Influence of hydration, hydroxylation, and doping of TiO₂(110) on the adsorption and decomposition of sarin: a density functional theory investigation

Abstract

Metal oxides are promising candidates for the adsorption and decomposition of chemical warfare agents (CWA) and can be the foundations of novel CWA destruction technologies. In this work, we use density functional theory (DFT) to explore how dry, wet and doped states of the metal oxide TiO₂(110), influence the adsorption and chemical dissociation of the nerve agent, sarin. The DFT calculations show that the dissociative adsorption of sarin is more energetically favored than the molecular adsorption for all dry, wet and doped states of TiO₂(110). The calculated energy barrier for the adsorption of sarin on dry TiO₂(110) showed that sarin is initially adsorbed with a molecular configuration, followed by the dissociative process. For the adsorption of sarin on wet TiO₂, under both hydrated and hydroxylated states, the 0.5 H₂O monolayer (ML) showed the lowest adsorption energy compared to the dry and other wetness levels explored. Finally, the adsorption of sarin on TiO₂(110) doped with Hf exhibited lower adsorption energy and higher charge transfer compared with TiO₂(110) doped with Zr and Ge, as well as the undoped systems. These results demonstrate how hydration, hydroxylation, and doping of TiO₂(110) significantly influence the adsorption and decomposition of sarin on this metal oxide.

---

1. Introduction

Nerve agents are highly toxic chemicals that continue to pose a threat to military and civilian personnel.¹ The nerve agents are organophosphate (OP) compounds, and their lethality comes from their ability to bind to the acetylcholinesterase (AChE) enzyme, preventing the enzymatic breakdown and regulation of the neurotransmitter acetylcholine (ACh). The resulting accumulation of acetylcholine causes continuous nerve impulses and muscle contractions, and this neurotoxicity can be acute and fatal.^1,2 Sarin (isopropyl methyl phosphono fluoridate, commonly designated as agent GB), is one example of a nerve agent which has been used in terrorist attacks and armed conflicts including the Syrian war, resulting in many fatalities and acute health effects for vast populations. Apart from such weaponization, the residual, undeclared, stockpiles of chemical weapons held by nations, that have not been destroyed despite it being 28 years since the Chemical Weapons Convention,³ pose an increasing environmental threat to land, water and civilian populations. Therefore, to protect soldier and civilian populations and to decontaminate equipment and infrastructure in the event of a chemical attack and to ensure remediation of environments affected by residual stockpiles of chemical agents, there is an imperative need to develop novel materials that can adsorb and dissociate nerve agents such as sarin into harmless chemicals.

Among the materials considered for the adsorption and dissociation of nerve agents are various metal oxides, which exhibit high stability, large numbers of reactive sites and good selectivity for applications in sensing and decontamination of nerve agents.^4–8 Titanium dioxide has proven to be the leading material for photocatalyst applications.^9–13 Furthermore, metal doping of TiO₂ has been shown to be an effective way to tune electronic and optical properties including decreasing the band gap and extending the absorption into the visible region, which is expected to improve the photocatalytic activity for degradation of chemical warfare agents.^14–17 Cation doping of TiO₂ is also considered a strategy for delaying the deactivation of TiO₂.

Experimental investigations of adsorption and dissociation of nerve agents on metal oxides are necessarily limited due to their high toxicity. To circumvent handling these toxic chemicals, one approach is to conduct experiments using simulants, which are molecules with structures and molecular features similar to the nerve agents but with lower toxicity.¹⁸ Many experimental studies have investigated the adsorption and dissociation of simulants for nerve agents, dimethyl methyl phosphonate (DMMP) being one of the most frequently selected simulants.^19–25 The other approach is to conduct theoretical studies on nerve agents, since this obviously does not involve any toxicity risk and allows one to explore the fundamentals of adsorption and decomposition of nerve agents and other toxic chemicals on different decontaminating material surfaces.

Among theoretical methods, density functional theory (DFT) is probably the most important and widely used tool that allows the exploration of how agent molecules interact with different types of potential decontaminating surfaces. DFT-based first principles methods are essential for exploring novel materials, enabling the theoretical calculation of materials properties and the simulation of interaction mechanisms. DFT computations provide fundamental insights complementing experimental findings by offering physical understanding and guiding experimental design, exploration and screening efforts. Some recent examples of diverse applications for DFT computations include calculations of electronic properties of perylene diimide derivatives to determine the response of electromechanical sensors,²⁶ the binding energy of amino acids on transition metal dichalcogenides for chemical sensor applications,²⁷ or the NO adsorption energy on metal chelates for controlling environmental NOx emissions.²⁸ Using DFT, one can determine the structure, stability and electronic properties of nerve agents^29–38 and simulants^39–51 on metal oxide surfaces, at atomic and molecular scales.

While the experimental and theoretical studies in the literature cited above have provided a wealth of information, there are still important unanswered questions as to how surface impurities such as dopants and factors such as the environmental conditions impact the adsorption and dissociation process of nerve agents on metal oxide surfaces. We address this knowledge gap in this work, selecting the metal oxide TiO₂(110) rutile, which is the most thermodynamically stable phase of TiO₂.^52,53 Using DFT methods, we can provide important insights into the ability of TiO₂ to adsorb and decompose the chemical warfare agent sarin. More importantly we can study the local structure modification through doping titania with different elements and the influence of the environmental conditions including H₂O adsorption on the surface and determine their potential impact on the catalytic activity of TiO₂ for the adsorption and decomposition of sarin.

Dopants can influence the surface properties of TiO₂ because the number of their valency electrons may be different from that of Ti⁴⁺, the steric effect related to the different size of the dopant compared to Ti⁴⁺ or the orbital effect due to the energy levels associated to the dopant not being present on the pristine TiO₂ surface. We selected isovalent dopants (eliminating influence of different numbers of valency electrons) that are of a size similar to Ti, so that the geometry perturbations on titania will be minimal when substitutional doping of a Ti atom with the dopant atom is done. The isovalent dopant elements Hf, Zr and Ge can adopt the same oxidation state to the one of Ti on the TiO₂ structure (4⁺) and possess a similar atomic radius⁵⁴ (155 pm for Hf and Zr and 125 pm for Ge) compared to Ti (140 pm). They have also been used in experimental studies.^16,17

We investigated the molecular and dissociative adsorption of sarin on dry, wet and doped TiO₂(110), its stability and structure. The molecular and dissociative adsorption energy of sarin on TiO₂ was calculated using density functional theory, while the nudge elastic band method was used to calculate the energy barrier for the dissociation of sarin on dry TiO₂. The charge transfer was calculated using the Bader charge analysis. We determined the effect of water presence and concentration on the TiO₂ surface for the adsorption and dissociation of sarin. By doping TiO₂ with different elements, we were able to determine the variation in the adsorption energy and charge transfer due to the presence of these dopants on the TiO₂ surface for the adsorption and dissociation of sarin. We found that the hydration, hydroxylation and doping of the TiO₂(110) surface cause significant variations in the adsorption and decomposition behavior of sarin on this metal oxide.

