This article was downloaded by: [RMIT University]On: 02 September 2013, At: 06:49Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Molecular SimulationPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/gmos20

Behaviour of hydrogen bonding and structureof poly(acrylic acid) in water–ethanol solutioninvestigated by explicit ion molecular dynamicssimulationsSriram Srikant a , Sulatha S. Muralidharan a & Upendra Natarajan aa Molecular Modeling and Simulation Lab, Department of Chemical Engineering , IndianInstitute of Technology (IIT) Madras , Chennai , 600036 , IndiaPublished online: 21 Sep 2012.

To cite this article: Sriram Srikant , Sulatha S. Muralidharan & Upendra Natarajan (2013) Behaviour of hydrogen bondingand structure of poly(acrylic acid) in water–ethanol solution investigated by explicit ion molecular dynamics simulations,Molecular Simulation, 39:2, 145-153, DOI: 10.1080/08927022.2012.708417

To link to this article: http://dx.doi.org/10.1080/08927022.2012.708417

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Behaviour of hydrogen bonding and structure of poly(acrylic acid) in water–ethanol solutioninvestigated by explicit ion molecular dynamics simulations

Sriram Srikant1, Sulatha S. Muralidharan2 and Upendra Natarajan*

Molecular Modeling and Simulation Lab, Department of Chemical Engineering, Indian Institute of Technology (IIT) Madras, Chennai600036, India

(Received 23 December 2011; final version received 26 June 2012)

Molecular dynamics simulations of polyelectrolyte poly(acrylic acid) at infinitely dilute conditions in aqueous solutionscontaining up to 6 vol% ethanol were performed with explicit solvent and counter-ion description for varying degree ofneutralisation. In the range of ethanol concentration employed, this study shows a swelling in the case of non-ionised chain,chain collapse for 25% ionised system, minimal swelling for 50% ionised chain and no significant change in conformationfor 75–100% ionised chains. Ethanol interacts with non-ionised residues via hydrogen bonding and is found to be dominantin the solvation cage surrounding them, whereas no such interaction is seen with ionised residues. Thus, chains with highercharge densities show a preference for the aqueous environment in comparison to ethanol and hydrogen bonding to water isunaltered. For chains with lower charge density, a reduction in the hydrogen bonding with water is seen with an increase inethanol concentration.

Keywords: anionic polyelectrolyte; aqueous; binary solvent; ethanol; hydrogen bonding

1. Introduction

Polyelectrolytes [1,2] constitute an interesting class of

macromolecules, which dissociate releasing counter-ions

in solution. Aqueous and polar solvent molecules interact

with these chains primarily through their functional groups

[3]. Binary solvent systems could interact differently

with the functional groups on a polymer chain, thereby

modifying the polymer chemistry and influencing the

physical properties. Polyelectrolytes find utility in cos-

metics and detergents and other chemical applications such

as biomedical systems [4] like dental adhesives, controlled

release devices and polymeric drugs all of which contain

polyelectrolytes.More recently, polyelectrolytes have been

used as fluorescent biosensors, polymer light-emitting

diodes and polymer solar cells [5], further expanding their

functional diversity. To supplement experimental efforts

for understanding the properties of such a system, it is

important to establish an understanding of the molecular

conformations and structural motifs of the system

elucidated through molecular simulation.

Several synthetic anionic polyelectrolytes constituted

with hydrophobic side chains and side groups, such as

poly(methacrylic acid) (PMA), copolymers of maleic acid

with n-butyl vinyl ether as well as with styrene and a-

methyl styrene and others have been known to show

conformational transitions that are very similar to the

conformational transition accompanying protein denatura-

tion. The transition in these synthetic polyelectrolytes, in

going from a compact coil to an extended chain, is driven by

changes in pH of the aqueous solution arising from the

degree of ionisation of the acid groups along the chain. The

conformational transition of the denaturation in natural

macromolecules such as proteins, in going from its natural

soluble coil state to a collapsed globule state, is known to

take place by the addition of hydrophobic organic

molecules such as urea. The effect of urea on charge-

induced transition from hypercoiled to extended state

conformation of alkyl vinyl ether–maleic acid copolymers

in NaCl solutions is a classic and well-known example. In

the case of natural macromolecules, the denaturant activity

of structurally and chemically very different compounds

such as alcohols, detergents, urea and guanidine-based

compounds is very well known and experimentally a

general phenomenon in biopolymers. The traditional long-

standing approach of investigations towards identification

of plausible sites of interaction and estimation of the

corresponding thermodynamic contributions to the dena-

turation process in biopolymers has been to study small

molecule model compounds, chemically similar to protein

moieties, interacting with protein molecules in solution [6].

