Hydrolysis of Dimethyl Methylphosphonate by the Cyclic Tetramer of

Sep 19, 2017 - U.S. Naval Research Laboratory, Code 6189, Washington, District of ... Ashley R Head , Roman Tsyshevsky , Lena Trotochaud , Yi Yu , Osm...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCA

Hydrolysis of Dimethyl Methylphosphonate by the Cyclic Tetramer of Zirconium Hydroxide Igor V. Schweigert* and Daniel Gunlycke U.S. Naval Research Laboratory, Code 6189, Washington, District of Columbia 20375, United States S Supporting Information *

ABSTRACT: We present hybrid density functional theory (DFT) calculations of hydrolysis of dimethyl methylphosphonate (DMMP) by the cyclic tetramer of zirconium hydroxide [Zr4(OH)16]. Various binding configurations of DMMP and its hydrolysis products on the tetramer as well as transition structures connecting them were explored using structure optimizations based on multiple, randomly selected initial structures. We find that DMMP can bind to the tetramer through the phosphoryl O, forming either a strong hydrogen bond to a bridging hydroxyl or a coordinate bond to a coordinatively unsaturated Zr atom. The resulting hydrogen-bonded complexes and Lewis adducts have similar energies. We also find that hydrolysis of a P−OCH 3 bond can occur either via an addition−elimination mechanism involving a same-site terminal hydroxyl or direct interchange between a terminal hydroxyl and a methoxy group of DMMP. The computed activation and reaction enthalpies show that the addition− elimination is both kinetically and thermodynamically favored over the direct interchange. Our findings support recent observations of the reactivity of amorphous zirconium hydroxide toward phosphonate esters including chemical warfare agents. nism of phosphonate ester hydrolysis.31−33 In alkaline aqueous solutions, the substitution proceeds via a two-step mechanism consisting of nucleophilic addition of a hydroxide anion to the phosphorus followed by elimination of a leaving group.34−37 The addition−elimination mechanism is characterized by a pentacoordinated P intermediate (P5), wherein the ligands around the P atom are arranged in a trigonal bipyramidal configuration. On hydroxylated surfaces of metal oxides, a metal-bound terminal hydroxyl can act as a nucleophile, resulting in a surface-bound, bidentate P5 intermediate.23,25,37,38 A recent theoretical study of hydrolysis of Sarin, a related phosphonate ester, on Zr-based MOFs showed that, in absence of bulk water, the elimination step proceeds via intramolecular proton transfer between the ligands in the P5 intermediate. While DMMP binding and reactivity on amorphous ZH can be expected to follow similar trends, the kinetic and thermodynamic accessibilities of various hydrolysis intermediates and products depend on the specific structural and electronic properties of the ZH surface and are the focus of this study. Unlike metal oxides and MOFs, the bulk structure of amorphous ZH is highly dependent on the material preparation and processing.4−10 Southon et al. determined that the basic unit of a sol−gel derived ZH is the cyclic tetramer, Zr4(OH)16, and proposed that the tetrameric units are fused into singlelayer sheets of varying sizes, which further stack to form disk-

1. INTRODUCTION Amorphous zirconium hydroxide (ZH) has recently been shown to efficiently hydrolyze the G- and V-type nerve agents,1 prompting its consideration for filtration and decontamination applications. Its reactivity was correlated to the presence of both acidic and basic sites on the surface, wherein coordinatively unsaturated (cus) metal sites and bridging hydroxyls participate in the initial adsorption, and terminal hydroxyls participate in the hydrolysis of the agents.2−4 The density and accessibility of these surface groups depend on how the material is prepared and processed, as evident in significant structural variations in sol−gel derived or wet-precipitated ZH.4−10 Better understanding of how phosphonate esters interact with surface metal sites and hydroxyl groups can aid in linking the material preparation and processing history to its performance in decontamination of chemical warfare agents. Among various phosphonates, dimethyl methylphosphonate (DMMP) has been extensively studied as a simulant for the G and V series of nerve agents. Both experimental and theoretical studies of its adsorption and reactivity on hydroxylated surfaces of metal oxides,11−26 silica,13,27−29 and Zr-based metal organic frameworks30 (MOFs) have been reported. At lower temperatures, DMMP adsorbs molecularly at acidic sites provided by bridging hydroxyl groups and cus metal atoms. At elevated temperatures, DMMP undergoes hydrolysis of the P-OCH3 bonds with a concurrent consumption of surface hydroxyls. Metal-bound phosphonates and methoxy groups, as well as volatile products including methanol, have been reported as the hydrolysis products.11,12,14−16,20,22−24 These products are consistent with the general nucleophilic substitution mechaThis article not subject to U.S. Copyright. Published XXXX by the American Chemical Society

