Implicit Solvent Model Research Articles

Comparing the influence of explicit and implicit solvation models on site-specific thermodynamic stability of proteins.

Understanding the molecular basis for protein stability requires a thermodynamic analysis of protein folding. Thermodynamic analysis is often performed by sampling many atomistic conformations using molecular simulations that employ either explicit or implicit water models. However, it remains unclear to what extent thermodynamic results from different solvation models are reliable at the molecular level. In this study, we quantify the influence of both solvation models on folding stability at the individual backbone and side chain resolutions. We assess the residue-specific folding free energy components of a β-sheet protein and a helical protein using trajectories resulting from TIP3P explicit and generalized Born/surface area implicit solvent simulations of model proteins. We found that the thermodynamic discrepancy due to the implicit solvent mostly originates from charged side chains, followed by the under-stabilized hydrophobic ones. In contrast, the contributions of backbone residue in both proteins were comparable for explicit and implicit water models. Our study lays out the foundation for detailed thermodynamic assessment of solvation models in the context of protein simulation.

A method for protein structure sampling under three-dimensional reference interaction site model solvation based on hybrid Monte Carlo framework

An efficient sampling method for protein structure using molecular dynamics (MD) simulation and a three-dimensional reference interaction-site model (3D-RISM) theory has been developed based on the hybrid Monte Carlo framework. In this method, new structures are generated by MD simulation with the generalized Born implicit solvent model and the Metropolis-like criterion to accept or reject the structure candidate is applied based on the free energy corresponding to the 3D-RISM Hamiltonian at each specified time propagation interval. The resulting ensemble of structures satisfies the 3D-RISM Hamiltonian-based equilibrium distribution. The method was applied to alanine dipeptide and methionine-enkephalin polypeptides in aqueous solution as test cases. The results demonstrated the efficiency of the method.

Spectroscopic behavior differences between lumazine and alloxazine in the DMSO-water mixture

The absorption and emission spectra were investigated for lumazine, alloxazine and their cyanated or fluorinated derivatives, respectively. The spectroscopic properties were modulated by varying water concentration in dimethyl sulfoxide (DMSO). Some intriguing experimental results were found for the samples containing 65 % of water and 35% of DMSO. This finding is consistent with previously published molecular dynamics (MD) simulations confirming the concept of the ‘local bulk’ model. In this case, a notable decrease in absorption and emission intensities was registered, even larger than the water quenching observed in other cases. The changes in midrange DMSO concentrations could be explained by the formation of local solvents structures as predicted by MD, specifically the formation of DMSO·2H2O dimers. Experimentally, the cyano-substituted lumazine has shown a remarkable sensitivity to DMSO concentration. The spectroscopic measurements were interpreted using the density functional theory where the implicit DMSO solvent model was combined with explicit water molecules. Together with its enhanced water solubility, the cyanated lumazine derivate could be used for non-destructive DMSO detection in vitro for applications such as drug uptake monitoring, since DMSO is often used in pharmaceutical practice.

Molecular Dynamics of Iron Porphyrin-Catalyzed C-H Hydroxylation of Ethylbenzene.

Quasi-classical molecular dynamics (MD) simulations were carried out to study the mechanism of iron porphyrin-catalyzed hydroxylation of ethylbenzene. The hydrogen atom abstraction from ethylbenzene by iron-oxo species is the rate-determining step, which generates the radical pair of iron-hydroxo species and the benzylic radical. In the subsequent radical rebound step, the iron-hydroxo species and benzylic radical recombine to form the hydroxylated product, which is barrierless on the doublet energy surface. In the gas-phase quasi-classical MD study on the doublet energy surface, 45% of the reactive trajectories lead directly to the hydroxylated product, and this increases to 56% in implicit solvent model simulations. The percentage of reactive trajectories leading to the separated radical pair is 98-100% on high-spin (quartet/sextet) energy surfaces. The low-spin state reactivity dominates in the hydroxylation of ethylbenzene, which is dynamically both concerted and stepwise, since the time gap between C-H bond cleavage and C-O bond formation ranges from 41 to 619 fs. By contrast, the high-spin state catalysis is an energetically stepwise process, which has a negligible contribution to the formation of hydroxylation products.

