Born-Oppenheimer Molecular Dynamics Simulations of a Bromate Ion

Subscriber access provided by READING UNIV

Article

Born-Oppenheimer Molecular Dynamics Simulations of a Bromate Ion in Water Reveal Its Dual Kosmotropic and Chaotropic Behavior Bikramjit Sharma, and Amalendu Chandra J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.7b09300 • Publication Date (Web): 27 Jan 2018 Downloaded from http://pubs.acs.org on January 29, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Born-Oppenheimer Molecular Dynamics Simulations of a Bromate Ion in Water Reveal its Dual Kosmotropic and Chaotropic Behavior

Bikramjit Sharma and Amalendu Chandra∗ Department of Chemistry, Indian Institute of Technology Kanpur, India 208016.

———————————————————— ∗ Corresponding author. E-mail: [email protected], Tel. +91 512 2597241

1 Environment ACS Paragon Plus

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The solvation structure and dynamics of a bromate (BrO− 3 ) ion in water are studied by means of Born-Oppenheimer molecular dynamics simulations at two different temperatures using the BLYP (Becke-Lee-Yang-Parr) functional with Grimme D3 dispersion corrections. The bromate ion possesses a pyramidal structure and it has two types of solvation sites, namely the bromine and the oxygen atoms. We have looked at different radial and orientational distributions of water molecules around the bromate ion and also investigated their hydrogen bonding properties. The solvation structure of the bromate ion is also compared with that of the iodate (IO− 3 ) ion which is structurally rather similar to the bromate ion and was found to have some unusual solvation properties in water. It is found that the bromate ion follows a similar trend as the iodate ion as far as the solvation structure is concerned. However, the effect of the former on surrounding water is found to be much weaker than that of the latter. On the dynamical side, we have looked at diffusion, residence dynamics and also the orientational and hydrogen bond relaxation of water molecules around the BrO− 3 ion and compared them with those of the bulk. Dynamical results are presented for both H2 O and D2 O around the BrO− 3 ion. Interpretation of the dynamical results in terms of structure making (kosmotropic)/breaking (chaotropic) properties of the BrO− 3 ion reveals that the bromine atom of this ion acts as a water structure breaker while the three oxygens act as water structure makers. Thus, in spite of being a single ion, the bromate ion has dual characteristics and the experimentally observed kosmotropic ability of this ion is actually a trade off between a chaotropic site (the bromine atom) and three kosmotropic sites (three oxygen atoms) that are present in the ion.


Page 2 of 42

Page 3 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Introduction

Polyoxy-anions play important roles in many chemical, biological as well as environmental processes. Water being the universal solvent, the hydration of polyoxy anions needs to be understood well in order to unveil the roles of these anions in various processes. The bromate (BrO− 3 ) ion is a polyoxy-anion with pyramidal shape. The three oxygen atoms of this anion occupy the three base points while the bromine atom occupies the pinnacle of the pyramid. In the present work, the solvation structure and dynamics of a BrO− 3 ion in water is studied by means of Born-Oppenheimer molecular dynamics 1 simulations. The BrO− 3 ion is known be a weak structure maker in water with a positive value of Jones-Dole B-coefficient 2 . The effects of BrO− 3 ion on surrounding water can be complicated because of its pyramidal shape with two types of solvation sites; namely the bromine and the oxygen atoms. Thus, deciphering the effects of the two types of solvation sites of the BrO− 3 ion on surrounding water will be interesting. Further, understanding the solvation properties of BrO− 3 ion can have important implications in human health. This anion is a disinfection byproduct of ozonation of water containing bromide ions. The BrO− 3 ion has been found to be carcinogenic and thus potentially fatal. Hence, removal of this anion from water is very important. Various methods have been developed for the removal of bromate ions from water e.g. semiconductor photocatalysis 3 , adsorption on activated granular carbon 4 , electrochemical reduction 5 etc. Understanding the solvation and dynamical behavior of BrO− 3 ion in water can provide information which may be crucial in designing methods for its removal from water. Another motivation for studying the solvation properties of BrO− 3 anion comes from the interesting properties of the closely related iodate ion in water 6 . Since the iodate and bromate ions have similar structure, it would be interesting to investigate the behavior of BrO− 3 ion in water. The iodate ion was found to show some unusual behavior in water, e.g. it provides a local cationic region due to the positively charged iodine atom 6 . The water molecules around the iodate ion arrange themselves in two distinct regions giving rise to two solvation shells of different characteristics: One due to the effect of the positively charged iodine atom while the other due to effects of the three negatively charged oxygen atoms of the iodate ion on the surrounding water molecules. Apart from similarity in 3 Environment ACS Paragon Plus


the shape, there are other matching properties of iodate and bromate ions. For example, − 2 both have positive values of Jones-Dole B-coefficient (0.009 for BrO− 3 and 0.140 for IO3 )

meaning that both the anions act as water structure makers. However, this behavior is − rather unexpected of such large anions. The BrO− 3 ion differs from the IO3 ion mainly

in terms of the properties of the bromine and iodine atoms. The bromine atom has a smaller size, higher electronegativity and lower polarizability as compared to the iodine atom. Thus, quantification of solvation properties of the bromate and iodate ions can provide useful information on the effects of these basic atomic properties on the solvation characteristics of these polyoxy-anions. There are few studies involving the BrO− 3 ion in bulk as well as interfacial aqueous systems. A recent study involving Monte-Carlo simulations 7 determined the surface tension of various electrolyte-air interfaces by modeling the solvation of ions at the interfaces as an ion inside a spherical shell at the position of the Gibbs dividing surface. The spherical shell represented both the water and air as uniform dielectric medium of permittivity of 80.0 and 1.0 respectively. The calculated value of the surface tension for the BrO− 3 ion was found to be consistent with the fact that this anion is weakly hydrated at the interface. Another study 8 involving quantum mechanical charge field (QMCF) and X-ray diffraction techniques looked at the solvation of BrO− 3 ion in bulk water. A series of polyoxy-anions were investigated to look at some of the structural and dynamical aspects of their solvation shells. The QMCF method used Hartree-Fock electronic structure calculations to represent the ions and water molecules around them while all other water molecules were treated with classical force field. In the current study, we have treated the entire system under the quantum density functional theory framework and looked at the solvation shell structure and dynamics. We have also drawn comparisons of the results of − 6 solvation structure of the BrO− 3 ion with those of IO3 ion as found in earlier study . Such

comparisons give information on the effects of atomic properties of the bromine and iodine atoms on the solvation structures of these pyramidal polyoxy-anions. The dynamics of water molecules around the BrO− 3 ion is also investigated in the current work and the findings are connected to its water structure making/breaking ability. Our results show that the bromine atom of the BrO− 3 ion accelerates the dynamics of its surrounding water while its oxygen atoms retard the dynamics of surrounding water molecules. Extrapolat-


Page 4 of 42

Page 5 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ing these dynamical trends to the effects of BrO− 3 ion on water structure suggests that the BrO− 3 ion has both kosmotropic (the oxygen atoms) and chaotropic (the bromine atom) sites. Thus, it has dual characteristics although overall it is a water structure maker as suggested by the positive value of the Jones-Dole B-coefficient.

The rest of the paper is organized as follows. The details of simulations are given in Section 2. Section 3 deals with the structure and dynamics of water molecules around the BrO− 3 ion. Dynamical results for both D2 O and H2 O are discussed in Section 3. In this section, we have also discussed the connection of dynamics of water around the BrO− 3 ion to its water structure making/breaking ability. Finally, our conclusions are summarized in Section 4. Apart from the results presented in the main text, some additional results of the binding energy, structure and dynamics are also included in the Supporting Information.