Received: June 29, 2017 Revised: September 14, 2017 Published: September 19, 2017 A

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

shape particles.6 Southon et al. also reported that in the gel the sheets stack to form disk-shape particles, which are several nanometers wide (50−100 Zr atoms per sheet) and a fraction of a nanometer thick (2−3 sheets). Mogilevksy et al. analyzed commercially available ZH and reported the ratio of bridging to terminal hydroxyls to be about 3−3.4 to 1.7 DeCoste et al. reported that the ratio remained the same upon crushing a sample from the same vendor, but a sample from a different vendor had a much higher ratio of 8 to 1.10 They interpreted the variations in the observed ratios between the bridging and terminal hydroxyls in terms of differences in the particle size distributions. Indeed, in a single-layer sheet, only the peripheral Zr atoms have terminal hydroxyls; hence, the ratio between the bridging and terminal hydroxyls depends on the particle size. However, it is also plausible that the ratio is affected by other factors, including contact interfaces among ZH particles, different stacking arrangements of single layers, and the presence of structural defects such as Zr vacancies or surfaceadsorbed oligomers. In this work, we used hybrid density functional theory (DFT) to study hydrolysis of DMMP by a stand-alone cyclic tetramer, Zr4(OH)16. In lieu of a better model for the ZH surface, the tetramer was selected as the simplest model that features the tetrameric unit. The reduced computational cost of this model enabled us to perform an exhaustive search for possible binding configurations of DMMP and its reaction intermediates as well as to locate transition structures mediating the hydrolysis of the P-OCH3 bonds. As discussed below, we found that within our model DMMP binds and reacts in the vicinity of a single Zr atom; therefore, the cyclic tetramer is best viewed as an approximation to an corner of a ZH particle or a structural defect with an exposed six-coordinated Zr atom. The details of our computational procedure are given in section 2. Calculated structures and binding enthalpies of DMMP to the tetramer, its hydrolysis products, and the transition structures that mediate the two hydrolysis mechanisms are presented and discussed in section 3. A summary of our findings is given in section 4.

Initial structures for various complexes and adducts on the tetramer were generated by randomly sampling different positions and orientations of each adsorbate species with respect to the tetramer. The optimized structure of the adsorbate was first superimposed on the tetramer structure so that the adsorbate’s center of mass coincided with one of the Zr atoms. A random rotation axis and angle were then chosen by selecting a random unit quaternion from the set of uniformly distributed quaternions. A random direction was chosen by selecting a point on a unit hemisphere centered at the selected Zr atom. The adsorbate was first rotated according to the selected quaternion and then translated along the selected direction until the shortest distance between any two atoms was greater than a predetermined value (2 Å, unless noted otherwise). Using this procedure, 50 initial structures for each complex or adduct were generated. The structures were optimized using the same convergence criteria as for individual species and confirmed as local minima. Transition structures were searched for using constrained structure optimizations wherein various bond lengths and/or angles were varied to approximate target transition states and product complexes. The approximate transition structures were subsequently optimized without any constraints using the default eigenvector-following algorithm available in Gaussian09 and confirmed as a first-order saddle point based on analytically computed eigenvalues of the Hessian matrices. Furthermore, a Hessian-based implementation of the intrinsic reaction coordinate (IRC) method was used to confirm each saddle point as a transition structure and to determine the corresponding reactant and product complexes. Thermodynamic functions including the vibrational zeropoint energy (ZPE), enthalpy content at 298.15 K, and absolute entropy at 298.15 K and 1 bar were computed using the idealgas, rigid-rotor, and unscaled harmonic-oscillator partition functions. We note that all structures computed in this work exhibit low-frequency, high-amplitude vibrations for which the harmonic-oscillator approximation is expected to perform poorly. As a result, the errors in the computed entropies due to the lack of anharmonic corrections are expected to be much higher than the errors in the computed enthalpies. Therefore, only enthalpies are reported throughout the paper.