Effects of angular momentum and thermostat on molecular dynamics simulation of single-chain conformation in an implicit solvent

The effect of angular momentum on single-chain conformation in an implicit solvent model was investigated by molecular dynamics simulation. The distribution of the end-to-end distance depends on whether the angular momentum of the entire molecule is fixed or thermalized, and these two distributions can be transformed from the one to the other through the reweighting method. In addition, depending on the thermostat, the thermalization of the angular momentum becomes insufficient even if no constraint is imposed on the angular momentum, resulting in the distortion of the distribution of chain conformation. These results suggest the importance of the choice of thermostat that enables the thermalization of the angular momentum.

Performance of the COSMO solvation model for photoacidity and basicity in water.

The possibilities and problems to predict excited-state acidities and basicities in water with electronic structure calculations combined with a continuum solvation model are investigated for a test set of photoacids and photobases. Different error sources, like errors in the ground-state values, the excitation energies in solution for the neutral and (de-)protonated species, basis set effects, and contributions beyond implicit solvation are investigated and their contributions to the total error in are discussed. Density functional theory in combination with the conductor like screening model for real solvents and an empirical linear Gibbs free energy relationship are used to predict the ground-state values. For the test set, this approach gives more accurate values for the acids than for the bases. Time-dependent density-functional theory (TD-DFT) and second-order wave function methods in combination with the conductor like screening model are applied to compute excitation energies in water. Some TD-DFT functionals fail for several species to predict correctly the order of the lowest excitations. Where experimental data for absorption maxima in water is available, the implicit solvation model leads with the applied electronic structure methods in most cases for the excitation energies in water to an overestimation for the protonated and to an underestimation for the deprotonated species. The magnitude and sign of the errors depend on the hydrogen bond donating and accepting ability of the solute. We find that for aqueous solution this results generally in an underestimation in the changes from the ground to the excited state for photoacids and an overestimation for photobases.

Effect of the solvent quadrupolarizability on the strength of the hydrogen bond: Theory vs data for the Gibbs energy and enthalpy of homo- and heteroassociation between carboxylic acids and water.

A cavity model of the effect of a solvent on thermodynamic parameters of dimerization of polar species in non-polar liquids has been developed and compared to experimental data. Bulk solution data have been collected for stearic acid in cyclohexane and in toluene to quantify the extent of self-association of the acid in terms of the dimer self-dissociation constant, Kd. Composition and temperature-dependent experimental data have been collected to determine Kd, the enthalpy of dissociation, and temperature-dependent infrared molar absorption coefficients. The interaction of stearic acid with small amounts of water present in non-aqueous solvents is also addressed and quantified with a hetero-dissociation (or dehydration) constant, Kh. Existing data for acetic acid are also considered. The model connects Kd and Kh to the vapor-phase association equilibria. Solute dipole-solvent quadrupole interactions are shown to have a major effect on Kd in quadrupolar liquids, such as toluene, benzene, and CS2. This work provides important background as a prelude to adsorption studies of these additives from non-aqueous solvents to solid surfaces with relevance to commercial fluids, such as oil-based corrosion inhibitors and friction modifiers. Moreover, the presented theory of the solvent effect on Kd is a first step to generalization of standard implicit solvent models in computational chemistry (such as the polarizable continuum model) to media of significant quadrupolar strength. This is expected to be particularly important for polar species dissolved in CO2 relevant for carbon capture and storage where appropriate models do not currently exist.

Theoretical Study on the Mechanism of CO* Electrochemical Reduction on Cu(111) under Constant Potential

In order to understand the mechanism of the electrochemical reduction of CO* on Cu(111), which competitively generates two intermediates, CHO* and COH*, we performed a first principles calculation on these two electrocatalytic reactions, including the solvent effect and imposing a constant potential. The transition states of the two reactions under constant potential conditions were located by the electrochemical nudged elastic band (eNEB) method and the charge effect in the two reactions was elucidated by charge correction in the constant potential model and Bader charge analysis. It was found that the two reactions have different potential dependencies and involve different degrees of partial charge transfer. In addition, we confirmed that the COH* reaction pathway exists only in a certain potential range when using the implicit solvent model.

