Advances in Physical Chemistry

Volume 2018, Article ID 1487982, 11 pages

https://doi.org/10.1155/2018/1487982

## Solution of the Rovibrational Schrödinger Equation of a Molecule Using the Volterra Integral Equation

^{1}Department of Physics, Beirut Arab University, P.O. Box 11-5020, Riad El Solh, Beirut 1107 2809, Lebanon^{2}Department of Physics, Khalifa University, P.O. Box 57, Abu Dhabi, UAE

Correspondence should be addressed to Mahmoud Korek; moc.oohay@kerokf

Received 22 May 2018; Revised 28 June 2018; Accepted 30 July 2018; Published 4 October 2018

Academic Editor: Joel Bowman

Copyright © 2018 Mahmoud Korek and Nayla El-Kork. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

#### Abstract

By using the Rayleigh-Schrödinger perturbation theory the rovibrational wave function is expanded in terms of the series of functions , where is the pure vibrational wave function and are the rotational harmonics. By replacing the Schrödinger differential equation by the Volterra integral equation the two canonical functions and are well defined for a given potential function. These functions allow the determination of (i) the values of the functions at any points; (ii) the eigenvalues of the eigenvalue equations of the functions which are, respectively, the vibrational energy , the rotational constant , and the large order centrifugal distortion constants . Based on these canonical functions and in the Born-Oppenheimer approximation these constants can be obtained with accurate estimates for the low and high excited electronic state and for any values of the vibrational and rotational quantum numbers v and J even near dissociation. As application, the calculations have been done for the potential energy curves: Morse, Lenard Jones, Reidberg-Klein-Rees (RKR), ab initio, Simon-Parr-Finlin, Kratzer, and Dunhum with a variable step for the empirical potentials. A program is available for these calculations free of charge with the corresponding author.

#### 1. Introduction

For the gas-phase of molecules, the rovibrational spectroscopy remains the most accurate and reliable source for understanding the molecular structure and the development of quantum mechanics. The theoretical determination of the rovibrational constants enables the prediction of line positions, which can be used in guiding experimental investigations and greatly facilitating the detection of unknown molecules. The fundamental atomic and molecular data are very important for interpreting observations and modelling the atmospheres of planetary and stellar objects. In the last two decades hundreds of extrasolar planets, cool stars, the atmospheres of exoplanets, and brown dwarfs have been detected where their other proposal uses molecular states.

Since diatomic molecules exhibit very long decoherence times, the molecular electronic states (electronic, vibrational, and/or rotational) are chosen recently to act as the quantum bits (qubits). A femtosecond laser pulse can be shaped to produce high fidelity binary shaped laser pulses to act as a quantum logic gate on the chosen qubits. By using the vibrational states of a linear molecule of N atoms as the qubit basis the number of vibrational degrees of freedom is 3N – 6. A huge quantity of spectroscopic data is required to implement experiments using shaped laser pulses for molecular quantum computing. However, these data are still not available in literature.

The quantum mechanical canonical function method is a powerful technique to compute the very large rotation–vibration constants. The vibration rotation energy of an electronic state of a diatomic molecule is commonly represented by , where , v and J are, respectively, the vibrational and rotational quantum numbers, is the pure vibrational energy, the rotational constant, and are the centrifugal distortion constants (CDC) related to a potential energy curve U(r) and v. In the conventional approach of the Raleigh-Schrödinger perturbation theory (RSPT) [1] Albritton et al. [2] derived the first analytical expressions of the centrifugal distortion constants (CDC). Because of the complexity and the tedious computation of these expressions, Hutson derived, by using the Numerov difference equation, an algorithm [3] that allows the determination of the constants , , , and only in terms of the vibrational wave function . However, difficulties appeared in this algorithm for some potentials (like the Lennard-Jones potential) in treating high vibrational level (near dissociation for example) [4]. Because of the development in spectroscopic techniques, higher orders are needed [5]. Tellinghuisen [4] introduced an improvement to Huston algorithm, but it was not sufficient to accede these levels and reach large order of the centrifugal distortion constants.

In this work, we present the canonical function method (superposition of difference equation) which is an advantageous technique with respect to other difference equation methods for solving the vibration-rotation eigenvalue problem [6]. Moreover, with this canonical approach with variable step [7], the large order of CDC and the calculation of the vibration-rotation energy levels, even near dissociation [8], for any potential energy curves (either RKR [9], ab initio [10–12], or empirical [13]) and for any electronic state can be done by one single and simple routine. A program is available for these calculations free of charge with the corresponding author, with applications on the six potential energy curves: Morse potential, Lenard Jones potential, RKR potential, ab initio potential, Simon-Parr-Finlin (Kratzer) potential, and Dunhum potential. Since the canonical function method is proven to be valid for any type of potential function, the program is structured in different subroutines for the different cited potentials in order to give the possibility to use the same program for other types of potentials that can be added.