2. COMPUTATIONAL METHODS All calculations reported here were done with the M06 metahybrid functional39 combined with the 6-31+G(d,p) allelectron, double-ζ set40−42 of atom-centered Gaussian-type orbitals for elements H through P, the LANL2DZ relativistic set43 of effective core potentials for Zr core electrons, and the modified double-ζ basis set44 for Zr valence electrons, as implemented in Gaussian09.45 Pruned grids with 99 radial shells around each atom with 590 angular points in each shell were used in these calculations. All species were calculated in their spin-singlet state using the restricted Kohn−Sham (KS) equations. The possibility of a lower-energy, open-shell electronic configuration was checked for each species using the orbital stability analysis,46 but no such configurations were found. Individual structures for the tetramer, DMMP, and various hydrolysis products were optimized without constraints until the maximum (root-mean-square) forces in internal coordinates were below 15 × 10−6 (10 × 10−6) Hartree/Bohr or Hartree/radian and maximum (root-mean-square) displacements were below 60 × 10−6 (40 × 10−6) Bohr or radian. Each structure was confirmed as a local minimum using analytically computed eigenvalues of the Hessian matrices.

3. RESULTS AND DISCUSSION Hydrogen-bonded complexes between DMMP and Zr4(OH)16. Previous studies of DMMP adsorption on hydroxylated surfaces of metal oxides suggested that DMMP binds by forming hydrogen bonds between the phosphoryl O and acidic hydroxyl groups.12,14−16,19−21,23 To determine the preferential binding configurations of DMMP on the tetramer, we optimized hydrogen-bonded complexes between DMMP and the tetramer starting with 50 randomly generated structures. The resulting three lowest-enthalpy structures (labeled 1a, 1b, and 1c in Figure 1) feature the phosphoryl O coordinated to a bridging hydroxyl, with the O···H distance of 1.89 Å. These complexes differ in how DMMP is positioned with respect to the tetramer, atop of the ring (1a and 1c) or aside the edge (1b), and in how the methyl group of DMMP is oriented with respect to the tetramer: parallel to the surface (1a) or pointing away (1b and 1c). All three complexes 1a, 1b, 1c are further stabilized by the proximity of a terminal hydroxyl with the O···H distances ranging from 2.0 to 2.2 Å. Lewis adducts between DMMP and Zr4(OH)16. The other reported binding motifs of DMMP to hydroxylated B

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Similarly, constrained geometry optimizations wherein the distances between the phosphoryl O and Zr in 2c were varied between 2.3 and 5.0 Å showed that this adduct is accessible from the hydrogen-bonded complex 1a with the highest-energy structure along the approximate reaction path lying at 30 kJ/ mol. While we were unable to obtain the precise transition structure for the reason described above, we interpret this finding as an indication that 2c and other Lewis adducts are readily accessible from hydrogen-bonded complexes at room temperature. This suggests that DMMP is mobile on ZH surfaces and can diffuse along the surface from a likely initial adsorption site such as atop a large particle to other, more reactive sites such as an exposed corner of a single-layer sheet modeled in this study. Product complexes of methanol elimination. Previous studies of DMMP decomposition on metal oxides identified methanol as the principal volatile product.11,12,15−17,20,22 The formation of methanol requires combination of a methoxide anion from DMMP and a proton from one of the surface hydroxyls. We assumed that demethoxylated DMMP binds to the deprotonated terminal or bridging hydroxyl on the surface and generated 100 initial, randomly oriented structures with demethoxylated DMMP placed near the deprotonated terminal or bridging oxygen. The resulting structures were fully optimized in the presence of methanol placed near the reaction site. The three lowest-enthalpy structures are shown in Figure 3. We note that the bidentate binuclear product 3a wherein a

Figure 1. Hydrogen-bonded complexes between DMMP and Zr4(OH)16. Zr atoms are shown in cyan, O atoms in red, C atoms in gray, and H atoms in white. Dashed lines depict hydrogen bonds. Values shown in parentheses are differences in enthalpies with respect to 1a.

surfaces of metal oxides are Lewis adducts between the phosphoryl O and cus metal atoms.12,14−16,19−21,23 To locate such adducts for ZH, we generated a set of 50 structures wherein the distances between the phosphoryl O and a specified Zr atom were restrained to be shorter than 2.5 Å. We note that three such initial structures resulted in erratic convergence and/or failures of the self-consistent field (SCF) Kohn−Sham equations. Among the 47 converged structures, the three with the lowest enthalpies (labeled 2a, 2b, and 2c in Figure 2) have the distances between Zr and the phosphoryl O

Figure 2. Lewis adducts between DMMP and Zr4(OH)16. Values shown in parentheses are differences in enthalpies with respect to 1a. Figure 3. Product complexes of the methanol elimination reaction.

