-Abelian vortex physics).
The Higgs vacuum is a condensate: the universe is permeated by a nonzero Higgs field. The mathematical language developed for condensates reappears in the vacuum of particle physics. This is the deep unity of physics, the same mathematical structure appearing at radically different scales, in radically different contexts, described by the same equations.
Coda
When Satyendra Nath Bose sent his paper to Einstein in 1924, he wrote: “I have ventured to send you the accompanying article for your perusal and opinion. I am quite unaware whether it contains anything of value.” He was an unknown physicist from a colonial outpost of the British Empire, writing to the most famous scientist in the world.
Einstein read it. He recognized it immediately. He wrote back, on a postcard dated 2 July 1924: “It signifies an important step forward and I liked it very much.” He translated it himself, submitted it to the best journal, then extended it, predicting a new state of matter that would not be observed for 70 years. Bose never received the Nobel Prize; Einstein received it for something else. The condensate bears both their names.
At 10:54 in the morning, June 5, 1995, in a lab at JILA in Boulder, Cornell and Wieman watched a cloud of about 2,000 rubidium atoms, cooled to roughly 20 nanokelvin, collapse into a single point of light, narrower than a human hair. “It really is a new form of matter,” Wieman said. Three months later, Ketterle’s group at MIT condensed a much larger cloud of sodium atoms.
This was it. Seventy years after Einstein’s prediction: a new state of matter, quantum mechanics made macroscopic, the most precise control of quantum systems ever achieved. And a window into the deep structure of reality, from rubidium atoms at tens of nanokelvin to the Higgs field that fills all of space to the quantum vacuum that seethes beneath everything.
The same physics. Different scales. One universe. (yeah, physics has no problem with being romantic!)
Essential reading
Bose, “Plancks Gesetz und Lichtquantenhypothese” (1924), Z. Phys. 26, 178, and Einstein, “Quantentheorie des einatomigen idealen Gases” (1924, 1925), Sitz. Ber. Preuss. Akad. Wiss. 1924, 261; 1925, 3: the original predictions.
Anderson et al., “Observation of Bose-Einstein Condensation in a Dilute Atomic Vapor” (1995), Science 269, 198, and Davis et al., “Bose-Einstein Condensation in a Gas of Sodium Atoms” (1995), Phys. Rev. Lett. 75, 3969: the experimental discoveries.
Pitaevskii & Stringari, Bose-Einstein Condensation and Superfluidity (2016), Oxford University Press: the definitive textbook.
Pethick & Smith, Bose-Einstein Condensation in Dilute Gases (2002), Cambridge University Press: excellent graduate text.
Leggett, Quantum Liquids (2006), Oxford University Press: BEC, superfluidity, superconductivity unified.
For reviews
Dalfovo et al., “Theory of Bose-Einstein condensation in trapped gases” (1999), Rev. Mod. Phys. 71, 463.
Bloch, Dalibard & Zwerger, “Many-body physics with ultracold gases” (2008), Rev. Mod. Phys. 80, 885: the comprehensive review of the field.
Enjoy Reading This Article?
Here are some more articles you might like to read next: