Time-Dependent Schrödinger Equation

Carved into the memorial to Erwin Schrödinger at the University of Vienna is a single line of mathematics: $i\hbar\,\dot\Psi = H\Psi$ . It has earned the stone. Newton's second law tells a planet how to move; this equation tells a quantum state how to become its future self. Everything dynamical in the nonrelativistic quantum world — the winding phases that make atoms emit sharp spectral lines, the leak of a wavefunction through a barrier, the gate pulses of a quantum computer — is this one sentence, applied. You arrive well armed. Wave–particle duality handed you the dictionary $E = \hbar\omega$ , $p = \hbar k$ ; de Broglie matter waves made the momentum–wavelength link quantitative; the wavefunction $\Psi$ became the object that carries a particle's state, with $|\Psi|^2$ its Born-rule probability map, normalized to one can of paint. What is missing is a law of motion. Given $\Psi$ now, what is it later? The time-dependent Schrödinger equation answers with three structural surprises: the law is first order in time, so the present wavefunction alone determines the future — no initial velocity required; it is linear, so superpositions evolve into superpositions and interference survives forever; and it wears an irremovable factor of $i$ , which turns what would otherwise be a diffusion equation into a wave equation whose solutions rotate rather than decay. This article builds the equation the honest way — as a postulate you are motivated into, not a theorem you derive — then squeezes out what its structure alone guarantees, before the separation-of-variables machinery takes over in the articles that follow.

Key takeaways

The law of quantum motion: $i\hbar\,\partial_t\Psi = \hat H\Psi$ , with $\hat H = -\frac{\hbar^2}{2m}\partial_x^2 + V(x)$ built from the classical energy by $p \to -i\hbar\,\partial_x$ .
It is first order in time: $\Psi(x,0)$ alone is a complete initial condition — wavefunction evolution is fully deterministic; chance enters only at measurement.
It is linear: superpositions evolve component by component, which is why interference is permanent and why "expand, evolve phases, resum" is the master strategy.
The $i$ is physics, not decoration: it makes solutions oscillate ( $e^{-iEt/\hbar}$ ) instead of diffusing away, and it forces $\Psi$ to be complex.
A free-particle plane wave solves it precisely when $\hbar\omega = \hbar^2k^2/2m$ — the de Broglie dispersion relation, recovered as a consistency check.
Energy eigenstates just spin their phase at rate $E/\hbar$ ; all visible time dependence comes from beats between different energies.
Key idea: the Hamiltonian is the generator of time translation — knowing the energy operator is knowing how every state turns in Hilbert space.

The core insight

The Hamiltonian is the generator of time translations: $i\hbar\,\partial_t\Psi = \hat H\Psi$ means that knowing the energy operator tells you exactly how the state evolves, and the Hermiticity of $\hat H$ guarantees that this evolution preserves total probability.

Where this lives today

Solving the TDSE numerically is a daily industrial activity. Quantum-chemistry codes propagate molecular wavepackets with split-operator and Crank–Nicolson schemes to watch bonds break in real time; the 2023 Nobel Prize in Physics went to attosecond light pulses that strobe electron motion on the $10^{-18}\ \mathrm{s}$ scale — experiments planned and interpreted by integrating exactly equation (1). Your worked example's femtosecond clock is their laboratory reality. Quantum computing is the equation turned into engineering. A logic gate is a Hamiltonian switched on for a precise interval, rotating the state by exactly the intended unitary $e^{-i\hat Ht/\hbar}$ ; calibrating a two-qubit gate means shaping $\hat H(t)$ until the TDSE's solution matches the target to a part in $10^4$ . The equation's structure supplies the guardrails: linearity underwrites superposition algorithms, and norm conservation is why error correction fights decoherence rather than probability leakage.

The physical picture

Read the equation as a machine with four working parts. On the right, the Hamiltonian $\hat H$ — the energy operator — interrogates the current wavefunction: it measures, point by point, how much curvature (kinetic energy) and how much potential energy the state is carrying. On the left, $i\hbar\,\partial_t$ converts that energy reading into a rotation rate for the wavefunction's complex phase. High-energy configurations spin fast, low-energy ones slowly, and $\hbar$ is the exchange rate between energy and angular speed. The $i$ steers the change perpendicular to $\Psi$ itself — a rotation, never a stretch — which is the geometric reason total probability holds steady while everything else churns.