equal to 2.28 Å. For comparison, the distances between Zr and terminal O atoms are 2.07 Å. The bond lengths between P and the phosphoryl O have values similar to those in hydrogenbonded complexes and stand-alone DMMP: 1.51 Å in 2a, 1.50 Å in 1a, and 1.48 Å in DMMP, indicating that the P−O bond retains its character. The three Lewis adducts differ in the orientation of the Zr−O bond with respect to the two same-site terminal hydroxyls and in the orientations of the methoxy and methyl groups with respect to the tetramer. Unlike the Lewis adducts previously simulated for metal oxides14,19,23,25,37 and Zr-based MOFs,38 the Lewis adducts computed here feature terminal hydroxyls bound to the same cus Zr atom that is bound to the phosphoryl O. Interconversion between hydrogen-bonded complexes and Lewis adducts. Constrained geometry optimizations (not shown) in 1a wherein the distances between the phosphoryl O and a bridging H were varied from 2.0 to 4.0 Å indicated that the barriers to interconversion between different conformers can be as low as 16 kJ/mol. Unfortunately, we were unable to converge precise transition structures connecting 1a to the other isomers because the conformational changes relating different isomers corresponded to low-frequency, largeamplitude normal modes that are numerically challenging for the eigenmode-following optimization algorithm.

deprotonated bridging O now binds to P is similar to the bidentate products of DMMP hydrolysis previously observed in DMMP hydrolysis on metal oxides.15,16,24 The monodentate mononuclear product 3b and bidentate mononuclear (chelating) product 3c have not been reported in the literature, but our calculations predict that they are similarly stable. In fact, we show below that 3c results from an additional elimination reaction involving a same-site hydroxyl and therefore is the product of methanol elimination in DMMP accessible from Lewis adducts such as 2c. Product complexes of direct hydroxyl-methoxy interchange. Previous studies have also reported the formation of surface-bound methoxy groups.15,16,23,24 A direct interchange between a metal-bound hydroxyl and a DMMPbound methoxy could directly yield a surface-bound methoxy group and methyl methylphosphonate (MMP). To our knowledge, this mechanism has not been reported in the literature. We assumed that MMP would initially be hydrogen bonded near the methoxylated metal site and generated 50 initial, randomly oriented structures with the methoxy substituting a terminal hydroxyl and another 50 with the methoxy substituting a bridging hydroxyl. The final, full optimized C

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

varied from 1.63 to 2.0 Å, and we found a very large barrier to the direct heterolytic cleavage of this bond. We subsequently performed optimizations wherein the P−OCH3 bond length of the equatorial methoxy group was varied from 1.68 to 2.0 Å and found that increased P−O distances resulted in a simultaneous increase in the O−H bond length in the added terminal hydroxyl, indicating a concerted elimination mechanism. The highest-energy structure along the approximate reaction path was then fully optimized without constraints to yield the final transition structure to elimination (labeled TS1 in Figure 5). TS1 has the equatorial P−OCH3 distance of 2.01 Å and the P− OH distance of 1.63 Å. Its enthalpy is 103.0 kJ/mol with respect to 1a and 72 kJ/mol with respect to 5. An IRC calculation showed that it connects 5 to the bidentate chelating product complex 3c (Figure 3). The enthalpy diagram summarizing these findings is shown in Figure 6. We note

structures with the lowest enthalpies are shown in Figure 4. We note that all three structures feature the methoxy group in the

Figure 4. Product complexes of the direct interchange reaction.

terminal position and the phosphoryl O of MMP being hydrogen bonded to a bridging hydroxyl. We also note that the lowest-enthalpy structure 4a is about 40 kJ/mol above the lowest product complex of methanol elimination 3a, indicating that the direct interchange mechanism is thermodynamically unfavorable compared to the methanol elimination. Detailed reaction path for the addition−elimination mechanism. Unlike hydroxylated surfaces of metal oxides or Zr-based MOFs, the tetramer model considered here features cus Zr atoms that have terminal hydroxyl. In fact, all three Lewis adducts 2a, 2b, and 2c (Figure 2) have same-site terminal hydroxyls in the vicinity of the P atom of DMMP (the O−P distances are equal to, respectively, 2.98, 3.04, and 2.84 Å), which could allow for a facile addition resulting in chelating P5 intermediates. We performed constrained geometry optimizations wherein the distances between the P atom of DMMP and the O atom of a same-site terminal hydroxyl in 2c were varied from 2.8 and 1.8 Å and found a bidentate, chelating P5 intermediate (labeled 5 in Figure 5). In 5, the distance between

Figure 6. Enthalpy diagram summarizing the intermediates and transition structures involved in the addition−elimination reaction.