2

Computational Details

Our simulation system consists of one BrO− 3 ion dissolved in 107 water molecules in a cubic box of edge length 14.8 ˚ A. BOMD simulations were carried out using the CP2K code 9 and the electronic structure calculations were done using the QUICKSTEP 10 module. The Kohn-Sham orbitals were expanded using Gaussian basis sets, namely the DZVPMOLOPT basis set for the bromine atom and TZV2P-MOLOPT basis set for the oxygen and hydrogen atoms. The plane waves used for calculation of the electron density were truncated at a kinetic energy cut-off of 400 Ry. It is worth mentioning that the dual basis approach of the Quickstep module has certain computational advantages. In this approach, the wave functions are expanded in atom-centered Gaussian-type basis while the density is expressed in terms of auxiliary plane wave basis 10 . The dual representation has the advantage of treating the electrostatic calculations in more efficient manner and it also enables linear scaling for the calculations of total energy and Kohn-Sham energy with respect to system size 10 . We note in this context the recent work of Raynaud et al. 11 who considered hydrogen fluoride clusters ((HF)n ) as prototypes to study the differences in results obtained by using the plane wave and localized atomic orbitals as basis functions. The localized atomic orbitals were found to give results in good agreement with the experimental results whereas, for the plane wave basis, the electronic energies and infrared frequencies were found converge slowly with increase of the cut-off 5 Environment ACS Paragon Plus


parameter. In another recent study, Miceli et al. 12 compared structural, dynamical and electronic properties of liquid water using both plane wave and atomic orbital basis sets at ambient conditions. Both types of basis sets with their chosen cut-offs were found to give similar results for liquid. Hence, it was concluded that either of the two types of basis functions can be applied for liquid water with appropriate cut-offs. The present plane wave cut-off of 400 Ry for the expansion of electron density has been established to produce converged results in the NVE and NVT ensembles and was employed in many earlier studies 13–17 . In order to further check the reliability of the current plane wave cutoff of 400 Ry, we performed a separate calculation of the geometry optimization of water dimer and calculated the binding energy for cut-offs of 400 and 500 Ry. These results of the binding energies are included in Table S1 of the Supporting Information. As can be seen, the results differ by less than 0.5% on increase of the basis set size from 400 to 500 Ry, hence we chose to stay with the cut-off of 400 Ry for our subsequent calculations. We note that the current simulations are done in the NVT and NVE ensembles like many of the earlier work where similar cut-off was used for simulations of aqueous systems using the same CP2K code 13–17 . It is also important to note in this context that in case of NPT ensemble, a larger cut-off is needed to produce the converged results 18,19 . We have used the BLYP-D3 functional which includes dispersion corrections at D3 level 20,21 into the BLYP functional 22,23 . It is known from earlier studies that the BLYP functional predicts somewhat over structured water with a slower dynamics at the experimental density 24–29 . The BLYP-D3 functional, on the other hand, was shown to provide significantly improved results for the structure, dynamics and surface properties of water 13–15,30,31 . Unlike BLYP, the BLYP-D3 functional was also shown to provide improved liquid density of water under normal pressure and room temperature 32 . The simulations of the current study were carried out at two different temperatures of 300 and 330 K. The higher temperature of 330 K was chosen because ab initio simulation of water at room temperature with BLYP-D3 functional is known to produce dynamical results which are, although much improved over the BLYP results, still slower compared to the experimental results 33,34 . However, ab initio simulation of water using a combination of BLYP-D3 functional and a temperature elevation of about 30 K above the room temperature was found to produce dynamical properties of water which are in better agreement with experiments at room temperature 13,14 . It may also be noted that BLYP-D3 at room temperature and


Page 6 of 42

Page 7 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


normal pressure gives a slightly higher density of 1.07 g cm−3 32,35 which again supports the use of a slightly elevated temperature for water simulation even when a dispersion corrected density functional such as BLYP-D3 is used. The deuterium mass was assigned to all the hydrogen atoms, thus the current simulation systems actually correspond to BrO− 3 ion in heavy water (D2 O). The systems were equilibrated for 15 ps at both the temperatures in the NVT ensemble where the temperature was maintained by applying the method of canonical sampling through velocity rescaling (CSVR) 36 using a time constant of 20 fs. Then, the production runs were carried out in the NVE ensemble for 72 ps for both the systems.

We note that the two simulation systems described above deal with heavy water (D2 O) rather than H2 O. The present Born-Oppenheimer molecular dynamics (BOMD) simulations do not include the nuclear quantum effects, hence the differences in the properties of D2 O and H2 O would arise from the difference in masses of the deuterium and hydrogen atoms in their classical dynamics. While this mass difference would not affect the structural properties at the current level of calculations provided simulations are run long enough, it would, of course, influence the dynamical properties to some extent. However, the difference in the dynamical properties would be systematic in nature and, hence, the essential physics revealed from the current simulations would remain the same for water and heavy water. For example, purely due to mass difference, the stretching frequency of OD discussed later in Sec. 3.2 should differ from that of OH by a factor of approximately 0.73 as can be obtained from the calculations of reduced mass. Thus, the results of the current study can be extended to understand the solvation of bromate ion in liquid H2 O also. Still, we have performed an additional simulation of the bromate ion in liquid H2 O at 330 K with the same system size, functional and time step as discussed above. This additional simulation was run for a shorter period of 28 ps in NVE ensemble after equilibration of 15 ps in NVT ensemble. We calculated some of the dynamical properties for water (H2 O) from the trajectory of this additional simulation and included them in the Supporting Information. The structural correlations between the bromate ion and oxygen and hydrogen atoms of water are also calculated and compared with those for D2 O to verify the equality of the calculated correlations for the two isotopic solvents. Further discussions of the results of H2 O are presented in later sections. Expectedly,



the results of H2 O are found to follow the same trend as D2 O. Hence, we have used the longer trajectories of BrO− 3 in D2 O systems for our main discussions of the solvation shell structure and dynamics of this polyoxy-anion in aqueous medium.

3 3.1

Results and Discussion Structural Properties of Solvation Shell Water Molecules

First, we have looked at the radial distribution of oxygen atoms of water (OW) with respect to the bromine atom (Br) of the BrO− 3 ion (Fig.1a). The Br-OW radial distribution function (RDF) is found to have a peak at 3.8 ˚ A and a small shoulder at around 3 ˚ A. However, in case of the iodate ion, two distinct well separated peaks were observed at 4˚ A and 2.9 ˚ A in the I-OW RDF 6 . Thus, there is a sharp contrast between the solvation shell structures of the bromate and iodate ions. In case of solvation of an iodate ion, the peak at 2.9 ˚ A was found to arise from the arrangement of water molecules adjacent to the positively charged iodine atom 6 . In order to shed light on the origin of the small shoulder at 3 ˚ A in the Br-OW RDF, we have compared it with the RDF of the center of mass (COM) of the BrO− 3 ion and OW atoms. Fig.1a shows this comparison where two differences are observed: First, there is no shoulder in the COM-OW RDF and, secondly, it starts at around 3 ˚ A while the Br-OW RDF starts at around 2.5 ˚ A. Thus, the shoulder in the Br-OW RDF is a characteristic feature of the correlation between the Br and OW atoms. The next question that arises is regarding which water molecules contribute to this shoulder. In order to find an answer, we have calculated the incremental radial distribution functions (IRDFs) (Fig.1b) for the first seven nearest water molecules from the Br atom. The IRDFs show that only the first and second nearest water molecules from the Br atom contribute appreciably to the shoulder region at around 3 ˚ A. This also confirms that the shoulder in the Br-OW RDF arises from those water molecules which directly interact with the bromine atom of the BrO− 3 ion. It may be noted that the bromine atom remains exposed to water due to pyramidal shape of the ion. The prominent peak at 3.8 ˚ A in the Br-OW RDF is found to be due to water molecules interacting with the oxygen atoms (OB) of the BrO− 3 ion. For the rest of the paper, we will be using “shoulder water” and “solvation shell water” to represent water molecules corresponding to the shoulder at 3 ˚ A and the peak at 3.8 ˚ A, respectively, in the Br-OW RDF. In order to separate the shoulder


Page 8 of 42

Page 9 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


region water molecules, we have used a cut-off distance of 3.3 ˚ A between the Br and OW atoms. We note here that this cutoff is decided not from the RDF as there is no well defined minimum, rather from the intersection point (3.3 ˚ A) of the Br-OW and COM-OW RDFs (Fig.1a). Also, this is approximately the distance from which the contribution of the third nearest water molecule from the bromine atom starts in the IRDF (Fig.1b). Since the third nearest water contributes primarily to the peak at 3.8 ˚ A (Fig.1b) i.e. the solvation shell region, it is reasonable to take 3.3 ˚ A as the cutoff distance for defining the shoulder region.

