Structure, Dynamics and Adsorption of Charged Guest within the

Subscriber access provided by Kaohsiung Medical University

Surfaces, Interfaces, and Applications

Structure, Dynamics and Adsorption of Charged Guest within the Nanocavity of Polymer Functionalized Neutral Macrocyclic Host Pooja Sahu, Sk. Musharaf Ali, Kalsanka Trivikram Shenoy, and Sadhana Mohan ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b03874 • Publication Date (Web): 31 May 2018 Downloaded from http://pubs.acs.org on May 31, 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 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 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.

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 32 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

ACS Applied Materials & Interfaces

Structure, Dynamics and Adsorption of Charged Guest within the Nanocavity of Polymer Functionalized Neutral Macrocyclic Host Pooja Sahu1,2, Sk. Musharaf Ali*1,2, Kalsanka Trivikram Shenoy1, and Sadhana Mohan1 1

Bhabha Atomic Research Center, Mumbai, Maharashtra, India 400085

2

Homi Bhabha National Institute, Mumbai, Maharashtra, India 400094

Abstract Host-guest encapsulation has been widely applied for purification and seizing of the metal ions. Macrocyclic crown ethers are one of the most popular hosts in the field of host-guest chemistry, which on functionalization with polymers are employed as an effective adsorbent. In spite of their vast applications, the microscopic information about their sensing mechanism towards cations/molecules is very scarce. Therefore, the present article is focused on the molecular insights of ion-exchange mechanism within the cavity of crown ether functionalized polymers, using molecular dynamics (MD) simulations. The present study investigates the molecular level events of chloromethylated polystyrene (CMPS) bearing dibenzo18-crown-6 (DB18C6) in the aqueous and acidic environment, which has been found to be particularly successful in sensing of various alkali and alkali earth metal ions. A strategy has been envisaged to design a crown ether based functionalized polymeric resin which exhibits good match of properties with the in-house synthesized resin. The MD studies well capture the experimentally observed Langmuir type adsorption isotherms of Li+ ions on crown ether grafted polymer resins. The presence of acid reduces the adsorption of Li+ ions due to the competition with H3O+ ions. In addition, the results revealed that the ‘adsorption in crown cavity’ follows a dual residence time function. To the best of our knowledge, this is the first report on the adsorption isotherm of functionalized crown ether using MD simulations. The structure and dynamics of binding sites were explored using radial distribution functions (RDFs) and diffusion coefficients. All these effects have been studied for different concentration of Li + ions, acid concentration and counter ions, different length of polymer chains and degree of polymerization. Overall, the present study provides insights and quantitative information for adsorption on the CMPS-DB18C6 resin, which might be useful in myriads of host-guest based adsorption experiments.

Keywords: functionalized crown ether, adsorption, lithium ion, MD simulation, host-guest

1|Page ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

Page 2 of 32

1. INTRODUCTION The incredible selectivity of crown ethers as demonstrated by Pedersen et al. has emerged an innovative eon on the investigations of host-guest type interactions1-3. Oxygen ends of cyclic polyether rings of crown ethers originate the micelle type structure, which on the other hand drive host-guest complex formation with the metal ions due to ion-dipole interactions4, 5.The potential selectivity of crown ethers not only for positively charged guests but also for neutral guest has introduced new host-guest chemistry principles for binding of crown ethers6, 7. In reality, the neutral guests are supposed to interact with crown ethers through the non-covalent interactions of Van der Waals forces and hydrogen bonding. The characterization of these crown ether complex has been made possible by the advance probe techniques8 such as IR-UV photo-dissociation spectroscopy9, laser-induced fluorescence spectroscopy10, and titration calorimetries11.Thanks to super-computing facilities, which have broadly explored the new host-guest principles in this area of supramolecular chemistry. The studies reveal the role of size and charge distribution of the guest molecule and their size similarity with the crown ether cavity. The selectivity of crown ethers is primarily decided by size matching criteria of crown ether-guest complex12, 13. Apart from these, the number of heteroatoms in the rings, the polarity of guest molecule and the nature of solvating medium are also observed to affect the binding stability of the crown-guest complex14-17. This has fueled the intense demand of crown ethers in the different fields of scientific and technical applications, such as potential amphiphiles in membranes18, ionophores, ion sensing and photochemical19 as well as electrochemical molecular switching and transport etc20, 21. Further, with the advent of functionalization, crown ethers have evolved numerous applications in the nuclear and medical industries20,

22

, which are well known to be associated with very challenging

separation needs. Crown ether-functionalized materials are particularly interesting for the applications involving chemical-exchange process due to their strong dipole interactions with the guest ions23, 24. Moreover, the implication of functionalized crown ethers as chemical exchange resins25, 26 has made the purification of incoming stream very efficient even with the low metal ion content. Grafting of crown ethers with different organic and inorganic polymers not only reduces its solubility loss in various solvents but also facilitates the fabrication of solid phase so that the distribution of desired species can be achieved between two phases. As a result, a variety of polymer units have been directly fused to the family of crown ethers through suitable spacers in order to make crown ether based resin materials27. Among the members of crown ether family, 15-crown-5, benzo-15-crown-5, 12-crown-4, benzo-18crown-6, and dibenzo-18-crown-6 have enjoyed an exceptional attention for large separation efficiency of various metal ions and isotopes due to their cavity fitting selectivity for a wide range of alkali and alkali earth metal ions26. After Sone and co-workers fabricated the first crown ether functionalized thiophenes in


Page 3 of 32 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


198928, numerous studies were devoted and several modifications were introduced in order to improve the sensing performance of different functionalized crown ethers and based polymer resins23, 29.The effective fabrication of crown ethers with appropriate binding agents has disclosed the plethora of applications in catalysis30, transistors31, and protein translocation experiments32. However, in spite of the numerous applications of functionalized crown ethers, there are only a few simulation studies by Jamali et al.33 and Zanuy et al.24, where the microscopic information about the sensing mechanism of crown-ethers is partially explored. Also, the important events related to the response of the binding agents still remains unknown. Besides, the interplay of interactions mechanism for adsorption capability of crown ether resins is still untouched and therefore need to be explored.