that the concerted elimination of HF via a proton transfer from an added hydroxyl to an equatorial F was recently reported by Troya for the P5 intermediate (ref 38) mediating Sarin hydrolysis on a Zr-based MOF. Detailed reaction path for the direct interchange mechanism. While a direct interchange between a Zr-bound hydroxyl and a methoxy moiety of DMMP has not been previously considered, it can directly lead to a Zr-bound methoxy group. We performed constrained geometry optimizations wherein the distances between a methoxy O and a Zr atom were varied between 4.3 and 2.3 Å and found an unstable intermediate (labeled 6 in Figure 7) with the methoxy O forming a Lewis adduct with the Zr atom. 6 is stabilized only by a hindered rotation of the second P−OCH3 bond, which corresponds to a shallow barrier below 3 kJ/mol on the electronic potential energy surface. We also performed constrained geometry optimizations wherein the distances between the same-site terminal hydroxyl

Figure 5. Bidentate chelating intermediate of terminal hydroxyl addition to DMMP 5 and transition structure TS1 to methanol elimination.

the phosphoryl O and P is 1.57 Å (a 0.06 Å increase compared to 2c) and the distance between the added terminal O and P is 1.86 Å. 5 lies 31 kJ/mol above 1a, and the enthalpy of the highest-energy structure along the approximate reaction path connecting 2c and 5 is 36 kJ/mol above 1a. We note that the stability of 5 with respect to 2c is similar to the stability of the P5 intermediate reported by Troya for Sarin on a Zr-based MOF: 31 kJ/mol in this work and 33 kJ/mol in ref 38 while the barrier is somewhat lower: 36 kJ/mol in this work versus 55 kJ/ mol.38 To determine the mechanism of the elimination step, we initially performed constrained geometry optimizations wherein the P−OCH3 bond length of the apical methoxy group in 5 was

Figure 7. Metastable intermediate of methoxy O addition to a Zr atom 6 and transition structure TS2 mediating a direct interchange between a terminal hydroxyl and a methoxy group. D

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

4. CONCLUDING REMARKS We have presented hybrid DFT calculations of the binding and reactivity of DMMP (Scheme 1) on the cyclic tetramer of ZH [Zr4(OH)16]. These calculations relied on the M06/6-31+G(d,p)/mod-LANL2DZ level of DFT and multiple, randomly generated initial structures to explore possible binding configurations of DMMP and its hydrolysis products to Zr4(OH)16. We also used transition structure optimization and intrinsic reaction coordinate calculations to identify the detailed reaction paths leading to the observed hydrolysis products. We found that DMMP initially forms hydrogen-bonded complexes with Zr4(OH)16, wherein the phosphoryl O is bound to bridging hydroxyls (reaction 1 in Scheme 1). The complexes can isomerize to Lewis adducts, wherein the phosphoryl O is bound to cus Zr atoms (reaction 2). The corresponding binding enthalpies are on the scale of 72 kJ/mol with respect to separated DMMP and Zr4(OH)16. We also found that hydrolysis of the P−OCH 3 bonds can occur via the addition−elimination mechanism and direct interchange mechanisms. The addition−elimination mechanism involves addition of a same-site hydroxyl to the P atom, followed by proton transfer and elimination of methanol (reaction 3 in Scheme 1). This reaction initiates from a Lewis adduct and yields a Zr-bound demethoxylated DMMP and weakly bound methanol. The direct interchange mechanism involves concerted formation of Zr-OCH3 and dissociation of Zr−OH bonds (reaction 4 in Scheme 1). This reaction initiates in a hydrogen-bonded complex and yields Zr-bound methoxy and weakly bound methyl methylphosphonate. The direct inter-

group and the phosphorus of DMMP in 6 were varied from 3.6 and 1.6 Å. We found that enforcing smaller P−OH distances directly lead to larger P−OCH3 distances, indicating that the direct interchange mechanism does not involve a P5 intermediate. The highest-energy structure along the approximate reaction path was then fully optimized without constraints to yield the final transition structure (labeled TS2 in Figure 7). TS2 has the P−OCH3 bond distance of 1.93 Å, the P−OH distance of 1.81 Å and lies 136.0 kJ/mol above 1a. A subsequent IRC calculation showed that TS2 connects 6 to the product complex 4a between MMP and Zr(OH)3OCH3. The enthalpy diagram summarizing these findings is shown in Figure 8.

Figure 8. Enthalpy diagram summarizing the intermediates and transition structures involved in the direct interchange reaction.

Scheme 1. Addition−Elimination and Direct Interchange Mechanisms of DMMP Hydrolysis by Zr4(OH)16.