The solvation shell water molecules are considered to be those whose oxygen atoms are within the distance range of 3.3 ˚ A−4.4 ˚ A from the bromine atom. We note that some water molecules in this region may not be directly hydrogen bonded to the oxygens of the BrO− 3 ion, but they are part of the solvation shell. Hence, we have considered all the water molecules (irrespective of whether they are hydrogen bonded to the BrO− 3 oxygens or not) in the solvation shell for calculations of different properties. Also, majority of the water molecules in the solvation shell are hydrogen bonded to the oxygen atoms of the BrO− 3 ion and thus these water molecules primarily govern the properties of the solvation shell. Indeed, as discussed in the next section, we have calculated some of the dynamical dynamical properties by both including and excluding the non-hydrogen bonded (to the BrO− 3 oxygens) water molecules in the solvation shell and found no significant differences. The spatial distribution function (SDF) of water molecules around the BrO− 3 ion is shown in Fig.2. The SDF shows that the water molecules in the shoulder region have their oxygen atoms (shown in cyan color) closer to the Br atom than their hydrogen atoms (shown in green). The proximity of the OW atoms with the Br atom in the shoulder region means that the Br atom is positively charged which is similar to the iodine atom of the iodate ion.

In order to see the effects of the bromine atom on the solvation of the oxygen atoms of the BrO− 3 ion or vice-versa, we have calculated the two dimensional radial distribution function or the 2D-RDF (g(r1 ,r2 )). It gives the distribution of a particular type of atoms simultaneously at distances of r1 from one solvation site and r2 from another solvation site of a polyatomic solute. We have calculated the 2D-RDF for the distribution of water oxygen (OW) with respect to the bromine (Br) and an oxygen (OB) atom of the BrO− 3 9 Environment ACS Paragon Plus


ion (Fig.3). The densities at ∼3.0 ˚ A and ∼4.8 ˚ A from the Br and OB atoms, respectively, represent the shoulder region. The diffused density around ∼3.8 ˚ A from the Br atom and ∼2.8 ˚ A from the OB atom corresponds to the solvation shell around the BrO− 3 ion. The 2D-RDF shows the solvation shell water density to be rather diffuse over a wide radial range. Also, the continuous density of OW in between the shoulder and solvation shell regions means that the two regions are interconnected by water.

We have also investigated other important structural properties like the orientational distributions (Figs.4a and 4b) of water molecules and hydrogen bond statistics (Figs.5a and 5b) in the shoulder and solvation shell regions. The orientational profile was investigated by looking at the angle between the dipole vector of the BrO− 3 ion and that of water molecules in the respective regions. It is found that water molecules in the shoulder region have a preferred orientation of ∼120° at both the temperatures. Such an orientation enables the lone pairs of the OW atoms of water molecules in the shoulder region to interact favorably with the positively charged Br atom of the BrO− 3 ion. The distributions of different types of hydrogen bonded water (Figs.5a and 5b) like double donor-double acceptor etc. show strong similarity between the solvation shell and bulk water molecules (Table 1). Available experimental results 37 of the hydrogen bond number of ambient water is also included in Table 1. hydrogen bond distribution differs strongly for water molecules in the shoulder region which exist mostly in the single acceptor-double donor hydrogen bonded state. This is also consistent with the picture that in the shoulder region, the oxygen atoms of water molecules point toward the positively charged Br atom and thus the two water hydrogens are available to donate hydrogen bonds to other water molecules in the bulk.

In order to make quantitative comparison between the solvation structures of the iodate and bromate ions, we have included the solvation shell properties of the two ions in Table 2. The numbers for the iodate ion are taken from Ref. 6 . From this Table, it can be concluded that although some of the solvation shell properties of the BrO− 3 ion follow qualitatively somewhat similar trends as of the IO− 3 ion, many significant differences are also found to exist regarding the structure and hydration numbers of the solvation shells.


Page 10 of 42

Page 11 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.2

Dynamical Properties of Water Molecules Around the Bromate Ion

The presence of ions is known to affect the dynamics of surrounding water molecules. Here we have calculated various dynamical properties of water molecules around the BrO− 3 ion and compared them with those of the bulk water. We have divided the OD bonds in wHB for the different regions into four major types and named them as: TB , TSh , TiHB Sol , TSol

OD modes in the bulk, shoulder region, solvation shell region forming hydrogen bonds with the ion and solvation shell region forming hydrogen bonds with water, respectively. These different types of OD bonds are shown in the illustration of Fig.S1 of the Supporting Information. 3.2.1

Dynamics of D2 O Molecules Around the BrO− 3 Ion

In this subsection, we present the dynamical results of the system of BrO− 3 ion in liquid D2 O. Although the solvent in the system is actually heavy water, we will continue to use the term ’water’ for convenience.

We have looked at the mean square displacement (MSD) as a function of time (Fig.6a) for water molecules around the BrO− 3 ion and in the bulk. The diffusion coefficient is related to the MSD by the relation E 1 D (r(t) − r(0))2 . t→∞ 6t

D = lim

(1)

The diffusion coefficients of water (D2 O) molecules around the BrO− 3 ion and in the bulk region were extracted from the slope of the MSD at long times (Table 3). Available experimental results 38 for bulk D2 O are also included in this Table. The diffusion coefficient of D2 O is found to be maximum for the bulk region followed by those in the solvation shell and then in the shoulder region. We have also calculated the MSD of solvation shell water by both including and excluding the non-hydrogen bonded (to the BrO− 3 oxygens) water molecules and the results are shown in Fig.S2 of the Supporting Information. Clearly, no significant difference is found between the two MSDs. The smaller values of diffusion coefficient of water molecules in the shoulder and solvation shell regions than the bulk may be associated with the longer residence times of water around the BrO− 3 ion. In order to confirm it, we have looked at the residence dynamics of water molecules 11 Environment ACS Paragon Plus


Page 12 of 42

39 around the BrO− 3 ion by following the prescription of Ref. . We define a survival variable

Pi (t0 , t, t∗ ) where Pi = 1 if the ith water molecule resides continuously in the solvation shell of a solute from time t0 to t with an allowance time of t∗ for moving out of the solvation shell in the interim period and Pi = 0 otherwise. The residence time correlation function is then defined as CRES (t, t∗ ) =

hPi (t0 , t, t∗ )ii,t0 , hPi (t0 , t0 , t∗ )ii,t0

(2)

where h....ii,t0 denotes averaging over water molecules and initial times. Although, it has been shown that the choice of t∗ may influence the calculated timescales of residence dynamics to some extent 40 , still the above approach has remained useful for studying the residence dynamics of water molecules in a given region and it has been used in a number of earlier studies 41–43 . Following earlier work, we have taken the allowance time (t∗ ) to be 2 ps for calculation of the residence time correlation function. The results of the residence time correlation function are shown in Fig.6b. We note that the MSD of the shoulder region water molecules was found to be smaller than the bulk water molecules. The slower diffusion of water molecules around a solute usually lead to slower escape from its solvation shell. However, we find that the diffusion of water molecules in the shoulder region is slower but their residence dynamics is faster than those of the bulk. The faster residence dynamics of water molecules in the shoulder region is, however, consistent with the absence of any well defined solvation shell. Rather this shoulder region arises from water molecules interacting directly with the positively charged bromine atom of the BrO− 3 ion. For the solvation shell water molecules, the residence dynamics is found to be slower (Fig.6b) than that of the bulk. This picture is consistent with the MSD trend as well. In order to reconcile for the apparent anomaly between the MSD and residence dynamics of water molecules in the shoulder region, we have examined the molecular dynamics trajectory and found that once a water molecule escapes from the shoulder region, it remains in the vicinity for sometime and then completely reorients itself to make one of its hydrogens available for hydrogen bonding with the oxygen atoms of the BrO− 3 ion. Figs.7a and 7b shows snapshots of such events. Thus, although water molecules escape easily from the shoulder region giving rise to faster residence dynamics, it does not translate far away from the BrO− 3 ion. Hence, the slower diffusion is observed for these water molecules. 12 Environment ACS Paragon Plus