With all these considerations, the present article is focused on the molecular insights of crown ether based ion-exchange mechanism by studying the structural and dynamical events of cations capture into the chloromethylated polystyrene (CMPS) bearing dibenzo18-crown-6 (DB18C6), which has been found to be quite successful in sensing of various alkali and alkali earth metal ions. In this context, the host-guest encapsulation of CMPS-DB18C6 has been explored with Li+ ions, as Li+ ions are widely used in many electronic devices, space applications, medical treatment and nuclear industries34-37. First, a computational modeling strategy has been envisaged to design a crown ether based functionalized polymeric resin followed by consistency of computational results with the properties of the in-house synthesized resin. Further, the separation abilities of the designed resin models have been explored by studying adsorption characteristics. The ability of the high-fidelity molecular simulations capture the experimentally observed adsorption isotherms, which are further explained in terms of structure and dynamics of the binding sites. The MD results provide molecular insights and microscopic understanding that can be applied to capture lots of quantitative information for the polymer-liquid interface. Moreover, the current studies on the Li+ ion embedded CMPS-DB18C6 will be of great practical use as they can be related to the myriads of hostguest based adsorption experiments.

2. SIMULATION PROTOCOL

In order to simulate the polymer functionalized crown ether, three resin models were considered, which differ in the number of repeating units, described as (I) one styrene strand at both end of crown ether, named as M1 (II) three styrene strands at each end of DB18C6 with linear linkage connections, named as M2, and (III) with multiple styrene strands at each end of DB18C6, having two more DB18C6 units in the cross-link position, named as M3, all shown in Fig.1. 3|Page ACS Paragon Plus Environment


Page 4 of 32

For all these resin models, the forcefield parameters for crown ether (DB18C6) were taken from our previous studies38, whereas the forcefield for the polymer part of the resin was borrowed from all-atom OPLS (Optimized Potential for Liquid Simulations) forcefield parameter models39. The resin monomer units were first optimized employing B3LYP density functional40 with SVP basis set41 using Turbomole package42 and then the optimized geometries, together with the generated Mulliken partial atomic charges were used as input for MD simulations. The results have been extended from single monomer unit in water box to cluster of monomer units in hydrated medium, using TIP4P/2005 water model43. Furthermore, the acid in the system was represented by dissociated ions of H3O+/NO3- ions, with forcefield parameters same as in our previous studies44, 45. The periodic boundary conditions were applied in all three directions in order to avoid surface effects. The interatomic interactions between the atoms were defined by equation1-8. The Van der Waals forces were simulated using Lennard-Jone (L-J) potential model, with a cut-off value of 12Å. The bond stretching and angle bending in the systems were modeled through Harmonic potential and dihedral angles in the systems were described using OPLS style. For unlike intermolecular interactions, Lorentz-Berthelot mixing rules (equation7-8) was used and the long-range interactions in the systems were applied using Particle-Particle-Particle mesh (PPPM) methods46.

Utotal  U vdw  U coul  Ubond  U angle  U dihedral

(1)

  12    6  N N   ij  ij U vdw (rij )    4 ij             r r i 1 j i 1 ij ij      

(2)

N N qq  U coul (rij )     i j    i 1 j i 1 rij 

(3)

U bond 

1 k a ( a  a0 ) 2 2

(4)

1 U angle  k (   0 ) 2 2

(5)

U dihedral  V11  cos( )  V2 1  cos(2 )  V11  cos(3 )  V11  cos(4 )

(6)

 ij   i   j  2

(7)

 ij   i j

(8)

where Uvdw, Ucoul, Ubond and Uangle and Udihedral represent the van der Waal interaction energy, electrostatic energy, bond energy, angle energy and dihedral energy respectively. σij and εij indicate the effective 4|Page ACS Paragon Plus Environment

Page 5 of 32 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


diameter and potential well depth respectively. q, Λ, ka, a, a0, kθ, θ, θ0, V, Ƴ are used to represent the charge of atom, dielectric constant of the medium, bond constant, bond length, equilibrium bond length, angle constant, angle bend, equilibrium angle bend, dihedral constants and dihedral coefficients respectively.

The sequence of the work follows first the computational modeling of CMPS-DB18C6 monomer units and then the validation of designed resin models by matching the MD results with the properties of the inhouse synthesized CMPS-DB18C6 resin. Further, the binding affinity of designed monomer units was tested for different concentration of Li+ ions in the aqueous phase as well as in the acidic medium. The charge neutrality of the systems was maintained by adding the desired number of counter-ions. Furthermore, the systems were extended for a cluster of monomer units in order to closely approach the adsorption experiments. The adsorption isotherms were generated using Li+ ion concentration from 0.2M to 1.2M. Further, the influence of aqueous phase acidity was incorporated by varying the acid concentration from 0.2M-1.0M. For each of these systems, the number of water molecules was fixed to 1500. The details of the simulation box for all the systems are provided in Table1. All the simulations reported herein were performed with LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) molecular dynamics package47 using NPT conditions (temperature 298 K and 1 atm of pressure) for equilibration run and NVT conditions (temperature 298 K) for production run carried for 50 ns and 10ns respectively. Furthermore, the images from the post analysis of simulation data were produced by VMD (visual molecular dynamics) graphical package48.

