P2 Operators#

What you need to know

  • For every experimental observable there is a corresponding operator in quantum mechanics.

  • Operators must be linear. Becasue they are dervied from Schrodinger equation which itself is linear.

  • Operatprs must be Hermitian becasue only Hermitian operators produce real eigenvalues.

  • Operatprs must produce real eigenvalues. Becasue eigenvalue are the only possible values that are measured in experiments.

  • Operators commutations show weather two experimental observables can be measured simulateneously. E.g can one simulatenoulsy and precisely determine the values of position and momentum of an electron.

  • Commuting operators share eignefunctions, non-commuiting operators have different eigenfunctions.

Operators#

  • In quantum mechanics, operators represent physical observables and are denoted by a hat symbol (\(\hat{}\)), which indicates a mathematical operation on functions.

  • For example, the momentum operator is differentiating the function with respect to \(x\), then multiplies the result by \(-i\hbar\).

    \[ \hat{p}_x = -i\hbar\frac{d}{dx} \]
  • When this operates on a function, it

  • The position operator simply multiplies the function by \(x\).

    \[ \hat{x} = x \]
  • In quantum mechanics we use a simple recepie to find operators: Take expressions from classical mechancis and replace positions and momentum by their respective operator expressions.

Linearity of Operators#

  • Operators in quantum mechanics are linear, meaning they satisfy:

\[ \hat{A}(\psi_1 + \psi_2) = \hat{A}\psi_1 + \hat{A}\psi_2 \]
\[ \hat{A}(c\psi) = c\hat{A}\psi \]
  • Where \(c\) is a constant, and \(\psi_1\), \(\psi_2\), and \(\psi\) are wavefunctions.

  • \(\hat{x}\), \(\hat{p_x}\), \(\hat{H}\) all satisfy this property

Commutations of operators#

Commutator \(\hat{A}\) and \(\hat{B}\)

\[{\left[\hat{A},\hat{B}\right]f = \left(\hat{A}\hat{B} - \hat{B}\hat{A}\right)f}\]
  • From linear algebra we know that order of matrix multiplicaiton matters and that \(AB\neq BA\) for two matrices \(A\) and \(B\)

  • Thus we also generally ecpect \(\hat{A}\hat{B} \neq \hat{B}\hat{A}\) for any two operators.

  • We can quantify relationship between two operators by computing the Commutator

    • If the commutator is zero, it means that order in multiplication of operators or matrices can be changed.

    • If the commutator is non-zero, the order matters and can not be changed!

Example

Prove that operators \(\hat{A} = x\) and \(\hat{B} = d/dx\) do not commute (i.e., \(\left[\hat{A}, \hat{B}\right] \ne 0\)).

Simple rules for commutators#

\[{\left[A,A\right] = \left[A,A^n\right] = \left[A^n,A\right] = 0}\]
  • This shows that operators always commuts with itself and its power. Now lets apply this. Would kinetic operator commute with the momentum operator?

\[{\left[A,B\right] = -\left[B,A\right]}\]
  • This shows that in comutator order is important. You swap the oeprators in places the sign changes.

Commutators and experimental measurements#

We have seen previously that operators may not always commute (i.e., \([A, B] \ne 0\)). An example of such operator pair is position \(\hat{x}\) and momentum \(\hat{p}_x\):

\[{\hat{p}_x\hat{x}\psi(x) = \hat{p}_x\left(x\psi(x)\right) = \left(\frac{\hbar}{i}\frac{d}{dx}\right)\left(x\psi(x)\right) = \frac{\hbar x}{i}\frac{d\psi(x)}{dx} + \frac{\hbar}{i}\psi(x)}\]
\[{\hat{x}\hat{p}_x\psi(x) = x\left(\frac{\hbar}{i}\frac{d\psi(x)}{dx}\right)}\]
\[{\Rightarrow \left[\hat{p}_x,\hat{x}\right]\psi(x) = \left(\hat{p}_x\hat{x} - \hat{x}\hat{p}_x\right)\psi(x) = \frac{\hbar}{i}\psi(x)}\]
\[{\Rightarrow \left[\hat{p}_x,\hat{x}\right] = \frac{\hbar}{i}}\]

In contrast, the kinetic energy operator and the momentum operators commute:

\[ {\left[\hat{T},\hat{p}_x\right] = \left[\frac{\hat{p}_x^2}{2m},\hat{p}_x\right] = \frac{p_x^3}{2m} - \frac{p_x^3}{2m} = 0} \]

We had the uncertainty principle for the position and momentum operators:

\[ \Delta x\Delta p_x \ge \frac{\hbar}{2} \]