Page 13 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The orientational relaxation of water molecules in the shoulder, solvation shell and bulk regions are investigated by looking at the decay of the orientational correlation functions of dipole (Fig.8a) and OD bond (Fig.8b) vectors of water molecules. The orientational correlation function is defined as Cl (t) =

hPl (ˆ e( t).ˆ e (0))i , hPl (ˆ e( 0).ˆ e (0))i

(3)

where Pl is the Legendre polynomial of rank l and ê(t) is the unit vector at time t along the direction of the vector whose rotation is being observed. The second order orientational time correlation functions of both the dipole and OD bond vectors show (Figs.8a and 8b) that the shoulder region water molecules relax faster than the solvation shell water molecules. The corresponding timescales are extracted by fitting the orientational time correlation functions to a bi-exponential function t t + (1 − a0 ) exp − . C2 (t) = ao exp − τ1 τ2

(4)

The two timescales (τ1 and τ2 ) and their respective weights are given in Table 4. The first timescale (τ1 ) corresponding to the initial fast decay of the orientational time correlation function arises from the librational and inertial motion while the second timescale (τ2 ) captures the slower rotation of water molecules. The weighted sum of these two timescales gives the overall time (τ 2;OD and τ 2;µ ) required for relaxation of C2 (t) (Table 4). In this Table, available experimental results 38,44,45 of orientational relaxation for bulk water (D2 O) are also included. Clearly, the OD bond reorientation is faster in the shoulder region than in the bulk. Usually, cations are known to have smaller effect on OD bond reorientation dynamics 46 . However, the appearance of only a shoulder in the Br-OW RDF suggests rather weak interaction between the positively charged bromine atom and oxygens of the surrounding water. This means that the bromine atom is weakly cationic and thus may not show the effects of typical cations. Indeed, the preferred orientation of ∼120° of water orientation (Figs.4a and 4b) and the hydrogen bond distribution (Figs.5a and 5b) in the shoulder region suggest that the Br-OW interaction is strong enough to make water molecules in this region loose one of their hydrogen bonds, but this interaction is not strong enough to show the effects of typical cations on the dynamics of surrounding water molecules. Thus, the loss of a hydrogen bond by water molecules in the shoulder region is not compensated by their interaction with the weakly cationic bromine atom. 13 Environment ACS Paragon Plus


Hence, water molecules in the shoulder region remain loosely bound and thus rotate somewhat faster. The rotation of OD bonds of water molecules in the solvation shell region is found to be slightly slower than those in the bulk. Like our calculations of MSDs, we have also calculated the orientational correlations of solvation shell water by both including and excluding the non-hydrogen bonded (to the BrO− 3 oxygens) water molecules and the results are shown in Fig.S3 of the Supporting Information. Again, no significant difference is found between the two correlations signifying the fact that majority of water molecules in the solvation shell are hydrogen bonded to the oxygens of the bromate ion.

The hydrogen bond dynamics is investigated by using the population time correlation function approach 47–51 . The integrated timescales of population correlation functions in different regions give the average hydrogen bond lifetimes which are included in Table 5. The decay of hydrogen bond correlation functions (Fig.9) shows that the water-water hydrogen bond dynamics in the shoulder region is faster than that of the bulk and solvation shell regions at 300 K. The lifetime of hydrogen bonds in a condensed phase mainly depends upon three factors: Hydrogen bond strength, orientation and diffusion of the molecules involved in the hydrogen bonding. Since we are considering the water-water hydrogen bond lifetimes, the hydrogen bond strength in different regions may not vary significantly. Out of orientation and diffusion aspects, the former is known to affect the hydrogen bond dynamics to a much greater degree than the latter. In the present case, water molecules in the shoulder region are found to have faster orientational dynamics and slower diffusion. Thus, it can be concluded that the faster orientational motion of water molecules in the shoulder region leads to the breaking of hydrogen bonds and thereby smaller lifetime. However, at 330 K, the water-water hydrogen bond dynamics in the shoulder region is found to be almost equal to that of the bulk with corresponding timescales of 1.27 and 1.25 ps (Table 5). Thus, the effect of enhanced temperature is found to be more on the water-water hydrogen bond dynamics in the bulk than in the shoulder region. This is because the increase in the temperature disrupts the hydrogen bonded network in the bulk, but no such network exists in the shoulder region. For the solvation shell, the water-water hydrogen bond dynamics is found to be slower than that in the other two regions at both the temperatures. This behavior can be attributed to the slower diffusion and orientational motion of water molecules in the solvation shell as


Page 14 of 42

Page 15 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


compared to those in the bulk and shoulder regions. Further, the solvation shell water molecules form hydrogen bonds with the OB atoms. We have also looked at the dynamics of these ion-water hydrogen bonds and found them to be much faster than water-water hydrogen bonds with lifetimes of 0.73 ps at 300 K and 0.68 ps at 330 K.

Vibrational spectral properties of water in aqueous solutions provide useful information about the local solvation structure. Such spectral properties are also measured experimentally through infrared spectroscopic techniques. Since we are not aware of any experimental vibrational spectroscopic studies of bromate ions in aqueous solutions, we have calculated the OD stretching frequencies of water molecules in different regions around the BrO− 3 anion. Our calculated spectral properties can give important insights into bromate ion solvation and also provide a starting point for future experimental studies of these solutions. The calculations of the instantaneous OD vibrational frequencies are carried out by using the method of wavelet transformation, details of which can be found in earlier work 52–54 . Fig.10 shows the distributions of OD vibrational frequencies in different regions at both the temperatures. The average frequency (Table 6) of OD bonds in the bulk and solvation shell that are hydrogen bonded to water are found to be very close. For the TSh type OD bonds, the average frequency is somewhat higher than that of the TB type. The reason for this shift in the average frequency is that the frequency distributions (Fig.10) for the TSh and TiHB Sol types are narrow as compared to the TB type OD bonds. This narrowing of distributions is seen towards the low frequency side and there is hardly any difference in the frequency distributions in the high frequency region. This suggests that some of the stronger hydrogen bonds are seen less for the ion-water pairs which are possible in the bulk. This also shows less environmental inhomogeneity in the solvation shell as compared to the bulk. Since the OD stretching frequencies correlate well with the strength of the associated hydrogen bonds, we can thus conclude that the ion-water hydrogen bonds are weaker than the water-water hydrogen bonds in the bulk on average. Also, this average weakening of the ion-water hydrogen bonds is because some stronger hydrogen bonded configurations do not occur in the solvation shell which are there for the water-water pairs in the bulk. We then looked at the vibrational spectral diffusion of different types of OD bonds and extracted the associated hydrogen bond dynamics. The vibrational spectral diffusion is investigated by calculating the frequency



Page 16 of 42

correlation function (Fig.11) which is defined as Cω (t) =

hδω (0) δω (t)i D

δω (0)2

E

,

(5)

where, δω (t) = ω (t) − hωi is the deviation from average frequency at time t. In order to extract the corresponding lifetime, the frequency correlation function was fitted to a biexponential function of the form given earlier (Eq. 4). The timescales obtained from the biexponential fits are given in Table 6. Available experimental timescales 44,45 of frequency correlation for bulk water are also included in this Table. From the longer timescales (τ2 ), which correspond to the associated hydrogen bond lifetimes, it is seen that the ion-water and water-water hydrogen bond dynamics in the solvation shell and shoulder regions, respectively, are faster than the bulk. This can be attributed to the weaker ion-water hydrogen bonds as suggested by the higher average vibrational frequencies of solvation shell OD modes. 3.2.2

Dynamics of H2 O Molecules Around the BrO− 3 Ion

First, we look at the BrO− 3 –H2 O structural correlations of Br-OW, Br-H, OB-OW and OB-H RDFs and compare them with those for D2 O because the corresponding pair of structural quantities are expected to be identical at the present level of calculations for − sufficiently long simulation trajectories. The results of BrO− 3 –H2 O and BrO3 –D2 O radial

distribution functions are shown in Fig.S4 of the Supporting Information. It can be seen from the plots of Fig.S4 that the solute-water structural correlations are essentially identical for H2 O and D2 O solvents. The minor discrepancies between the two sets of correlations, particularly a noticeable difference in the shoulder region at ∼ 3˚ A in the BrOW correlation, are believed to arise from unequal lengths of simulation trajectories for the two systems. Next, we discuss the dynamical properties of H2 O around the bromate ion and compare the results with those of D2 O. The results of the time dependence of MSD and orientational correlations of H2 O around the BrO− 3 ion are shown in Figs. S5 and S6 of the Supporting Information. Results are also presented for water in the bulk region in these figures. The values of the diffusion coefficients, integrated orientational relaxation times and also hydrogen bond lifetimes are included in Tables S2-S4. In these Tables, the corresponding values for D2 O are also included for comparison. The differences in the calculated dynamics of H2 O and 16 Environment ACS Paragon Plus

Page 17 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


D2 O molecules can be attributed to their mass difference, i.e. to their isotopic effects. As expected, the dynamics of H2 O is found to be generally faster than that of D2 O due to lighter mass of H2 O. However, the extent of isotope effects is found to be not the same for water in the solvation shell and bulk regions which shows the importance of intermolecular interactions in determining the magnitude of isotope effects in dynamical properties.