Macrocyclic host (DB18C6) Resin model M1

Resin model M2

Resin model M3 Figure 1. Macrocyclic host (DB18C6) and designed monomer unit models for functionalized DB18C6-CMPS resin ACS Paragon Plus Environment

5|Page


Page 6 of 32

Table 1. Details of the simulation box.

System name

NResin[model]

NLi

Ncounter-ion

Nwater

Nacid

Boxsize [ÅxÅxÅ]

---

38.225x29.731x29.731 40.107x31.194x31.194 40.368x31.397x31.397 38.396x29.864x29.864 40.376x31.404x31.404 40.368x31.302x31.302 59.848x29.924x29.924 59.957x29.978x29.978 62.790x31.395x31.395

Single monomer units A1 A1/ A1// A2 A2/ A2// A3 A3/ A3//

1 [M1]

1

1 [M2]

1

1 [M3]

3

--1 [Cl-] 1 [NO3-] --1 [Cl-] 1 [NO3-] --3 [Cl-] 3 [NO3-]

1500

Five monomer units B1 [Li+ ion concentration effect] 5 [M1]

B1/ [Acid concentration effect]

5 7 10 12 15 5 7 10 12 15

--------5 [Cl-] 7 [Cl-] 10 [Cl-] 12 [Cl-] 15 [Cl-]

10

10

30 50 70 90

30 50 70 90

90

90

30 50 70 90

30 50 70 90

90

90

30 50 70 90

30 50 70 90

90

90

--1500

12 25 40

73.733x18.645x18.645 73.861x18.677x18.677 73.879x18.682x18.682 74.159x18.753x18.753 74.162x18.784x18.784 73.737x18.646x18.646 73.845x18.673x18.673 74.008x18.714x18.714 74.015x18.721x18.721 74.087x18.735x18.735 66.876x19.983x19.983 67.406x20.144x20.144 66.409x20.609x20.609

Monomer cluster C1 [Li+ ion concentration effect] 90 [M1] C1/ [Acid concentration effect] C2 [Li+ ion concentration effect] 90 [M2] C2/ [Acid concentration effect] C3 [Li+ ion concentration effect] 30 [M3] C3/ [Acid concentration effect]

--1500 30 55 100 --1500 30 55 100 --1500 30 55 100

50.301x50.301x50.301 50.551x50.551x50.551 50.754x50.754x50.754 50.851x50.851x50.851 50.988x50.988x50.988 51.076x51.076x51.076 51.568x51.568x51.568 57.538x57.538x57.538 57.797x57.797x57.797 57.709x57.709x57.709 58.008x58.008x58.008 57.898x57.898x57.898 58.124x58.124x58.124 58.379x58.379x58.379 56.249x56.249x56.249 56.275x56.275x56.275 56.513x56.513x56.513 56.648x56.648x56.648 56.713x56.713x56.713 56.934x56.934x56.934 57.307x57.307x57.307


Page 7 of 32 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. RESULTS AND DISCUSSION 3.1 Experimental Validation of Resin Models Earlier, quantum simulation studies were performed to find out the suitability of the DB18C6 for Li+ ion adsorption, which was also corroborated by solvent extraction experiments. Furthermore, in order to reduce the aqueous phase solubility and to enhance the adsorption capacity, DB18C6 was anchored on suitable polymer resin22. Though macroscopic experimental parameters were generated, no microscopic investigations were attempted on the behavior of metal ions within the polymer functionalized crown ether. This prompt us to investigate the structural and dynamical behavior of ion adsorption for functionalized crown ether in the aqueous and acidic medium. The limitations of system size in Quantum studies, inclined our focus towards the MD studies of DB18C6 embedded crown ether in CMPS framework. Next, the question arises on length and structure of implanted monomer units. To know the effect of polymer chain length, 3 resin models were selected as shown in Fig.1. We validate the selection of our resin models with the shortest chain length of simulated DB18C6-CMPS resin model of Fig.1 by preparing the amorphous polymer from M1 monomer units. The results were compared with in-house synthesized CMPS-DB18C6 resin. The synthesis details of this resin are available in our earlier study 22.

In order to create the amorphous polymer of DB18C6-CMPS, first, free polymer surface was prepared by elongating one of the box dimensions to three times of original box dimension and then equilibrating the periodic system in isometric-isothermal (NVT) conditions at elevated temperature followed by stepwise cooling up to room temperature. The density of designed polymer was estimated to be 1.08 g/cm3, which was quite close to the experimental value of 1.18 g/cm3. Fig.2(a) shows the density distribution of simulated resin in x, y and z directions, which was calculated by binning the box into the rectangular bins with bin-width of 1Å in that particular direction. For example, to determine the density distribution in the x-direction, the box was distributed into the rectangular bins of dimension 1 x Ly x Lz [Å x Åx Å], where Ly and Lz are the box dimension in the y-direction and z-direction respectively. Further, the density of resin was estimated for each bin and plotted with respect to corresponding x-position of the bin. Similar to this, the density distribution for the y-direction and z-direction were considered. The results illustrate that the resin density is sufficiently uniform in all the dimensions. The variation of density is within 0.4% to 0.6%, which is within an acceptable range according to the literatures49 and therefore, the considered chain length of monomer unit can be assumed to be well suitable for further use. Additionally, the glass transition temperature of the constructed amorphous polymer was determined from the plot of specific volume vs. temperature over a range of temperature corresponding from rubbery state to glass state of



Page 8 of 32

polymer50 as shown in Fig.2(b). Furthermore, the least square fit to these data points was used to determine the volumetric thermal expansion coefficients as per equation.9 and corresponding results of Tg and α are shown in Table.2, which are in excellent agreement with the experimental observations.