Figure

Anatomy of the master equation — each symbol is a design decision

Now picture what the machine does to an energy eigenstate — a state where $\hat H$ returns one sharp value $E$ . The interrogation is the same everywhere, so the whole function rotates its phase in unison: $\Psi = \psi(x)\,e^{-iEt/\hbar}$ , a rigid clock hand of length $|\psi|$ spinning at angular rate $E/\hbar$ . Nothing about $|\Psi|^2$ moves. All visible dynamics — a density that actually sloshes, an electron that actually goes somewhere — comes from superposing at least two energies, whose clock hands rotate at different rates and drift in and out of alignment. The relative phase $e^{-i(E_2-E_1)t/\hbar}$ beats at the Bohr frequency $(E_2-E_1)/\hbar$ , and the density breathes at exactly that rate.

Figure

All visible motion is beats — a two-energy superposition sloshing

Hold that clock picture. It converts the scariest-looking equation of the course into bookkeeping: expand the state in energy eigenstates, attach a clock to each, let them spin, and resum whenever you want the density. Dynamics becomes phase accounting.

The law, stated honestly

Here is the postulate, in one spatial dimension (the three-dimensional version merely swaps $\partial_x^2$ for $\nabla^2$ ). Every quantum system you will meet this semester obeys it:

i\hbar\,\frac{\partial \Psi(x,t)}{\partial t} = \hat H\,\Psi(x,t) \qquad (1)

where: $\Psi(x,t)$ — the wavefunction; $\hbar = 1.055\times10^{-34}\ \mathrm{J\,s}$ — the reduced Planck constant; $\hat H$ — the Hamiltonian operator, the observable for total energy; $i$ — the imaginary unit, doing real work. Read it as a promise: hand the equation your state now, and it hands back the state's rate of change — no further input, no memory of the past, no initial velocity. For a single particle of mass $m$ in a potential $V(x)$ , the Hamiltonian is the classical energy $p^2/2m + V$ promoted to operators:

\hat H = -\frac{\hbar^2}{2m}\frac{\partial^2}{\partial x^2} + V(x) \qquad (2)

where: the first term is kinetic energy, born from the substitution $p \to -i\hbar\,\partial_x$ (so $p^2 \to -\hbar^2\partial_x^2$ ); $V(x)$ acts by plain multiplication. The kinetic term reads curvature: where $\Psi$ bends sharply, the state is momentum-rich and its phase will spin fast. Where does the substitution rule come from? That is exactly what the first derivation below is for — and the Momentum Operator article, later in this chapter, gives the operator its own careful treatment. Two structural properties of (1) do an enormous amount of unglamorous work. Linearity: if $\Psi_1$ and $\Psi_2$ are solutions, so is $\alpha\Psi_1 + \beta\Psi_2$ — the superposition principle survives dynamics, amplitudes interfere at every later time, and the two-slit logic from Wave–Particle Duality stays intact forever. First order in time: one time derivative means one initial condition, $\Psi(x,0)$ . Contrast a classical wave, whose second-order equation demands both the initial shape and the initial velocity of the string. A quantum state's present is its complete dossier. There is one more guarantee worth naming now: equation (1) conserves the normalization $\int|\Psi|^2dx = 1$ exactly, for any real potential — the paint supply from Probability Density and Normalization is never created or destroyed by the dynamics. The two-line proof, and the probability current that makes the conservation local, are owned by Probability Current and the Continuity Equation and Conservation of Normalization, two articles ahead. Likewise, when $\hat H$ has no explicit time dependence, product solutions $\psi(x)T(t)$ exist and turn (1) into an eigenvalue problem — but that separation ritual, the doorway to everything, is performed in Time-Independent Schrödinger Equation and Stationary States and Separation of Variables, and practiced in one dimension by Time-Independent Schrödinger Equation in One Dimension. This article's job is the law itself.