2. Computational methods

Density functional theory calculations were developed using the Perdew–Burke–Ernzerhof generalized gradient approximation (PBE-GGA)⁵⁵ and the projected-augmented wave (PAW)⁵⁶ approach as implemented in the Quantum Espresso software.⁵⁷ van der Waals interactions were also considered through the DFT-D2 method of Grimme.⁵⁸ The cutoff kinetic energy for the charge density and wave functions were 350 and 30 Ry, respectively (1 Ry = 1312.75 kJ mol⁻¹), as suggested in the pseudopotentials for the different atom types used in this research.

The charge transfer for the sarin doped TiO₂ system was calculated using the Bader charge analysis.⁵⁹ In Bader's approach, space is divided into regions by surfaces that run through minima in the charge density. Specifically, at a point on a dividing surface the gradient of the electron density has no component normal to the surface. The regions bounded by such dividing surfaces are referred to as Bader regions. Each Bader region often contains one nucleus, but this is not necessarily so, sometimes no nucleus is found within a Bader region. By integrating the electronic density within the Bader region where an atom's nucleus is located and possibly adding the electronic charge in nearby regions that do not include a nucleus, the total charge on an atom can be estimated. The computations are carried out using the algorithm developed by Henkelman.^60,61 Using this analysis, the charge for the Sarin molecule, the doped TiO₂ surface, and the doped interface were calculated, with the subsequent charge transfer determined.

2.1. Sarin isolated molecule and TiO₂ slab

To determine the most stable structure of the isolated sarin (C₄H₁₀FO₂P) molecule, three initial conformations were selected based on the geometries reported in previous investigations as the most stable structures of the molecule^62–66 and verifying the geometric parameters for the different conformations. We used a 20 × 20 × 20 ų supercell, relaxing the structures until the parameters reached the convergence criteria to obtain the most stable configuration, while the energy was calculated simultaneously.

The TiO₂ rutile bulk system was modeled as a tetragonal crystal lattice with a mesh of 8 × 8 × 10 k-points. The final a and c lattice parameters for the bulk rutile were obtained as 4.6408 Å and 2.9683 Å, respectively, in agreement with previous theoretical and experimental studies.^39,64 The TiO₂(110) rutile slab structure was built with a surface area of 2 × 4 (13.126 Å × 11.873 Å) with 5 TiO₂ perpendicular layers (corresponding to 80 TiO₂ groups or 240 total atoms) and a 15 Å vacuum region and is hereafter referred to as a 2 × 4 × 5 supercell. The TiO₂(110) rutile structure has been identified as the most thermodynamically stable among the TiO₂ phases with a large number of reactive sites.⁶⁷ The dimensions of the 2 × 4 × 5 supercell structure of TiO₂ selected in this study allow us to model the adsorption of the sarin molecule. Such large slab systems prevent modification of the molecular geometry associated with large intermolecular interaction effects and provide better accuracy compared to cluster systems.

2.2. Sarin adsorbed on dry TiO₂(110)

The adsorption of sarin on the TiO₂(110) surface was modeled with a 0.125 monolayer (ML) coverage or 1 molecule of sarin per 8 available Ti atoms. The sarin molecule was placed on top of the TiO₂(110) surface forming an asymmetrical configuration for which a dipole moment correction was required to account for the artificial field in the system. A vacuum region of 15 Å was selected after testing different vacuum lengths and corroborating the stability in the geometry and energy of the sarin–TiO₂ system. The initial adsorption geometries of the sarin molecule on the TiO₂(110) surface were built based on information from previous theoretical and experimental studies that reported the possible structure and geometry of the sarin–TiO₂ system.^{13,29,62,64,68} For dissociative adsorption of sarin, molecular groups generated from P–F and P–O bond breakages were considered. In addition, parameters such as bond angles, bond lengths, distance of the molecule to the surface, and the possible formation of monodentate configurations, bidentate configurations or secondary bonds were taken into consideration to model the initial structure (see the ESI† Section A, Fig. S1–S3 for examples of the initial sarin conformations considered).

2.3. Sarin adsorbed on wet TiO₂(110)

The adsorption of sarin on wet TiO₂(110) was explored for hydrated and hydroxylated systems with different water coverages. In the hydrated system, the O of the H₂O molecule bonds with a Ti atom on the surface, while in the hydroxylated system the oxygen atom of an OH⁻ group bonds with a Ti atom and the remaining H⁺ interacts with an O on the TiO₂ surface. The geometries of the TiO₂ hydrated and hydroxylated surfaces were selected to have three different water coverages 0.25, 0.5 and 0.75 H₂O ML, corresponding to 2, 4 and 6 water molecules per 8 available Ti atoms on the top TiO₂ layer. The adsorption of the sarin molecule on the hydrated and hydroxylated surfaces with different water coverage was modeled and then relaxed to obtain the most stable conformation. The initial geometries for the adsorption of sarin on the hydrated and hydroxylated TiO₂ were considered based on the results reported elsewhere⁶² for the adsorption of sarin on 0.5 H₂O ML coverage for the hydrated and hydroxylated TiO₂ (see the ESI† Section A, Fig. S4 for examples of the initial sarin conformations considered).

2.4. Sarin adsorbed on doped TiO₂(110)

The adsorption of sarin on TiO₂ doped with Hf, Zr and Ge as a single point dopant was evaluated. One Ti atom from TiO₂ was substituted by a single dopant atom of the elements mentioned. In the atomic position substituted in the first layer, the Ti atom is bonding with the sarin molecule. The single dopant atom was substituted in each one of the five TiO₂ layers at the Ti atom with the same or closer to the xy position to the Ti atom of the first layer. This procedure was repeated for each one of the three dopant elements, Hf, Zr and Ge. Additional individual dopant positions in the first TiO₂ layer were also considered with the Hf element substituting the Ti atoms closer to the sarin molecule. The initial conformations for molecular and dissociated sarin were selected as those for the dopant-free system and then subjected to optimization.

3. Results and discussion

3.1. Isolated sarin molecule and TiO₂(110) slab

Three different geometries of the isolated sarin molecule were studied in this work as initial conformations. These geometries were selected based on previous studies which described them as the stable configuration of sarin,^62–66 with some discrepancies about bond angles, bond lengths and in general atomic arrangement of the most stable configuration in the molecule. For the geometries of sarin molecule with conformation 1,^62–64 conformation 2,⁶⁵ and conformation 3,⁶⁶ the final energies were calculated. The final structures and geometrical parameters for the different conformations of sarin are shown in Fig. 1. The lowest total energy for the sarin molecule was obtained for the conformation 1. Conformation 2 and conformation 3 had a higher total energy with respect to conformation 1, of 0.0709 and 0.0388 eV, respectively. Based on the analysis of the total energy, conformation 1 of the sarin molecule has the highest stability.

Fig. 1 Final geometry and structures of isolated sarin molecules for three starting conformations. Conformation 1 is determined to be the most stable structure.

In the most preferable conformation of sarin, the P atom is a chiral center, while the F atom and the C_CH₃ group are close to a perpendicular direction with respect to the P–O_C–C_B plane. The F–P–O_C and C_C–C_B–O_C bond angles were obtained as 103° and 109°, respectively, and the P–F, P–O_C and P–O_M bond lengths were 1.59, 1.47 and 1.59 Å, respectively. The final geometric parameters of the most stable conformation 1 of isolated sarin molecules, including some bond lengths and bond angles are reported in Table 1.

