Spotlight #66: Condensed Matter Physics — Ising, BEC, Superconductors & Spin Glasses

Here is the surprising fact: roughly 40% of all published physics research is condensed matter physics. More Nobel prizes have been awarded in this field than in any other area of physics. Yet the name is almost unknown outside academia, because what condensed matter physicists study tends to go by other names when it reaches the public: superconductivity, semiconductors, magnetism, superfluidity, topological insulators. The simulations in this spotlight let you explore five foundational phenomena that underpin this vast and consequential field.

I. The Ising Model — How Order Emerges from Chaos

The Ising model is deceptively simple: place a grid of spins, each of which can point either up (+1) or down (−1), on a lattice. Each spin interacts only with its immediate neighbours. Yet this toy model captures the essential physics of ferromagnetic phase transitions — the spontaneous emergence of macroscopic order from microscopic randomness when temperature falls below a critical value.

The energy of the system is governed by the Hamiltonian:

H = -J * sum_{} s_i * s_j - mu * B * sum_i s_i

  J    : coupling constant (J > 0 ferromagnetic, J < 0 antiferromagnetic)
  s_i  : spin at site i (+1 or -1)
   : sum over nearest-neighbour pairs
  mu   : magnetic moment
  B    : external magnetic field

Critical temperature (2D square lattice, exact Onsager solution):
  k_B * T_c = 2J / ln(1 + sqrt(2))  ≈ 2.269 J

Order parameter: magnetisation m = (1/N) * sum_i s_i
  m ≈ 0   for T >> T_c  (disordered, paramagnetic)
  m ≈ ±1  for T << T_c  (ordered, ferromagnetic)

The phase transition at T_c is a second-order transition: the magnetisation rises continuously from zero as temperature falls below the critical point. Near T_c, the system exhibits critical phenomena — scale-invariant fluctuations, power-law correlations, and diverging susceptibility — that are described by universal critical exponents independent of microscopic details. The 2D Ising model is one of the very few exactly solvable models in statistical mechanics, solved by Lars Onsager in 1944 in a tour de force of mathematical physics.

The simulation uses the Metropolis–Hastings algorithm to sample the Boltzmann distribution: propose a random spin flip, accept it if it lowers energy, accept it with probability exp(-ΔE / k_B T) if it raises energy. This Monte Carlo approach naturally produces thermal fluctuations, domain walls, and the gradual nucleation of magnetic order as you drag the temperature slider below T_c.

II. Bose–Einstein Condensation — The Fifth State of Matter

In 1924, Satyendra Nath Bose and Albert Einstein predicted that bosons — particles with integer spin — would, at sufficiently low temperatures, all collapse into the single lowest-energy quantum state available to them. Unlike fermions (which obey the Pauli exclusion principle and must occupy different states), bosons actively prefer to crowd into the same state. The result is a macroscopic quantum object: the Bose–Einstein condensate (BEC).

The transition occurs when the thermal de Broglie wavelength of the particles becomes comparable to the inter-particle spacing — roughly when quantum wavefunctions begin to overlap:

Thermal de Broglie wavelength:
  lambda_dB = h / sqrt(2 * pi * m * k_B * T)

Critical temperature (3D ideal Bose gas):
  T_c = (h^2 / (2 * pi * m * k_B)) * (n / zeta(3/2))^(2/3)
  zeta(3/2) ≈ 2.612  (Riemann zeta function)
  n : number density

Condensate fraction below T_c:
  N_0 / N = 1 - (T / T_c)^3

The BEC is described by a single macroscopic wavefunction (order parameter):
  Psi(r, t) = sqrt(n_0) * exp(i * phi(r, t))
  n_0 : condensate density
  phi : macroscopic phase (coherent across the whole condensate)

BECs were first created in the laboratory in 1995 by Eric Cornell and Carl Wieman (working with rubidium atoms) and by Wolfgang Ketterle (working with sodium atoms), earning the 2001 Nobel Prize in Physics. The temperatures involved — around 100 nanokelvin, the coldest places in the known universe — require sophisticated laser cooling and magnetic trapping techniques.