  1      T  

  

(9)

Where ρ is density in g/cm3, T represents temperature and α correspond to volumetric thermal expansion coefficient. In addition, the other important parameters such as cohesive energy density51 (ECED) and solubility parameters52 (δ) for the designed polymer model are also reported in Table.2. ECED quantifies the intermolecular interactions between polymer chains, which is estimated using equation.10.

ECED  U vdw  U Q  V

(10)

where Uvdw and UQ represent the van der Waal energy and electrostatic energy respectively and V is the volume of the system.

Higher the ECED, stronger will be the interaction among polymer chains. ECED is fundamental property which can be used for calculation of various other properties, the example includes solubility parameter δ i.e. equal to (ECED)1/2. Solubility parameter δ enumerates solute-solvent interactions, which can be further used to estimate important information regarding swelling, permeation and other properties of the considered polymer system.

Figure 2. (a) Mass density distribution profiles and (b) Specific volume vs. Temperature for DB18C6-CMPS resin. 8|Page ACS Paragon Plus Environment

Page 9 of 32 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 2. Density (ρ), glass transition temperature (Tg) and expansion coefficients for the rubbery state (αr) and glass state (αg) Property (unit)

MD results

Exp. results

1.08

1.18

439.00

423-433

αr (10-4 K-1)

5.93

---

αg (10-4 K-1)

2.68

---

ECED(x108cal/m3)

87.88

---

δ (x104 (cal/m3)1/2)

9.37

----

ρ (g/cm3) Tg (K)

The excellent match of density and glass transition temperature for the simulated polymer of amorphous CMPS-DB18C6 with the in-house synthesized CMPS-DB18C6 resin, validate our selection of resin models. Further, these resin models have been implemented for a molecular level understanding of ionpolymer binding, which has been discussed in forthcoming sections, considering the Li+ ions in the aqueous phase as well as in the acidic environment.

3.2 Binding with Polymer functionalized crown ether Driven by the experimental support of designed resin model, we investigated the binding behavior of these designed monomer models by solvating them in a cubic box, containing 1500 water molecules, with Li+ ions near the crown ether cavity, named as system A1, system A2 and system A3 for model M1, model M2 and model M3 respectively. In these systems, the effect of counterion was incorporated by introducing (I) chloride ions, and (II) nitrate ions, specified by the suffix of single prime (/) for Cl- ions and double prime (//) for NO3- ions (refer to Table 1). The final snapshots of these systems extracted at t=60ns are shown in Fig. 3(a)-3(c); (a) in absence of any counter ion (b) in presence of chloride counter ions and (c) in presence of nitrate counterions respectively. The initial structures of these systems at t=0 ns are provided in the supporting information (Fig.S1). The results show that during the simulation, the Li+ ions remain intact within the crown ether cavity for model M1 and M2. However, as far as the M3 model is concerned, it was found that the end crown ethers remain occupied by Li + ions, whereas the Li+ from the middle crown ether was diffused away from the cavity. Also, for all the resin models, the DB18C6 cavity was observed to be highly folded to capture the Li+ ion, which is similar to the case when DB18C6 alone was considered in the aqueous phase (Fig.S2). Interestingly, the adsorption of Li+ ions 9|Page ACS Paragon Plus Environment


Page 10 of 32

was found to be unaffected by the presence of counter ions. The snapshots in Fig.3(a)-3(c) show that the counter-ions remain dispersed in the aqueous phase rather than binding with the Li+ ions attached to the crown ether. However, some of the nitrate and chloride ions were found to form an ion pair with the free Li+ ions. Also, to be noted, for M3 model in all the cases, only end DB18C6 were occupied and the middle one was always empty, which might be because of strong steric effect sensed by the middle crown ether than the end crown ethers, which on the other hand reduces the Li+-DB18C6 interaction for the middle one. Further, the binding strength of DB18C6-Li+ was estimated by calculating the interaction energy of Li+ODB18C6 pairs using LAMMPs package. In this context, it might be worthwhile to mention that since M3 model had total three DB18C6 rings unlike the M1 and M2 models which had only one DB18C6 ring, therefore the interaction energy obtained in M3 model was divided by three so that the binding energy for M3 model could be compared with the M1 and M2 models. In addition, the MD simulation for DB18C6/Li+ in aqueous medium was conducted to estimate the binding strength of DB18C6-Li+ in absence of any polymer chain. The corresponding results shown in Table.3, report the binding strength of Li+ to DB18C6 in the order of DB18C6>M1>M2 >M3. This indicates that the addition of polymer chain to DB18C6 reduces its interaction strength with the metal ions. In a real experiment, embedding of the polymer chain disperses the dipole strength of crown cavity, this, on the other hand, reduces the interaction strength of the ion-dipole type of interaction between the considered metal ion and the crown ether. Studies indicate that the presence of counter ions reduces the interaction strength slightly. However, the interactions energies of Li+-crown with both the Cl- ions, and NO3- ions were observed to be quite close for all the resin systems studied here.