In general, it turns out that for operators \(\hat{A}\) and \(\hat{B}\) that do not commute, the uncertainty principle applies in the following form:

\[{\Delta A\Delta B \ge \frac{1}{2}\left|\left<\left[\hat{A},\hat{B}\right]\right>\right|}\]
  • Let’s check this relation on the example of momentum and position operators

  • Denote \(\hat{A} = \hat{x}\) and \(\hat{B} = \hat{p}_x\).

\[\frac{1}{2}\left|\left<\left[\hat{A},\hat{B}\right]\right>\right| = \frac{1}{2}\left|\left<\left[\hat{x},\hat{p}_x\right]\right>\right| = \frac{1}{2}\left|\left<\frac{\hbar}{i}\right>\right| = \frac{1}{2}\left|\left<\psi\left|\frac{\hbar}{i}\right|\psi\right>\right| = \frac{1}{2}\left|\frac{\hbar}{i}\underbrace{\left<\psi\left|\psi\right.\right>}_{=1}\right| = \frac{\hbar}{2}\]
\[\Rightarrow \Delta x\Delta p_x \ge \frac{\hbar}{2}\]
  • We find that we can not measure precise values of position or momentum simulatneously.

Commuting operators and simultaneous measurments#

Commuting operators share eigenfunction

\[[\hat{A},\hat{B}]=0\]
\[\hat{A}\phi_k = a_k \phi_k\]
\[\hat{B}\phi_k = b_k \phi_k\]
  • If opertors commute that means we can simultaneously measure corresponding observables in a single experiment.

  • For instance operatos of kinetic energy and momentum commute. We can measure momentum and kinetic energy. But we can not do the same for momentum and position.

  • If we measure observables \(A\) and \(B\) desribed by a common eigenfunction \(\phi_k\) we find the observables to be the corresponding eigenvalues \(a_k\) and \(b_k\)

Expectation expression#

  • The expectation value of an observable \(\hat{A}\), which gives the average outcome of measurements, is computed as:

    \[ \langle A \rangle = \int \psi^* \hat{A} \psi \, d\tau \]
  • Special Case: If the wavefunction \(\psi\) is an eigenfunction of the operator \(\hat{A}\), with eigenvalue \(a\):

    \[ \hat{A}\psi = a\psi \]

    Then the expectation value simplifies to:

    \[ \langle A \rangle = \int \psi^* a \psi \, d\tau = a \int \psi^*\psi \, d\tau = a \]

    Since \(\int \psi^*\psi \, d\tau = 1\) (normalization), the expectation value is simply the eigenvalue \(a\).

Dirac Notation#

To express quantum states and operators more compactly, we use Dirac (bra–ket) notation.

  • A state is written as a ket, \(|\psi\rangle\), and its complex conjugate (dual) is the bra, \(\langle\psi|\).

  • The inner product between two states corresponds to the integral over space:

\[ \langle \phi | \psi \rangle = \int \phi^*(r), \psi(r), d\tau \]
  • In this notation, the expectation value of an operator \(\hat{A}\) becomes simply:

\[ \langle A \rangle = \langle \psi | \hat{A} | \psi \rangle \]
  • This form is elegant and general—it applies to all quantum systems, independent of the particular representation (position, momentum, etc.).

Hermitian Property of Operators#

  • In quantum mechanics, operators often act on complex-valued functions, so we need a notion of “complex conjugate” that applies not just to numbers, but to operators. This leads to the concept of the adjoint operator.

The Adjoint (Conjugate Transpose)#

  • For complex numbers we take the complex conjugate: \((3 + 2i)^* = 3 - 2i\).

  • For matrices or linear operators, the corresponding operation is the adjoint, denoted by the dagger symbol \((\dagger)\).

Definition of Adjoint

  • For an operator, the adjoint is defined through the inner product:

\[ \langle \phi | \hat{A}\psi \rangle = \langle \hat{A}^\dagger \phi | \psi \rangle. \]
  • This means that moving an operator from one side of an inner product to the other requires taking its adjoint (and thus a complex conjugate).

  • For a matrix, the adjoint is its conjugate transpose:

\[ A^\dagger = (A^T)^* \]

That is, swap rows and columns, then take the complex conjugate of every entry:

\[ (A^\dagger)_{jk} = A_{kj}^*. \]

In matrix element form taking the adjoint generates different elements

\[ a_{jk} = \langle \psi_j | \hat{A} | \psi_k \rangle \quad \Rightarrow \quad a^*_{kj} = \langle \psi_k | \hat{A}^\dagger | \psi_j \rangle. \]

Hermitian (Self-Adjoint) Operators#

An operator is Hermitian if it equals its own adjoint:

\[ \hat{A} = \hat{A}^\dagger. \]

This means the operator behaves the same way when acting on either side of the inner product.