Derivation ABuilding the equation from de Broglie waves

No one derives the Schrödinger equation from deeper principles — it is a postulate, and experiment is its judge. But Schrödinger did not guess blindly, and the reconstruction of his reasoning is the best way to own the equation. The strategy: take the one quantum state we fully trust — the free-particle plane wave with its de Broglie credentials — find the differential equation it obeys, and then promote that equation to a law for all states, potential included. Watch for the honesty clause at the end: the final move is a leap, not a proof.

Step 1. Start from the trusted state. A free particle of sharp momentum $p$ and energy $E$ is the plane wave $\Psi = e^{i(kx - \omega t)}$ , with $p = \hbar k$ from de Broglie Matter Waves and $E = \hbar\omega$ from the Planck–Einstein relation in Wave–Particle Duality. Why: this is the one wavefunction whose physical meaning the 1924–1927 experiments had already certified — everything else is built to be consistent with it.

Step 2. Find derivatives that read off E and p. Differentiate: $i\hbar\,\partial_t\Psi = \hbar\omega\,\Psi = E\Psi$ , and $-i\hbar\,\partial_x\Psi = \hbar k\,\Psi = p\Psi$ . Why: each derivative pulls down the conserved quantity as an eigenvalue — the operators $i\hbar\,\partial_t$ and $-i\hbar\,\partial_x$ are "energy readers" and "momentum readers" on plane waves. This is the origin of the substitution rules.

Step 3. Impose the classical energy budget. A free nonrelativistic particle has $E = p^2/2m$ . Written with the operator readers acting on $\Psi$ : $i\hbar\,\partial_t\Psi = \frac{1}{2m}(-i\hbar\,\partial_x)^2\Psi = -\frac{\hbar^2}{2m}\partial_x^2\Psi$ . Why: $(-i\hbar)^2 = -\hbar^2$ — the free Schrödinger equation is the dispersion relation $E=p^2/2m$ transcribed into calculus.

Step 4. Admit forces. For a particle in a potential the energy budget is $E = p^2/2m + V(x)$ ; promoting term by term gives

i\hbar\,\frac{\partial\Psi}{\partial t} = \left(-\frac{\hbar^2}{2m}\frac{\partial^2}{\partial x^2} + V(x)\right)\Psi \qquad (3)

Why: this is the leap. Plane waves no longer solve the equation once $V$ varies, so nothing forces this form — it is the simplest equation that reduces correctly when $V=0$ , and its warrant is a century of confirmed predictions, hydrogen first among them.

Step 5. State the postulate. Equation (3) — abstractly, $i\hbar\,\partial_t\Psi = \hat H\Psi$ — is adopted as the universal law of motion for nonrelativistic quantum states. Why: physics keeps score by consequences: energy quantization, tunnelling, atomic spectra, chemistry. The equation has never lost.

Go deeper · Derivation B · 1 minThe plane-wave consistency check

Derivation BThe plane-wave consistency check

A law built from plane waves had better accept them back. We verify that $\Psi = e^{i(kx-\omega t)}$ solves the free equation, and extract the condition it imposes — which turns out to be the de Broglie dispersion relation, closing the loop with the matter-wave articles. The calculation is two derivatives long; its payoff is a reading of why quantum wave packets spread, and a number you can check against a real experiment.

Step 1. Insert the plane wave into the free equation. With $V = 0$ : left side $i\hbar\,\partial_t\Psi = \hbar\omega\,\Psi$ ; right side $-\frac{\hbar^2}{2m}\partial_x^2\Psi = \frac{\hbar^2k^2}{2m}\Psi$ . Why: exponentials are eigenfunctions of differentiation — each derivative is a multiplication.

Step 2. Read off the dispersion relation. Equality for all $x,t$ demands

\hbar\omega = \frac{\hbar^2 k^2}{2m} \qquad (4)

Why: this is exactly $E = p^2/2m$ dressed in wave clothing — the equation accepts precisely those plane waves whose frequency honors the classical energy budget.

Step 3. Note the curvature of the dispersion. $\omega \propto k^2$ , not $\omega \propto k$ : different wavelengths travel at different speeds, so any localized packet — a sum of many $k$ 's — must disperse. Why: for light in vacuum $\omega = ck$ is straight and pulses hold their shape; the quadratic law (4) is the mathematical seed of wave-packet spreading, developed in Free Particle and Wave Packets and the velocity story in Group and Phase Velocity.