Considerable evidence exists from model compound

studies to the effect that urea interacts strongly with peptide

backbone groups thus weakening inter-peptide hydrogen

bonds [7,8] and enhancing the solubility of non-polar

groups thereby reducing the hydrophobic interactions

either by altering the bulk properties of the solvent

ISSN 0892-7022 print/ISSN 1029-0435 online

q 2013 Taylor & Francis

http://dx.doi.org/10.1080/08927022.2012.708417

http://www.tandfonline.com

*Corresponding author. Email: unatarajan@iitm.ac.in

Molecular Simulation

Vol. 39, No. 2, February 2013, 145–153

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

or through more localised effects [7,9,10]. However,

relative to natural macromolecules, it is easier to assess

the relative significance of such effects in studies with

synthetic polymers.

The work presented here follows past literature

involving simulations of polyelectrolytes directed towards

understanding their behaviour in solution and that of the

solvent shell surrounding the chain under dilute conditions

[11–13,14]. These earlier studies have focused on the

behaviour of the chain under dilute conditions with a

single molecule of the polyelectrolyte. Few other studies

have described interactions between two or more chains

and the resulting supra-molecular structures [15,16]. The

work presented in this study deals with a situation in which

a different molecule containing hydrophobic groups is

present and for which the interaction with the polyelec-

trolyte competes with that of water.

The motivation for the work here is to investigate the

influence of aliphatic hydrophobic residues (CH2 groups)

in perturbing the hydrogen bonding structure near the

polyelectrolyte chain and to its conformations in a mixed

solvent system. In particular emphasis, we have therefore

investigated the ethanol–water mixture, with variations in

ethanol content and charge density of the chain, via

atomistic molecular dynamics (MD) simulations. The

properties of ethanol–water solutions can be accurately

represented, as was shown by MD simulations [17–19].

Experimental measurements and simulations pertain-

ing to chain length versus hydrodynamic radius have shown

that short poly(acrylic acid) (PAA) chains can be used to

calculate and relate to properties of longer chains (or

macroscopic properties) [12,13]. Scaling behaviour of

sodium polyacrylate (Na-PAA), as measured by light

scattering experiments, was found to be in excellent

agreement with the results from coarse-grained computer

simulations, with the developed coarse-grained model

based on the atomistic MD simulation for a 23 repeat unit

chain. The details of (a) polyelectrolyte conformations

(global and local), (b) non-bonded interactions with the

solvent, (c) the structure of the solvation shell and (d) the

interaction with counter-ions are available from that study

[12,13]. Our recent atomistic MD simulation study of 20

repeat unit PAA and PMA chains in dilute aqueous solution

has shown that a PMA chain in water exhibits stronger

structural correlationwithNaþ counter-ions as compared to

PAA [20]. This arises due to water being a relatively poor

solvent for PMA as compared to PAA.

To our knowledge, the present paper is the first

atomistic simulation study testing the influence of

hydrophobicity on the structure of a polyelectrolyte in a

binary solvent. An earlier atomistic MD study of solvation

structure and local dihedral angle and hydrogen-bonding

lifetime dynamics of a polar polymer [poly(vinyl alcohol)

(PVA)] in water–ethanol mixtures has shown important

aspects of polymer behaviour [21]. It becomes important to

investigate water–ethanol mixtures for separation science

and technology and other practical applications. Funda-

mentally, the issue of the hydrophobic effect comes into

play in such a system. The case of polyelectrolytes is

complicated by the variability in the degree of ionisation as

well as by the presence of ionised groups and counter-ions.

