Abstract

Organophosphorus compounds are organic compounds widely employed in agriculture as well as in chemical weapons. The use in agriculture is due to their insecticidal properties. However, in chemical warfare, the use of organophosphorus is associated with acetylcholinesterase inhibition, which promotes the cholinergic syndromes. In this line, the fast detection of this class of compound is crucial for the determination of environmental exposure. This improved detection will naturally allow for more prompt courses of treatment depending on the contaminant findings. In this perspective, the dipyrrinone oxime (1) was employed for the detection of organophosphorus compounds that are employed as nerve agents, such as cyclosarin, sarin, soman, diethyl chlorophosphate, diisopropylfluorophosphate, 2-(dimethylamino)ethyl N,N-dimethylphosphoramidofluoridate, O-ethyl-S-[2-(diethylamino)ethyl]methylphosphonothioate, O-ethyl-S-[2(diisopropylamino)ethyl] methylphosphonothioate, and O,O-diethyl-S-[2-(diethylamino)ethyl] phosphorothioate, through fluorescent emission. The thermodynamics and kinetic parameters as well as spectroscopic properties of the complexes formed for 1 and all organophosphorus compounds previously cited were investigated by means of theoretical calculations. From our findings, only the diethyl chlorophosphate, 2-(dimethylamino)ethyl N,N-dimethylphosphoramidofluoridate, and O,O-diethyl-S-[2-(diethylamino)ethyl] phosphorothioate emitted fluorescence in the hexane, toluene, chloroform, dichloromethane, methanol, acetonitrile, water, and dimethyl sulfoxide solvents. The study of the absorption wavelength with the most polar solvent showed higher values compared to apolar solvents. In the same solvent, for instance, soman in hexane showed the lowest absorption wavelength value, 324.5 nm, and DCP the highest value, 330.8 nm. This behavior was observed in other tested solvents. The thermodynamic parameters indicate negative Gibbs free energy (ΔG) values for the O-ethyl-S-[2(diisopropylamino)ethyl] methylphosphonothioate with 1 reaction. On the other hand, the sarin and cyclosarin revealed the lowest Gibbs free energy (ΔG) values, being kinetically favorable and presenting more reactivity.

1. Introduction

Organophosphorus (OP) compounds are organic compounds that have, at least, one carbon-phosphorus chemical bond. The first OP compounds were prepared in the Middle Ages, but in the XIX century, their study became more intense by Michaelis [13]. In 1930, toxic and insecticidal properties of these compounds were reported [4]. Since then, there has been growing interest in the study of this class of compounds due to their possible application as agrochemicals and chemical weapons [5, 6]. The high toxicity of the OP compounds is associated with high affinity for the acetylcholinesterase enzyme [7, 8]. Actually, it is well known that OP compounds can form a stable chemical bond with the serine amino acid residue present in the active site of the acetylcholinesterase enzyme (AChE) [8, 9] Among the main nerve agents, it is possible to highlight the compounds cyclosarin, sarin, soman, DCP (diethyl chlorophosphate), DFP (diisopropylfluorophosphate), GV (2-(dimethylamino)ethyl N,N-dimethylphosphoramidofluoridate), VM (O-ethyl-S-[2-(diethylamino)ethyl]methylphosphonothioate), VX (O-ethyl-S-[2(diisopropylamino)ethyl] methylphosphonothioate), and VG (O,O-diethyl-S-[2-(diethylamino)ethyl] phosphorothioate) (Figure 1).

It should be kept in mind, however, that the reactivation reaction can occur by spontaneous hydrolysis, which is a very slow process, or by using oximes, which are considered AChE reactivators [10, 11].

To date, the use of oximes is the most efficient approach aiming at AChE reactivation [12, 13]. The AChE inhibition process can be illustrated in Scheme 1 [14].

Although the use of OP compounds as chemical weapons has called much attention and generated wide discussion in the international community [15, 16], it is important to keep in mind that the abusive use of OP compounds, as agrochemicals, is dangerous [16]. Currently, the use and misuse of OP compounds in agriculture are responsible for the majority of poisonings in Brazil [4, 16].