3.3

Comments on the Kosmotropic Properties of BrO− 3 Ion

The BrO− 3 ion has a slightly positive value of the Jones-Dole B-coefficient meaning that it is a structure maker in aqueous solution or a kosmotrope. However, this anion has two different kinds of interaction sites: The positively charged bromine atom (Br) and the negatively charged oxygen atoms (OB). The question that naturally arises is whether both the Br and OB atoms of the BrO− 3 ion act as kosmotropic sites and what is the extent of their structure making ability. We have connected the solvation shell dynamics of the BrO− 3 ion to its kosmotropic behavior. The solvation shell water dynamics can in fact be taken as a useful parameter to connect to kosmotropic/chaotropic behavior of solutes rather than the solvation shell structural parameters 55 . The general trends usually observed for kosmotropic (chaotropic) ions is that they tend to slow down (accelerate) the solvation shell water dynamics. For example, Na+ ion, which is a kosmotrope, slows down water diffusion while the chaotropic K+ ion makes the diffusion of the solvation shell water faster 56 . The extrapolation of solvation shell water dynamics to kosmotropic/chaotropic ability of polyoxy-anions was found to be even more complicated 57 . Gao and co-workers 57 showed that, while assigning water structure maker or breaker tag to an ion based on solvation shell dynamics, one also needs to mention the property under investigation. This conclusion was based on their study for solvation of the nitrate ion in water which was found to act as water structure maker “in affecting the translational motion of hydration water” 57 and as water structure breaker “in influencing the reorientation relaxation of hydration water” 57 . We note here that the ion-water hydrogen bond dynamics has not been linked to the water structure making/breaking ability of the ion as it may not be a suitable parameter. For example, the chloride-water hydrogen bond dynamics 53 is slower than that of water-water although chloride ion is a weak structure breaker. On the other hand, even though the iodide ion is also a water structure breaker (strong), the iodide-water hydrogen bond dynamics 58 shows the reverse trend. Thus, there is no clear



correlation between the water structure making/breaking ability of an ion and the ionwater hydrogen bond dynamics. From the comparison of MSDs, water molecules in the solvation shell and shoulder regions around the bromate ion are found to be translationally slower than that of the bulk. However, the slower diffusion of water molecules in the shoulder region is not because of the influence of the bromine atom, rather more due to the presence of the oxygen atoms of the BrO− 3 ion. Further, the residence dynamics of water in the shoulder region is found to be faster while that of the solvation shell water is slower than the bulk. Similar trend is also found for water-water hydrogen bond dynamics in different regions around the BrO− 3 ion although the timescales differ only marginally. The reorientational dynamics of water molecules in the shoulder region is clearly faster while it is almost similar for the solvation shell water in comparison to the bulk as shown by the corresponding timescales (Table 4). We emphasize here that water molecules in the shoulder and solvation shell regions feel influences primarily of the positively charged bromine atom and the negatively charged oxygen atoms of the BrO− 3 ion, respectively. The bromine atom accelerates while the negatively charged oxygen atoms slow down the motion of neighboring water molecules as compared to the bulk water. Thus, we can conclude based on the solvation shell dynamics of the BrO− 3 ion that the although its a single ion, it contains one chaotropic (the Br atom) and three weak kosmotropic sites (the OB atoms). The overall weak water structure making ability of the BrO− 3 ion is actually a result of the interplay between one chaotropic and three kosmotropic sites. Thus, it is important to treat different solvation sites of polyoxy-anions separately for better understandings of their positions in the Hofmeister series.

4

Conclusions The solvation shell structure and dynamics of a BrO− 3 ion in water have been in-

vestigated by means of Born-Oppenheimer molecular dynamics simulations at 300 and 330 K. The BrO− 3 ion is a pyramidal ion which can interact with the surrounding water molecules through its positively charged bromine and negatively charged oxygen atoms. We have calculated different structural and dynamical properties of water molecules interacting with these different sites. We have also drawn comparisons between the solvation shell structures of the bromate and iodate ions. The iodate ion was earlier found to have some solvation shell structural features 6 which were unexpected of an anionic solute. The 18 Environment ACS Paragon Plus

Page 18 of 42

Page 19 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


− BrO− 3 ion is similar to the IO3 ion apart from the differences in the size, electronegativity

and polarizability of the bromine and iodine atoms. Thus, a comparison of the solvation structures of the two ions sheds light on the effects of these molecular properties on the surrounding water structure of these pyramidal trioxy-anions in aqueous solutions. We have found that the interaction of water molecules with the positively charged bromine atom of the BrO− 3 ion gives rise to only a shoulder in the ion-water radial distribution function rather than a well defined hydration shell. However, the negatively charged oxygen atoms interact more strongly with the surrounding water molecules to give rise to a well defined solvation shell. In case of the iodate ion, a strong solvation shell was found containing water molecules interacting with the iodine atom apart from the other solvation shell around the negatively charged oxygens 6 . Thus, it is seen that as one moves − from the bromine atom (of BrO− 3 ) to the iodine atom (of IO3 ), the surrounding water

structure changes from a shoulder to a well defined solvation shell. The distribution of the hydrogen bonding states of water molecules in the shoulder region around the BrO− 3 ion is found to be very different from that of the bulk and solvation shell regions. Majority of water molecules in the shoulder region exist as double donor-single acceptor species with average number of hydrogen bonds per water molecule being 3.18 and 2.93, respectively, at 300 and 330 K. This is because the oxygen atoms of water molecules in the shoulder region point towards the bromine atom and take an orientation which maximizes their hydrogen bonds and at the same time oxygen lone pair interacts favorably with the bromine atom. The average angle between the dipole vectors of water molecules and that of the BrO− 3 ion in the shoulder region is found to be ∼120° . On the dynamical side, we have looked at the diffusion, residence dynamics, orientational motion and also hydrogen bond relaxation of water around the bromate ion. The water molecules in the shoulder region are generally found to have faster dynamics while those in the solvation shell region show slower dynamics as compared to the bulk region. The diffusion of water molecules in the shoulder region, however is found to be slower than that of the bulk water, even though the residence dynamics of the former is faster. The reason for this contrasting diffusive behavior of the shoulder region can be attributed to the fact that the water molecules in this region escape on a fast timescale but once they move out, they form hydrogen-bonds with the oxygen atoms of the BrO− 3 ion. Thus, the water molecules in the shoulder region do not move very far away from the ion giving rise



to smaller mean square displacement. We have also connected the solvation shell dynamical trends to the water structure making/breaking properties of the BrO− 3 ion. Based on the generally accepted view that water structure breaking/making ions make the solvation shell water dynamics faster/slower, one can conclude that the positively charged bromine and the negatively charged oxygen atoms of the BrO− 3 ion act as water structure breaker and maker, respectively. Thus, in spite of being parts of the same ion, these two types of solvation sites have opposite effects on the surrounding water. We note in this context that such dual characteristics for some complex ions like nitrate, ammonium etc. have been reported experimentally 59 . Thus, one needs to look at all possible major solvation sites of multi-atom solutes in water in order to understand the positions of such solutes in the Hofmeister series at a molecular level. In the current study, we have considered a single bromate ion in water. It would be worthwhile to study the effects of counterions and salt concentration on the structural and dynamical behavior of aqueous metal bromate solutions from ab initio simulations. Also, in the current study, we have employed the dispersion corrected BLYP-D3 functional 21 . This particular functional was used in many earlier ab initio simulations of aqueous systems 13–15,30,31 and was found to provide significantly improved description of the structure, dynamics and phase diagram of aqueous systems than the corresponding BLYP functional 22,23 without any dispersion correction. We also note in this context that other dispersion correction schemes such as DFT-vdw 60 , DFT-TS 61 have also been used in ab initio simulations of water and other hydrogen bonded systems. In addition to to the use of BLYP or similar functionals belonging to the so-called GGA (generalized gradient approximation) category combined with dispersion corrections, hybrid functionals with dispersion corrections have also been used for simulations of aqueous systems in recent years 19,62 . The hybrid functionals are composed of a fraction of the exact Hartree-Fock exchange part and a chosen linear combination of semi-local or local functionals 31 . Like GGA, hybrid functionals also lack the ability to properly describe the dispersion interactions arising from non-local dynamical electron correlations. However, hybrid functionals combined with van der Waals dispersion corrections have been found to describe the structure and dynamics of liquid water quite well under ambient conditions. Although calculations with such hybrid functionals are computationally more expensive, it will be worthwhile to use such functionals for simulations of aqueous ionic solutions in future. It