Table 1 Geometric parameters for the most stable isolated sarin molecule and the TiO₂(110) 2 × 4 × 5 supercell (labels are depicted in Fig. 1 and Fig. 2, respectively)

Isolated sarin molecule		TiO2(110) 2 × 4 × 5 supercell
Bond lengths (Å)		Bond lengths (Å)
P–F	1.59	Ti5fc–O3fc	1.96
P–OM	1.47	Ti6fc–O2fc	1.83
P–CA	1.80	Ti6fc–O3fc-1	2.11
P–OC	1.59
OC–CB	1.48
CB–CC, CB–CD	1.52

---

Bond angles (°)		Bond angles (°)
F–P–OM	112	Ti6fc–O2fc–Ti6fc	108
F–P–OC	103	Ti6fc–O3fc-1–Ti6fc	89
OM–P–CA	118	Ti5fc–O3fc–Ti5fc	98
CA–P–OC	103
P–OC–CB	121
CC–CB–OC	109

---

For the TiO₂(110) 2 × 4 × 5 supercell, the two bottom layers of the metal oxide supercell were fixed at the TiO₂ bulk position while the three top layers were allowed to relax. The geometrical parameters of the TiO₂(110) 2 × 4 × 5 supercell are outlined in Table 1 and the corresponding labels are shown in Fig. 2.

Fig. 2 Final structure of the TiO₂(110) rutile 2 × 4 × 5 supercell. The lattice parameters of the slab are a_slab = 6.563 Å and b_slab = 2.968 Å. The top and bottom layers have two Ti atom types, five-fold coordinated (Ti_5fc) and six-fold coordinated (Ti_6fc) and two O atom types, the two-fold coordinated (O_2fc) and the three-fold coordinated (O_3fc).

In the TiO₂(110) surface, the top and bottom layers have two Ti atom types, the Ti five-fold coordinated (Ti_5fc) and the Ti six-fold coordinated (Ti_6fc), as well as two O atom types, the O two-fold coordinated (O_2fc) and the O three-fold coordinated (O_3fc). In Ti_5fc and Ti_6fc the Ti atom bonds with 5 and 6 O atoms, respectively, while in O_2fc and O_3fc, the O atom bonds with 2 and 3 Ti atoms, respectively.

3.2. Adsorption and dissociation of sarin on dry TiO₂

The molecular and dissociative adsorption of sarin on the TiO₂(110) 2 × 4 × 5 supercell structure was initially investigated for the clean surface (without impurities or water adsorption). The initial structures were selected based on the information reported in a similar work⁶² for the adsorption of sarin on TiO₂. The molecular and dissociative adsorptions of sarin on TiO₂(110) were studied with different initial possible configurations of sarin on rutile TiO₂(110) in which the sarin molecule interacts with the TiO₂ surface by the formation of a monodentate configuration with Ti–O or F–Ti bonds or a bidentate configuration with the Ti–O and F–Ti bonds. We determined the preferable configuration of the sarin-TiO₂ system through calculation of the adsorption energy for the different conformations explored. The adsorption energy (E_ads) for the sarin molecule adsorbed on TiO₂(110) is defined as:

E_ads = E_{Sarin-TiO2(110)} − E_TiO2(110) − E_Sarin (1)

where E_{Sarin-TiO2(110)} is the energy of the sarin–TiO₂(110) system after adsorption, E_TiO2(110) is the energy of the clean TiO₂(110) surface and E_Sarin is the energy of the gas phase isolated sarin molecule. The E_ads for the most stable adsorbed geometry of sarin on TiO₂(110) was obtained as −1.84 eV and is characterized by the formation of a monodenate covalent bond between the P=O group from sarin and the Ti_5fc atom from the TiO₂(110) surface with the formation of several secondary bonds between sarin and the O_2fc surface atom of TiO₂(110). The E_ads and structure of the most stable configuration for the molecular adsorption of sarin on TiO₂(110) is shown on Fig. 3.

Fig. 3 Calculated adsorption energy and activation energy for the dissociation of sarin on dry TiO₂(110). Inserted in the figure are sarin conformations in the initial state (IS) of molecular adsorption and the final state (FS) of dissociative adsorption. The conformation in the transition state (TS) preceding dissociation is also shown.

We also considered the dissociative adsorption of sarin on TiO₂(110) with different dissociative products, which was also described in previous reported work.⁶² The E_ads for the most stable dissociative adsorption of sarin on TiO₂(110) was obtained as −2.70 eV. The final geometry for the dissociative adsorption of the sarin–TiO₂ system was obtained as the dissociation of the F atom from the surface and its adsorption on the TiO₂ surface by the formation of a F–Ti bond. The remaining C₄H₁₀O₂P⁺ molecule interacts with an O_2fc atom on the surface forming a bidentate configuration. This final geometry, as well as the E_ads for the dissociative adsorption of sarin on TiO₂ are also shown on Fig. 3. The comparison between the E_ads values obtained for the most stable configurations of the molecular and dissociative adsorption of sarin on TiO₂ indicates that the dissociative adsorption of sarin is preferred over molecular adsorption. These results, along with the final structures obtained, could indicate the possible mechanism being an initial molecular adsorption of sarin on TiO₂ followed by the dissociation of the molecule, as also suggested by previous experimental studies.^12,13

To clearly establish the adsorption and dissociative process of sarin on TiO₂(110), the energy barrier between the molecular and dissociative adsorption of the molecule was calculated using the nudge elastic band method. The final geometry for the molecular adsorption of sarin on TiO₂ is considered as the initial state (IS) while the dissociative adsorption configuration of sarin on TiO₂ was identified as the final state (FS). We found that the dissociation of sarin on TiO₂(110) is exothermic with an energy barrier of 0.35 eV at the transition state (TS). The energy barrier and the structure of IS, TS and FS obtained for the adsorption process of sarin on TiO₂ are shown in Fig. 3. The rotation and orientation of the isopropyl group and the F atom in the sarin molecule from the initial state to the transition state, before the dissociation of F, are the main contributions to the energy barrier for sarin dissociation on TiO₂(110). Given the small energy barrier obtained for the dissociation of sarin on TiO₂ and its exothermic nature, we suggest the adsorption mechanism of sarin on TiO₂ as an initial molecular adsorption (observed in IS) with a subsequent dissociation of the F atom and the formation of a bidentate configuration by the remaining molecule as observed in FS in Fig. 3. The adsorption followed by the dissociation of sarin on TiO₂(110) is shown in Fig. 4.

Fig. 4 Molecular adsorption followed by the dissociation of sarin at the P–F bond on dry TiO₂.

Our computational results agree with previous theoretical and experimental studies on the adsorption of sarin on different surfaces. The theoretical investigations on the adsorption of sarin on the anatase TiO₂(001) and Al₂O₃ clusters determined the E_ads values to be −2.00 eV,⁶⁶ and −2.23 eV,²⁹ respectively. A combined theoretical and experimental work on the adsorption of sarin on SiO₂ also reported values for E_ads of −0.82 and −0.95 eV from the experiments and calculations, respectively.⁶⁹ A previous experimental work⁶⁸ investigating the adsorption of sarin on hydroxylated TiO₂ found that dissociative adsorption of the molecule is preferable with the F atom from sarin interacting with the hydroxyl groups from the surface, facilitating the final dissociation of the fluorine atom and the subsequent formation of a bidentate configuration from the remaining sarin molecule.