In view of this scenario, the detection and quantification of OP compounds in the environment are crucial [2]. In 2013, Walton et al. suggested the use of fluorescent techniques to detect OP compounds [17]. In fact, when OP compounds bind to a fluorescent probe, such as dipyrrinone oxime (2-ethyl-7-[(Z)-(hydroxyimino)methyl]-1-methyl-3H-dipyrrolo[1,2-c:2′,1′-f]pyrimidine-3,5-dione), the color changes from red to yellow (Scheme 2) [17].

Despite several efforts to apply fluorescent probes for OP detection, there is still a dearth of systematic understanding of the thermodynamics of guest-responsive hybrid frameworks. In this perspective, the theoretical work could be an ally in the rationalization of thermodynamics and spectroscopic properties of these materials [18]. Furthermore, theoretical approaches provide an economically viable route for applications, as these methods can also open a new avenue to study the structural modifications, which could optimize the detection of OP compounds [19].

Thus, aiming at combining efficiency and selectivity in the OP compound detection process, this paper is devoted to the theoretical study of spectroscopic properties of nerve agents as fluorescent probes.

2. Methodology

The oxime-organophosphorus complex calculations were carried out in the ground and the excited states using the DFT and TD-DFT levels [20, 21], respectively. Six different functionals were tested (B3LYP, CAM-B3LYP, B3PW91, MPW1PW91, PBEPBE, and ωB97XD) [22, 23]. The functional that best approached the wave function theoretical result with respect to the experimental result was selected for performing the calculations together with the DGTZVP basis set. For the potential energy surface, the surface response was employed that was developed by selecting two α1 (C26, C28, C29, and C31) and α2 (C29, C31, C32, and C33) dihedral angles according to Figure 2. The energy for the ground and excited states was calculated, and a potential energy surface scan of dihedral angles α1 and α2 from −180 to 180° was performed, according to Table 1. The effect of the methanol solvent on the surface response was carried out by employing IEFPCM [24] for each point of the surface.

All compounds involved in the reaction described in Scheme 2 were optimized and the frequency was calculated at the DFT/B3LYP/DGTZVP level. The transition state was evaluated using Spartan quantum software [25] The imaginary frequency for all transition states was determined at the DFT/B3LYP/DGTZVP level [26, 27]. The relaxation of the solvent was investigated using the Tamm–Dancoff approximation (TDA) [2830] for the nine organophosphorus compounds (cyclosarin, DCP, DFP, GV, sarin, soman, VG, VM, and VX) complexed with the dipyrrinone oxime (1), taking into account the effect of the solvents (such as hexane, acetonitrile, methanol, toluene, chloroform, dichloromethane, dimethyl sulfoxide, and water). The TDA is a computationally simple method for molecular excited states, which can be used for predicting vibrationally resolved absorption and emission spectra of diverse molecules [2830]. These calculations were performed in conjugation with the TD-DFT method for the solvent relaxation and spectroscopic property calculation investigation for each oxime/organophosphorus compound. All spectroscopic calculations were carried out in Orca quantum chemistry [31].

3. Results and Discussion

3.1. Selection of the DFT Functional for Spectroscopic Calculations

The spectroscopic properties of the complexes formed by 1 and the organophosphorus compounds studied (Figure 2) were evaluated by both DFT and TD-DFT methods. At the first moment, the selection of the best DFT functional was performed by error calculation in the absorption wavelength value for the oxime (1)/DFP complex. This complex showed an absorption wavelength at 400 nm.

TD-DFT/DGTZVP was employed in conjugation with the DFT functionals B3LYP, CAM-B3LYP, B3PW91, MPW1PW91, PBE, and ωB97XD with the IEFPCM model for methanol, which was used in the experimental part. These functionals were tested for behold different families: meta-GGA, hybrid functional, and long-range hybrid functional. The compound geometric properties showed similarity in the DFT and TD-DFT methods, which allowed us to apply the same geometries in the ground and excited states. The bond lengths are similar for geometries generated by both methods, for instance, the P atom bounded with the three oxygen atoms in VG, DFP, and DCP and the P atom bonded to N, C, or S atoms in VG, VM, sarin, and soman. The distances between P and these atoms are reported in Supplementary Material, Table S3. In the same cases, the distance between P = O was around 1.46 Å, and the other O atoms showed values between 1.58 and 1.71 Å. The distance between P and F or Cl atoms was 1.68 Å, and the distance between P and S atom was 2.70 and 1.61 Å. After the reaction with oxime, all distances showed the same values, but the atoms bound with P atom were changed. In the same cases, the P atom linked with four O atoms with 1.80 at 1.64 Å.