E

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

(2) Peterson, G. W.; Karwacki, C. J.; Feaver, W. B.; Rossin, J. A. Zirconium Hydroxide as a Reactive Substrate for the Removal of Sulfur Dioxide. Ind. Eng. Chem. Res. 2009, 48, 1694−1698. (3) Peterson, G. W.; Wagner, G. W.; Keller, J. H.; Rossin, J. A. Enhanced Cyanogen Chloride Removal by the Reactive Zirconium Hydroxide Substrate. Ind. Eng. Chem. Res. 2010, 49, 11182−11187. (4) Peterson, G. W.; Rossin, J. A.; Karwacki, C. J.; Glover, T. G. Surface Chemistry and Morphology of Zirconia Polymorphs and the Influence on Sulfur Dioxide Removal. J. Phys. Chem. C 2011, 115, 9644−9650. (5) Jung, K. T.; Shul, Y. G.; Bell, A. T. The Preparation and Surface Characterization of Zirconia Polymorphs. Korean J. Chem. Eng. 2001, 18, 992−999. (6) Southon, P. D.; Bartlett, J. R.; Woolfrey, J. L.; Ben-Nissan, B. Formation and Characterization of an Aqueous Zirconium Hydroxide Colloid. Chem. Mater. 2002, 14, 4313−4319. (7) Mogilevsky, G.; Karwacki, C. J.; Peterson, G. W.; Wagner, G. W. Surface Hydroxyl Concentration on Zr(Oh)(4) Quantified by H-1 Mas Nmr. Chem. Phys. Lett. 2011, 511, 384−388. (8) Kamimura, Y.; Endo, A. Co2 Adsorption-Desorption Performance of Mesoporous Zirconium Hydroxide with Robust Water Durability. Phys. Chem. Chem. Phys. 2016, 18, 2699−2709. (9) Chitrakar, R.; Tezuka, S.; Sonoda, A.; Sakane, K.; Ooi, K.; Hirotsu, T. Selective Adsorption of Phosphate from Seawater and Wastewater by Amorphous Zirconium Hydroxide. J. Colloid Interface Sci. 2006, 297, 426−433. (10) DeCoste, J. B.; Glover, T. G.; Mogilevsky, G.; Peterson, G. W.; Wagner, G. W. Trifluoroethanol and 19f Magic Angle Spinning Nuclear Magnetic Resonance as a Basic Surface Hydroxyl Reactivity Probe for Zirconium(Iv) Hydroxide Structures. Langmuir 2011, 27, 9458−9464. (11) Templeton, M. K.; Weinberg, W. H. Adsorption and Decomposition of Dimethyl Methylphosphonate on an AluminumOxide Surface. J. Am. Chem. Soc. 1985, 107, 97−108. (12) Templeton, M. K.; Weinberg, W. H. Decomposition of Phosphonate Esters Adsorbed on Aluminum-Oxide. J. Am. Chem. Soc. 1985, 107, 774−779. (13) Henderson, M. A.; Jin, T.; White, J. M. A Tpd/Aes Study of the Interaction of Dimethyl Methylphosphonate with Alpha-Fe2o3 and Sio2. J. Phys. Chem. 1986, 90, 4607−4611. (14) Zhanpeisov, N. U.; Zhidomirov, G. M.; Yudanov, I. V.; Klabunde, K. J. Cluster Quantum-Chemical Study of the Interaction of Dimethyl Methylphosphonate with Magnesium-Oxide. J. Phys. Chem. 1994, 98, 10032−10035. (15) Mitchell, M. B.; Sheinker, V. N.; Mintz, E. A. Adsorption and Decomposition of Dimethyl Methylphosphonate on Metal Oxides. J. Phys. Chem. B 1997, 101, 11192−11203. (16) Rusu, C. N.; Yates, J. T. Adsorption and Decomposition of Dimethyl Methylphosphonate on Tio2. J. Phys. Chem. B 2000, 104, 12292−12298. (17) Mitchell, M. B.; Sheinker, V. N.; Cox, W. W.; Gatimu, E. N.; Tesfamichael, A. B. The Room Temperature Decomposition Mechanism of Dimethyl Methylphosphonate (Dmmp) on AluminaSupported Cerium Oxide - Participation of Nano-Sized Cerium Oxide Domains. J. Phys. Chem. B 2004, 108, 1634−1645. (18) Michalkova, A.; Ilchenko, M.; Gorb, L.; Leszczynski, J. Theoretical Study of the Adsorption and Decomposition of Sarin on Magnesium Oxide. J. Phys. Chem. B 2004, 108, 5294−5303. (19) Bermudez, V. M. Quantum-Chemical Study of the Adsorption of Dmmp and Sarin on Gamma-Al2o3. J. Phys. Chem. C 2007, 111, 3719−3728. (20) Gordon, W. O.; Tissue, B. M.; Morris, J. R. Adsorption and Decomposition of Dimethyl Methylphosphonate on Y2o3 Nanoparticles. J. Phys. Chem. C 2007, 111, 3233−3240. (21) Panayotov, D. A.; Morris, J. R. Uptake of a Chemical Warfare Agent Simulant (Dmmp) on Tio2: Reactive Adsorption and Active Site Poisoning. Langmuir 2009, 25, 3652−3658. (22) Panayotov, D. A.; Morris, J. R. Thermal Decomposition of a Chemical Warfare Agent Simulant (Dmmp) on Tio2: Adsorbate