Table 3. Binding Strength of DB18C6-Li+ in kcal/mol for water solvated DB18C6/resin monomer and Li+ ions. BE(DB18C6-Li) (kcal/mol)

DB18C6

M1

M2

M3

In absence of counter-ions

-312.7

-294.8

-265.9

-247.5

In presence of Cl- ions

-310.8

-290.6

-263.2

-244.3

In presence of NO3- ions

-310.7

-290.7

-263.0

-243.6

10 | P a g e ACS Paragon Plus Environment

Page 11 of 32 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 3. Adsorption of Li+ ions (yellow balls) at t=60ns on single monomer unit (presented by dynamic bonds, bonds of DB18C6 are shown bold) of model M1, model M2 and model M3 (a) In absence of counter ion (b) In presence of Chloride ions (green balls) and (c) In presence of nitrate ions (N: blue, O: red balls) (water solvent is shown in ice blue)

Furthermore, 5 monomer units of M1 resin model were considered with the total number of Li+ ions varied from 5,7,10, 12 and 15, corresponding to Li+ ion concentration of 0.32M, 0.45M, 0.64M, 0.76M and 0.95M and DB18C6:Li+ ratio to be 1:1, 1:1.4, 1:2, 1:2.4 and 1:3 respectively in the systems B. In each case, the results show only 1:1 type adsorption of DB18C6:Li+. In other words, only one Li+ ion was seen to be adsorbed into single crown ether cavity. However, the equilibration time (time required for all crown ethers to be occupied with Li+ ions) was seen to be reduced with an increase in Li+ ion concentration (data shown in Fig.S3). Also, it was noticed that the binding strength of DB18C6-Li+ is reduced for a higher concentration of Li+ ions, which might be due to a possible exchange of bounded Li+ ions with the unbounded Li+ ions (data not shown). The simulations were conducted initially in the 11 | P a g e ACS Paragon Plus Environment


Page 12 of 32

absence of counter-ions and then in the presence of equivalent chloride ions. Together both, show that the equilibration period is achieved at an earlier stage in the absence of counter-ions as shown in Fig.S4. Furthermore, the above simulation results with 0.45 M concentration of Li+ ions were extended to study the effect of acid concentration in the system B1/. In this case, the number of hydronium (H3O+) ions was increased from 12 to 25 and then 40 corresponding to the acid concentration of 0.75 M, 1.5 M and 2.5 M respectively. Results showed the reduced adsorption of the Li+ ions in presence of H3O+ ions. The quantity of free Li+ ions was increased with increase in the acid concentration as shown in Fig.4. With 0.75 N acid concentration, three out of five monomer units were found to be adsorbed by H3O+ ions whereas remaining two were Li+ ion bounded. On the other hand, with a higher concentration of acid i.e. 1.5 M and 2.5 M, only one crown ether was observed to be Li+ ion bounded and remaining four were adsorbed by H3O+ ions. Also, with an increase in acid concentration, the counter ion pairing of Li+ ions

Figure 4. Adsorption of Li+ ions (yellow balls) and H3O+ (O: red balls and H: white balls) for model M1 (dynamic bond, DB18C6 is shown bold) at acid concentration of (I) 0.7M (II) 1.5M and (III) 2.4M (Water and counter ions are not shown to make images clearer)


Page 13 of 32 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


was increased as expected.

Overall, the above simulation results demonstrate that presence of counter ions will increase the adsorption time for Li+ ions (data shown in Fig.S4). Besides, the reduced Li+ ion adsorption in presence of high acid concentration is the indication of regeneration procedure in the experiments, where the resin is regenerated by desorbing the Li+ ions with high concentration of acidic solution (experimental details are shown in Fig.S5). In the regeneration process, the H3O+ ions replace the Li+ from the crown ether cavity and the same has been captured by our MD simulations.

3.3 Adsorption behavior of resin clusters During the experiment, the purification of Li+ ions using grafted-resin is carried out in a column chromatography mode, where the resin is filled inside the column and the lithium-chloride solution is passed through to adsorb Li+ ions in column bed, which is in general known as elution chromatography. In order to understand the adsorption mechanism of Li+ ions by crown ether functionalized materials, a part of resin bed was simulated by considering the cluster of monomer units such that each system had 90 crown ether units. Therefore, total 90, 90 and 30 monomer units were considered for model M1, M2, and M3 respectively (refer to system C in Table.1). Each of them was solvated with 1500 water molecules. The final snapshot for M1 resin model is shown in Fig.5. The systems were simulated for 0.2M-1.0M range of Li+ ion concentration in order to demonstrate the adsorption isotherms. Also, the effect of acid

+

+

Figure 5. 3D image of system C1 with 1.0M Li concentration and snapshot showing the adsorption of Li + + ions and H3O ions in DB18C6 cavity [Li and Cl are presented by yellow and green balls, bold dynamic 13 | Pbonds age shows DB18C6 cavity, whereas polymer backbone is shown by light dynamic bonds, water is represented by ice blue clouds] ACS Paragon Plus Environment


Page 14 of 32

concentration was established for the adsorption of Li+ ions by utilizing the acid concentration in the range of 0.2M-1.0M.

