Secular equations in the variational method

Secular equations in the variational method are linear equations obtained by minimising the energy with respect to trial-function coefficients, yielding a matrix eigenvalue problem whose solutions give approximate energy levels and corresponding wavefunctions.

Consider the eigenvalue equation \(\hat{H}\psi=E\psi\), where \(\psi=\sum_jc_j\phi_j\) and the coefficients \(c_j\) are real. Multiplying the eigenvalue equation on the left by \(\psi^*\) and integrating over all space gives:

$$E\sum_i\sum_jc_ic_j\langle\phi_i\vert\phi_j\rangle=\sum_i\sum_jc_ic_jH_{ij}$$

where \(H_{ij}=\langle\phi_i\vert\hat{H}\vert\phi_j\rangle\).

Using the orthonormality condition \(\langle\phi_i\vert\phi_j\rangle=\delta_{ij}\) results in:

$$E\sum_ic^2_i=\sum_i\sum_jc_ic_jH_{ij}$$

Taking the partial derivative of \(E\) with respect to each coefficient \(c_k\) gives:

$$\frac{\partial E}{\partial c_k}\sum_ic^2_i+E\frac{\partial}{\partial c_k}\sum_ic^2_i=\frac{\partial}{\partial c_k}\sum_i\sum_jc_ic_jH_{ij}$$

Applying the variational method by setting \(\frac{\partial E}{\partial c_k}=0\) and computing the remaining derivatives yield:

$$2Ec_k=\sum_i\sum_j(\delta_{ik}c_jH_{ij}+c_i\delta_{jk}H_{ij})$$

$$2Ec_k=\sum_jc_jH_{kj}+\sum_ic_iH_{ik}$$

Assuming that the expectation values are real and the Hamiltonian is Hermitian, \(H_{ik}=H_{ki}\), and

$$2Ec_k=\sum_jc_jH_{kj}+\sum_ic_iH_{ki}$$

Relabelling the second summation index from \(i\) to \(j\), we have \(Ec_k=\sum_jc_jH_{kj}\) or equivalently,

$$Ec_k=\sum_{j\neq k}c_jH_{kj}+c_kH_{kk}$$

which rearranges to the secular equations:

$$(H_{ii}-E)c_i+\sum_{j\neq i}H_{ij}c_j=0$$

where we have relabelled the index \(k\) to \(i\).

We can express the secular equations in the following matrix form:

$$\begin{pmatrix}H_{11}-E&H_{12}&\cdots&H_{1n}\\H_{21}&H_{22}-E&\cdots&H_{2n}\\\vdots&\vdots&\ddots&\vdots\\H_{n1}&H_{n2}&\cdots&H_{nn}-E\end{pmatrix}\begin{pmatrix}c_1\\c_2\\\vdots\\c_n\end{pmatrix}=0$$

This is a linear homogeneous equation with non-trivial solutions only if

$$\begin{vmatrix}H_{11}-E&H_{12}&\cdots&H_{1n}\\H_{21}&H_{22}-E&\cdots&H_{2n}\\\vdots&\vdots&\ddots&\vdots\\H_{n1}&H_{n2}&\cdots&H_{nn}-E\end{vmatrix}=0$$

or equivalently, if

$$det(\boldsymbol{H}-E\boldsymbol{I})=0$$

where \(\boldsymbol{H}\) is the Hamiltonian matrix with elements \(H_{ij}=\langle\phi_i\vert\hat{H}\vert\phi_j\rangle\) and \(\boldsymbol{I}\) is the \(n\times n\) identity matrix.

Expanding the determinant gives the characteristic equation, which can then be solved for its roots.

Secular equations are used to determine the allowed energy levels and wavefunctions of electrons in molecules. They are fundamental in solving eigenvalue problems in many aspects of chemistry, including:

    1. Degenerate perturbation theory
    2. The Hückel method
    3. The Stark effect and the Zeeman effect
    4. Vibration of polyatomic molecules
    5. The Hartree-Fock-Roothaan procedure

 

Next article: Overlap integral
Previous article: The variational method
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 *

Mono Quiz