Step 4. Attach numbers. For the $54\ \mathrm{eV}$ Davisson–Germer electron ( $\lambda = 0.167\ \mathrm{nm}$ ): $k = 2\pi/\lambda = 3.76\times10^{10}\ \mathrm{m^{-1}}$ , and (4) gives $\omega = \hbar k^2/2m = 8.19\times10^{16}\ \mathrm{rad/s}$ — whence $\hbar\omega = 54.0\ \mathrm{eV}$ . Why: the loop closes on real data — the same electron whose diffraction confirmed $p=\hbar k$ carries a phase clock whose rate is set by its kinetic energy.

Go deeper · Derivation C · 2 minFirst order in time — why the wavefunction must be complex and the future is fixed

Derivation CFirst order in time — why the wavefunction must be complex and the future is fixed

Two of quantum mechanics' most quoted oddities — the irreducibly complex wavefunction and the perfectly deterministic evolution of a probabilistic theory — are not separate mysteries. Both fall out of one design feature: the single time derivative on the left of the master equation. We extract each in turn, then see what the $i$ itself prevents. Nothing here needs more than inspection of the equation; the point is learning to read structure.

Step 1. One derivative, one initial condition. A first-order equation in $t$ means: give $\Psi(x,0)$ , and (1) supplies $\partial_t\Psi$ at $t=0$ , hence $\Psi$ an instant later, and so on — the future is constructed from the present alone. Why: Newton's law is second order, needing position and velocity; a classical wave needs shape and shake. The quantum state needs only itself: there is no such thing as "the initial velocity of a wavefunction."

Step 2. Determinism, correctly placed. Evolution between measurements is exactly as deterministic as classical mechanics — more rigidly so, since linearity forbids chaos in the state itself. Why: the celebrated quantum randomness enters only when a measurement converts $\Psi$ into an outcome via the Born rule — the dynamics of that step is the Formalism chapter's business (Measurement and Wavefunction Collapse). Keep the two regimes distinct and half the "quantum is random" confusions dissolve.

Step 3. Now suppose $\Psi$ were real. Then $\hat H\Psi$ is real (real potential, real derivatives), while the left side $i\hbar\,\partial_t\Psi$ is purely imaginary. A real number equals an imaginary one only if both vanish: $\partial_t\Psi = 0$ and $\hat H\Psi = 0$ . Why: a real wavefunction cannot move at all — the equation's two sides live in perpendicular directions of the complex plane. Complexity is not notation; it is load-bearing. (Compare the classical wave equation, second order, which is perfectly content with real solutions.)

Step 4. See what the i forbids. Delete it: $\hbar\,\partial_t\Psi = -\frac{\hbar^2}{2m}\partial_x^2\Psi$ is the heat equation — its plane-wave solutions decay as $e^{-Dk^2t}$ , smearing structure away irreversibly. With the $i$ , the same mode oscillates, $e^{-i\omega t}$ , at constant amplitude. Why: the $i$ converts dissipation into rotation — probability is preserved, information is preserved, evolution is reversible. Quantum mechanics is, in this precise sense, the heat equation rotated into imaginary time. *One can confirm directly: $e^{ikx}e^{-Dk^2t}$ solves the $i$ -less (heat) equation and decays, while $e^{i(kx-\omega t)}$ solves equation (1) and oscillates at constant amplitude — the two solutions make the contrast between diffusion and rotation explicit.*

Limiting cases

Case	Result	Meaning
time-independent $\hat H$	$\Psi = \sum_n c_n\psi_n(x)e^{-iE_nt/\hbar}$	eigenstates with winding clocks — the expansion built in Stationary States and Separation of Variables
free particle, $V=0$	plane waves with $\hbar\omega = \hbar^2k^2/2m$	de Broglie dispersion; quadratic law ⇒ packets spread
single energy eigenstate	$\lvert\Psi\rvert^2$ frozen in time	a "stationary state" — the phase spins, the density does not
narrow packet, smooth $V$	$\tfrac{d\langle p\rangle}{dt} = -\langle\partial_x V\rangle$	Newton recovered on average — Ehrenfest's Theorem, later this chapter
$\hat H$ depends on time	no single winding phase; time-ordered evolution	driven systems (lasers, gates); the honest tools arrive in Quantum Dynamics (QM3)