3.3. Adsorption and dissociation of sarin on wet TiO₂ with different H₂O ML coverage

The results obtained for the adsorption of sarin on dry TiO₂ indicate that theoretically, the dissociation of the molecule on a dry surface is possible. However, a clean and dry TiO₂ surface is difficult to obtain in practice, under common experimental conditions. Previous experimental study⁶⁸ reported the adsorption of sarin on a hydroxylated TiO₂(110) as an initial molecular adsorption process followed by the dissociation of the sarin molecule on the hydroxylated TiO₂(110) surface. Based on a previous investigation⁶² on the hydration and hydroxylation of sarin on TiO₂(110) with 0.5 H₂O ML coverage, we selected the most stable molecular and dissociative configurations of sarin on a TiO₂ surface and considered additional levels of H₂O coverage.

To determine the effect of water content on the TiO₂ surface for the adsorption and dissociation of sarin, we compared the E_ads for the molecular and dissociative adsorption of sarin on a hydrated and hydroxylated TiO₂(110) surface for three H₂O coverages: 0.25 ML, 0.5ML and 0.75 ML (corresponding to 2, 4 and 6 H₂O molecules of water, respectively, per 8 available Ti_5fc atoms on top of the TiO₂ slab). The molecular and dissociative adsorption of sarin on the 0.5 H₂O ML for the hydrated and hydroxylated TiO₂ had the lowest E_ads values among the different H₂O coverages, as shown in Fig. 5. In addition, E_ads for the dissociative adsorption of sarin on both the hydrated and hydroxylated TiO₂ surfaces were lower compared to E_ads for the molecular adsorption, showing a similar tendency to that observed for the adsorption of sarin on dry TiO₂. The comparison of E_ads on dry and wet sarin–TiO₂(110) systems showed that the dissociative adsorption of sarin on 0.5 H₂O ML TiO₂, specifically a hydrated surface is the preferable configuration, is in general agreement with the suggestion proposed in a previous experimental study.⁶⁸

Fig. 5 Calculated adsorption energies for molecular and dissociative adsorption of sarin on dry, hydrated (H₂O) and hydroxylated (OH⁻ + H⁺) TiO₂(110). The open symbols correspond to molecular adsorption while the filled symbols correspond to dissociative adsorption.

The final geometries for the most stable molecular and dissociative adsorption of sarin on wet (hydrated and hydroxylated) TiO₂ systems are shown on Fig. 6 and these are similar to the ones observed for the dry systems. A schematic of the mechanism of adsorption of sarin on dry and wet TiO₂ for hydrated (H₂O) and hydroxylated (OH + H⁺) surfaces is provided in the ESI† Section B, Fig. S5.

Fig. 6 Structure and geometry of molecularly and dissociatively adsorbed sarin on dry, hydrated (H₂O) and hydroxylated (OH⁻ + H⁺) TiO₂(110).

For the molecular adsorbed structure, the formation of a P=O bond with the TiO₂ surface was more stable, while in the dissociatively adsorbed structure, the F atom dissociates from sarin and is adsorbed on the TiO₂ surface while the remaining molecule forms a bidentate configuration with a O_2fc from TiO₂. Additional secondary bonds formed between sarin and the water molecules were observed for the hydrated and hydroxylated systems. The formation of these secondary bonds could have a relevant impact on the stability of the molecular and dissociative adsorption of sarin on wet TiO₂.

A comparison of the number of secondary bonds formed between the different H₂O ML suggests that in the case of the lower H₂O coverage (0.25 ML) the number of secondary bonds between sarin and H₂O or OH groups decreased compared to the 0.5 H₂O ML. In contrast, in the case of higher H₂O coverage (0.75 ML) the presence of additional H₂O or OH groups saturate the Ti_5fc available positions preventing the bond formation between Ti and the sarin molecule.

At 0.25 H₂O ML, the chemisorption of sarin on TiO₂ is observed with weak interatomic and intermolecular interactions occurring between sarin and H₂O. At higher degrees of hydration, specifically 0.5 H₂O ML, sarin still exhibits chemisorption on TiO₂ with strong bonds formed between the molecule and the surface. Additionally, there are interatomic interactions occurring between H₂O and the sarin molecules, which could be mostly identified as van der Waals interactions or hydrogen bonding, resulting in lower E_ads. As the degree of hydration increases for 0.75 H₂O ML, the H₂O molecules start to compete with sarin for adsorption sites on the titania surface with sarin showing now weaker interaction and mostly physisorption on the TiO₂ surface and consequently higher E_ads.

3.4. Adsorption and dissociation of sarin on substitutionally doped TiO₂ with X (X = Hf, Zr or Ge)

Previous experimental studies^4,70–72 have suggested that doping TiO₂ with elements such as Hf, Zr and Ge can increase the degradation rate of chemical warfare agents including soman, VX and sulfur mustard. In the case of sarin, there are no available theoretical or experimental studies, to the best of our knowledge, on the adsorption of sarin on doped TiO₂. To understand the influence of doping TiO₂ on the adsorption and decomposition of sarin, we selected three dopant elements Hf, Zr and Ge to act as substitutional dopants. These elements can adopt the same oxidation state to the one of Ti on the TiO₂ structure (4+) and possess a similar atomic radius⁵² (155 pm for Hf and Zr and 125 pm for Ge) compared to Ti (140 pm).

Different orientations of sarin on the TiO₂ surface were considered to account for the possible adsorption configurations of the molecule both in the molecular and dissociative configuration (see the ESI† Section A, Fig. S1–S3 for examples of initial sarin conformations considered, also for the doped TiO₂ system). The specific configurations were selected based on different geometrical parameters for the interface including bond angle, bond length, possible secondary bonds formed, as well as distance and orientation of the molecule respect to the surface.⁶² The selection of the most stable adsorption configuration was determined based on the adsorption energy (E_ads) values calculated for each system.

3.4.1. Adsorption energy when the dopant atom is in different layers of the supercell. Initially, we doped a single position at a Ti site for each of the five TiO₂ layers when sarin is molecularly and dissociatively adsorbed on the surface. The dopant site at the first layer was selected as the position at which the X dopant (X = Hf, Zr or Ge) is at the Ti_6fc site of the TiO₂ surface forming a bond with the sarin molecule. Therefore, the dopant element X is bonding with the O_M atom (oxygen atom of sarin forming a bond with the surface) of the sarin molecule. The dopant positions for the other layers (between the 2nd and the 5th layer) were selected in the same or closer to the xy atomic position chosen for the dopant at the first layer. The configurations obtained as the most stable structures for the molecular and dissociative adsorption of sarin on clean and dry TiO₂, were considered as the initial configurations for sarin on X-doped TiO₂ systems.

The E_ads values for the molecular and dissociative adsorption of sarin on doped TiO₂ with the dopants Hf, Zr and Ge were calculated with a similar relationship to eqn (1) and are shown in Fig. 7. The final E_ads values are lower when the dopants Hf and Zr are at the first layer of the TiO₂ surface forming a bond with the sarin molecule. The E_ads values for the molecular and dissociative adsorption of sarin on Hf-doped TiO₂ were −2.40 and −3.22 eV, respectively. For the case of sarin on Zr-doped TiO₂ the molecular and dissociative adsorption had values of −2.18 and −3.03 eV, respectively, as shown in Fig. 7. When these dopants are positioned at the other TiO₂ layers, the E_ads values increase (between 0.3 and 0.5 eV larger) and in some cases are even higher than the undoped systems. In the case of the molecular and dissociative adsorption of sarin on Ge-doped TiO₂, the E_ads shows in general a constant tendency with higher values than those of sarin on Hf and Zr-doped TiO₂, as well as the undoped systems.