Though all the ions were initially allocated randomly within the box during the simulation, they were migrated to the cavity. DB18C6 has cavity size of 2.82Å, which is sufficient enough to accommodate Li+ (ionic radius of 0.94Å) in their hydrophilic cavity as shown in Fig.5. Furthermore, the structural distribution of surrounding atoms/molecules for the adsorbed ions was estimated using the radial distribution functions (RDFs) and Coordination number (CN) respectively53. Interestingly, the results in Fig. 6(a) indicate that RDF of Li+-ODB18C6 has peak position at 2.12 Å for model M1, which has been shifted to 2.17 Å for a linear and cross-linked model of M2 and M3. Therefore, it might be stated that though polymer chains do not directly interact with the Li+ ion, yet have a minor effect on the ion-crown ether interaction. Furthermore, the average coordination of Li+ ions to DB18C6 cavity was considered, which was found to be reduced from 0.67 to 0.58 and 0.51 while moving from model M1 to M2 and M3 respectively. This also supports that the addition of polymer chains have an effect on the binding of Li+ to DB18C6 cavity. As far as the adsorption of H3O+ ions is concerned, the results show the OH3O+- ODB18C6 RDF peak at 2.63 Å for M1 and at 2.68 Å for M2 and M3. The reason for a larger distance of H3O+ ion from crown oxygen as compared to the Li+ ions is supposed to be related to the nature of binding. Li+DB18C6 has metal ion-dipole type interaction, which is considered to be stronger than the reduced iondipole type interaction in between H3O+-DB18C6. Also, the dipole orientation in H3O+ is such that their H atoms point towards the crown cavity, leading to a comparatively long distance of OH3O+-ODB18C6 than the Li+-ODB18C6. To be noted, no difference in peak positions for hydration of adsorbed Li+/H3O+ as well as for ion-pairs of Li+-Cl- was observed for simulated resin models (see Fig.6 (a)). Nevertheless, the difference in heights of the RDF peaks represents the dissimilarities in the coordination of ions for the different resin models. The results demonstrate that the number of Li+-Cl- pairs are higher for M2 and M3 models than the M1 model. Also, the adsorbed Li+/H3O+ ions were found comparatively more water dehydrated for M1 model than the M2 and M3 models.


Page 15 of 32 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). Radial distribution functions (RDFs) for considered resin models



Page 16 of 32

Fig.6(b) represents the effect of Li+ ion concentration for M1 model. The results show that the peak positions remain constant with the increase in Li+ ion concentration. However, the peak heights are changed depending on the availability of ion pairs. It was observed that the peak heights of Li+-DB18C6, Li+-Cl-and Li+-Ow are increased with increase in Li+ ion concentration, which demonstrates that with the increase in Li+ ion concentration, the availability of both the free Li+ ions as well as the total number of adsorbed Li+ ions in the systems is increased. Enhancement of free Li+ ions would increase the number of Li+-Cl- pairs, and hydration of Li+ ions. On the other hand, increase in a number of adsorbed Li+ ions in the systems intensifies Li+-DB18C6 coordination. The same has been discussed and made clear in later sections of the article. To the surprise, at Li+ ion concentration of 0.4M, the first coordinated shell of chloride ions to Li+ was observed to be absent as the first RDF peak appeared at 5.28Å for model M1 and M3. This might be related to the orientation of DB18C6 in the polymer backbone of resin model, which plays an important role, especially for the very low concentration of Li+ ions. In particular, at the very low concentration of Li+ ions, almost all the Li+ ions are adsorbed by DB18C6 cavities, leaving negligible free Li+ ions to be paired with the chloride ions. Furthermore, the pairing of chloride ions with the adsorbed

Figure 6(b). Radial distribution functions (RDFs) profiles for M1 model at different Li+ ion concentration [L1, L2, L3 and L4 represent the number of Li+ ions to be 30, 50, 70 and 90 respectively. Li+ ions depends on the orientation of DB18C6 in the polymer backbone, which on the other hand results 16 | P a g e ACS Paragon Plus Environment

Page 17 of 32 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


into the dissimilar Li+-Cl- pairs for different resin models at the low concentration of Li+ ions in the system.

The effect of acid concentration on the radial distribution functions has been displayed in Fig.6(c). All these systems were simulated with fixed Li+ ion concentration, maintaining 1:1 ratio of DB18C6:Li+ (refer to systems named as C/ in Table.1) The results show that the peak intensities of Li+-DB18C6 are reduced with increased acid concentration, which represents the decreasing Li+ adsorption with the increase in acid concentration. Reverse to it, the H3O+ adsorption was observed to be increased with increase in acid concentration as seen from the increasing peak intensity of H3O+-DB18C6 with higher acid concentration. This trend is in accordance with the experimental observations, followed during the regeneration step of the resin in column chromatography. In addition, it has been found that the coordination of water to both the Li+ and H3O+ is increased with increase in acid concentration. This seems to be related to the rise in the number of free (unbound) Li+ and H3O+ ions with an increase in acid concentration, which is estimated and discussed furthermore in impending sections.