This study focuses on the influence of solvent composition

and degree of ionisation of the polyelectrolyte on the chain

conformations and solvation cage. The conformationally

averaged values of radius of gyration (Rg), the radial

distribution functions (RDFs) of the constituent groups and

the extent and type of hydrogen bonding would aid in

providing an understanding of structural changes that

occur in PAA chains due to variations in charge density and

ethanol concentration. We focus attention on a specific

counter-ion (Naþ) with explicit ion description.

2. Simulation methodology

Test simulations using GROMOS 53a6 force field [22]

showed that Rg values of an atactic chain and a syndiotactic

chain with 20 repeat units (n ¼ 20) were comparable (1 nm

for fully ionised chains). For 50% ionised chain (for

n ¼ 20), these values were 0.87 nm (syndiotactic) and

0.88 nm (atactic). It is known from experimental

measurements that solution properties of atactic and

syndiotactic PAA are indistinguishable [23,24]. Therefore,

here syndiotactic chains were used in order to present

stereo-chemically well-structured chains for probing the

structure and influence of solvent polarity of the mixed-

solvent system.

The polymer chains were generated with a ‘fully

atomistic’ description given by Flory, Rotational Isomeric

State Metropolis Monte Carlo method, to produce a chain

with different dihedral (torsion) angles at the backbone

bonds. This was followed by potential energy minimis-

ation to give the chain conformation, which was used to

prepare the dilute solution with water in the periodic box.

The tacticity (stereochemical sequencing) of the chain was

specified to be syndiotactic.

PAAwas modelled as one syndiotactic polymer strand,

with 20 repeat units solvated in 4000 water molecules

(concentration excluding counter-ions, 0.014 mol/l). The

acid form of the residue has the carboxyl group as COOH

with an explicit hydrogen, and the fully ionised residue has

the carboxyl group in the COO2 form. The degree of

ionisation was varied for the polyelectrolyte by the

replacement of COOH groups with COO2 along the chain.

Five different ionised forms were studied separately in

solution by having 0, 5, 10, 15 and 20 ionised carboxyl

(COO2) groups along the chain (charge density,

f ¼ Nc/N ¼ 0, 0.25, 0.5, 0.75 and 1, respectively, where

Nc is the number of ionised residues and N is the total

number of residues). An adequate number of Naþ ions

were added to the solvated system in order to maintain

S. Srikant et al.146

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

overall charge neutrality. In the partially ionised

polyacrylates, the ionised residues are distributed at the

chain ends and at the centre of the chain, i.e. residues 1, 2,

11, 19 and 20 in the 25% ionised polymer; residues 1, 2, 3,

4, 10, 11, 17, 18, 19 and 20 in the 50% ionised polymer and

residues 1, 2, 3, 4, 5, 6, 10, 11, 12, 15, 16, 17, 18, 19 and 20

in the 75% ionised polymer. The chains are solvated in

ethanol–water system of varying composition, i.e. 0, 20,

40 and 80 molecules of ethanol in 4000 molecules of

water. This corresponded to 0, 1.6, 3.1 and 6.1 vol% of

ethanol, respectively.

The aliphatic carbon atoms are treated as united atoms

and polar hydrogens explicitly in the simulations. The

initial configuration was generated by placing a fully

stretched polyelectrolyte chain in a cubic periodic box

along with randomly distributed Naþ counter-ions, a pre-

determined number of ethanol molecules and SPC water

molecules [25]. The representation of ethanol was made

with an explicit hydroxyl hydrogen and united atoms for

the alkyl chain. Water geometry was constrained using the

SETTLE [26] algorithm. All minimisation and MD

simulations were performed using GROMACS (version

4.0.7) [27]. The GROMOS 53a6 parameter set [22] was

used in the MD simulations.

Lennard-Jones interactions described by a 6-12

potential form were truncated and shifted at 0.9 nm.

