A Hermitian operator satisfies on its operator domain. Here
Two integration by parts operations, together with the boundary conditions in the Hamiltonian domain, show that the kinetic term is Hermitian; multiplication by the real function is Hermitian as well. Thus is Hermitian.
For a time-independent operator , the Schrödinger equation and its adjoint give the Ehrenfest theorem
Using the canonical commutation relation ,
Consequently
The starting point of a conforming finite element method is a weak formulation in a Hilbert space . For example, a homogeneous Dirichlet boundary condition on a bounded Lipschitz domain gives . One seeks
where is a bounded bilinear form, , and is a bounded linear functional. A conforming finite element space is finite dimensional and is typically built from functions that are polynomial on each element of a finite element mesh. With a basis , the Galerkin method imposes the same identity only for , giving
The stiffness matrix is sparse when basis supports overlap only locally. Local element integrals assemble its entries; the mass matrix represents an term. This separates the choice of trial functions from the variational principle determining their coefficients.
The fundamental well-posedness theorem is the Lax-Milgram theorem: if for some , there is a unique solution for each , with . The same theorem on gives a unique discrete solution. The constants must be independent of the mesh for uniform stability of a numerical method. Boundary conditions are built into the trial space when they are essential constraints; flux conditions arise through boundary terms in integration by parts and are natural constraints in the weak formulation.
When is symmetric and a coercive bilinear form, the Ritz method minimizes
over . Its first variation is exactly the Galerkin method, so Ritz minimization and Galerkin testing coincide for symmetric coercive problems. The energy norm turns the approximation into an orthogonal projection. Indeed, subtracting the exact and discrete weak equations gives Galerkin orthogonality, . For every the Pythagorean identity becomes
so
This exact best-approximation property is the central advantage of the Ritz method.
Symmetry is unnecessary for the Galerkin method. For any bounded, coercive form, the Céa lemma gives
To prove it, put . Galerkin orthogonality yields , so coercivity and boundedness give . Divide when and take the infimum. Thus approximation capability plus uniform coercivity gives convergence. In contrast, minimizing for a nonsymmetric form differentiates its symmetric part and generally does not solve the original weak equation.
A concrete symmetric example is the one-dimensional reaction-diffusion problem from Question 5. Continuous piecewise-linear hat functions give the diffusion stiffness matrix with diagonal and neighbouring entries , and the mass matrix with diagonal and neighbouring entries . Therefore and . Its positive-definite matrix is the coordinate form of the continuous energy. The exact smooth solution makes this an explicit test of the method, not just an abstract existence result.
For a nonsymmetric example, take positive , constant real , and
Its form is . Since , it has coercivity constant in the full norm, although it is not symmetric when . The Galerkin method remains well posed. On the uniform hat basis, , where , and . The skew-symmetric matrix cancels from , proving nonsingularity. A Ritz energy would omit precisely this advection contribution. Small may nevertheless make the approximation constant large and produce poorly resolved layers; algebraic solvability is not an accuracy guarantee.
Quantitative convergence uses a finite element interpolation estimate. On a shape-regular mesh, for piecewise-linear conforming elements and in the usual low-dimensional setting,
The Céa lemma therefore gives an energy error. The Aubin–Nitsche duality argument improves the L2 norm error when the dual elliptic problem has regularity: solve and use
Consequently
with constants depending on the solution and regularity bounds. The second estimate is conditional on dual regularity; corners or insufficient data regularity can reduce the rate.
The same Galerkin method handles evolution equations. For the homogeneous heat equation, the semidiscrete finite element heat equation is . Because is a positive-definite matrix and is a positive semidefinite matrix,
The decreasing quadratic form is exactly the L2 norm squared of the finite-element function. This separates spatial stability of a numerical method from the subsequent time integrator, just as in the Schrödinger equation example. Conformity, coercivity and approximation estimates together explain convergence.
Choose any basis of two independent local solutions and collect their coefficients into a column immediately to the right of the th cell boundary. The Floquet matrix is the one-period matrix defined by
Equivalently, it advances the Cauchy data by one period. The Wronskian of two solutions of the stationary Schrödinger equation is constant between the delta functions, and the derivative jump at each delta function also has determinant one. Consequently
For a real potential the Floquet matrix in the real Cauchy-data basis is real, so its trace is real in every basis. Its Floquet multipliers satisfy
By Bloch theorem, an energy is in the interior of an allowed energy band when are distinct unit-modulus numbers. Since is real, this is equivalent to
For the attractive delta-comb Kronig-Penney model, integrating the Schrödinger equation across a lattice point gives
At negative energy write and use the local basis . Free propagation through one cell followed by the derivative jump gives
Thus
The band condition is therefore
Using and gives the requested inequality
At positive energy write and use . The corresponding matrix is
so
Hence the positive-energy bands are exactly the values of for which
Finally, the zero-energy limit has . If , then , and therefore zero energy lies in a forbidden gap.
Inside the infinite square well, the stationary Schrödinger equation for unit mass is
Its normalized eigenstates and energies are
The general normalized time-dependent solution is
Immediately after removing the barrier, the normalized wavefunction is
Its Fourier sine series coefficients in the full-well basis are
They are
The Born rule therefore gives
For a wavefunction , the probability density and probability current are
Multiply the Schrödinger equation by , multiply its complex conjugate by , and subtract. The real potential energy terms cancel, leaving
which is the probability continuity equation .