Fig. 7 Molecular (A) and dissociative (B) adsorption energy values for sarin on X-doped TiO₂ systems for X = Hf, Zr or Ge. A single point dopant on each of the five different layers was considered. The dopant position and the corresponding conformation of the adsorbed sarin are also shown.

In terms of the different types of adsorption, the E_ads values calculated for the dissociative adsorption of sarin on X-doped TiO₂ are lower than for the molecular adsorption, for all the cases considered. The comparison between the doped and undoped systems showed that the adsorption of sarin on Hf and Zr-doped TiO₂ where the dopant is placed at the first layer is more stable compared with the undoped systems, for the cases of both molecular and dissociative adsorption.

In addition, as the dopant position moves farther away from the surface, and from the sarin molecule, the E_ads values are closer to those of the undoped system, indicating that the effect of the dopant becomes more tenuous, as can be expected. For the case of Hf as the dopant, the asymptotic adsorption energy for both molecular and dissociative adsorption, as the layer depth increases is still a little lower than the adsorption energy for the undoped TiO₂. While dopants located deeply within the titania might have a less direct impact compared to surface dopants, they can still influence catalytic activity through electronic and strain effects that extend to the surface. The extent of this influence depends on the material properties, dopant characteristics, and the depth of the dopant within the titania structure. Considering the ionic radius, lattice strain and electronegativity of the different dopant elements Hf, Zr and Ge, compared with Ti in the rutile structure, the combined properties of Hf (small radius and electronegativity) can lead to a more stable structure when doping the rutile TiO₂ material and therefore be a better candidate for the adsorption of sarin compared to the other doped materials. The comparison in the direction of the E_ads observed for the three dopant elements indicate that in the case of Hf a surface reconstruction or strain can be present which could lead to variation in the properties compared to the other dopants.

3.4.2. Charge transfer when the dopant atom is in different layers of the supercell. To determine the relationship between the electronic properties and the stability of the sarin on doped TiO₂ systems, we calculated the charge transfer at the sarin molecule and the dopant atoms when they are at the first layer forming a bond.

The charge transfer (ΔQ) was calculated for each atom of the sarin X-doped TiO₂ systems (X = Hf, Zr or Ge), using the following expression:

ΔQ = Q_{Sarin-X-doped TiO2} − Q_{X-doped TiO2} − Q_Sarin (2)

where Q_{Sarin-X-doped TiO2} is the charge of the sarin X-doped TiO₂ system, Q_{X-doped TiO2} is the charge of the X-doped TiO₂ surface with X = Hf, Zr or Ge being at the same position as for the sarin X-doped TiO₂ system and Q_Sarin is the charge of the sarin molecule. All these quantities are evaluated at each atom of the specific system. Fig. 8 shows ΔQ at some relevant atoms for the molecularly and dissociatively adsorbed sarin on doped TiO₂ systems for the Hf, Zr and Ge dopants. The atomic positions and labels are depicted at the sarin on the doped TiO₂ structure.

Fig. 8 Charge transfer calculated at different atoms of the molecularly and dissociatively adsorbed sarin on X-doped TiO₂ systems for X = Hf, Zr or Ge. The dopant is on the surface layer.

Among the ΔQ values obtained for the different atoms, the results for the dopants (X) and OM draw special attention. The ΔQ values calculated at the dopant atom showed a higher value for sarin on Hf-doped TiO₂ compared with the systems with Zr or Ge. These results apply to both the molecular as well as the dissociative adsorption. In addition, there is a clear difference in the charge transfer direction for the sarin on X-doped TiO₂, when the system is doped with Ge compared to Hf or Zr. In the case of sarin on Hf and Zr-doped TiO₂ the charge transfer is positive, while the same property displays a negative value for sarin on Ge-doped TiO₂.

These results can be interpreted as due to the difference in electronegativities between the different dopants. The electronegativities of Zr and Hf (1.33 and 1.3, respectively) are lower than the electronegativity of Ti (1.54), while the opposite is true for Ge (2.01). The Ge dopant atom has the tendency to attract more electrons when it substitutes the Ti atom of the TiO₂ surface, while Hf and Zr show the opposite effect. A similar comparison for the ΔQ values at the O_M atom of sarin on TiO₂ doped with the different elements show that ΔQ is more negative for the system doped with Hf compared to Zr and Ge. In the X–O_M bond, the charge has a preferred direction toward O due to its higher electronegativity^3,50 and becomes even more negative when the TiO₂ surface is doped with Hf compared to Zr and Ge. Once more, the difference in electronegativity between Ti and the dopant atoms play a role in the charge transfer direction. In addition, the analysis of ΔQ for the P atom shows a different charge transfer direction for the molecular and dissociative configurations with positive and negative values, respectively.

In the molecularly adsorbed configuration, the F atom forms a P–F bond in the sarin molecule, while in the dissociatively adsorbed configuration the F atom dissociates and bonds with a Ti atom at the surface. The breakage of the P–F bond causes a variation not only in the ΔQ value but also in the charge transfer direction at the P atom.

The comparison for sarin on X-doped TiO₂ between the different dopants indicates that the trend observed for the adsorption energy in terms of increasing stability is also observed qualitatively for the charge difference. Sarin on the Hf-doped TiO₂ system with the Hf dopant at the first layer had the lowest E_ads (the highest stability) and showed a higher charge transfer, while the E_ads had the lowest value for sarin on the Ge-doped TiO₂ system with charge transfer having a lower value and also an opposite direction compared with the Hf-doped system, which suggest that the trend observed for the E_ads could have the same origin in the electronic properties of sarin on doped systems.

3.4.3. Adsorption energy and charge transfer when the dopant atom is in different locations on the top layer of the supercell. It was noted earlier in Fig. 7 that the lowest E_ads values were obtained for sarin on Hf-doped TiO₂ when the dopant atom is at the first layer, bonding with the adsorbed sarin molecule. Taking this into consideration, we selected the Hf dopant to further explore the adsorption of sarin on Hf-doped TiO₂ with additional dopant positions specifically at the first TiO₂ layer. Fig. 9 shows the additional dopant sites selected at the first layer for the molecular and dissociative adsorption of sarin on Hf-doped TiO₂ and the E_ads values for those specific dopant positions.

Fig. 9 Calculated adsorption energies for molecular and dissociative adsorption of sarin on Hf-doped TiO₂(110) with Hf at different positions on the first TiO₂ layer (Hf atoms in blue).

In the case of the molecular adsorption of sarin on Hf-doped TiO₂, we observe that the E_ads values for the additional dopant positions considered (1B, 1C, 1D, 1E, 1F and 1G) are between −1.90 and −2.10 eV which are similar to the E_ads of the sarin-TiO₂ undoped system (−1.85 eV) and higher than the value of −2.4 eV for the dopant position Hf-1 in which case sarin bonds with the dopant Hf. In the case of the dissociative adsorption of sarin on Hf-doped TiO₂, the E_ads values range between −3.20 and −2.80 eV with 2 systems being more stable and having the same E_ads value (−3.20 eV). These E_ads are obtained for sarin on Hf-doped TiO₂ systems when Hf is bonding with the O_M atom (Hf-1, as also described previously) and when Hf is bonding with the F atom dissociated from sarin, corresponding to 1B in Fig. 9.