Electrostatic interactions were computed using a Coulomb

potential (cut-off at 1.4 nm) along with a reaction-field

correction, with the reaction field dielectric 1RF equal to

the value for water (1RF ¼ 78.5) [28]. Bond lengths were

held constant using the SHAKE procedure [29]. Equations

of motion were integrated using a leap-frog algorithm

using a time step of 2 fs. Other simulation parameters

included the following: a neighbour list, which was

updated every 10 steps with a list cut-off of 1 nm; weak

coupling [30] to temperature (T ¼ 300 K) and pressure

( p ¼ 1 atm) baths with coupling times tt ¼ 0.1 ps and

tp ¼ 0.5 ps (water compressibility 4.5 £ 10210 kPa21).

The solvated system was subjected to energy minimisation

without any constraints using steepest descent method.

100 ps NVT simulation run was performed, followed by a

100 ps NPT run with position restraints on the polymeric

chain, with no constraints on the water molecules, so that

temperature and pressure are at their equilibrium values.

This was then followed by NPT MD simulation for 15 ns,

and the data were recorded at every 500 steps.

3. Results and discussion

The PAA chain interacts with the surrounding ethanol,

water and Naþ ions primarily via the carboxyl groups. The

interactions of water and ethanol with the PAA chain are

presented in Figure 1 as taken from the simulated structure.

The spatial picture of the intra-chain hydrogen bonding and

of the hydrogen bonding between PAA, water and ethanol

molecules, as obtained from the MD simulation, is also

shown in Figure 1. There is a considerable interaction

between the carboxylic acid groups of the non-ionised

chain ( f ¼ 0) with ethanol molecules, in addition to their

interaction with water.

3.1 Chain conformations

The conformational state of PAA is described by the

distribution of radius of gyration (Rg) and is shown in

Figure 2. The non-ionised PAA chain ( f ¼ 0) is in a fully

coiled state in water. Increase in charge density along the

chain results in an enhancement of repulsive electrostatic

interactions between the COO2 groups, thereby leading to

the unfolding of the coil as shown by the relatively higher

values of Rg upon ionisation in aqueous solution. The

chain expansion due to increase in charge density as

observed in aqueous solution is seen in systems with

ethanol–water mixtures above a charge density of 0.5. The

non-ionised PAA chain swells with an increase in ethanol

concentration, while for Na-PAA ( f ¼ 0.25) the chain

collapses as shown by reduction in Rg values with an

increase in ethanol concentration. A minimal swelling is

observed in the case of Na-PAA ( f ¼ 0.5). For ionised

chains ( f ¼ 0.75 and above), the conformations are rather

unchanged in the range of ethanol concentration studied.

For Na-PAA ( f ¼ 0.25), we carried out additional

simulations on chains with different arrangements of the

ionised COO2 groups along the chain, and here also chain

coiling is observed with an addition of ethanol.

Weakly charged polyelectrolyte gels as well as single

polyelectrolyte chains in solution are known to collapse

with an increase in degree of ionisation, especially in less

polar solvents. This unusual behaviour with increasing

charge density has been probed by experiments and theory

Figure 1. (Colour online) (A)Snapshot showing the localisation ofethanol (with greenCatoms) andwater (stick representation) arounda PAA residue (with grey C atom). Oxygen atoms and hydrogenatoms participating in hydrogen bonding are shown explicitly in redand white, respectively. (B) Snapshot from the system containingPAA chain in 3.1 vol% ethanol–water mixture. A case of intra-molecular hydrogen bonding is highlighted in a blue circle.