Page 20 of 42

Page 21 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


would also be worthwhile to perform calculations of nuclear quantum effects which are missing in the current study. We note in this context the recent work of DiStasio et al. 62 where the temperature of the simulation system was raised by 30 K to take into account approximately the nuclear quantum effects. Their calculations clearly showed the importance of the collective influence of exact exchange, dispersion interaction and choice of temperature on the equilibrium properties of liquid water. Thus, it would be worthwhile to carry out a study of the comparative performance of different functionals for aqueous solutions of bromate and other polyoxy-anions.

Supporting Information The Supporting Information provided with this Paper contains results of binding energy calculations of water dimer for two different basis set cut-off and also additional results of some of the dynamical properties of D2 O and H2 O.

Acknowledgment Financial support through a J.C. Bose Fellowship to A.C. from the Science and Engineering Research Board, a statuary body of the Department of Science and Technology, and Council of Scientific and Industrial Research (through a Junior/Senior Research Fellowship to B.S.), Government of India, are gratefully acknowledged. Part of the calculations were done at the High Performance Computing Facility at Computer Center, IIT Kanpur.



References [1] Marx, D.; Hutter, J. Ab Initio Molecular Dynamics: Basic Theory and Advanced Methods; Cambridge University Press: Cambridge, 2009. [2] Jenkins, H. D. B.; Marcus, Y. Viscosity B-Coefficients of Ions in Solution. Chem. Rev. 1995, 95, 2695-2724. [3] Mills, A.; Belghazi, A.; Rodman, D. Bromate Removal from Drinking Water by Semiconductor Photocatalysis. Wat. Res. 1996, 30, 1973-1978. [4] Bao, M. L.; Griffini, O.; Santiani, D.; Barbieri, K.; Burrini, D.; Pantiani, F. Removal of Bromate Ion from Water Using Granular Activated Carbon. Wat. Res. 1999, 33, 2959-2970. [5] Kishimoto, N.; Matsuda, N. Bromate Ion Reoval by Electrochemical Reduction Using an Activated Carbon Felt Electrode. Environ. Sci. Technol. 2009, 43, 20542059. [6] Baer, M. D.; Pham, V.; Fulton, J. L.; Schetner, G. K.; Balasubramanian, M.; Mundy, C. J. Is Iodate a Strongly Hydrated Cation? J. Phys. Chem. Lett. 2011, 2, 2650-2654. [7] Diehl, A.; dos Santos, A. P.; Levin, Y. Surface Tension of an ElectrolyteAir Interface: A Monte Carlo Study. J. Phys.: Condens. Matter 2012, 24, 284115. [8] Eklund, L.; Hofer, T. S.; Persson, I. Structure and Water Exchange Dynamics of Hydrated Oxo Halo Ions in Aqueous Solution Using QMCF MD Simulation, Large Angle X-Ray Scattering and EXAFS. Dalton Trans. 2015, 44, 1816. [9] The CP2K Developers Group, 2014. http://www.cp2k.org/. [10] VandeVondele, J.; Krack, M.; Mohamed, F.; Parrinello, M.; Chassaing, T.; Hutter, J. Quickstep: Fast and Accurate Density Functional Calculations Using a Mixed Gaussian and Plane Waves Approach. Comput. Phys. Commun. 2005, 167, 103-128. 22 Environment ACS Paragon Plus

Page 22 of 42

Page 23 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[11] Raynaud, C.; Maron, L.; Jolibois, F.; Daudey, J.; Esteves, P. M.; Ram´ırez-Sol´ıs, A. Ab Initio Molecular Dynamics: Plane Waves vs. Local Basis The Role of Energy Cutoff on the Convergence of Molecular Properties. Chem. Phys. Lett.2005, 414, 161-165. [12] Miceli, G.; Hutter, J.; Pasquarello, A. Liquid Water through Density-Functional Molecular Dynamics: Plane-Wave vs Atomic-Orbital Basis Set. J. Chem. Theory Comput. 2016, 12, 3456-3462. [13] Jonchiere, R.; Seitsonen, A.P.; Ferlat, G.; Saitta, A.M.; Vuilleumier, R. Van der Waals Effects in ab initio Water at Ambient and Supercritical Conditions. J. Chem. Phys. 2011, 135, 154503. [14] Bankura, A.; Karmakar, A.; Carnevale, V.; Chandra, A.; Klein M. L. Structure, Dynamics and Spectral Diffusion of Water Molecules from First Principles Molecular Dynamics. J. Phys. Chem. C 2014, 118, 29401-29411. [15] Ohto, T.; Usul, K.; Hasegawa, T.; Bonn, M. Nagata, Y. Toward ab initio Molecular Dynamics Modeling for Sum-Frequency Generation Spectra; An Efficient Algorithm Based on Surface-Specific Velocity-Velocity Correlation Function. J. Chem. Phys. 2015, 143, 124702. [16] K¨ uhne, T.D.; Pascal, T.A.; Kaxiras, E.; Jung, Y. New Insights into the Structure of the Vapor/Water Interface from Large-Scale First-Principles Simulations. J. Phys. chem. Lett. 2011, 2, 105-113. [17] Sulpizi, M.; Salanne, M.; Sprik, M.; Gaigeot, M.-P. Vibrational Sum Frequency Generation Spectroscopy of the Water Liquid-Vapor Interface from Density Functional Theory-Based Molecular Dynamics Simulations. J. Phys. Chem. Lett. 2012, 4, 83-87. [18] Galib, M.; Duignan, T.T.; Misteli, Y.; Baer, M.D.; Schenter, G.K.; Hutter, J.; Mundy, C.J. J. Chem. Phys. 2017, 146, 244501. [19] Ambrosio, F.; Miceli, G.; Pasquarello, A. Structural, Dynamical, and Electronic Properties of Liquid Water: A Hybrid Functional Study. J. Phys. Chem. B. 2016, 120, 7456-7470. 23 Environment ACS Paragon Plus


[20] Grimme S. Semiempirical GGA-Type Density Functional Constructed with A LongRange Dispersion Correction. J. Comput. Chem. 2006, 27, 1787-1799. [21] Grimme, S.; Antony, J.; Ehrlich, S.; Kreig, H. Consistent and Accurate Ab Initio Parameterization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132, 154104. [22] Becke A. D. Density-Functional Exchange-Energy Approximation with Correct Asymptotic Behavior. Phys. Rev. A 1988, 38, 3098-3100. [23] Lee C.; Yang W.; Parr R. G. Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron Density. Phys. Rev. B 1988, 37, 785-789. [24] Sprik, M.; Hutter, J.; Parrinello, M. Ab initio Molecular Dynamics Simulation of Liquid Water: Comparison of Three Gradient Corrected Density Functionals. J. Chem. Phys. 1996, 105, 1142-1152. [25] VandeVondele, J.; Mohamed, F.; Krack, M.; Hutter, J.; Sprik, M.; Parrinello, M. The Influence of Temperature and Density Functional Models in ab initio Molecular Dynamics Simulation of Liquid Water. J. Chem. Phys. 2005, 122, 014515. [26] Kuo, I.-F. W.; Mundy C. J. An ab initio Molecular Dynamics Study of the Aqueous Liquid-Vapor Interface. Science, 2004, 303, 658-660. [27] McGrath, M. J.; Siepmann, J. I.; Kuo, I.-F. J.; Mundy, C. J.; VandeVondele, J.; Hutter, J.; Mohamed, F.; Krack, M. Isobaric-Isothermal Monte Carlo Simulations from First Principles: Application to Liquid Water at Ambient Conditions. ChemPhysChem 2005, 6, 1894-1901. [28] Lee, H. -S.; Tuckerman, M. E. Dynamical Properties of Liquid Water from ab initio Molecular Dynamics Performed in the Complete Basis Set Limit. J. Chem. Phys. 2007, 126, 164501. [29] Chakraborty, D.; Chandra, A. Hydrogen Bonded Structure and Dynamics of LiquidVapor Interface of Water-Ammonia Mixture: An ab initio Molecular Dynamics Study” J. Chem. Phys. 2011, 135, 114510.