Pharmacophore Oriented MP2 Characterization of Charge Distribution for Anti-SARS-CoV-2 Inhibitor Nirmatrelvir

Quantum mechanical second order Møller–Plesset (MP2) perturbation theory and density functional theory (DFT) Becke, 3-parameter, Lee-Yang-Parr (B3LYP) and Minnesota 2006 local functional (M06L) calculations were performed to optimize structure of nirmatrelvir and compute the Merz-Kollman electrostatic potential (MK ESP), natural population analysis (NPA), Hirshfeld, charge model 5 (CM5), and mulliken partial charges. The mulliken partial charge distribution of nirmatrelvir exhibits a poor correlation with the MK ESP charges in MP2, B3LYP, and M06L calculations respectively. The NPA, Hirshfeld, and CM5 partial charge scheme of nirmatrelvir indicate a reasonable correlation with MK ESP charge assignments in B3LYP and M06L calculations. The above correlations were not improved by the inclusion of implicit solvation model. The MK ESP and CM5 partial charges show a strong correlation between the results of MP2 and two DFT methods. The three optimized structures present a certain degree of differences from the crystal bioactive conformation of nirmatrelvir, suggesting the nirmatrelvir-enzyme complex is formed in the induced-fit model. The Reactivity of warhead electrophilic nitrile is justified by the relatively weaker strength of π bonds in the MP2 calculations. The nirmatrelvir hydrogen bond acceptors consistently show strong delocalization of lone pair electrons in three calculations, whereas hydrogen bond donors are found to have high polarization on the heavy nitrogen atoms in MP2 computations. This work helps to parametrize the force field of nirmatrelvir and improve accuracy of molecular docking and rational inhibitor design.

The shape of water – How cluster formation provides a unifying explanation of water’s anomalous properties

Water, essential to life on Earth, has many anomalous properties, that are not yet fully understood. We used density functional theory, the COSMO-RS implicit solvent model and statistical thermodynamics to calculate the dynamic equilibrium of formation of water clusters in water. Clusters of certain shape are predicted to be thermodynamically stable at ambient down to supercooled conditions and their presence almost quantitatively explains water’s anomalous properties as a function of temperature and pressure. The dodecahedron is the dominant cluster below ∼235 K and below ∼150 MPa and forms a miscibility gap with water below a critical point at ∼230 K. The hexagonal torus is the dominant cluster above ∼235 K and below ∼75 MPa. Both the dodecahedron and the torus are low-density clusters and their presence explains the density maximum of pure water at 277 K (+4 °C). The waters in these clusters are tetrahedrally coordinated, their structures are consistent with experimental data and their presence explains the observed tetrahedral fraction as a function of temperature. A significant part of the overall thermodynamic stability comes from the configurational entropy of the hydrogen bond network in the clusters, so dynamic cluster formation and reformation would be observed as fluctuations in the water structure, rather than stable clusters. At pressures above ∼100 MPa, many of water’s anomalous properties diminish or vanish, which is also the pressure above which high-density water clusters become more stable and prevalent. Our simple unifying 3-cluster theory reproduces water’s anomalous properties from supercooled to ambient temperatures and from ambient pressure up to at least 400 MPa, including density anomalies, two-liquid behavior, compressibility, and heat capacity.

An Electrostatically Embedded QM/MM Scheme for Electrified Interfaces.

Atomistic modeling of electrified interfaces remains a major issue for detailed insights in electrocatalysis, corrosion, electrodeposition, batteries, and related devices such as pseudocapacitors. In these domains, the use of grand-canonical density functional theory (GC-DFT) in combination with implicit solvation models has become popular. GC-DFT can be conveniently applied not only to metallic surfaces but also to semiconducting oxides and sulfides and is, furthermore, sufficiently robust to achieve a consistent description of reaction pathways. However, the accuracy of implicit solvation models for solvation effects at interfaces is in general unknown. One promising way to overcome the limitations of implicit solvents is going toward hybrid quantum mechanical (QM)/molecular mechanics (MM) models. For capturing the electrochemical potential dependence, the key quantity is the capacitance, i.e., the relation between the surface charge and the electrochemical potential. In order to retrieve the electrochemical potential from a QM/MM hybrid scheme, an electrostatic embedding is required. Furthermore, the charge of the surface and of the solvent regions has to be strictly opposite in order to consistently simulate charge-neutral unit cells in MM and in QM. To achieve such a QM/MM scheme, we present the implementation of electrostatic embedding in the VASP code. This scheme is broadly applicable to any neutral or charged solid/liquid interface. Here, we demonstrate its use in the context of GC-DFT for the hydrogen evolution reaction (HER) over a noble-metal-free electrocatalyst, MoS2. We investigate the effect of electrostatic embedding compared to the implicit solvent model for three contrasting active sites on MoS2: (i) the sulfur vacancy defect, which is rather apolar; (ii) a Mo antisite defect, where the active site is a surface bound highly polar OH group; and (iii) a reconstructed edge site, which is generally believed to be responsible for most of the catalytic activity. According to our results, the electrostatic embedding leads to almost indistinguishable results compared to the implicit solvent for the apolar system but has a significant effect on polar sites. This demonstrates the reliability of the hybrid QM/MM, electrostatically embedded solvation model for electrified interfaces.