Molecular Simulation 147

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

[31–37]. The collapse in less polar solvents has been

ascribed to the formation of ion pairs between ionised

polyelectrolyte and counter-ions, which can aggregate due

to dipole–dipole attraction. For PAA chains, this collapse

was observed in methanol using a variety of experimental

techniques in which the polymer first expands and then

collapses with an increase in the degree of neutralisation

[34]. Morawetz et al. [35] reinvestigated this phenomenon

with titration of PAA and PMA in methanol and ethanol

through solution viscosity and NMR measurements. On

the basis of that seminal study it was concluded that the

polyion collapse occurs due to attraction between the ion

pairs formed in a medium with relatively low dielectric

constant, which also inhibits the conformational mobility

of the chains, as revealed from the NMR spectra. This

unusual behaviour of chain shrinkage at higher charge

density has been theoretically explained based on the

energy gain in the formation of ion pairs in a collapsed

state of low polarity, which competes with the swollen

state where most of the counter-ions are mobile in the

solution [33]. From theoretical analysis provided by

Khokhlov and Kramarenko [33], the results obtained for

polyelectrolyte gels are applicable for weakly charged

polyelectrolyte macromolecules. Swelling of PMA gels

accompanying an increase in charge density with various

organic and aqueous-organic solvent mixtures was found

to depend on the dielectric constant of the medium in

agreement with theoretical derivations. The swelling

behaviour of PAA gels in water–ethanol mixtures has

been investigated experimentally as a function of charge

Figure 2. Distribution of polyacrylate radius of gyration, Rg for with varying ethanol concentration. The graphs are shown for the variouscharge densities of the polyacrylate chain.

S. Srikant et al.148

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

density [35,36]. An increase in the degree of swelling in

water (0% ethanol) with an increase in pH was observed;

however, it remains constant beyond pH ¼ 5. A similar

behaviour was observed for systems containing up to 30

vol% ethanol. In mixtures with 40–60 vol% ethanol, a

sharp deswelling (chain collapse) was observed for

pH . 4. In the PAA gel (pH 3, degree of ionisation

estimated to be 0, even though a few COOH groups are

dissociated), swelling ratios were constant for 0–80 vol%

of ethanol, with a significant deswelling observed above

80 vol%.

No experimental reports exist, exactly in the mixed

solvent concentration range investigated here, to make

direct comparison with our simulation results for PAA

chains. The available experimental data on PAA gels do

not provide details regarding the interactions at the

microscopic scale between the polyelectrolyte chain and

the solvent molecules. This study indicates a swelling of

non-ionised PAA and chain collapse for 25% ionised

system with added ethanol, whereas chain above 50%

ionisation remains unaffected. This study, although at low

ethanol concentration, sheds light on the preferential

interaction between the COOH groups and ethanol as

compared to the COO2 groups as given by the analysis

presented in the following sections.

3.2 Hydrogen bonding

The ionised COO2 group is capable of participating only

as a hydrogen bond acceptor. Therefore, in systems except

those containing completely ionised PAA, there is a

possibility of intra-molecular hydrogen bonding which

could stabilise the polymer conformation. The COOH

groups show propensity for the formation of hydrogen

bonds as either acceptors or donors. We used the criterion

that a hydrogen bond is formed provided the two oxygens

are within a distance of 0.35 nm and the OZHZO angle is

.1308. Addition of ethanol to the non-ionised system

leads to an increase in intra-molecular hydrogen bonding.

The number of hydrogen bonds present at any time during

the simulation in the case of non-ionised PAA increases

from 6.96 (0 vol% ethanol) to 10.5 (6.1 vol% ethanol) (the

mean error in the number of hydrogen bonds is ,5% in all

the cases reported here). For ionised polyacrylates

( f ¼ 0.25, 0.5, 0.75), the number of intra-molecular

hydrogen bonds corresponds to 11.4, 9.2 and 5,

respectively, which remain unchanged with an increase

in ethanol concentration. Another possible interaction is

the inter-molecular hydrogen bonding between the

carboxyl group and water. The number of inter-molecular

hydrogen bonds between carboxyl groups and water

increases with ionisation in water, and this trend is retained

in the presence of ethanol as well. This number varies as

31.1, 43.6, 72, 105.3 and 141 (for f ¼ 0, 0.25, 0.5, 0.75 and

1, respectively). In range f ¼ 0.25–1, it is fairly constant

irrespective of ethanol content. This indicates that there is

no disruption of carboxylate–water hydrogen bonds due to

the presence of ethanol. However, in the case of non-

ionised PAA ( f ¼ 0), the number of inter-molecular

hydrogen bonds with water molecules decreases with an

increase in ethanol content. Hydrogen bonding between

the PAA residues and ethanol is the highest for the non-

ionised chain followed by chains with f ¼ 0.25 and 0.5

which increases with ethanol concentration. No such

hydrogen bonding interaction is seen for chains with

f ¼ 0.75 and 1. For hydrogen bonds between ethanol and

water, at specified ethanol content, there occurs more

hydrogen bonding between ethanol and water in presence

of ionised chains than in presence of a non-ionised chain.