From our findings, the best functional for the spectroscopic properties investigated was the B3LYP hybrid functional. The error for the B3LYP functional was 7.1 nm, while for B3PW91 it was –9.31 nm and for MPW1PW91 it was –13.22 nm (Table 1). On the other hand, the CAM-B3LYP and ωB97XD long-range hybrid functional [3235] showed higher error values, –52.87 nm and –54.75 nm, respectively. The PBE meta-GGA functional showed an error value of 78.82 nm, the highest error for the spectroscopic properties (Table 1).

The ANQ geometry for the enol and keto forms was optimized at the DFT level employing the six DFT functionals cited above, the B3LYP showed results with lower deviation between and and a good agreement with hybrid functional and long-range hybrid functional (CAM-B3LYP and ωB97XD) for the conjugation [28, 3638].

The B3LYP functional selected in the study was employed for the optimization in the ground state as well as in the theoretical calculations of twenty excited states for all compounds because the first 20 excited states give a good description of the electronic parameters for organophosphorus compounds [36, 3941].

3.2. Response Surface Analysis

It is well known that spectroscopic properties are modulated by structural features [4246]. Thus, to take into account the geometric effects on the nerve agents, we have used chemometric techniques, such as surface response methodology [47].

The response surface (RS) is a design of experiment (DOE) methodology employed for optimization (in this case minimization) of response [48, 49] The RS is based on the factorial design with a central point and using the quadratic equation provides for minimum response with the least number of experiments (in this case theoretically calculated) [50, 51]. The response surface was used to obtain the energy minimum in relation to the two α1 and α2 dihedral angles (Figure 2). The dihedral angles were modified from −180 to 180° according to that described in Supplementary Information Tables S1 and S2. The surface response in the gas phase is presented in Figure 3, which describes the optimization obtained using the employed response surface that connected the variation of α1 and α2 angles (see Figure 2) with the energy variation calculated by the B3LYP/DGTZVP method. The results indicated that the variation of angles in the studied compounds modified the calculated energy. The optimization generated the contour graph, the 2D dimension plot, as well as the contours map described in the surface and put in evidence the minimum energy values for dihedral angles [52, 53]. In the gas phase, the energy minimum is associated with α1 dihedral angle of 25° and with α2 dihedral angle of 0°. On the other hand, Figure 4 shows the same angles for the compounds in the solvent, which also describes the optimization obtained using the response surface, which connected the variation of angles α1 and α2 to the variation of energy calculated by the same method mentioned above; however, the oxime/organophosphate complex is in methanol.

From the dihedral angles (Figure 3), the spectroscopic properties were obtained for all analyzed compounds. The OPs analyzed were sarin, soman, DCP, DFP, cyclosarin, GV, VX, VM, and VG. Interestingly, all complexes formed between OP and the spectroscopic probe 1 showed the same dihedral angles for the energy minimum. Thus, the dihedral angles were fixed for all compounds for the TD-DFT/DGTZVP/B3LYP spectroscopic property calculations in the gas phase, as well as in the other eight solvents (hexane, chloroform, dichloromethane, toluene, methanol, acetonitrile, water, and DMSO).

3.3. Thermodynamics and Kinetic Parameters for the Reaction between Oxime (1) and Nerve Agents

The thermodynamic and kinetic parameters for the colorimetric reaction described in Scheme 2 were calculated using the DFT methodology. The 1 reacted with each OP compound forming the oxime/OP complex, having the leaving group (R4) according to the OP employed. The SN2 reaction mechanism was employed for this reaction [1] The enthalpy (ΔH), Gibbs free energy for reaction (ΔG), transition state Gibbs free energy (ΔG) values, and rate constants were calculated for the analysis of the spontaneity of this reaction. The thermodynamic and kinetic parameters are shown in Table 2.