#### 2. The Theory

##### 2.1. Vibration-Rotation Canonical Functions

In the Born-Oppenheimer approximation, the vibration-rotation motion of a diatomic molecule is described by the wave function and the energy that are, respectively, the eigenfunction and the eigenvalue of the radial Schrödinger equation [18]where , *μ* is the reduced mass, is the Planck constant, v and J are the vibrational and rotational quantum numbers, respectively, and r is the internuclear distance. This equation can be simply represented bywhere is the value of r at the equilibrium, andwith . Equation (1) is equivalent to the Voltera integral equation [19]in the sense that any solution of (4) is solution of (1). By inserting by its expression many times in the integral, one can find:where and are two canonical functions [20, 21] defined bywith the well determined initial valuesThe initial value for the unnormalized wave function can be deduced from and by using (5), on one hand, and on the other hand the boundary conditions [5]We find

##### 2.2. The Rotational Schrödinger Equations

In the Rayleigh-Schrödinger perturbation theory (RSPT), the eigenvalue and the eigenfunction of (1) are given, respectively, bywhere is the pure vibrational energy, is the rotational constant, , are the centrifugal distortion constants, is the pure vibrational wave function, and is the rotational correction.

Therefore, the energy factor can be written aswhere From this expression of , the functions and are given by By replacing , by their expressions and in (5) we findA set of Schrödinger equations is obtained by replacing the wave function by its expression in (2) to obtain a set of Schrödinger equations where the first of these equations is the radial Schrödinger equation of pure vibration. All the other equations are called the rotational Schrödinger equations.

##### 2.3. Analytic Expressions of the Rotation Harmonics

###### 2.3.1. Pure Vibration

For one electronic state and for a given potential function, the solution of the vibrational Schrödinger equation is given by [21]where and are the pure vibration canonical functions defined by (5) and in which we replace by (i.e., we take J=0).

###### 2.3.2. Calculation of the Rotational Harmonics

A rotational Schrödinger equation , , , is given byIf we multiply this equation by and integrate it between zero and x, it will becomeBy replacing by its expression , , in , , , we obtainwherefor x = 0Thereforefor the derivation of order iandFor the unnormalized wave function, we chose , and by using we obtainBy replacing by its value in we obtainOn the other hand, the rotation harmonics must be vanished at the boundaries ; thus givesand the rotation harmonic is given by [7]This expression is valid without any restriction for the form of a given potential function.

##### 2.4. Numerical Method

###### 2.4.1. Calculation of the Vibrational Wave Function

For one electronic state and for a given potential function, the vibrational wave function is given byThe determination of requires the calculation of , , and .

*(1) Calculation of ** and *. For a given potential, and on one interval, has the polynomial formThe canonical functions and are particular solutions of , because is expanded in polynomial [8]; and also can be expanded asBy representing and by the same function y(r) for a given potential U(r) and energy E, the function y(r) is given byBy using , we obtain the following recursion relation:withThe initial values and are given bywhere y(0) = 1, for function and y(0) = 0, for function

Therefore, the canonical functions and are well determined at any point r.

*(2) Calculation of *. From equation (5), the wave function is given byBy using the boundary conditions one can findFor the unnormalized wave function , the vibration function is determined at any point r.

##### 2.5. Diatomic Centrifugal Distortion Constants (CDC)

The rotational equations , , , and are all of the formWe multiply this equation by and integrate between and . By making use of , , , and , we obtainThen we make use of the boundary conditions for and z (at ) on one hand, and of on the other hand, and we findand similarly, for the other boundary condition (at ),The continuity condition for s(r) implies the equality of given by ; thereforeThis equation gives for the successive value of s These equations give simple expression of in terms of whereOnce is given, the determination of is reduced to that of simple definite integrals and depending on [19].

#### 3. Results and Discussion

##### 3.1. Calculated Data

In order to test the accuracy and the validity of the present method for the calculation of the different rovibrational parameters we used as examples for the empirical potentials Morse potential [14, 15], Lenard Jones potential [4, 8, 22], Dunhum Potential [23], and Kratzer potential [24]. For the investigation of the ab initio potential [17], we employed the state averaged Complete Active Space Self Consistent Field (CASSCF) followed by Multireference Configuration Interaction (MRCI) method with Davidson correction (+Q), single and double excitations. In these entire calculations, we used the computational chemistry program MOLPRO [25] with the advantage of the graphical user interface GABEDIT [26]. The CASSCF configuration space was used as the reference in the MRCI calculations. The potential energy U(r) is determined in terms of the internuclear distance r with a step equal to 0.02Å. The Schrödinger equation is solved by using the cubic spline interpolation between every two consecutive points. A Rydberg-Klein-Rees (RKR) potential is obtained experimentally and defined by their turning points [27–29]. The calculation has been done by an interpolation between every two consecutive points also by using the cubic spline interpolation. If the number of the turning points is limited we extrapolate the potential function curve to the right by [30] U(x) = a/r^{6} +b/r^{8} +c/r^{10} and to the left by U(x) = d/r^{15} where the constants a, b, c can be obtained from literature (if available) and d is calculated using the used potential energy curve.

The comparison of our calculated values of the vibrational energy levels , by using the empirical Morse potential for the molecule CO, with those given in [14, 15] shows, respectively, excellent agreements with the relative differences and (Table 1). Similar excellent agreement = 0.00% is obtained by comparing our calculated values of the different vibrational levels with those given by Martin et al. [16] of the RKR potential of the molecule I_{2} (Table 2). Our method is also tested in case of an* ab initio* potential for the molecule ZnBr [17] (Table 3). Recently [31], we calculated, by using the present method, the rovibrational parameters , , of the molecule LiSr by using an* ab initio* potential. The comparison of the calculated values of and with those available in literature shows the good agreement with the relative differences for and acceptable agreement for . However, the comparison of our calculated values of , by the present method, with those given in literature shows a very good agreement with the relative difference for the molecule NaSr [31].