This again confirms that the non-ionised PAA interacts

favourably with ethanol as compared to ionised PAA. This

leads to favourable bonding between ethanol and water in

the latter cases. This is further verified by the penetration

of ethanol into the solvation shell of the chain, as seen in

Figure 3(A) which shows the RDF between the centre of

mass of the PAA residues and ethanol molecules.

As seen in Figure 3(A), the first peak reveals that

ethanol molecules are distributed at 0.5 nm from the

centre-of-mass of the PAA residues. For the non-ionised

chain ( f ¼ 0), considerable localisation of ethanol is

observed, and this is significantly higher as compared to

the ionised cases. For chains with f ¼ 0–0.5, a distinct

solvation shell with ethanol is seen which is absent in the

case of chains (0.75 and 1). Irrespective of the ethanol

content, it is observed that ethanol is localised next to the

non-ionised residues of the chain and is excluded

completely in the case of a fully ionised chain. For fully

ionised PAA chain, there is no localisation of the ethanol

molecules in close proximity to the chain and ethanol

molecules self-aggregate in the aqueous solution.

The number of ethanol molecules in the solvation shell

of PAA is 1.8, 3.6 and 4.6 in systems containing non-

ionised chain for 1.6, 3.1 and 6.1 vol% of ethanol,

respectively. For systems with 1.6 wt% ethanol, the

number of ethanol molecules distributed around the chains

varies as 1.8, 0.5 and 0.1 (corresponding to f ¼ 0, 0.25,

0.5, respectively). In the case of 6.1 vol% ethanol mixture,

the corresponding values are 4.6, 1.0 and 0.1, respectively.

The influence of ethanol concentration is insignificant in

systems containing completely ionised polyacrylates, as

shown in Figure 3(A). In the absence of ethanol, the water

molecules occupy the solvation cage, which can be

identified by the peak located at 0.35 nm as shown in

Figure 3(B). The fully ionised chain interacts strongly with

water as compared to the non-ionised chain, rationalised

on the basis that COO2 groups being hydrophilic are able

to attract more water than the COOH groups. Thus, in a

polar environment with charged groups (COO2), the chain

interacts favourably with water than with ethanol. The

COOH groups interact strongly with the ethanol

Molecular Simulation 149

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

Figure 3. RDF between the centre of mass of the polyacrylate residues and the (A) ethanol molecules (for each graph, the curvescorrespond to: (i) ______, 1.6; (ii) _ _ _ _, 3.1; (iii) ........., 6.1 vol% of ethanol) and (B) water (for each graph, the curves correspond to:(i)______, 0; (ii)_ _ _ _, 1.6; (iii) ........., 3.1; (iv) ........., 6.1 vol% of ethanol for chains at different charge densities).

S. Srikant et al.150

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

molecules. Addition of ethanol to the aqueous system

having chains with lower charge density results in the

displacement of water molecules from the first hydration

shell. The PAA–water RDF (Figure 3(B)) peak shows a

reduction in its intensity with an increase in ethanol

concentration for the non-ionised chain, thereby indicating

a decrease in the contribution of water to the solvation

shell as a result of the displacement by ethanol molecules.

In the case of ionised chains, the first hydration shell with

respect to the PAA chain is unaffected by ethanol.

The population of water molecules in the solvation

cage increases with degree of ionisation of PAA. This is

verified by the solvation numbers of water as obtained by

RDF peak integration up to 0.45 nm from Figure 3(B).

These are 3.4, 4.5, 5.6, 6.9 and 8.2, respectively ( f ¼ 0,

0.25, 0.5, 0.75, 1). We find that for the non-ionised chain

the solvation number decreases with an increase in ethanol

content – 3.4, 3.1, 2.3 and 1.8 (for 0, 1.6, 3.1 and 6.1 vol%

of ethanol, respectively). For ionised chains, the solvation

numbers are 4.5, 5.6, 6.9 and 8.2, respectively, for f in the

range 0.25–1 and these remain unchanged as a function of

ethanol concentration. This shows that for ionised chains

there is no change in the localisation of water as a function

of ethanol concentration. The unionised residues (COOH

groups) by their interaction with ethanol molecules lead to

a perturbation of the local organisation of water around the

chain backbone.