change mechanism is both kinetically and thermodynamically less favorable than the addition−elimination mechanism. Our findings are consistent with the previously reported studies of binding and reactivity of DMMP on hydroxylated metal surfaces and Zr-based MOFs.11,12,15−17,19−25,28−30 However, the proximity of terminal hydroxyls to bridging hydroxyls and cus Zr sites on amorphous ZH surfaces lead to subtle differences in their reactivity toward phosphonate esters. The addition−elimination reaction proceeds via addition of a same-site rather than a different-site hydroxyl, resulting in chelating rather than bridging intermediates. It is possible that the resulting monodentate product of methanol elimination can undergo further decomposition prior to isomerization to more stable bridging products. Such secondary reactions are currently being investigated. The direct interchange reaction requires the presence of a terminal hydroxyl on a cus Zr site and is assisted by the strong hydrogen bond between the phosphoryl O and an adjacent bridging hydroxyl. This reaction may be unique to ZH. Since it involves O atoms from the surface, isotopic scrambling in reaction products from 18O-labeled DMMP could validate this mechanism. The presented calculations did not include water or other possible atmospheric species, which are known to be present on amorphous ZH. Water is expected to compete with phosphonates for accessible cus Zr atoms and bridging hydroxyls, thus inhibiting DMMP binding to the surface. Acid-forming gases such as carbon and sulfur dioxides can react with terminal hydroxyls, forming sulfites and carbonates. Future work will address the effects of coadsorbed atmospheric species on the energetics of the addition−elimination and direct interchange mechanisms.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.7b06403.

■

(XYZ)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Igor V. Schweigert: 0000-0002-9474-7767 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank V. M. Bermudez, A. Balboa, I. Iordanov, C. Knox, R. Balow, J. Lundin, D. Barlow, J. Wynne, M. McEntee, W. Gordon, G. Peterson, and P. E. Pehrsson for useful discussions. This work was supported by the Defense Threat Reduction Agency (DTRA, project CB10123) and by the Office of Naval Research (ONR) through the U.S. Naval Research Laboratory (NRL).

■

REFERENCES

(1) Bandosz, T. J.; Laskoski, M.; Mahle, J.; Mogilevsky, G.; Peterson, G. W.; Rossin, J. A.; Wagner, G. W. Reactions of Vx, Gd, and Hd with Zr(Oh)4: Near Instantaneous Decontamination of Vx. J. Phys. Chem. C 2012, 116, 11606−11614. F