Page 24 of 42

Page 25 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[30] Ohto, T.; Backus, E.H.G.; Mizukami, W.; Hunger, J.; Bonn, M. Nagata, Y. Unveiling the Amphiphilic Nature of TMAO by Vibrational Sum Frequency Generation Spectroscopy. J. Phys. Chem. C 2016, 120, 17435-17443. [31] Gillan M. J.; Alfé D.; Michaelides A. Perspective: How Good is DFT for Water? J. Chem. Phys. 2016, 144, 130901. [32] Del Ben, M.; Schönherr, M.; Hutter, J.; VandeVondele, J. Bulk Liquid Water at Ambient Temperature and Pressure from MP2 Theory. J. Phys. Chem. Lett. 2013, 4, 3753-3759. [33] Schwegler, E.; Grossman, J. C.; Gygi, F.; Galli, G. Towards an Assessment of the Accuracy of Density Functional Theory for First Principles Simulations of Water. II J. Chem. Phys. 2004, 121, 5400-5409. [34] Kuo, I.-F. W.; Mundy, C. J.; McGrath, M. J.; Siepmann, J. I.; VandeVondele, J.; Sprik, M.; Hutter, J.; Chen, B.; Klein, M. L.; Mohamed, F., et al. Liquid Water from First Principles: Investigation of Different Sampling Approaches. J. Phys. Chem. B 2004, 108, 12990-12998. [35] Ma, Z.; Zhang, Y.; Tuckerman, M.E. Ab Initio Molecular Dynamics Study of Water at Constant Pressure Using Converged Basis Sets and Empirical Dispersion Corrections. J. Chem. Phys. 2012, 137, 044506. [36] Bussi, G.; Donadio, D.; Parrinello, M. Canonical Sampling Through Velocity Rescaling. J. Chem. Phys. 2002, 126, 014101. [37] Soper, A.K.; Bruni, F.; Ricci, M.A. Site-Site Pair Correlation Functions of Water from 25 to 400° C: Revised Analysis of New and Old Diffraction Data. J. Chem. Phys. 1997, 106, 247-254. [38] Hardy, E.H.; Zygar, A.; Zeidler, M.D.; Holz, M.; Sacher, F.D. Isotope Effect on the Translational and Rotational Motion in Liquid Water and Ammonia. J. Chem. Phys. 2001, 114, 3174-3181. [39] Impey, R. W.; Madden, P. A.; McDonald, I. R. Hydration and Mobility of Ions in Solution. J. Phys. Chem. 1983, 87, 5071-5083.



[40] Lagge, D.; Hynes, J. T. On the Residence Time for Water in a Solute Hydration Shell: Application to Halide Solutions. J. Phys. Chem. B 2008, 112, 7697-7701. [41] Roy, S.; Baer, M. D.; Mundy, C. J.; Schenter, G. K. Reaction Rate Theory in Coordination Number Space: An Application to Ion Solvation. J. Phys. Chem. C 2016, 120, 7597-7605. [42] Loganathan, N.; Yazaydin, A. O.; Bowers, G. M.; Kalinichev, A. G.; Kirkpatrick, R. J. Cation and Water Structure, Dynamics, and Energetics in Smectite Clays: A Molecular Dynamics Study of Ca-Hectorite. J. Phys. Chem. C 2016, 120, 1242912439. [43] Savage, J.; Tse, Y. S.; Voth, G. A. Proton Transport Mechanism of Perfluorosulfonic Acid Membranes. J. Phys. Chem. C 2014, 118, 17436-17445. [44] Fecko, C.J.; Loparo, J.J.; Roberts, S.T.; Tokmakoff, A. Local Hydrogen Bonding Dynamics and Collective Reorganization in Water: Ultrafast Infrared Spectroscopy of HOD/D2 O. J. Chem. Phys. 2005, 122, 054506. [45] Park, S.; Fayer, M.D. Hydrogen Bond Dynamics in Aqueous NaBr Solutions. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16731-16738. [46] van der Post, S. T.; Tielrooij, K.; Hunger, J.; Backus, E. H. G.; Bakker, H. B. Femtosecond Study of the Effects of Ions and Hydrophobes on the Dynamics of Water. Faraday Discuss. 2013, 160, 171-189. [47] Rapaport, D. C. Hydrogen Bonds in Water: Network Organization and Lifetimes. Mol. Phys., 1983, 50, 1151-1162. [48] Luzar, A.; Resolving the Hydrogen Bond Dynamics Conundrum. J. Chem. Phys. 2000, 113, 10663-10675. [49] Balasubramanian, S.; Pal S.; Bagchi, B. Hydrogen-Bond Dynamics near a Micellar Surface: Origin of the Universal Slow Relaxation at Complex Aqueous Interfaces. Phys. Rev. Lett., 2002, 89, 115505. [50] Chandra A. Effects of ion Atmosphere on Hydrogen-Bond Dynamics in Aqueous Electrolyte Solutions. Phys. Rev. Lett., 2000, 85, 768-771. 26 Environment ACS Paragon Plus

Page 26 of 42

Page 27 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[51] Das, S.; Biswas, R.; Mukherjee, B. Collective Dynamic Dipole Moment and Orientation Fluctuations, Cooperative Hydrogen Bond Relaxations, and their Connections to Dielectric Relaxation in Ionic Acetamide Deep Eutectics: Microscopic Insight from Simulations. J. Chem. Phys., 2016, 145, 084504. [52] Vela-Arevelo, L. V.; Wiggins, S. Time-Frequency Analysis of Classical Trajectories of Polyatomic Molecules. Int. J. Bifurcat. Chaos 2001, 11, 1359-1380. [53] Mallik, B. S.; Semparithi, A.; Chandra, A. A first principle theoretical study of vibrational spectral diffusion and hydrogen bond dynamics in aqueous ionic solution: D2 O in hydration shells of Cl− ions. J. Chem. Phys. 2008, 129, 194512. [54] Mallik, B. S.; Chandra, A. Vibrational Spectral Diffusion in Supercritical D2 O from the First Principles: An Interplay Between the Dynamics of Hydrogen Bonds, Dangling OD Groups and Inertial Rotation. J. Phys. Chem. 2008, 112, 13518-13527. [55] Stirnemann, G.; Wernersson, E.; Jungwirth, P.; Laage, D. Mechanisms of Acceleration and Retardation of Water Dynamics by Ions. J. Am. Chem. Soc. 2013, 135, 11824-11831. [56] Ishai, P. B.; Mamontov, E.; Nickels, J. D.; Sokolov, A. P. Influence of Ions on Water Diffusion-A Neutron Scattering Study. J. Phys. Chem. B 2013, 117, 7724-772. [57] Xie W. J.; Yang Y. I.; Gao Y. Q. Dual Orientational Relaxation Routes of Water Molecules in Oxyanion’s Hydration Shell: A Molecular Geometry Perspective. J. Chem. Phys. 2015, 143, 224504. [58] Chowdhuri, S.; Chandra, A. Dynamics of Halide Ion-Water Hydrogen Bonds in Aqueous Solutions: Dependence on Ion Size and Temperature. J. Phys. Chem. B 2006, 110, 9674-9680. [59] Conte, P. Effects of Ions on Water Structure: A Low Field HT-1 NMR Relaxometry Approach. Magn. Reson. Chem. 2015, 53, 711-718. [60] Dion, M.; Rydberg, H.; Schröder, E.; Langreth, D.C.; Lundqvist, B.I. Van der Waals Density Functional for General Geometries. Phys. Rev. Lett., 2004, 92, 246401.