We calculated the charge transfer at different atomic positions and when the Hf dopant is at the 1B position, we found that ΔQ at Hf is small (0.02 e), while ΔQ at the F atom is negative with a value close to −0.11. In the Hf–F bond, the charge could go in the direction toward the F atom which could be the reason for the small ΔQ value obtained at the Hf atom. The values for ΔQ at O_M and P are around −0.16 and 0.04 eV with the positive charge difference in P due to the dissociation of F from sarin and the charge rearrangement in the remaining molecule. The additional dopant positions also have a lower E_ads compared with the undoped system (−2.70 eV), with two positions having values close to −3.0 eV. In these two positions, the O_2fc atom bonding with the remaining sarin molecule after dissociation is also forms X–O bonds with the dopant, corresponding to positions 1C and 1D in Fig. 9. The results described here show that not only the dopant element but also the dopant position plays a decisive role in the stability and dissociation of the sarin–TiO₂ system.

4. Conclusions

The molecular and dissociative adsorption of sarin on dry, wet and doped TiO₂(110) surfaces was studied using density functional theory. We found that

(i) For the different cases considered here, the dissociative adsorption of sarin had lower adsorption energy compared against the molecular adsorption, with the F atom dissociating from the sarin molecule and forming a bond with a Ti_5fc atom from the surface.

(ii) The energy barrier for the dissociation process of sarin on dry TiO₂ was obtained as 0.35 eV in which an initial molecular adsorption was followed by a dissociative adsorption of sarin on the dry TiO₂ surface.

(iii) The calculation of the adsorption energy for sarin on wet TiO₂ surfaces showed that the dissociative adsorption of sarin on a hydrated TiO₂ with 0.5 H₂O ML coverage was the preferred configuration among the different wet and dry systems considered.

(iv) The adsorption and dissociation of sarin on doped TiO₂ with Hf, Zr and Ge as single point dopants at different layers of the TiO₂ slab was also studied and the calculations showed higher stability for the dissociative adsorption of sarin on Hf-doped TiO₂ compared to Zr and Ge doped systems. The dopant position also had an effect on the stability of sarin on doped TiO₂ systems.

(v) The calculation of the charge transfer at the dopant atom and the sarin molecule for sarin on doped TiO₂ showed the highest value for the system with the highest stability, namely, sarin on Hf-doped TiO₂.

The results obtained in this work suggest that methods such as the hydration, hydroxylation and doping of TiO₂ could represent effective ways to improve the adsorption and dissociation of sarin on this metal oxide and more importantly could facilitate the decontamination process of this highly toxic molecule.

Author contributions

Conceptualization, Y. C. Q., R. N.; methodology and investigation, Y. C. Q.; formal analysis, Y. C. Q., R. N.; writing – original draft preparation, Y. C. Q.; writing – review and editing, Y. C. Q., R. N.; project administration, R. N. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors declare no conflict of interest.

Abbreviations

Ach	Acetylcholine
AChE	Acetylcholine esterase
CWA	Chemical warfare agent
DFT	Density functional theory
DMMP	Dimethyl methyl phosphonate
GB	Sarin, Isopropyl methyl phosphono fluoridate
ML	Monolayer
OP	Organophosphate
PAW	Projected augmented wave
PBE-GGA	Perdew–Burke–Ernzerhof generalized gradient approximation

Data availability

All original data from this study are included in the article itself and further inquiries can be directed to the corresponding author. Example of an input file for the sarin–TiO₂ system with molecular adsorption is included in the ESI† Section C.

Acknowledgements

This research was conducted as part of the core research at the U.S. Army Combat Capabilities Development Command Soldier Center. The authors acknowledge computational support through the DoD High-Performance Computing Modernization Program. Yenny Cardona Quintero was a National Research Council Postdoctoral Fellow during this project.