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Reactions with Lattice Oxygen as Studied by Infrared Spectroscopy. J. Phys. Chem. C 2009, 113, 15684−15691. (23) Bermudez, V. M. 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. (24) Chen, D. A.; Ratliff, J. S.; Hu, X. F.; Gordon, W. O.; Senanayake, S. D.; Mullins, D. R. Dimethyl Methylphosphonate Decomposition on Fully Oxidized and Partially Reduced Ceria Thin Films. Surf. Sci. 2010, 604, 574−587. (25) Bermudez, V. M. 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. (26) Bermudez, V. M. Investigation of the Interaction of GammaAl2o3 with Aqueous Solutions of Dimethyl Methylphosphonate Using Infrared Multiple Internal Reflection Spectroscopy. Langmuir 2013, 29, 1483−1489. (27) Bermudez, V. M. Computational Study of the Adsorption of Trichlorophosphate, Dimethyl Methylphosphonate, and Sarin on Amorphous Sio2. J. Phys. Chem. C 2007, 111, 9314−9323. (28) Wilmsmeyer, A. R.; Uzarski, J.; Barrie, P. J.; Morris, J. R. Interactions and Binding Energies of Dimethyl Methylphosphonate and Dimethyl Chlorophosphate with Amorphous Silica. Langmuir 2012, 28, 10962−10967. (29) Troya, D.; Edwards, A. C.; Morris, J. R. Theoretical Study of the Adsorption of Organophosphorous Compounds to Models of a Silica Surface. J. Phys. Chem. C 2013, 117, 14625−14634. (30) Plonka, A. M.; Wang, Q.; Gordon, W. O.; Balboa, A.; Troya, D.; Guo, W.; Sharp, C. H.; Senanayake, S. D.; Morris, J. R.; Hill, C. L.; et al. In Situ Probes of Capture and Decomposition of Chemical Warfare Agent Simulants by Zr-Based Metal Organic Frameworks. J. Am. Chem. Soc. 2017, 139, 599−602. (31) Hudson, R. F.; Keay, L. The Hydrolysis of Phosphonate Esters. J. Chem. Soc. 1956, 2463−2469. (32) Westheimer, F. H.; Huang, S.; Covitz, F. Rates and Mechanisms of Hydrolysis of Esters of Phosphorous acid. J. Am. Chem. Soc. 1988, 110, 181−185. (33) Yang, Y. C.; Baker, J. A.; Ward, J. R. Decontamination of Chemical Warfare Agents. Chem. Rev. 1992, 92, 1729−1743. (34) Zheng, F.; Zhan, C. G.; Ornstein, R. L. Theoretical Studies of Reaction Pathways and Energy Barriers for Alkaline Hydrolysis of Phosphotriesterase Substrates Paraoxon and Related Toxic Phosphofluoridate Nerve Agents. J. Chem. Soc. Perkin Trans. 2 2001, 2355− 2363. (35) Seckute, J.; Menke, J. L.; Emnett, R. J.; Patterson, E. V.; Cramer, C. J. Ab Initio Molecular Orbital and Density Functional Studies on the Solvolysis of Sarin and O,S-Dimethyl Methylphosphonothiolate, a VX-Like Compound. J. Org. Chem. 2005, 70, 8649−8660. (36) Ashkenazi, N.; Segall, Y.; Chen, R.; Sod-Moriah, G.; Fattal, E. The Mechanism of Nucleophilic Displacements at Phosphorus in Chloro-Substituted Methylphosphonate Esters: P-O vs P-C Bond Cleavage: A Dft Study. J. Org. Chem. 2010, 75, 1917−1926. (37) Chapleski, R. C.; Musaev, D. G.; Hill, C. L.; Troya, D. Reaction Mechanism of Nerve-Agent Hydrolysis with the Cs8Nb6O19 Lindqvist Hexaniobate Catalyst. J. Phys. Chem. C 2016, 120, 16822− 16830. (38) Troya, D. Reaction Mechanism of Nerve-Agent Decomposition with Zr-Based Metal Organic Frameworks. J. Phys. Chem. C 2016, 120, 29312−29323. (39) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group Thermochemistry, Thermochemical Kinetics, Noncovalent Interactions, Excited States, and Transition Elements: Two New Functionals and Systematic Testing of Four M06-Class Functionals and 12 Other Functionals. Theor. Chem. Acc. 2008, 120, 215−241. (40) Hehre, W. J.; Ditchfield, R.; Pople, J. A. Self-Consistent Molecular-Orbital Methods 0.12. Further Extensions of Gaussian-Type Basis Sets for Use in Molecular-Orbital Studies of Organic-Molecules. J. Chem. Phys. 1972, 56, 2257.

(41) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; Defrees, D. J.; Pople, J. A. Self-Consistent Molecular-Orbital Methods 0.23. A Polarization-Type Basis Set for 2nd-Row Elements. J. Chem. Phys. 1982, 77, 3654−3665. (42) Frisch, M. J.; Pople, J. A.; Binkley, J. S. Self-Consistent Molecular-Orbital Methods 0.25. Supplementary Functions for Gaussian-Basis Sets. J. Chem. Phys. 1984, 80, 3265−3269. (43) Hay, P. J.; Wadt, W. R. Abinitio Effective Core Potentials for Molecular Calculations - Potentials for the Transition-Metal Atoms Sc to Hg. J. Chem. Phys. 1985, 82, 270−283. (44) Couty, M.; Hall, M. B. Basis Sets for Transition Metals: Optimized Outer P Functions. J. Comput. Chem. 1996, 17, 1359−1370. (45) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A., et al. Gaussian 09; Gaussian, Inc.: Wallingford, CT, USA, 2009. (46) Seeger, R.; Pople, J. A. Self-Consistent Molecular-Orbital Methods 0.18. Constraints and Stability in Hartree-Fock Theory. J. Chem. Phys. 1977, 66, 3045−3050.

G

DOI: 10.1021/acs.jpca.7b06403 J. Phys. Chem. A XXXX, XXX, XXX−XXX