Selection of semi-empirical calculation methods for insecticide development

Background: Insecticides are substances used to control, repel, or eradicate troublesome organisms, particularly insect-based plant pests. The discovery of new insecticide compounds fuels the ongoing development of insecticides. The integration of computational chemistry into the development of insecticidal chemicals was beneficial. Objective: This study aims to identify the most suitable method among 12 available semiempirical calculation methods in the Hyperchem application. Methods: The selection process involved comparing experimental data of the infra-red spectrum of chlorpyrifos with corresponding calculation data. Results: The largest Predicted Residual of Sum Squares (PRESS) value was observed in the INDO method of 55466.3856. Conversely, the smallest PRESS value was observed in the AM1, measuring 3242.6549. The AM1 semiempirical method yields the smallest value. Conclusion: The results indicated that the calculation method chosen was the AM1 semiempirical method.


Introduction
Insecticides play a significant role in increasing agricultural production, particularly in controlling plant pests.However, insecticides also possess toxic properties.In Indonesia, most insecticides utilized belong to the organophosphate group, with chlorpyrifos being one example [1].Chlorpyrifos is a highly toxic compound with an LC 50 value of 0.024 g/L in fish [2].This value is used to determine the toxicity of insecticides in a given concentration and its potential to cause the death of test animals [3,4].Due to the various functional groups present, the use of insecticide combinations offers a variety of reaction mechanisms [5].
In the 1950s, the development of computer technology started molecular modeling.Techniques invented by artificial intelligence (AI) developing computational scientists have been mainly applied to drug design in recent years.These methods are called de novo or rational drug design [6].The general method is used to identify the active functional group and enter the desired functional group to interact with other functional groups.This method also studies toxicological and anti-inflammatory effects [7].Some researchers studied a compound, namely chlorpyrifos.
Chlorpyrifos is a white solid with a sharp odor.If chlorpyrifos enters the body waters, it will kill aquatic biotas such as fish and shrimp.This chlorpyrifos insecticide is non-systemic and works when it comes in contact with the skin, is ingested, and is inhaled [8].Its molecular formula is C 9 H 11 Cl 3 NO 3 PS, with a molecular weight of 350.59 g/mol.Chlorpyrifos has a melting point of 42 °C and a specific gravity of 1.4 g/cm³.It belongs to the organothiophosphate group, as illustrated in Figure 1.
The present study employed a semi-empirical calculation method, which was carried out using the Hyperchem software.There are 12 different semiempirical methods available for calculation [9], and the method selected for chlorpyrifos calculation was based on experimental infrared (IR) spectra data [10,11].Semiempirical methods utilize the principles of quantum mechanics [12] (HyperCube, 2007).The choice of the method was determined by comparing the results with the experimental IR spectra [11].This study aims to select one method of 12 semi-empirical calculation methods available in the Hyperchem application.

Methods Equipment
This study was theoretical research conducted on a computer with the following specifications: Intel(R) Core (TM) i5-6500 CPU @ 3.20GHz, 8.00 GB RAM, Windows 10 64-bit operating system, x64 processor, and Hyperchem 8.0 Program.The research used the molecular model of chlorpyrifos and selected the semiempirical method based on the infrared spectrum (IR) obtained from previous studies [4].

Molecular modeling of chlorpyrifos
After the chlorpyrifos molecule reached stable energy, the infrared spectrum was calculated using the Hyperchem Program.The calculation was performed by selecting the Compute menu and the vibrational spectrum in the available options.

Infrared spectrum analysis
The infrared spectrum was calculated using the Hyperchem Program by selecting the Compute menu and the vibrational spectrum from the available options.

Method selection and analysis
The data were obtained from the IR spectrum calculation method for each method used.The best calculation method was determined by the Predicted Residual of Sum Squares (PRESS) method, where the smallest value was chosen as the result.

Results
A molecular of chlorpyrifos was generated, then the model was transformed into a three-dimensional (3D) image and subjected to geometry optimization (Table 1).Subsequent to this, we calculate the infrared spectrum of the optimized geometry.To accomplish this, various calculation methods were employed to determine the infrared spectrum.The outcome of these calculations are presented in Figure 2 and summarized in Table 1.
Experimental infrared spectrum data, sourced from previous studies (referred to as experimental data in Table 1) [4], were used for comparison.The calculation results of the infrared spectrum, obtained using the Hyperchem version 8 program, are outlined in Table 2 [9].The comparison between the calculated and experimental data for the chlorpyrifos infrared spectrum is provided in Table 3.
Identifying the best semi-empirical calculation method was based on the criterion of achieving the lowest Predicted Residual of Sum Squares (PRESS) value [11].The final results of this calculation are provided in Table 4.

Discussion
The initial step involves modeling the chlorpyrifos molecule and optimizing its geometry, as depicted in Figure 1.Subsequently, the next stage encompasses calculating the infrared spectrum.After obtaining the infrared spectrum value, a comparison was drawn between these values and the closest experimental data [11].
Table 1 shows that among these semi-empirical calculation methods, the CNDO approach yields the lowest energy value of -128624.2295725kcal/mol.Conversely, the MNDO semi-empirical calculation method has the highest value of -90619.6154333kcal/ mol.Notably, there semi-empirical calculation methods are incapable of conducting molecular modeling for chlorpyrifos compounds [12].These methods are Extended Huckle, MINDO3, and ZINDO-S.
The foundational CNDO method simplifies calculations by considering only s and p orbitals, represented through linear combinations of the Slater function [13].Conversely, the integral MNDO method lacks parameterization but approximates calculations using the classical multipole technique [14].Extended Huckle, MINDO3, and ZINDO-S aim to increase the applicability of theories such as INDO, rendering them more versatile [12].
Experimental data for the infrared spectrum was obtained from previous research, herein referred to as experimental data [4].This data is presented as spectrum values and wave numbers, as listed in Table 2.
The research data involves computed infrared spectrum results using the Hyperchem version 8 program (Table 3).Notably, three semi-empirical methods did not produce an infrared spectrum, suggesting these methods cannot be optimized in geometry.
Table 3 shows the proximity between experimental and calculated values.The lowest value of the Predicted Residual of Sum Squares (PRESS) is the value of the selected semi-empirical calculation method [11].This analysis leads to Table 4. Conversely, similar to the empirical method, the AM1 approach exhibits the lowest PRESS value of 3242.6549[11].Based on these outcomes, it can be concluded that the AM1 semi-empirical method is the most suitable for insecticide development.

Conclusions
The AM1 semi-empirical calculation method was the calculation method chosen and suitable for use in the development of insecticidal compounds.

Table 4 ,
the INDO semi-empirical method yields the largest PRESS value of 55466.3856.

Table 1 .
Geometry optimization of chlorpyrifos compound through semiempirical methods

Table 2 .
Experimental data of infrared spectrum and wave

Table 3 .
Data calculation and experimental data of chlorpyrifos infrared spectrum

Table 4 .
Predicted residual of sum squares (PRESS) data for all semiempirical calculation methods