Postulates of quantum mechanics

The postulates of quantum mechanics are fundamental mathematical statements that cannot be proven. Nevertheless, they are statements that everyone agrees with.

Examples of other postulates of science and mathematics are Newton’s 2nd law, F=ma, and the Euclidean statement that a line is defined by two points respectively.

Generally, the postulates of quantum mechanics are expressed as the following 6 statements:

1) The state of a physical system at a particular time t is represented by a vector in a Hilbert space.

We call such a vector, a wavefunction \psi(x,t), which is a function of time and space . A wavefunction contains all assessable physical information about a system in a particular state. For example, the energy of the stationary state of a system is obtained by solving the time-independent Schrodinger equation \hat{H}\psi=E\psi.

2) Every measurable physical property of a system is described by a Hermitian operator \hat{O} that acts on a wavefunction representing the state of the system.

The most well-known Hermitian operator in quantum mechanics is the Hamiltonian \hat{H}, which is the energy operator. The wavefunctions that \hat{H} acts on are called eigenfunctions. Eigenfunctions of a Hermitian operator in quantum mechanics are further postulated to form a complete set. Other Hermitian operators frequently encountered in quantum mechanics are the momentum operator \hat{p} and the position operator \hat{x}.

3) The result of the measurement of a physical property of a system must be one of the eigenvalues of an operator \hat{O}.

The state of a system is expressed as a wavefunction, which can be a single basis wavefunction or a linear combination of a complete set of basis wavefunctions. Since basis wavefunctions of a Hermitian operator form a complete set, all wavefunctions can be written as a linear combination of basis wavefunctions.



What about a system described by a stationary state?


The wavefunction can be expressed, though trivially, as \psi=\sum_{n=0}^{\infty}c_n\phi_n, where c_i=1 and c_{n\not\equiv i}=0 (we have assumed that the spectrum is discrete, i.e. the eigenvalues are separated from one another).


It is generally accepted by scientists that the values of the coefficients c_n are unknown prior to a measurement. Upon measurement, the result obtained is an eigenvalue associated with one of the eigenfunctions, and hence the phrase ‘the initial wavefunction collapses to one of the eigenfunctions’. This implies that the measurement alters the initial wavefunction such that a 2nd measurement, if made quickly, yields that same result (this obviously refers to wavefunctions describing non-stationary states, as wavefunctions of stationary states are independent of time). If we prepare a large number of identical systems and simultaneously measure them, the values of c_n are found; with \sum_{n=0}^{\infty}\vert c_n\vert^{2}=1, and the expectation value of the measurements being \langle\psi\vert\hat{O}\vert\psi\rangle.

4) The probability of obtaining an eigenvalue E_i upon measuring a system is given by the square of the inner product of the normalised \psi with the orthonormal eigenfunction \phi_i.

In other words,

\vert\langle\phi_i\vert\psi\rangle \vert^{2}=\vert\langle\phi_i\vert\sum_{n=0}^{\infty}c_n\phi_n\rangle \vert^{2}=\vert\langle c_i\phi_i\vert\phi_i\rangle \vert^{2}=\vert c_i\vert^{2}

If the spectrum is continuous, \psi=\int_{-\infty}^{\infty}c_n\phi_n\, dn and the probability of obtaining an eigenvalue in the range dn is \vert\langle\phi_n\vert\psi\rangle \vert^{2}\, dn.

5) The state of a system immediately after a measurement yielding the eigenvalue E_i is described by the normalised eigenfunction \phi_i.

We have explained in the postulate 3 that this is commonly known as the collapse of the wavefunction \psi to one of the eigenfunctions \phi_i. It is also known as the projection of \psi onto \phi_i, i.e. \hat{P}_i\vert\psi\rangle; or if \psi not is normalised,




Show that \sqrt{\langle\psi\vert\hat{P}_i\vert\psi\rangle} is the normalisation constant.


To normalise a wavefunction,

NN\int c_{i}^{*}\phi_{i}^{*}c_i\phi_i\, d\tau=1\; \; \; \Rightarrow \; \; \; N^{2}=\frac{1}{\vert c_i\vert^{2}\langle\phi_i\vert\phi_i\rangle}\; \; \; \Rightarrow \; \; \; N=\frac{1}{ c_i\sqrt{\langle\phi_i\vert\phi_i\rangle}}


\frac{1}{ \sqrt{\langle\psi\vert\hat{P}_i\vert\psi\rangle}}=\frac{1}{ \sqrt{\langle\psi\vert c_i\phi_i\rangle}}=\frac{1}{ \sqrt{\langle c_i\phi_i\vert\psi\rangle^{*}}}=\frac{1}{ \sqrt{(c_{i}^{*}c_i\langle \phi_i\vert\phi_i\rangle)^{*}}}

=\frac{1}{ \sqrt{c_{i}^{*}c_i\langle \phi_i\vert\phi_i\rangle}}=\frac{1}{ c_i\sqrt{\langle \phi_i\vert\phi_i\rangle}}


6) The time evolution of the wavefunction \psi(t) is governed the time-dependent Schrodinger equation i\hbar\frac{d}{dt}\vert\psi(t)\rangle=\hat{H}(t)\vert\psi(t)\rangle.


Next article: one particle, time-independent schrodinger equation
Previous article: linear combination of wavefunctions
Content page of quantum mechanics
Content page of advanced chemistry
Main content page

Leave a Reply

Your email address will not be published. Required fields are marked *