Figure 6(c). Radial distribution functions (RDFs) profiles for M1 at different acid concentrations [A1, A2 and A3 correspond to number of acid ions to be 30, 55 and 100 respectively, L4 means the number of Li+ ions is fixed to be 90. 17 | P a g e ACS Paragon Plus Environment


Page 18 of 32

3.4 Adsorption isotherms Further, the separation abilities of the designed resin models have been explored by adsorption isotherms. The system C1, system C2, and system C3 were used for the adsorption studies of model M1, model M2 and model M3 under the varied concentration of Li+ ions. Also, the effect of acid concentration on the adsorption of Li+ ions was studied by using the system C1/, system C2/ and system C3/ for model M1, model M2 and model M3 respectively (refer to Table.1). For all these systems, the number of Li+ ions and H3O+ ions adsorbed to DB18C6 cavity were counted and converted into the unit of mol/mol [i.e. moles of ions adsorbed per mole of DB18C6 cavity], named qe as in equation.11., Further, the data of C (concentration of ions in the systems in mol/L] and qe was plotted in Fig.7 and also fitted to Langmuir adsorption isotherm (equation.11) to get the maximum adsorption capacity qmax of corresponding resin model. To the best of our knowledge, this is the first report of adsorption isotherm of functionalized crown ethers by means of MD simulations. The MD results well captured the experimentally observed Langmuir type adsorption isotherms (Fig.S6) as the adsorption data fit equation.11 extremely well, with a correlation coefficient (R2) close to one.

qe 

qmax BC 1  BC

(11)

Where qe is the moles of Li+ ion adsorbed per mole of the crown ether, C is a concentration of Li+ ions in the system in mol/L. qmax and B are the fitted parameters where qmax represents the maximum moles of Li+ ions adsorbed per mole of crown ether and B is the equilibrium constant in the unit of L/mol.

The qmax obtained from fitting of adsorption isotherms are summarized in Table 4. The results indicate the highest adsorption of Li+ ions with M1 model and minimum with the M3 model. The results are in line to the simulation results carried with single monomer unit of each resin model in the cubical box of water, where the interaction of Li+-DB18C6 was found to be strongest for M1 model and weakest for M3 model. The results show that the net adsorption of Li+ ions is decreased with increase in the polymer chain. In other words, the Li+ ions are supposed to adsorb better by DB18C6 alone than the DB18C6 grafted with a polymer chain. Also, the maximum adsorption capacity will decrease with increase in length or branching of the added polymer chain. In the case of free crown ether, the adsorption is supposed to be higher because of their direct exposure to Li+ ions. However, while grafting, few Li+DB18C6 interactions might be shielded due to the presence of polymer backbone. With an increase in polymer branching or polymer chain length, the shielding thickness increases, leading to reduced Li+DB18C6 interaction. However, the overall adsorption of Li+ ions by crown ether grafted polymer resin


Page 19 of 32 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


seems to follow the Langmuir model in all the cases and the same is expected with other monovalent or divalent alkali and alkali earth metal ions.

Figure 7. Adsorption isotherms for Li+ ions for (a) Model M1, (b) Model M2 and (c) Model M3 in aqueous phase

Further, the adsorption/desorption behavior of Li+ ions was studied in the presence of hydronium ions. The number of H3O+ taken were 30, 55 and 100 corresponding to the acid concentration of 0.2M-1.2M. All these acidic systems were simulated with 1M concentration of Li+ ions. Similar to Li+ ions, the results demonstrate the Langmuir type adsorption isotherms for H3O+ions as shown in Fig.8. Together, they lead to Langmuir type adsorption isotherms for a total number of ions (Li++H3O+) adsorbed to the different resin models.



Page 20 of 32

Table 4. Parameters qmax [mol/mol] for Langmuir adsorption Isotherm (eqn.11) Li

H3O

Total ions (Li+H3O)

Resin

qmax

qmax

qmax

M1

1.35±0.03

1.87±0.16

1.06±0.03

M2

1.26±0.04

1.63±0.17

1.01±0.01

M3

0.99±0.15

0.76±0.05

0.86±0.04

The results illustrate an increase in adsorption of H3O+, on the other hand, a decrease in adsorption of Li+ ions with an increase in the acid concentration as many of the adsorbed Li+ ions are replaced with H3O+ ions. The same was also validated by RDFs of Li+-DB18C6 and H3O+-DB18C6 for various concentration of acid ions in the systems, where with the increase in acid concentration, the RDF peak intensities for Li+-DB18C6 were observed to be reduced while increased for H3O+-DB18C6 (see Fig.6c). Nevertheless, the total number of ions (Li+ and H3O+) adsorbed by resin was always increased with increase in acid concentration for all the considered resin systems. The results of decreasing Li+ adsorption with an increase in acid concentration are in line with the experimental observations (Fig.S5). In fact, in presence of acid, many crown ether cavities are occupied by H3O+ ions and hereby remain unavailable to adsorb Li+ ions. As a result, net adsorption of Li+ ions is decreased. However, among the resin models, the adsorption of Li+ ions was found to be highest for M1 model for all the considered acid concentration, which is in accordance to our previous observations from RDFs of Li+-DB18C6 and the interaction strength for the pairing of Li+-DB18C6. Furthermore, the results showed a decreasing adsorption profile for Li + ions with increase in acid concentration. On the other hand, an increasing adsorption profile for H3O+ ions was observed with the increase in acid concentration. Interestingly, the results of qmax (including error bar) indicate that the Li+ ion adsorption highly depends on the selection of resin model, where H3O+ adsorption seems to be comparatively model independent. The reason might be related to the best-fit match of the DB18C6 cavity with binding ions of Li+ and H3O+. The ionic radius of Li+ and H3O+ ions are 0.94Å and 1.15Å respectively. Size of H3O+ matches better with DB18C6 cavity of size 2.82Å, whereas Li+ ions on the hand, being smaller in size requires the distortion of the cavity to be adsorbed as shown in Fig.4. Since, the distortion of crown ether cavity is very much dependent on nature, size, and orientation of the added polymer, therefore lead to model (of resin) dependent adsorption of Li+ ions.