Common trap“Stationary states don't change, and the equation needs an initial velocity”

Both halves of that sentence are wrong, and both cost marks every year. First: the equation is first order in time — you specify $\Psi(x,0)$ and nothing else. Students trained on $F=ma$ reflexively hunt for a second initial condition; there is none, and supplying one is a category error. Second: a stationary state is stationary only in its density. The full wavefunction $\Psi = \psi\,e^{-iEt/\hbar}$ never stops moving — its phase spins at $E/\hbar$ eternally — and that motion is one superposition away from visibility: mix two energies and the relative phase beats at the Bohr frequency, sloshing the density. "The phase doesn't matter" is true only of one global phase; relative phases between energy components are the entire mechanism of quantum dynamics.

Worked examples

1 · The quantum clock rate (confidence builder)

How fast does the phase of an energy eigenstate actually spin? Take a state with $E = 1.00\ \mathrm{eV}$ — a chemistry-scale energy — and find the angular rate $\omega$ , the cycle frequency $\nu$ , and the period of one full phase revolution.

Step 1. Angular rate: $\omega = \dfrac{E}{\hbar} = \dfrac{1.602\times10^{-19}}{1.055\times10^{-34}} = \boxed{1.52\times10^{15}\ \mathrm{rad/s}}$ .

Step 2. Cycle frequency: $\nu = \dfrac{E}{h} = \dfrac{1.602\times10^{-19}}{6.626\times10^{-34}} = \boxed{2.42\times10^{14}\ \mathrm{Hz}}$ — comparable to the frequency of visible light ( $600\ \mathrm{nm}$ light oscillates at $5.0\times10^{14}\ \mathrm{Hz}$ ), and that is no coincidence: optical photons carry eV-scale energies.

Step 3. Period: $T = 1/\nu = \boxed{4.14\ \mathrm{fs}}$ . Every eV of energy is a phase clock ticking a quarter-million times per nanosecond.

Step 4. The observable version: superpose two states $1\ \mathrm{eV}$ apart and the relative phase — the only part nature displays — completes a cycle every $4.14\ \mathrm{fs}$ . An oscillating charge distribution at $2.42\times10^{14}\ \mathrm{Hz}$ radiates at exactly that frequency: this is why level differences, never levels, appear in spectra.

2 · The Davisson–Germer electron as a wave in time (exam level)

The 54 eV electrons that confirmed matter waves have $\lambda = 0.167\ \mathrm{nm}$ . Treating one as a plane wave, find $k$ , $\omega$ , the phase velocity $v_p = \omega/k$ , and the group velocity $v_g = d\omega/dk$ . Compare both to the classical particle speed.

Step 1. Wavenumber: $k = \dfrac{2\pi}{\lambda} = \dfrac{2\pi}{1.67\times10^{-10}} = \boxed{3.76\times10^{10}\ \mathrm{m^{-1}}}$ .

Step 2. From the dispersion relation (4): $\omega = \dfrac{\hbar k^2}{2m} = \dfrac{(1.055\times10^{-34})(3.76\times10^{10})^2}{2(9.11\times10^{-31})} = \boxed{8.19\times10^{16}\ \mathrm{rad/s}}$ . Check: $\hbar\omega = 8.64\times10^{-18}\ \mathrm{J} = 54.0\ \mathrm{eV}$ — the energy comes back intact.

Step 3. Phase velocity: $v_p = \omega/k = \dfrac{\hbar k}{2m} = \boxed{2.18\times10^6\ \mathrm{m/s}}$ .

Step 4. Group velocity: $v_g = \dfrac{d\omega}{dk} = \dfrac{\hbar k}{m} = \boxed{4.36\times10^6\ \mathrm{m/s}}$ — and the classical check $v = \sqrt{2E/m} = 4.36\times10^6\ \mathrm{m/s}$ agrees exactly.