References

1. Y. J. Jang, K. Kim, O. G. Tsay, D. A. Atwood and D. G. Churchill, Update 1 of: Destruction and detection of chemical warfare agents, Chem. Rev., 2015, 115, PR1–PR76.
2. R. Stone, How to defeat a nerve agent, Science, 2018, 359, 23.
3. Chemical Weapons Convention, Organization for the Prohibition of Chemical Weapons, 2025, https://www.opcw.org/chemical-weapons-convention.
4. V. Štengl, J. Henych, P. Janoš and M. Skoumal, Nanostructured metal oxides for stoichiometric degradation of chemical warfare agents, Rev. Environ. Contam. Toxicol., 2016, 236, 239–258.
5. N. Sharma and R. Kakkar, Recent Advancements on warfare agents/metal oxides surface chemistry and their simulation study, Adv. Mater. Lett., 2013, 4, 508–521.
6. E. D. Davis, W. O. Gordon, A. R. Wilmsmeyer, D. Troya and J. R. Morris, Chemical warfare agent surface adsorption: Hydrogen bonding of sarin and soman to amorphous silica, J. Phys. Chem. Lett., 2014, 5, 1393–1399.
7. G. W. Wagner, O. B. Koper, E. Lucas, S. Decker and K. J. Klabunde, Reactions of VX, GD, and HD with nanosize CaO: Autocatalytic dehydrohalogenation of HD, J. Phys. Chem. C, 2000, 104, 5118–5123.
8. W. Zhang, A. M. Asiri, D. Liu, D. Du and Y. Lin, Nanomaterial-based biosensors for environmental and biological monitoring of organophosphorus pesticides, Trends Anal. Chem., 2014, 54, 1–10.
9. T. Li, R. Tsyshevsky, M. McEntee, E. M. Durke, C. Karwacki, E. E. Rodriguez and M. M. Kuklja, Titania nanomaterials for sarin decomposition: Understanding fundamentals, ACS Appl. Nano Mater., 2022, 5, 6659–6670.
10. C. Tesvara, C. J. Karwacki and P. Sautet, Decomposition of the toxic nerve agent sarin on oxygen vacancy sites of rutile TiO₂(110), J. Phys. Chem. C, 2023, 127, 8006–8015.
11. G. Bazargan, I. V. Schweigert and D. Gunlycke, Adsorption of organophosphate nerve agent VX on the (101) surface of anatase titanium dioxide, Surf. Sci., 2022, 716, 121957.
12. T. Hirakawa, K. Sato, A. Komano, S. Kishi, C. K. Nishimoto, N. Mera, M. Kugishima, T. Sano, H. Ichinose, N. Negishi, Y. Seto and K. Takeuchi, Experimental study on adsorption and photocatalytic decomposition of isopropyl methylphosphonofluoridate at surface of TiO₂ photocatalyst, J. Phys. Chem., 2010, 114, 2305–2314.
13. K. Sato, T. Hirakawa, A. Komano, S. Kishi, C. K. Nishimoto, N. Mera, M. Kugishima, T. Sano, H. Ichinose, N. Negishi, Y. Seto and K. Takeuchi, Titanium dioxide photocatalysis to decompose isopropyl methylphosphonofluoridate (GB) in gas phase, Appl. Catal., B, 2011, 106, 316–322.
14. D. R. Sengele, N. Keller, V. Keller, A. Herissan and C. Colbeau-Justin, Ta-doped TiO₂ as photocatalyst for UV-A activated elimination of chemical warfare agent simulant, J. Catal., 2016, 334, 129–141.
15. C. S. Cao, W. Liu, A. Ma, X. Jiao, Y. Yang, J. Li and F. Fu, Single-atom Pt-decorated TiO₂ nanotubes for boosted photocatalytic degradation of chemical warfare agents, ACS Sustainable Chem. Eng., 2025, 13, 6379–6387.
16. V. Štengl, T. M. Grygar, F. Opluštil and T. Němec, Ge⁴⁺ doped TiO₂ for stoichiometric degradation of warfare agents, J. Hazard. Mater., 2012, 227, 62–67.
17. Z. Shen, J. Y. Zhong, N. N. Chai, X. He, J. Z. Zang, H. Xu, X. Y. Han and P. Zhang, Evolution of microstructural defects of TiO2 nanocrystals by Zr⁴⁺ or/and Ge⁴⁺ doping lead to high disinfection efficiency for CWAs, Chem. Phys. Lett., 2017, 678, 146–152.
18. J. Lavoie, R. Srinivasan and R. Nagarajan, Using cheminformatics to find simulants for chemical warfare agents, J. Hazard. Mater., 2011, 194, 85–91.
19. C. N. Rusu and J. T. Yates, Adsorption and decomposition of dimethyl methylphosphonate on TiO₂, J. Phys. Chem. B, 2000, 104, 12292–12298.
20. D. A. Panayotov and J. R. Morris, Uptake of a chemical warfare agent simulant (DMMP) on TiO₂: Reactive adsorption and active site poisoning, Langmuir, 2009, 25, 3652–3658.
21. D. A. Panayotov and J. R. Morris, Thermal decomposition of a chemical warfare agent simulant (DMMP) on TiO₂: Adsorbate reactions with lattice oxygen as studied by infrared spectroscopy, J. Phys. Chem. C, 2009, 113, 15684–15691.
22. J. Zhou, S. Ma, Y. C. Kang and D. A. Chen, Dimethyl methylphosphonate decomposition on titania-supported Ni clusters and films: A comparison of chemical activity on different Ni surfaces, J. Phys. Chem. B, 2004, 108, 11633–11644.
23. J. Zhou, K. Varazo, J. E. Reddic, M. L. Myrick and D. A. Chen, Decomposition of dimethyl methylphosphonate on TiO₂ (110): Principal component analysis applied to X-ray photoelectron spectroscopy, Anal. Chim. Acta, 2003, 496, 289–300.
24. A. R. Wilmsmeyer, W. O. Gordon, E. D. Davis, D. Troya, B. A. Mantooth, T. A. Lalain and J. R. Morris, Infrared spectra and binding energies of chemical warfare nerve agent simulants on the surface of amorphous silica, J. Phys. Chem. C, 2013, 117, 15685–15697.
25. P. Janos, P. Kuran, M. Kormunda, V. Stengl, T. M. Grygar, M. Dosek, M. Stastny, J. Ederer, V. Pilarova and L. Vrtoch, Cerium dioxide as a new reactive sorbent for fast degradation of parathion methyl and some other organophosphates, J. Rare Earths, 2014, 32, 360–370.
26. A. Tiwari, V. Adepu, R. S. Fernandes, N. Dey, P. Sahatiya and S. Kanungo, Perylene diimide architecture-based electromechanical sensors: a systematic experimental and theoretical framework for the comparative analysis and study of the transduction mechanism, Mater. Adv., 2024, 5, 9243–9258.
27. A. Tiwari, H. V. Barkale, N. Harshini, S. Chennareddy, N. Dey and S. Kanungo, Phenylalanine and tyrosine interactions with the WSe2 nanosheet: A comparative analysis based on DFT calculation and experimental confirmation, J. Phys. Chem. C, 2025, 129, 1769–1778.
28. H. J. Cho, S. H. Kim, J. Kang, K. Cho, H. Y. Lee, H. Kim, H. K. Ju, J. W. Choi, S. G. Kim, S. G. Oh, C. W. Lee and H. C. Yoon, Experimental and DFT studies on the equilibrium properties, kinetics, and mechanism of nitric oxide removal using metal-EDTA and ferrous thiochelates, Chem. Eng. J., 2022, 431, 134010.
29. V. M. Bermudez, Computational study of environmental effects in the adsorption of DMMP, Sarin, and VX on γ-Al₂O₃: Photolysis and surface hydroxylation, J. Phys. Chem. C, 2009, 113, 1917–1930.
30. A. Michalkova, M. Ilchenko, L. Gorb and J. Leszczynski, Theoretical study of the adsorption and decomposition of sarin on magnesium oxide, J. Phys. Chem. B, 2004, 108, 5294–5303.
31. A. Michalkova and J. Leszczynski, Interactions of nerve agents with model surfaces: Computational approach, J. Vac. Sci. Technol., A, 2010, 28, 1010–1017.
32. B. N. Papas, I. D. Petsalakis, G. Theodorakopoulos and J. L. Whitten, CI and DFT studies of the adsorption of the nerve agent sarin on surfaces, J. Phys. Chem. C, 2014, 118, 23042–23048.
33. R. Wang, B. Wang, A. S. Abdullahi and H. Fan, Understanding the prototype catalyst TiO₂ surface with the help of density functional theory calculation, Comput. Mol. Sci., 2024, 14, e1686.
34. P. Janos, J. Henych, O. Pelant, V. Pilarova, L. Vrtoch, M. Kormunda, K. Mazanec and V. Stengl, Cerium oxide for the destruction of chemical warfare agents: A comparison of synthetic routes, J. Hazard. Mater., 2016, 304, 259–268.
35. V. S. Vaiss, I. Borges and A. A. Leit, Sarin degradation using brucite, J. Phys. Chem. C, 2011, 115, 24937–24944.