The most striking property of a BEC is its superfluidity: below the condensation temperature, the quantum coherence of the condensate suppresses dissipation, and the fluid flows with zero viscosity. Stir a superfluid fast enough and it generates quantised vortices — vortex lines around which the phase of the order parameter winds by exactly 2π, each carrying one quantum of angular momentum.

III. Josephson Junctions — Quantum Tunnelling at Macroscopic Scale

A Josephson junction is one of the most remarkable quantum devices ever conceived: two superconductors separated by a thin insulating barrier, across which Cooper pairs (bound pairs of electrons responsible for superconductivity) tunnel quantum mechanically. The entire junction is described by just two numbers — the phase difference φ between the macroscopic wavefunctions of the two superconductors, and the critical current I_c. Yet from these two numbers emerge two profound phenomena.

DC Josephson effect (zero voltage, non-zero current):
  I = I_c * sin(phi)
  phi : phase difference across the junction

AC Josephson effect (non-zero voltage produces oscillating current):
  d(phi)/dt = (2e / hbar) * V
  f_J = (2e / h) * V  (Josephson frequency)
  f_J ≈ 483.6 GHz / mV  (exact, used as voltage standard)

RCSJ model (Resistively and Capacitively Shunted Junction):
  C * d^2(phi)/dt^2 + (1/R) * d(phi)/dt + I_c * sin(phi) = I_bias

  This is mathematically identical to a damped driven pendulum:
  m * d^2(theta)/dt^2 + b * d(theta)/dt + (mg/L) * sin(theta) = tau

Critical current suppression in magnetic field (Fraunhofer pattern):
  I_c(Phi) = I_c0 * |sin(pi * Phi / Phi_0) / (pi * Phi / Phi_0)|
  Phi_0 = h / 2e ≈ 2.07e-15 Wb  (magnetic flux quantum)

The DC Josephson effect is extraordinary: a supercurrent flows with zero voltage applied, purely driven by the quantum phase difference. The AC Josephson effect — whereby a DC voltage produces an oscillating supercurrent at a precise frequency — is so accurate that it now defines the volt in SI units. Modern voltage standards worldwide are based on Josephson junction arrays.

IV. Spin Glasses — Frustration and Frozen Disorder

What happens when you take an Ising model and make the coupling constants J_{ij} random — sometimes ferromagnetic, sometimes antiferromagnetic, at each bond chosen randomly? The result is a spin glass: a system so frustrated that it cannot find a simple ordered ground state and instead freezes into one of an astronomically large number of metastable configurations, all with nearly the same energy.

Spin glass Hamiltonian:
  H = -sum_{} J_ij * s_i * s_j

Edwards-Anderson order parameter (measures local freezing, not global order):
  q_EA = lim_{t -> inf} (1/N) * sum_i 
  q_EA = 0  in paramagnetic phase (spins fluctuate freely)
  q_EA > 0  in spin glass phase (spins frozen, but randomly oriented)

Frustration: a triangle of antiferromagnetic bonds (J < 0) is frustrated
  because it is impossible to satisfy all three bonds simultaneously.
  On a triangular lattice with J < 0 everywhere:
    any configuration leaves at least one bond violated.

Free energy landscape:
  Exponentially many local minima separated by barriers proportional to sqrt(N).
  Equilibration time ~ exp(alpha * N^sigma): grows faster than any polynomial.

Spin glasses exhibit ageing — the older the sample, the more slowly it responds to a perturbation — and memory effects, whereby a spin glass “remembers” its thermal history. These properties make spin glasses a prototype for a much broader class of problems: any complex system with many competing constraints and an exponentially rugged energy landscape. Protein folding, neural networks, optimisation heuristics, and even social dynamics have been analysed using spin glass theory.