Step 5. Read the factor of two: $v_g = 2v_p$ for every free matter wave, straight from $\omega \propto k^2$ . The envelope of a packet — the thing that carries the electron — moves at the particle speed $v_g$ ; the ripples inside it lag at half that. Which velocity "is" the electron's, and why the phase speed carries no signal, is the business of Group and Phase Velocity; the packet construction itself is Free Particle and Wave Packets. Here you have the raw numbers that make the distinction unavoidable.

Origins · 1 minRichard Feynman and how this was discovered

Richard Feynman

1918–1988 · Princeton, then Caltech

Feynman never doubted the Schrödinger equation — he rebuilt the road to it. As a Princeton graduate student he seized on a half-page remark of Dirac's from 1933, that the quantum amplitude for a short time step "corresponds to" $e^{iL\,dt/\hbar}$ with $L$ the classical Lagrangian, and asked what Dirac left hanging: corresponds — or equals? Working it out, he found that summing $e^{iS/\hbar}$ over every conceivable path reproduces quantum mechanics exactly — and that demanding consistency of the sum for short times yields precisely $i\hbar\,\partial_t\Psi = \hat H\Psi$ . His 1942 thesis and 1948 paper made it rigorous: a second, independent road to the same law, with classical mechanics visible inside it (the stationary-action path dominates as $\hbar \to 0$ ). One equation, two derivations, zero disagreements. The path-integral machinery itself returns in the Quantum Dynamics chapter.

History

1924

— de Broglie attaches a wave to every particle; the wave still lacks an equation of motion.

1925

— Heisenberg, Born and Jordan build matrix mechanics: dynamics as evolving arrays, no wavefunction yet.

1926

— Schrödinger, prompted by Debye's remark that a proper wave needs a wave equation, publishes the TDSE and solves hydrogen with it.

1926

— Born reads

\lvert\Psi\rvert^2

as probability, giving the equation's solutions their meaning.

1933

— Dirac notes the Lagrangian's shadow in quantum amplitudes over short times.

1948

— Feynman turns that note into the path integral: the same equation, re-derived from summing over histories.

Going further

Texts. Griffiths & Schroeter, Introduction to Quantum Mechanics, 3rd ed., Ch. 1–2 · Sakurai & Napolitano, Modern Quantum Mechanics, 2nd ed., Ch. 2 · Shankar, Principles of Quantum Mechanics, 2nd ed., Ch. 4 · Cohen-Tannoudji, Diu & Laloë, Quantum Mechanics, Vol. I, Ch. III · Townsend, A Modern Approach to Quantum Mechanics, 2nd ed., Ch. 2 & 6. The reduced Planck constant used throughout is $\hbar\approx1.055\times10^{-34}\,\mathrm{J\,s}$ (CODATA). Key numerical results: $\omega = 1.52\times10^{15}\ \mathrm{rad/s}$ , $\nu = 2.42\times10^{14}\ \mathrm{Hz}$ , $T = 4.14\ \mathrm{fs}$ for $E = 1\ \mathrm{eV}$ ; and $k = 3.76\times10^{10}\ \mathrm{m^{-1}}$ , $\omega = 8.19\times10^{16}\ \mathrm{rad/s}$ , $v_p = 2.18\times10^6\ \mathrm{m/s}$ , $v_g = 4.36\times10^6\ \mathrm{m/s}$ for the 54 eV electron.

Prerequisites. The Wavefunction · Born Statistical Interpretation · de Broglie Matter Waves · Probability Density and Normalization. Leads to. Time-Independent Schrödinger Equation · Stationary States and Separation of Variables · Probability Current and the Continuity Equation — then the whole one-dimensional laboratory, beginning with the Time-Independent Schrödinger Equation in One Dimension.

Is the Schrödinger equation derived from anything, or just guessed?

It is a postulate — the derivation card above is a motivation, not a proof, and its final step is honestly labeled a leap. What makes the postulate science rather than decree is its scorecard: hydrogen's spectrum, the periodic table, tunnelling rates, chemical bond energies, semiconductor band structures. Feynman's path integral later re-derived the same equation from a completely different starting point (sum over histories), which raises confidence that the structure is forced rather than lucky. If you want the axiomatic view — evolution must be unitary, unitaries have Hermitian generators, the generator is the energy — that formulation arrives in the Formalism chapter. All three roads meet at the same equation.