Reaction I, employing the cyclosarin compound as substrate and fluorine as a leaving group, showed the ΔH value of –2.49 kcal/mol. This process was exothermic and the activation Gibbs free energy (ΔG) for reaching the transition state was 15.68 kcal/mol. It was possible to observe that reaction II showed higher ΔG, ΔH, and ΔG values than reaction I. The reaction II employed the DCP compound and the chlorine as leaving group. The ΔH for this reaction was the highest among the endothermic reactions (see Table 2).

Now, by analyzing the other reactions having the fluorine as leaving group, it is possible to notice that reaction III, employing DFP as OP, showed 1.80 and 6.49 kcal/mol for ΔH and ΔG, respectively. Thus, this reaction is endothermic and is not spontaneous. The reactions IV, V, and VI were exothermic with ΔH values of −0.51, −2.50, and −2.53 kcal/mol, respectively. Thus, having the fluorine as the leaving group, only reaction III was endothermic. The ΔG values for the reactions IV, V, and VI were 6.44, 2.68, and 3.08 kcal/mol, respectively. Similarly, all reactions were not spontaneous. The other four reactions exhibit the sulfur group as the leaving group. The reactions VII and VIII showed the 2-[di(propan-2-yl)amino]ethanethiolate compound as the leaving group, while the reactions IX exhibit the 2-(diethylamino)ethanethiolate as the leaving group. The reactions VII and VIII showed ΔH values of 0.37 and 0.28 kcal/mol while the ΔG values were 0.69 and 0.40 kcal/mol, respectively. Although the reaction showed positive ΔH and ΔG values, these systems were the most promising among the studied reactions without the sulfur group as the leaving group. The reaction IX showed the ΔH value of 0.08 kcal/mol and the ΔG value of −0.23 kcal/mol. The reaction IX was endothermic and spontaneous.

Turning now to kinetic studies, the ΔG values (Figure 5) were positive for all studied reactions, which is according to the transition state theory. Surprisingly, the halogen derivatives indicated the lowest ΔG values, probably due to lower steric hindrance. This fact was observed in reactions VII, VIII, and IX, where there is the presence of more bulky substituents as leaving groups. In fact, it is well known that steric factors are crucial for SN2 reactions.

Furthermore, the rate constant (k) was also calculated employing the Eyring equation [5457] according to equation (1). The rate constant demonstrates that sarin is more reactive, while VM was less reactive among the OPs. Therefore, the reactivity order was sarin > cyclosarin > DFP > DCP > GV > VG > VM > VX. The chemical reactions for VX were thermodynamically favorable, which can be characterized by negative ΔG values. The reaction with cyclosarin and sarin was the most kinetically favorable among all, revealing the lowest ΔG value. It should be kept in mind that all OP compounds were kinetically, but not thermodynamically, favorable for the colorimetric detection reaction of nerve agents (Scheme 2):

3.4. Spectroscopic Properties of the Oxime/Organophosphate Complex

The complexes formed between the oxime derivatives and OPs are displayed in Scheme 2. The compounds formed by the reaction of nine OPs with the oxime (Table 3) also have their spectroscopic properties calculated.

The absorption wavelengths of all complexes in eight solvents are reported in Table 4. The spectroscopic parameters were calculated by using the ORCA program at the TD-DFT/DGTZVP/B3LYP level, incorporating the Tamm–Dancoff approximation for evaluating the solvent relaxation as well as the absorption and emission energy for all compounds.

The study of the absorption wavelength with the most polar solvent showed higher values compared to apolar solvents. In the same solvent, for instance, soman in hexane showed the lowest absorption wavelength value (324.5 nm), followed by DFP (325.9 nm), sarin (326.2 nm), VM (326.4 nm), VG (327.7 nm), VX (327.9 nm), cyclosarin (328 nm), GV (329.4 nm), and DCP (330.8 nm). This behavior was observed in other tested solvents. For the same OPs, for example, GV, the absorption wavelength was according to the following order: hexane (329.4 nm), toluene (329.3 nm), chloroform (326.2 nm), dichloromethane (326.5 nm), methanol (325.7 nm), acetonitrile (325.8 nm), water (325.7 nm), and DMSO (326.4 nm). All values for the other compounds are described in Table 4. In fact, the constant dielectric (δ) value of the studied solvents influences the absorption wavelength. The solvent hexane (δ = 1.88) always showed lower values compared to water (δ = 80), which shows the highest constant dielectric value in the studied series. Other solvent properties, such as the dipolar moment (), can also be taken. From our findings, hexane ( = 0D) showed lowest absorption wavelength value compared to DMSO ( = 3.96D), according to Table 3. In this line, the dipolar moment can, in principle, rationalize the behavior observed in OP compounds cyclosarin, DCP, GV, sarin, soman, VG, VM, and VX, which showed higher absorption wavelength values compared to water ( = 1.85D). It is important to notice that DFP was the only OP that showed an inversion of the behavior, and in water solvent, it showed the highest absorption wavelength values compared to DMSO.