[61] Tkatchenko, A.; Scheffler, M. Accurate Molecular van der Waals Interactions from Ground-State Electron Density and Free-Atom Reference Data. Phys. Rev. Lett., 2009, 102, 073005. [62] DiStasio, Jr. R.A.; Santra, B.; Li, Z.; Wu, X.; Car, R. The Individual and Collective Effects of Exact Exchange and Dispersion Interactions on the Ab Initio Structure of Liquid Water. J. Chem. Phys. 2014, 141, 084502.


Page 28 of 42

Page 29 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TABLE 1. Average number of hydrogen bonds per water molecule in the shoulder region, solvation shell and the bulk. Temperature Shoulder region Solvation shell Bulk Experiment 300 K 3.18 3.58 3.60 3.58a 330 K 2.93 3.49 3.53 a

The experimental value is inferred from neutron diffraction data of bulk water at ambient

condition. See Ref. 37 .

TABLE 2. Comparison of solvation structures of the iodate and bromate ions. Properties Nature of the RDF

Iodate 6 Sharp peaks at 2.9 ˚ A (SIO I ) and 4˚ A (SIO II )

Bromate Shoulder at 3 ˚ A BO (SH ) and sharp peak at 4 ˚ A (SBO )

No. of water molecules

IO SIO I =3, SII =9

BO SBO =6 H =1, S

Average no. of hydrogen bonds per water

Data not available

BO =3.59 SBO H =2.23, S

− ˚ SIO I =First solvation shell around the IO3 ion corresponding to the first peak at 2.9 A. − IO A. SII =Second solvation shell around IO3 ion corresponding to the second peak at 4 ˚ − BO A. SH =Shoulder region around the BrO3 ion corresponding to the shoulder at around 3 ˚ − BO ˚ S =Solvation shell around the BrO3 ion corresponding to the peak at 4 A.



Page 30 of 42

TABLE 3. Diffusion coefficients (D) of D2 O molecules in different regions (in 10−5 × cm2 /s). Temperature

300 K 330 K a

Water in the shoulder region water 0.64 1.60

Water in the solvation shell water 1.13 2.0

Bulk water Experiment

1.43 2.05

1.96a 3.78a

Self-diffusion coefficient of D2 O deduced from experimental data as described by Eq.3

of Ref. 38 .

TABLE 4.

Timescales (ps) obtained from the biexponential fits (τ1 and τ2 ) of the

second-rank orientational correlation functions of D2 O in different regions. The integrated orientational relaxation times (τ 2 ) of D2 O are also included.

Temperature

Quantity

300 K

330 K

τ1OD (τ1µ ) τ2OD (τ2µ ) τ 2;OD (τ 2;µ )

Water in the shoulder region 0.09 (0.11) 2.83 (2.80) 2.12 (1.78)

Water in the solvation shell 0.11 (0.09) 3.78 (3.05) 2.67 (2.07)

τ1OD (τ1µ ) τ2OD (τ2µ ) τ 2;OD (τ 2;µ )

0.08 (0.06) 2.10 (1.50) 1.50 (0.98)

0.07 (0.07) 2.34 (1.95) 1.65 (1.26)

Bulk water Experiment 0.10 (0.10) 3.68 (3.14) 2.91 (2.08)

3.0a , 2.6b 1.82c

0.08 (0.08) 2.32 (1.92) 1.68 (1.30)

0.93c

a

Long-time component of the orientational relaxation of HOD in D2 O from Ref. 44 .

b

Long-time component of the orientational relaxation of HOD in H2 O from Ref. 45 .

c

Integrated second-rank orientational relaxation time of D2 O deduced from experimental

data as described by Eq.3 of Ref. 38 combined with appropriate parameters for rotational relaxation given in the same reference.


Page 31 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TABLE 5. Timescales (ps) obtained from integration of the continuous hydrogen bond population correlation functions of D2 O in different regions. Temperature 300 K 330 K

TABLE 6.

Shoulder region water-water 1.49 1.27

Solvation shell water-water 1.75 1.40

Ion-water Bulk water-water 0.73 0.68

1.67 1.25

Average stretching frequencies of OD bonds in different environments and

parameters obtained from biexponential fits of the frequency time correlation function, Cω (t). Type of OD bond T (K) TSh 300 300 TwHB Sol iHB 300 TSol TB 300 TSh 330 wHB TSol 330 330 TiHB Sol TB 330

ωavg (cm−1 ) 2500 2485 2521 2485 2515 2500 2520 2493

a0 τ1 (ps) τ2 (ps) Experiment (τ2 ) 0.71 0.08 1.05 0.71 0.08 1.25 0.78 0.08 0.60 0.69 0.08 1.15 1.4±0.2a , 1.7±0.5b 0.67 0.07 0.85 0.71 0.08 1.20 0.80 0.07 0.52 0.67 0.07 0.96

a

Long-time component of the frequency correlation of OH mode of HOD in D2 O from Ref. 44 . b Long-time component of the frequency correlation of OD mode of HOD in H2 O from Ref. 45 .



Figure 1: (a) The radial distribution functions for the Br-OW pair and center of mass (COM) of the bromate ion and water oxygen pair, (b) The incremental radial distribution functions for the Br atom and its nth (n=1-7) nearest OW atom.


Page 32 of 42

Page 33 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2: The spatial distribution function (SDF) for the distributions of the hydrogen and oxygen atoms of water molecules around the BrO− 3 ion. The green color represents the hydrogen atoms (HW) (isovalue=70.0) and the cyan color represents the oxygen atoms (OW) (isovalue=99.0) of water respectively. The two lobes in the left having the OWs (cyan color) closer to the bromine atom represents the shoulder region. In all other regions, the HWs (green color) are closer to the BrO− 3 ion than the OW atoms.



Figure 3: Two dimensional radial distribution function (2D-RDF) for the distribution of water oxygen (OW) with respect to the bromine atom (Br) and one of the oxygen atoms (OB) of the bromate ion.


Page 34 of 42

Page 35 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4: Distributions of angle between the dipole axis of the BrO− 3 ion and that of water molecules in the shoulder and solvation shell regions.



Figure 5: Fraction of different hydrogen bonded water molecules in the bulk, shoulder and solvation shell regions. Panel (a) is for 300 K and panel (b) is for 330 K. Here, NIL=no hydrogen bond, A=single acceptor, AA=double acceptor, D=single donor, DD=double donor, AD=single acceptor-single donor, AAD=double acceptor-single donor, ADD=single acceptor-double donor, AADD=double acceptor-double donor.


Page 36 of 42

Page 37 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6: (a) Mean square displacements (MSDs) of D2 O molecules in the bulk, shoulder and solvation shell regions, (b) Residence dynamics of D2 O molecules in the bulk, shoulder and solvation shell regions at 300 and 330 K



(a)

(b)

(c)

Figure 7: (a) Snapshot of a D2 O molecule initially in the shoulder region under the influence of the BrO− 3 ion, and (b) The same D2 O molecule after forming hydrogen bond with one of the oxygen atoms of the BrO− 3 ion, (c) Distance between the D2 O oxygen and the bromine atom as a function of time steps (blue color) to show that the D2 O molecule does not move very far away from the bromine atom even after escaping from the shoulder region. The red color indicates the cutoff (3.3 ˚ A) for the shoulder region.


Page 38 of 42

Page 39 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8: Orientational relaxation of dipole vectors (upper panel) and OD bond vectors (lower panel) of D2 O molecules in different regions around the BrO− 3 ion at 300 and 330 K.



Figure 9: Relaxation of continuous hydrogen bond population correlation function of D2 O in different regions at 300 and 330 K.


Page 40 of 42

Page 41 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10: Distribution of stretching frequencies of OD modes in different environments at 300 and 330 K.

Figure 11: Relaxation of frequency time correlation function, Cω (t), of OD modes in different environments at 300 and 330 K.



Figure 12: TOC Graphic


Page 42 of 42

Born-Oppenheimer Molecular Dynamics Simulations of a Bromate Ion

Recommend Documents