The Sherrington–Kirkpatrick (SK) model extends this to infinite-range interactions and was solved by Giorgio Parisi in the 1980s using the replica method — one of the most technically demanding calculations in theoretical physics, and the work for which Parisi received the 2021 Nobel Prize in Physics.

V. Topological Edge States — The SSH Model

The discovery that quantum states of matter can be classified not just by symmetry breaking (as Landau theory classifies ferromagnets and superconductors) but by topology — a global property of the wavefunction that cannot be changed by smooth deformations — was one of the landmark discoveries of late-twentieth-century physics, earning the 2016 Nobel Prize for Thouless, Haldane, and Kosterlitz. The simplest model that captures topological physics is the Su–Schrieffer–Heeger (SSH) chain.

SSH Hamiltonian (1D chain of dimers):
  H = -sum_i [ t1 * c_{i,A}^dag * c_{i,B} + t2 * c_{i,B}^dag * c_{i+1,A} ] + h.c.

  t1 : intra-cell hopping amplitude
  t2 : inter-cell hopping amplitude

Bloch Hamiltonian in momentum space:
  H(k) = (t1 + t2 * cos(k)) * sigma_x + t2 * sin(k) * sigma_y

Winding number (topological invariant):
  nu = (1 / 2pi) * integral_BZ d(phi(k))
  phi(k) = arg(t1 + t2 * exp(ik))

  nu = 0  if |t1| > |t2|  (trivial phase, no edge states)
  nu = 1  if |t1| < |t2|  (topological phase, zero-energy edge states)

Bulk-boundary correspondence:
  A chain in the topological phase (nu = 1) with open boundary conditions
  hosts exactly two zero-energy states localised at the two ends,
  regardless of chain length (for large enough N).

The topological phase transition occurs at t1 = t2: the bulk band gap closes and reopens, and the winding number jumps from 0 to 1 (or back). The edge states are “topologically protected” — they cannot be removed by any perturbation that preserves the symmetries of the Hamiltonian and does not close the bulk gap. In real materials, this robustness means that edge conductance in topological insulators is immune to impurities, disorder, and surface imperfections.

The SSH model was originally proposed to describe conducting polymers (polyacetylene) in 1979 and helped explain why doped polyacetylene could carry electrical current despite having a nominal band gap. Today, SSH physics appears in cold atoms in optical lattices, photonic crystals, acoustic metamaterials, and the edge states of two-dimensional topological insulators such as HgTe quantum wells — all described by the same topological invariant and the same bulk-boundary correspondence.

Try It Yourself

The five simulations covered in this spotlight span the central concepts of condensed matter physics. We suggest exploring them in the order presented: the Ising model builds intuition for phase transitions; the BEC simulation shows what happens when quantum statistics, rather than interactions, drive a phase transition; the Josephson junction demonstrates quantum coherence at macroscopic scale; the spin glass reveals what frustration does to a system's ability to order; and the SSH model introduces the topological perspective that has reshaped the field since the 1980s.

• Ising Model — Magnetic Phase Transition — drag the temperature across T_c and watch domains nucleate
• Bose–Einstein Condensate — cool below the critical temperature and observe the condensate fraction grow
• Josephson Junction — increase the bias current past the critical current and observe the voltage jump
• Spin Glass — compare cooling rate effects on the frozen configuration
• Topological Insulator (SSH) — tune t1/t2 across the phase boundary and watch edge states appear

Closing Thought

Condensed matter physics is sometimes called “the physics of everything else” — everything, that is, that is not an elementary particle or a star. The phenomena in this spotlight are not curiosities confined to ultra-low-temperature laboratories. Josephson junctions underlie the most sensitive magnetic field detectors ever built (SQUIDs), the voltage standard used by every national metrology institute, and the qubit architectures being developed for quantum computing. Topological insulators are being investigated as platforms for fault-tolerant quantum computation via Majorana fermions. Spin glass mathematics appears in the theory of deep learning. The Ising model is used to model everything from protein folding to financial market crashes. Understanding these simulations is not just physics for its own sake — it is physics that shapes the technology of the next decade.