The emission wavelength value was observed in three OPs, DCP, GV, and VG (Table 5). The compounds showed four O atoms bonded with the P atom, and this feature was favorable for charge delocalization and promotion of the fluorescence emission. The other OP compounds dissipated the energy without fluorescence emission. The wavelength in the tested solvents was constant showing less variance. DCP emitted fluorescence at 734 nm, which corresponds to red color emission; GV emitted fluorescence at 420 nm, which corresponds to violet color emission; and VG emitted fluorescence at 570 nm, which corresponds to green color emission.

The main transitions in the absorption process in apolar solvents (hexane, toluene, and chloroform) were from HOMO − 1 to LUMO, HOMO − 1 to LUMO + 1, and HOMO to LUMO. On the other hand, in polar solvents (water and DMSO), only the transition HOMO-LUMO was observed for the absorption process. This effect was observed for all OPs, and the HOMO orbitals involved in the excitation were localized just in the 1 structure, while the LUMO orbitals are delocalized in 1 as well as in other organophosphorus compounds. The angles observed on the response surface can help in the delocalized orbitals between the organophosphorus and 1 as observed in Figure 6. This feature demonstrates that the conformational structure generated on the response surface was important for orbital delocalizations as well as for the emission process.

4. Conclusion

The DFT and TD-DFT were satisfactory for this study, and the hybrid functional family showed the lowest error when compared with other families. The B3LYP functional exhibited the error inside its family (7.1), being the functional selected for the optimization step as well as absorption and emission property calculations. The surface response analysis showed the energy minimum when angle α1 was 25° and α2 was 0°. The solvent analysis showed less variation in the absorption values. The dipolar moment and dielectric constant showed influence on the absorption properties. The reaction between 1 and the OP compounds showed to be kinetically favorable for all compounds, Sarin and cyclosarin are more reactive. The OPs with a sulfur leaving group demonstrated to be more thermodynamically favorable (VX, VG, and VM). The VX reaction showed –73 kcal/mol for ΔG and –98 kcal/mol for ΔH. The TDA was significant because this methodology describes the solvent effects better [2830]. Only three compounds showed emission values (DCP, VG, and GV). The emission was influenced by HOMO – 1-LUMO, HOMO-LUMO, and HOMO-LUMO + 1 transitions. The conformational structure generated on the response surface was important for orbital delocalization between oxime and organophosphates. The understanding of the nature of the emission and the solvent influence was important for this study because it allows proposing new ways for compound structural modifications in order to build more efficient fluorescent probes.

Data Availability

Values for angles and energies of ground states (GS) and excited states (ES) in gas as well as in water solution are included within the supporting information. The other data used to support the finding of this study are included in the article, and if further data related to analysis are required, the corresponding author can provide upon request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors thank the Brazilian agencies FAPEMIG, CAPES, and CNPq for the financial support of this research and UFLA for infrastructure and encouragement in this work. This work was also supported by Excellence project FIM and by the University of Hradec Kralove (Faculty of Science, VT2019-2021).

Supplementary Materials

S1. Surface response. Table S1: values for angles, energies of ground state (GS) and excited state (ES), and wavelength calculated for combination of angles employed for building surface response. Table S2: values for angles, energies of ground state (GS) and excited state (ES), and wavelength calculated for combination of angles employed for building surface response in water solvent. Table S3: values for geometric properties from organophosphorus compounds and the complex with oxime (distance). (Supplementary Materials)