Why does energy, of all observables, get to run time evolution?

Because energy and time are conjugate in exactly the way momentum and position are. In classical mechanics the Hamiltonian already generates time translation — Hamilton's equations say "to know the future, differentiate $H$ ." Quantum mechanics inherits the arrangement: the operator version of the energy tells every component of the state how fast to turn its phase, $e^{-iEt/\hbar}$ . A state of sharp energy therefore evolves trivially (one global clock), while a spread of energies evolves richly (many clocks beating). This is also why energy conservation and time-translation symmetry are the same fact — a thread picked up properly by Noether-style arguments in the Formalism and Dynamics chapters.

If quantum evolution is deterministic, where exactly does randomness enter?

At measurement, and only there. Between observations, $\Psi$ evolves under the TDSE as rigidly as a planet under gravity — given $\Psi(x,0)$ , the state at any later time is fixed to the last decimal. When you measure, the Born rule converts the deterministic $\Psi$ into probabilistic outcomes, and the post-measurement state is updated discontinuously. The interface between the two regimes — why measurement looks so different from evolution, and whether it truly is — is the measurement problem, treated honestly in the Formalism chapter and the Foundations chapter at the end of the track. For calculational purposes the split is clean: evolve deterministically, then bet with $|\Psi|^2$ .

What does it even mean for the wavefunction's phase to spin if phase is unobservable?

One phase is unobservable: the global one. Multiply the whole state by $e^{i\theta}$ and no experiment notices — states are rays. But the TDSE spins different energy components at different rates, and those relative phases are brutally observable: they are why the two-hump superposition in the figure sloshes, why spectral lines sit at Bohr frequencies $(E_2-E_1)/\hbar$ , and why an interferometer's output flickers when one arm is delayed. Think of the global phase as the choice of when to call $t=0$ , and relative phases as the actual clockwork.

Why is the equation first order in time when the classical wave equation is second order?

Because $\Psi$ is complex, it packs two real functions into one object — effectively carrying both "position-like" and "velocity-like" information at once. A real classical wave needs its shape and its velocity specified separately, hence two time derivatives. The complex first-order equation is the tighter design: one function, one derivative, one initial condition. This is not a cosmetic difference — it is why quantum states cannot be "started at rest" and why the theory's information accounting (unitarity) works out so cleanly.

Does the TDSE apply when the potential itself changes with time?

Yes — equation (1) holds verbatim with $V(x,t)$ ; what breaks is the solution strategy, not the law. With a time-dependent $\hat H$ there are no stationary states to expand in: no single set of clocks, because the clockmaker keeps changing the mechanism. The workhorse tools for that regime — the time-ordered evolution operator, the interaction picture, time-dependent perturbation theory, Rabi oscillations — are the core of the Quantum Dynamics and Time-Dependent Phenomena chapters. Everything in this article that used only linearity and first-orderness survives untouched; only the phase-winding expansion assumed a time-independent $\hat H$ .

Is the Schrödinger equation relativistic? What happens at high speeds?

It is not — the giveaway is the lopsided treatment of space and time: second derivative in $x$ , first in $t$ , and pure-Newton dispersion $E = p^2/2m$ . Schrödinger actually tried the relativistic version first (now called the Klein–Gordon equation), found it disagreed with hydrogen's fine structure, and shelved it for the nonrelativistic equation that worked. The proper treatments — Klein–Gordon for spin-0, Dirac's equation for the electron — are the Relativistic chapter's story. Rule of thumb: with kinetic energy well under the rest energy $mc^2$ , equation (1) is quantitatively excellent.

The equation looks like a diffusion equation with imaginary time — is that meaningful or a coincidence?

Deeply meaningful, and heavily exploited. Substitute $t \to -i\tau$ and the TDSE becomes the heat equation; reverse it and diffusive decay becomes quantum oscillation. Numerically, evolving in imaginary time is a standard trick for finding ground states: every excited component decays faster than the ground state, so the ground state survives the purge. In quantum statistical mechanics the correspondence runs deeper — the thermal weight $e^{-\hat H/k_BT}$ is evolution through imaginary time $\hbar/k_BT$ . A curiosity about the letter $i$ becomes a working computational bridge.