Page 21 of 32 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. Adsorption isotherms for H3O+ions for (a) Model M1, (b) Model M2, and (c) Model M3; (d) adsorption isotherm of Li+ ions in acidic medium and (e) adsorption isotherm of total ions (Li++H3O+) for different resin models in acidic medium

Furthermore, the residence time of these ions in the crown cavity was estimated using the time correlation function RT(t), which is expressed as equation.12 21 | P a g e ACS Paragon Plus Environment


  (t ) (t  t )   (t ) (t )

Page 22 of 32

N

RT (t ) 

i 1 i

0

i

0

N

i 1 i

0

i

(12)

0

Where ξi(t0) was taken to be 1 if ion was bound to crown ether cavity and 0 otherwise. Further, the correlation function was fitted to an exponential curve to obtain the residence time. Fig.9(a) represent the time correlation function for the different resin models with Li+ ion concentration of 0.5M. Results show the very fast decay of time correlation function for a small time interval and very slow decay for a long time interval. Initial fast decay is supposed to be related with the movement of ions from a free state to bound (with crown ether) state whereas the slow decay at a longer time represents the correlation when ions are already in the bound state. Interestingly, it was observed that this correlation functions better fit to a sum of two exponential functions rather than single exponent curve. This is because once the ions are adsorbed, they are not able to come out of the cavity very easily, leading to the very long residence time of ions in the crown cavity. On the other hand, the events when ions are moved from free to bound state or vice versa ((0 1)), are rare and happens within the very short span of time, therefore have short residence time. Together both, make ‘adsorption in crown cavity’ to be a dual residence time function with a residence time of τ1 and τ2, where τ1 represents the event time when ions are moved free to bound state or vice versa and τ2 is the residence time when the ion is already bound inside the crown ether cavity. The initial decay was found to be faster for a higher concentration of Li+ ions and the same has been indicated by the data points of residence time τ1in Fig.9(b). Also, for the same Li+ ion concentration, residence time τ1 was observed to be decreased with increase in acid concentration. This might be due to the increase in the availability of a total number of ions to be exchanged from free to bound state and vice versa with the rise in either of Li+ or H3O+ concentration and therefore lead to a comparatively faster decay. This can also be linked with the hydration of ions, which was observed to be increased for both the Li+ ions as well as H3O+ ions with an increase in acid concentration due to rise in free ions (both the Li+ and H3O+). Furthermore, the results show a decreasing decay rate of the correlation function with increase in time. In case of τ2, it was difficult to find a pattern in residence time of Li + and H3O+. On an average, τ1 was observed to be in the range of 0.5-1.0ps, whereas τ2 was ~200ps. Such dual residence time behavior of crown ether has also been reported by earlier studies54, 55.


Page 23 of 32 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 9. (a) Time correlation function for occupancy of Li+ ions for different resin models using 0.5M concentration of Li+ ions in the aqueous medium

Figure9 (b). Residence time for adsorption of Li+ ions and H3O+ ions [Filled symbols relate to Left Y axis, whereas the lined symbols relate with Right Y axis]

3.5 Dynamics of binding sites

Furthermore, the estimated MD results are supported by the dynamical behavior of binding sites, explored with well-known Einstein relation (equation13)

D

1 d lim R (t )  R (0) 2 t  2d dt

(13) 23 | P a g e ACS Paragon Plus Environment


Page 24 of 32

Where R(t) represents the position of atom/molecule at time t, denotes the mean square displacement profile, d and D respectively indicates the dimension of the system and the diffusion coefficient. It has been observed that the diffusivity of Li+ ions is largely suppressed in presence of grafted resin, which is due to their confinement in DB18C6 cavities. Surprisingly, on the contrary of bulk systems, the diffusion coefficient of Li+ ions in presence of resin model, was observed to be increased with the increase in Li+ concentration as shown in Fig.10(a) and Fig.10(b) respectively. In particular, while the increase in Li+ ion concentration from 0.2M to 1.2M, the percentage of Li+ ions confined by DB18C6 is reduced from 90% to 50%, leading to an enhanced population of free Li+ ions, as a result, the average diffusion coefficient of Li+ ions is increased with increase in Li+ ion concentration. However, for all the considered concentrations, the average diffusion coefficient was always lesser than the estimated values in absence of resin. Furthermore, it was observed that the diffusivity of Li+ ions was minimum for the M1 model and very close in value for other two models. The results are in line with the number of unbound (free) Li+ ions, which was found to be minimum for M1 followed by M3 and M2 with minor differences. In case of acidic systems, the diffusivity of both the Li+ ions as well as H3O+ ions was observed to be increased with increase in acid concentration as shown in Fig.10(c). However, the enhancement in diffusion coefficient was considerably high beyond the acid concentration of 0.7M. One should borne in mind that for such systems, the diffusion coefficient can’t be solitarily described by mass but is a complicated function of (i) number of free ions (ii) the mass of ions and (iii) counter ion pairing. Therefore, it is difficult to compare the diffusion coefficient of Li+ ions and H3O+ions for the simulated resin systems in the acidic environment. As shown in Fig.10(d), for aqueous systems, the dynamics of associated water channel was observed to follow order M1

Structure, Dynamics and Adsorption of Charged Guest within the

Recommend Documents