Infrared and vibrational circular dichroism spectra of methyl β-D-glucopyranose in water: The application of the quantum cluster growth and clusters-in-a-liquid solvation models.

The infrared (IR) and vibrational circular dichroism (VCD) spectra of methyl β-D-glucopyranose in water were measured. Both implicit and explicit solvation models were utilized to explain the observed spectra. The vast body of existing experimental and theoretical data suggested that about eight explicit water molecules are needed to account for the solvent effects, supported by the current Quantum Cluster Growth (QCG) analysis. Extensive manual and systematic conformational searches of the molecular target and its water clusters were carried out by using a recently developed conformational searching tool, conformer-rotamer ensemble sampling tool (CREST), and the microsolvation model in the associated QCG code. The Boltzmann averaged IR and VCD spectra of the methyl β-D-glucopyranose-(water)n (n = 8) conformers in the PCM of water provide better agreement with the experimental ones than those with n = 0, 1, and 2. The explicit solvation with eight water molecules was shown to greatly modify the conformational preference of methyl β-D-glucopyranose from its monomeric form. Further analyses show that the result is consistent with the existence of long-lived methyl β-D-glucopyranose monohydrates with the additional explicit water effects being accounted for with the quantum mechanical treatment of the other seven close-by water molecules in the PCM of water.

Entropic Origin of Ionic Interactions in Polar Solvents.

Implicit solvent models that reduce solvent degrees of freedom into effective interaction potentials are widely used in the study of soft materials and biophysical systems. For electrolyte and polyelectrolyte solutions, coarse-graining the solvent degrees of freedom into an effective dielectric constant embeds entropic contributions into the temperature dependence of the dielectric constant. Properly accounting for this electrostatic entropy is essential to discern whether a free energy change is enthalpically or entropically driven. We address the entropic origin of electrostatic interactions in a dipolar solvent and provide a clarified physical picture of the solvent dielectric response. We calculate the potential of mean force (PMF) between two oppositely charged ions in a dipolar solvent using molecular dynamics and dipolar self-consistent field theory. We find with both techniques that the PMF is dominated by the entropy gain from the dipole release, owing to the diminished orientational polarization of the solvent. We also find that the relative contribution of the entropy to the free energy change is nonmonotonic with temperature. We expect that our conclusions are applicable to a broad range of problems involving ionic interactions in polar solvents.

Accurate Boundary Integral Formulations for the Calculation of Electrostatic Forces with an Implicit-Solvent Model

An accurate force calculation with the Poisson-Boltzmann equation is challenging, as it requires the electric field on the molecular surface. Here we present a calculation of the electric field on the solute-solvent interface that is exact for piecewise linear variations of the potential and analyze four different alternatives to compute the force using a boundary element method. We performed a verification exercise for two cases: the isolated and two interacting molecules. Our results suggest that the boundary element method outperforms the finite difference method, as the latter needs a much finer mesh than in solvation energy calculations to get acceptable accuracy in the force, whereas the same surface mesh as in a standard energy calculation is appropriate for the boundary element method. Among the four evaluated alternatives of force calculation, we saw that the most accurate one is based on the Maxwell stress tensor. However, for a realistic application, like the barnase-barstar complex, the approach based on variations of the energy functional, which is less accurate, gives equivalent results. This analysis is useful toward using the Poisson-Boltzmann equation for force calculations in applications where high accuracy is key, for example, to feed molecular dynamics models or to enable the study of the interaction between large molecular structures, like viruses adsorbed onto substrates.