From an earlier report on MD simulation study [21] of

the solvation of poly(vinyl alcohol), a polar polymer in

water, ethanol and equimolar mixture of water–ethanol

solution, a preferential solvation of the polymer chain by

water molecules was observed, in which hydrogen

bonding with water is shown to be more than with

ethanol. A reduction in hydrogen bonding with water was

also observed in the presence of ethanol in this study. In

the present work, the maximum concentration of ethanol

employed is 6.1 vol%. Ionised PAA chains with higher

charge density show a preference for aqueous environment

in comparison with ethanol. The difference in the

interaction towards ethanol arises due to the charge

density of the PAA chain. The ionised chains show

negligible interaction with ethanol (in the range studied)

and the hydrogen bonding to water remains unchanged in

presence of ethanol. For the non-ionised chain, a reduction

in the number of hydrogen bonds to water is observed with

an increase in ethanol concentration.

Wu and Parris [38] have investigated the interactions of

anionic acrylic polymers with a series of alcohols by means

of surface tension, viscosity and solubility measurements.

Their study revealed that higher hydrophobicity of both

alcohol and polymer results in a decrease in the alcohol

activity. This was attributed to the association between

polymer and alcohol driven by hydrophobic forces.

These experimental observations support our simulation

results, in that the PAA chains with lower charge density

( f ¼ 0–0.5) interact favourably with ethanol when

compared to chains with higher charge density. This is

Figure 4. Radial distribution functions for the pair of counter-ion (Naþ) and polyacrylate chain repeat units. For each graph, the curvescorrespond to: (i)______, 0; (ii)_ _ _ _, 1.6; (iii) ........., 3.1; (iv) ........., 6.1 vol% of ethanol for chains at different charge densities.

Molecular Simulation 151

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

driven by the presence of appreciable number COOH

groups along the chains of lower charge density.

3.3 Interaction with counter-ions

The presence of Naþ counter-ions in the simulations

allows the systems to be equilibrated in a chemically

realistic form, permitting fruitful interpretation and to

relate the results of the simulations to experimental

evidence. The Naþ counter-ions are majorly distributed at

0.5 nm from the centre of mass of the residues (Figure 4).

Another peak at 1 nm is observed which corresponds to

solvent-separated ions. For chains having lower charge

density ( f ¼ 0.25), there is a slight increase in intensity

corresponding to the peak at 0.35 nm (condensed counter-

ions) with an increase in ethanol concentration (this peak

is negligible for higher charge density chains). An increase

in ethanol concentration gives a slight increase in the

strength of the spatial correlation between counter-ions

and the monomer residues specifically for chains

( f ¼ 0.25–0.75) as seen from Figure 4. It is known that

when solvent quality changes from good to poor, there is

an enhanced correlation between the counter-ions and

polyelectrolyte, eventually leading to chain collapse [3]. In

agreement with this observation, as the solvent polarity

decreases with an increase in ethanol content, it is

expected that counter-ions should tend to be closer to the

chain forming stable ion pairs, and this is shown to be true

by our simulations. Specifically with respect to PAA chain

( f ¼ 0.25), we find an increase in the distribution of

counter-ions near the chain backbone with an increase in

ethanol concentration, leading to the chain collapse

observed in this particular case.

4. Conclusions

MD simulations with explicit solvent molecules and

counter-ion description were performed for PAA in

ethanol–water binary solvent. The effect of the degree

of ionisation of PAA solvated in 0, 1.6, 3.1 and 6.1 vol%

ethanol–water mixture on inter-molecular structure and

hydrogen bonding was investigated. The polyacrylate

conformation varies significantly with change in degree of

ionisation, PAA chain interacts more strongly with water

and opens up from a coiled structure to an extended one. In

the range of ethanol concentration studied here, the

simulations show (i) swelling of PAA ( f ¼ 0), (ii) coiling

of ionised chain ( f ¼ 0.25), (iii) slight swelling in the case

of f ¼ 0.5 and (iv) for ionised systems ( f ¼ 0.75–1)

no significant change in the conformation is observed.