Hermitian Matrix

\[ A = A^\dagger, \quad a_{jk} = a_{kj}^*. \]

Hermitian Operator

\[ \langle \phi | \hat{A}\psi \rangle = \langle \hat{A}\phi | \psi \rangle, \qquad \text{or equivalently,} \qquad \langle j| \hat{A}|k\rangle = \langle k| \hat{A}|j\rangle^*. \]

In integral form:

\[ \int \psi_j^*(x), [\hat{A}\psi_k(x)],dx = \int \psi_k(x), [\hat{A}\psi_j(x)]^*,dx. \]

Why Hermitian Operators Matter#

  1. Eigenvalues are real: Observables in quantum mechanics (energy, momentum, position, etc.) are represented by Hermitian operators, ensuring all measurement outcomes are real numbers.

\[ \hat{A}|\psi\rangle = a|\psi\rangle \implies a \in \mathbb{R}. \]
  1. Eigenfunctions are orthogonal:

\[ \langle \psi_m | \psi_n \rangle = 0 \quad (m \ne n). \]

Example of Hermitian Matrix

Which of these matricies is Hermitian?

\(\begin{pmatrix} 1 & 2 \\ 3 & 4 \end{pmatrix}\), \(\begin{pmatrix} i & 0 \\ 0 & 1 \end{pmatrix}\), \(\begin{pmatrix} -1 & -3i \\ 3i & 8 \end{pmatrix}\), \(\begin{pmatrix} 1 & 2i \\ 2i & 3 \end{pmatrix}\)

  • To see that Differentiation operators are Hermitian requires a little more work.

  • A trick that helps see it is integration by parts where the constant term is zero because wavefunction decays to zero at boundaries (postulate 1, keeping probability finite)!

\[\int \psi_1 d\psi_2 =- \int \psi_2d\psi_1 + \psi_1\psi_2\Big|_{x_{min}}^{x_{max}} =- \int \psi_2d\psi_1\]

Example of Hermitian Operator

Prove that the momentum operator (in one dimension) is Hermitian.

Geometric Intuition Hermitian operators are the analog of symmetric matrices in real vector spaces. They represent linear transformations that do not rotate vectors into complex directions—only stretch or compress them along real axes.

Problems#

Problem-1: Is \(xd/dx\) operator Hermitian#

Check weather the operator \(\hat{A} = xd/dx\) is Hermitian.

  • You can test weather the following condition holds:

\[ \int_{a}^{b} \psi_1^*(x) \left( x \frac{d}{dx} \psi_2(x) \right) \, dx = \int_{a}^{b} \left( x \frac{d}{dx} \psi_1(x) \right)^* \psi_2(x) \, dx \]
  • Note how complex conjugation applies to an expression with an operator inside!

  • But since our operator contains no imaginary number complex conjugation only applies to the wavefunction

Problem-2: Is \(d^2/dx^2\) operator Hermitian?#

  • You can test weather the following condition holds

\[ \int_{a}^{b} \psi_1^*(x) \left( \frac{d^2}{dx^2} \psi_2(x) \right) \, dx = \int_{a}^{b} \psi_2(x)\left( \frac{d^2}{dx^2} \psi_1(x) \right)^* \, dx \]
  • Note how complec conjugation applies to an expression with operator inside. but since our operator contains no imaginary numbers it will only apply to wavefunction

Problem-3: Is \(id^2/dx^2\) operator Hermitian?#

  • You can test weather the following condition holds

\[ \int_{a}^{b} \psi_1^*(x) \left( \frac{d^2}{dx^2} \psi_2(x) \right) \, dx = \int_{a}^{b} \psi_2(x)\left( \frac{d^2}{dx^2} \psi_1(x) \right)^* \, dx \]

Problem-4: Identify Hermitian Matrices#

\[\begin{split} A = \begin{pmatrix} 1 & 2 \\ 2 & 3 \end{pmatrix} \end{split}\]
\[\begin{split} B = \begin{pmatrix} i & 1 \\ -1 & -i \end{pmatrix} \end{split}\]
\[\begin{split} C = \begin{pmatrix} 2 & i \\ -i & 2 \end{pmatrix} \end{split}\]

Problem-5 Momentum Matrix#

Show how the momentum operator looks in matrix form using a finite-dimensional example where you evaluate wavefunction onf 4 points which will correspond to \(4 \times 4\) matrix.