Photophysical and photochemical properties of 3-hydroxyflavone in ethanol solution: Implicit vs explicit solvent models

Photophysical and photochemical properties of 3-hydroxyflavone in gas-phase and solution were studied using implicit (Polarizable Continuum Model, PCM) and explicit solvent (Averaged Solvent Electrostatic Potential from Molecular Dynamics calculations, ASEP/MD) models. The conformational equilibrium in the excited state between the normal (N*) and tautomeric (T*) forms and their absorption and fluorescence spectra was studied with Time-Dependent Density Functional Theory (TD-DFT) and three different functionals. The calculated transitions are in good agreement with the spectroscopic data. It was found that solvent effects on the absorption and fluorescence spectral bands are negligible. Still, they can modify appreciably the relative stability of the N* minimum with respect to the Franck-Condon position, which affects the kinetics of the reaction. This fact, together with the increase in the barrier height in protic solvents, permits us to explain the reduction of the emission signal from T* and the increase of the N* signal in ethanol solution.

Interfacial water molecules contribute to antibody binding to the receptor-binding domain of SARS-CoV-2 spike protein

Antibodies that recognize the spike protein of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), especially the neutralizing antibodies, carry great hope in the treatment and final elimination of COVID-19. Driven by a synchronized global effort, thousands of antibodies against the spike protein have been identified during the past two years, with the structural information available at atomistic detail for hundreds of these antibodies. We developed an improved molecular mechanics/Poisson–Boltzmann surface area (MM/PBSA) method including explicitly treated interfacial water to calculate the binding free energy between representative antibodies and the receptor binding domain (RBD) domain of SARS-COV-2 spike proteins. We discovered that explicit treatment of water molecules located at the interface between RBD and antibody effectively improves the results for the WT and variants of concern (VOC) systems. Interfacial water molecules, together with surface and internal water molecules, behave drastically from bulk water and exert peculiar impacts on protein dynamics and energy, and thus warrant explicit treatment to complement implicit solvent models. Our results illustrate the importance of including interfacial water molecules to approach efficient and reliable prediction of binding free energy. Communicated by Ramaswamy H. Sarma

Machine learning of an implicit solvent for dynamic Monte Carlo simulations.

The Bond Fluctuation Model (BFM) is a highly efficient and versatile method for simulating polymers, membranes, and soft matter. Due to its coarse-grained nature, the BFM is employed to understand the universal properties of polymers. Solvent effects are often mediated by explicit solvent particles, while implicit solvent models have had limited use as they may lead to frozen states and, thus, ergodicity-related problems. In simulation setups, such as coagulated multiple homopolymers chains, explicit solvent models are computationally expensive because the region of interest can be localized in a small space compared to the dimension of the periodic box. We introduce an implicit solvent model based on an artificial neural network (NN) that was trained with BFM simulation data for single homopolymers in an explicit solvent. We demonstrate that NN-based simulations that take into account only the information of the local environment of monomers reproduce the expected universal macroscopic properties of the polymer under varying solvent conditions. The homopolymer chains simulated using the NN reproduce the coil-globule transition, the static and dynamic bond autocorrelation, and the mean square displacement of chain monomers. We show that the learned parameters from a single chain system can be transferred to a system containing multiple homopolymers, indicating that the learned parameters are transferable to considerably different systems.

Comprehensive Evaluation of End-Point Free Energy Techniques in Carboxylated-Pillar[6]arene Host-Guest Binding: III. Force-Field Comparison, Three-Trajectory Realization and Further Dielectric Augmentation.