For a stationary state, , so and is constant in space. A normalizable wavefunction has vanishing current at spatial infinity, forcing this constant to be zero. By contrast the nonnormalizable plane wave
has stationary density and nonzero current
when .
Write and . Using
two differentiations of
give
Thus it solves the time-independent Schrödinger equation in the stated Pöschl-Teller potential, with
As and , respectively,
There is no left-moving term, so this is a reflectionless potential:
The incident and transmitted amplitudes have equal modulus because
Hence
Now set with . Then
At this decays for every positive . At , however, the factor tends to while grows. Normalizability therefore requires . In that case
which decays at both ends. The bound-state energy is
For a real short-range spherically symmetric potential, the stationary wavefunction must solve the Schrödinger equation, be regular at the origin whenever the potential is finite there, and satisfy the scattering boundary condition
Here is the incident plane wave, the second term is an outgoing spherical wave, and its coefficient is the scattering amplitude.
Comparing the stated asymptotic expression with the partial-wave expansion of a scattering amplitude gives
where is a Legendre polynomial. The outgoing-to-incoming coefficient in channel is
Partial-wave unitarity requires
or equivalently
The real scattering phase shift is defined modulo by
For a one-dimensional wavefunction,
The probability continuity equation is
For a stationary state , the probability density is independent of time, so and the probability current is constant in space.
For the plane wave
one obtains
This is a momentum eigenstate with momentum and, when it satisfies the free Schrödinger equation, energy . Its constant density and current describe a spatially uniform beam carrying probability in the sign of . It is not a normalizable wavefunction, so it represents an idealized state rather than a localized particle.
The equation can be written
For the squared L2 norm, periodic integration by parts and the reality of give
Thus the Schrödinger equation generates a unitary flow and is constant.
Inside the infinite square well, the time-independent Schrödinger equation has Dirichlet boundary conditions at and . Its normalized energy eigenstates and energy eigenvalues are
Substitute into the radial Schrödinger equation and equate the coefficients of , , and the constant term. Regularity selects
The corresponding bound-state energy is
Let denote the particle positions. In the Ghirardi--Rimini--Weber model, the wavefunction obeys the ordinary Schrödinger equation
between random collapses. Each particle has an independent Poisson process of collapse times with rate . At such a time the state jumps according to
where the random centre has probability density . The normalization is arranged so that .
The original GRW scales are approximately
An isolated microscopic particle is therefore exceedingly unlikely to collapse during a laboratory experiment. A macroscopic pointer containing about relevant particles has total collapse rate and collapse time about . After a measurement interaction correlates different microscopic outcomes with pointer positions separated by much more than , one constituent's localization suppresses all incompatible pointer branches. This GRW amplification mechanism produces one definite macroscopic outcome with probabilities given by the Born rule, while leaving ordinary microscopic unitary time evolution almost unchanged.
For one spatial coordinate, average over the random centre of one collapse. The resulting density operator has position-space kernel
Its diagonal is unchanged, so and are unchanged. The first derivative of the Gaussian factor vanishes at , so is also unchanged. Its second derivative does not vanish, and the position representation of the momentum operator gives
In three dimensions the increase in total is , corresponding to kinetic-energy increase per collapse. These are ensemble statements: conditioning on one specified collapse centre can shift the position moments. Repetition at rate predicts GRW spontaneous heating, so precision searches for anomalous bulk heating, spontaneous radiation, momentum diffusion, and loss of matter-wave interference test the model.
Write . The real potential and periodic boundary condition make a self-adjoint operator on the periodic domain. Since ,
because the inner product is real. Thus the Schrödinger equation preserves the norm.
For ,
The stationary Schrödinger equation becomes
Hence, choosing the positive root and a positive normalization constant,
The state is even, so . The Gaussian integrals give
and therefore . These values are collected in gaussian eigenstate for a quadratic potential.
Finally, equality in the derivation of the uncertainty relation requires the centred vectors to be linearly dependent with a purely imaginary proportionality constant. Thus for some ,
In position space this says
whose normalizable solutions are
Thus every saturating state is Gaussian up to translation, a plane-wave factor, and an overall phase, as in the equality case of the Heisenberg uncertainty relation.
The Schrödinger equation and its complex conjugate are
Because the potential is real, its two contributions cancel when differentiating the probability density . Therefore
The expression in square brackets is the probability current , so
This probability continuity equation says that probability can leave a region only through the current across its boundary.
Suppose and are two bound-state solutions with the same energy. Their Wronskian
satisfies because the Schrödinger equation contains no first-derivative term. Hence is constant. Both wavefunctions and their derivatives decay at infinity, so there and therefore . The two solutions are linearly dependent. Thus every one-dimensional bound-state energy is nondegenerate.
Write the expectation value as
Differentiating and retaining the possible explicit time dependence of the observable gives
The Schrödinger equation and its adjoint are
where the Hamiltonian operator is Hermitian. Substitution yields
This is the proof of Ehrenfest theorem from the Schrodinger equation.
Differentiating and using the Schrödinger equation and its adjoint gives the Ehrenfest theorem
For ,
Substitution into the Schrödinger equation and comparison of the constant and coefficients gives