36. A. Michalkova, J. Martinez, O. A. Zhikol, L. Gorb, O. V. Shishkin, D. Leszczynska and J. Leszczynski, Theoretical study of adsorption of sarin and soman on tetrahedral edge clay mineral fragments, J. Phys. Chem. B, 2006, 110, 21175–21183.
37. A. Michalkova, Y. Paukku, D. Majumdar and J. Leszczynski, Theoretical study of adsorption of tabun on calcium oxide clusters, Chem. Phys. Lett., 2007, 438, 72–77.
38. L. Yang, D. Tunega, L. Xu, N. Govind, R. Sun, R. Taylor, H. Lischka, W. A. Dejong and W. L. Hase, Comparison of cluster, slab, and analytic potential models for the dimethyl methylphosphonate (DMMP)/TiO₂(110) intermolecular interaction, J. Phys. Chem. C, 2013, 117, 17613–17622.
39. V. M. Bermudez, Ab initio study of the interaction of dimethyl methylphosphonate with rutile (110) and anatase (101) TiO₂ surfaces, J. Phys. Chem. C, 2010, 114, 3063–3074.
40. V. M. Bermudez, First-principles study of adsorption of dimethyl methylphosphonate on the TiO₂ anatase (001) surface: Formation of a stable titanyl (Ti = O) site, J. Phys. Chem. C, 2011, 115, 6741–6747.
41. V. M. Bermudez, Computational study of the adsorption of dimethyl methylphosphonate (DMMP) on the (010) surface of anatase TiO₂ with and without faceting, Surf. Sci., 2010, 604, 706–712.
42. Y. Paukku, A. Michalkova and J. Leszczynski, Adsorption of dimethyl methylphosphonate and trimethyl phosphate on calcium oxide: An ab initio study, Struct. Chem., 2008, 19, 307–320.
43. J. Quenneville, R. S. Taylor and A. C. T. van Duin, Reactive molecular dynamics studies of DMMP adsorption and reactivity on amorphous silica surfaces, J. Phys. Chem. C, 2010, 114, 18894–18902.
44. M. R. Housaindokht and N. Zamand, A DFT study of associative and dissociative chemical adsorption of DMMP onto SnO₂(110) surface nanocluster, Struct. Chem., 2015, 26, 87–96.
45. D. A. Trubitsyn and A. V. Vorontsov, Molecular and reactive adsorption of dimethyl methylphosphonate over (001) and (100) anatase clusters, Comput. Theor. Chem., 2013, 1020, 63–71.
46. A. Kiselev, A. Mattson, M. Andersson, A. E. C. Palmqvist and L. Osterlund, Adsorption and photocatalytic degradation of diisopropyl fluorophosphate and dimethyl methylphosphonate over dry and wet rutile TiO₂, J. Photochem. Photobiol., A, 2006, 184, 125–134.
47. M. B. Mitchell, V. N. Sheinker and E. A. Mintz, Adsorption and decomposition of dimethyl methylphosphonate on metal oxides, J. Phys. Chem. B, 1997, 101, 11192–11203.
48. A. R. Head, R. Tsyshevsky, L. Trotochaud, Y. Yu, L. Kyhl, O. Karslıoǧlu, M. M. Kuklja and H. Bluhm, Adsorption of dimethyl methylphosphonate on MoO₃: The role of oxygen vacancies, J. Phys. Chem. C, 2016, 120, 29077–29088.
49. L. Trotochaud, R. Tsyshevsky, S. Holdren, K. Fears, A. R. Head, Y. Yu, S. Pletincx, B. Eichhorn, M. Zachariah and M. M. Kuklja, et al., Spectroscopic and computational investigation of room-temperature decomposition of a chemical warfare agent simulant on polycrystalline cupric oxide, Chem. Mater., 2017, 29, 7483–7496.
50. I. V. Schweigert and D. Gunlycke, Hydrolysis of dimethyl methylphosphonate by the cyclic tetramer of zirconium hydroxide, J. Phys. Chem. A, 2017, 121, 7690–7696.
51. G. Wang, C. Sharp, A. M. Plonka, Q. Wang, A. I. Frenkel, W. Guo, C. Hill, C. Smith, J. Kollar, D. Troya and J. R. Morris, Mechanism and kinetics for reaction of the chemical warfare agent simulant, DMMP(g), with Zirconium (IV) MOFs: An ultrahigh-vacuum and DFT study, J. Phys. Chem. C, 2017, 121, 11261–11272.
52. S. J. Smith, R. Stevens, S. Liu, G. Li, A. Navrotsky, J. Boerio-Goates and B. F. Woodfield, Heat capacities and thermodynamic functions of TiO₂ anatase and rutile: Analysis of phase stability, Am. Mineral., 2009, 94, 236–243.
53. Z. H. Cui, F. Wu and H. Jiang, First-principles study of relative stability of rutile and anatase TiO₂ using the random phase approximation, Phys. Chem. Chem. Phys., 2016, 18, 29914–29922.
54. J. C. Slater, Atomic radii in crystals, J. Chem. Phys., 1964, 41, 3199–3205.
55. J. Perdew, J. Chevary, S. Vosko, K. Jackson, M. Pederson, D. Singh and C. Fiolhais, Atoms, molecules, solids, and surfaces: Applications of the generalized gradient approximation for exchange and correlation, Phys. Rev. B: Condens. Matter Mater. Phys., 1992, 46, 6671–6687.
56. P. E. Blöchl, Projector augmented-wave method, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
57. P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni and I. Dabo, et al., QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials, J. Phys.: Condens. Matter, 2009, 21, 395502.
58. S. Grimme, Semiempirical GGA-type density functional constructed with a long-range dispersion correction, J. Comput. Chem., 2006, 27, 1787–1799.
59. R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Clarendon Press, Oxford, 1994.
60. G. Henkelman, A. Arnaldsson and H. Jónsson, A fast and robust algorithm for Bader decomposition of charge density, Comput. Mater. Sci., 2006, 36, 354–360.
61. Henkelman Group Code for Bader Charge Analysis, 2023. https://theory.cm.utexas.edu/henkelman/code/bader/.
62. Y. Cardona Quintero and R. Nagarajan, Molecular and dissociative adsorption of DMMP, sarin and soman on dry and wet TiO₂(110) using density functional theory, Surf. Sci., 2018, 675, 26–35.
63. A. Kaczmarek, L. Gorb, A. J. Sadlej and J. Leszczynski, Sarin and soman: Structure and properties, Struct. Chem., 2004, 15, 517–525.
64. X. Shan, J. C. Vincent, S. Kirkpatrick, M. D. Walker, M. R. Sambrook and D. C. Clary, A combined theoretical and experimental study of sarin (GB), J. Phys. Chem. A, 2017, 121, 6200–6210.
65. T. Ash, T. Debnath, T. Banu and A. K. Das, Exploration of unimolecular gas-phase detoxication pathways of sarin and soman: A computational study from the perspective of reaction energetics and kinetics, Chem. Res. Toxicol., 2016, 29, 1439–1457.
66. N. Q. Le, C. E. Ekuma, B. I. Dunlap and D. Gunlycke, First-principles calculations of sarin adsorption on anatase surfaces, J. Phys. Chem. C, 2018, 122, 2832–2839.
67. S. Tan and B. Wang, Active Sites for Adsorption and reaction of molecules on rutile TiO₂(110) and anatase TiO₂(001) surfaces, Chin. J. Chem. Phys., 2015, 28, 383–395.
68. C. J. Howard, T. M. Sabine and F. Dickson, Structural and thermal parameters for rutile and anatase, Acta Crystallogr., Sect. B, 1991, 47, 462–468.
69. D. C. Taylor, K. Runge, M. G. Cory, D. S. Burns, J. L. Vasey, J. D. Hearn, K. Griffith and M. V. Henley, Surface binding of organophosphates on silica: Comparing experiment and theory, J. Phys. Chem. C, 2013, 117, 2699–2708.
70. V. Štengl, S. Bakardjieva, N. Murafa and F. Opluštil, Zirconium doped titania: Destruction of warfare agents and photocatalytic degradation of orange 2 dye, Open Process Chem. J., 2008, 1, 1–7.
71. V. Štengl, S. Bakardjieva, N. Murafa, F. Opluštil, L. Österlund, A. Mattsson and P. O. Andersson, Warfare agents degradation on zirconium doped titania, Microsc. Microanal., 2009, 15, 1038–1039.
72. V. Štengl, T. M. Grygar, F. Opluštil and T. Nemec, Ge⁴⁺ doped TiO₂ for stoichiometric degradation of warfare agents, J. Hazard. Mater., 2012, 227–228, 62–67.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5cp01092f

‡ Current address: Naval Undersea Warfare Center, Division Newport, Newport, RI 02841, USA.

---