Host-guest binding, despite the relatively simple structural and chemical features of individual components, still poses a challenge in computational modelling. The extreme underperformance of standard end-point methods in host-guest binding makes them practically useless. In the current work, we explore a potentially promising modification of the three-trajectory realization. The alteration couples the binding-induced structural reorganization into free energy estimation and suffers from dramatic fluctuations in internal energies in protein-ligand situations. Fortunately, the relatively small size of host-guest systems minimizes the magnitude of internal fluctuations and makes the three-trajectory realization practically suitable. Due to the incorporation of intra-molecular interactions in free energy estimation, a strong dependence on the force field parameters could be incurred. Thus, a term-specific investigation of transferable GAFF derivatives is presented, and noticeable differences in many aspects are identified between commonly applied GAFF and GAFF2. These force-field differences lead to different dynamic behaviors of the macrocyclic host, which ultimately would influence the end-point sampling and binding thermodynamics. Therefore, the three-trajectory end-point free energy calculations are performed with both GAFF versions. Additionally, due to the noticeable differences between host dynamics under GAFF and GAFF2, we add additional benchmarks of the single-trajectory end-point calculations. When only the ranks of binding affinities are pursued, the three-trajectory realization performs very well, comparable to and even better than the regressed PBSA_E scoring function and the dielectric constant-variable regime. With the GAFF parameter set, the TIP3P water in explicit solvent sampling and either PB or GB implicit solvent model in free energy estimation, the predictive power of the three-trajectory realization in ranking calculations surpasses all existing end-point methods on this dataset. We further combine the three-trajectory realization with another promising modified end-point regime of varying the interior dielectric constant. The combined regime does not incur sizable improvements for ranks and deviations from experiment exhibit non-monotonic variations.

Theoretical and experimental study of the diastereoisomers (2S) and (2R)-naringenin-6-C-β-D-glucopyranoside obtained from Clitoria guianensis

In this work the diastereoisomers (2S) and (2R)-naringenin-6-C-β-D-glucopyroside, isolated for the first time from Clitoria guianensis, were studied using the density functional theory. The frontier molecular orbitals and structural properties showed that the diastereoisomers exhibit the same energy gap 166.61kcalmol-1 and structural properties different, where in the S diastereoisomer, the bond length between the chiral carbon and the phenolic group is greater (difference of 0.0126Å). The HPLC data showed that the retention time of the S-diastereoisomer (16.7min) is shorter than that of R, suggesting that the S compound is more polar than R. The HPLC results corroborates with the molecular electrostatic potential which showed that in the S configuration, the electronegative density was more intense overall, particularly in the glucose molecule. The reactivity indices showed that the diastereoisomers are good electrophiles and reactive species. Finally, the absolute configuration of the diastereoisomers were determined using electronic circular dichroism (ECD) spectroscopy and the theoretical spectra were similar to the experimental. METHODS : All calculations of Density Functional Theory (DFT) and Time-Dependent Density Functional Theory (TDDFT) were performed using the program Gaussian 09 and the structures of the diastereoisomers were generated and analyzed using the GaussView program. The optimization and vibrational frequency calculations were performed using the functional CAM-B3LYP and 6-311 + + G(2d,2p) basis set. Conformational searches were performed for R configuration, by molecular mechanics using the MM + , MMFF, and OPLS05 force fields; the entire molecular mechanics simulation was performed using the Maestro/MacroModel software. The calculations for the simulations of the ECD spectra were performed for the eight lowest energy conformers obtained in the geometric optimization step, and the TDDFT at the CAM-B3LYP/6-311 + + G(2d,2p) theory level used. The effects of methanol and chloroform were calculated using the SMD implicit solvent model.

Solvation effects on glyphosate protonation and deprotonation states evaluated by mass spectrometry and explicit solvation simulations.

Glyphosate is a widely used herbicide, and its protonation and deprotonation sites are fundamental to understanding its properties. In this work, the sodiated, protonated, and deprotonated glyphosate were evaluated in the gas phase by infrared multiple photon dissociation spectroscopy to determine the exact nature of these coordination, protonation, and deprotonation states in the gas phase. In this context, Natural Bond Orbital analyses were carried out to unravel interactions that govern glyphosate (de)protonation states in the gas phase. The solvent effect on the protonation/deprotonation equilibria was also investigated by implicit (Solvation Model Based on Density and polarizable continuum models) and explicit solvation models (Monte Carlo and Molecular Dynamics simulations). These results show that glyphosate is protonated in the phosphonate group in the gas phase because of the strong hydrogen bond between the carboxylic oxygen (O7) and the protonated phosphonate group (O8-H19), while the most stable species in water is protonated at the amino group because of the preferential interaction of the NH2 + group and the solvent water molecules. Similarly, deprotonated glyphosate [Glyp-H]- was shown to be deprotonated at the phosphonate group in the gas phase but not in solution, also because of the preferential solvation of the NH2 + group present in the other deprotomers. Therefore, these results show that the stabilization of the protonated amino group by the solvent molecules is the governing factor of the (de)protonation equilibrium of glyphosate in water.