In systems with low degrees of ionisation, the addition

of ethanol leads to an increase in intra-molecular

hydrogen bonding, resulting in a narrow distribution of

conformations of the polyacrylate chain. The simulations

identify localisation of ethanol around non-ionised

residues of the chain, their proximity indicating hydrogen

bond interactions. Ethanol is competitively dominant in

the solvation cage surrounding the non-ionised residues.

Inter-molecular hydrogen bonding between PAA and

ethanol is observed, whereas this type of interaction is

absent in the case of ionised polyacrylate chains. COO2

groups are more hydrophilic than COOH groups and so

interact strongly with water molecules, whereas COOH

has favourable interaction with ethanol. Thus, the

simulations presented here clearly distinguish the distinct

interactions between COO2 and COOH groups in the PAA

chain in a solvent of lower polarity. The simulations show

that there is no change in the localisation of water as a

function of ethanol concentration for ionised chains, and

therefore, it is the non-ionised residues (COOH groups)

which by means of their interaction with ethanol

molecules lead to a perturbation of the local organisation

of water around the chain backbone.

Here, the simulation results, which show that non-

ionised PAA residues interact favourably with ethanol

when compared to the ionised residues, are in agreement

with experimental observations in literature, which

attribute molecular association between polymer and

alcohol driven by hydrophobic interactions [38]. As the

solvent polarity decreases with an increase in ethanol

content, it is expected that counter-ions should tend to be

closer to the chain forming stable ion pairs based on

experimental observations. The inter-molecular structural

features arising due to the preferential interaction between

the functional groups in the polyelectrolyte chain and the

solvent molecules are brought forth from the present

simulations. We plan to extend this study to higher

concentration of ethanol in the solvent mixture in order to

obtain insights into the mechanisms leading to polyelec-

trolyte chain collapse in less polar solvents as a function of

charge density.

Acknowledgements

MSS thanks DST, New Delhi, for a research fellowship under thegrant DST-WOS-A/CS-63/2008.

Notes

1. Present address: Graduate Program in Engineering andPhysical Biology, Department of Molecular and CellularBiology, Harvard University, Cambridge, MA, USA.

2. Email: mssulatha@iitm.ac.in.

References

[1] H. Dautzenburg and B. Philipp, Polyelectrolytes: Formation,characterisation, and application, Hanser Publishers, Munich,1994.

[2] M. Hara, Polyelectrolytes: Science and Technology, Marcel Dekker,New York, 1993.

S. Srikant et al.152

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013

[3] A.V. Dobrynin, Theory and simulations of charged polymers: Fromsolution properties to polymeric nanomaterials, Curr. Opin. Coll.Interf. Sci. 13 (2008), p. 376–388.

[4] A.B. Scranton, B. Rangarajan, and J. Klier, Biomedical applicationsof polyelectrolytes, Adv. Polym. Sci. 122 (1995), p. 1–54.

[5] H. Jiang, P. Taranekar, J.R. Reynolds, and K.S. Schanze,Conjugated polyelectrolytes: Synthesis, photophysics and appli-cations, Angew. Chem. Int. Ed. 48 (2009), p. 4300–4316.

[6] C. Tanford, Protein denaturation. Part C. Theoretical models for themechanism of denaturation, Advan. Protein Chem. 24 (1970), p. 1–97.

[7] Y. Nozaki and C. Tanford, The solubility of amino acids and relatedcompounds in aqueous urea solutions, J. Biol. Chem. 238 (1963),p. 4074–4081.

[8] D.R. Robinson and W.P. Jencks, The effect of compounds of theurea-guanidinium class on the activity coefficient of acetyltetragly-cine ethyl ester and related compounds, J. Am. Chem. Soc. 87(1965), p. 2462–2470.

[9] D.B. Wetlaufer, S.K. Malik, L. Stoller, and R.L. Coffin, Nonpolargroup participation in the denaturation of proteins by urea andguanidinium salts. Model compound studies, J. Am. Chem. Soc. 86(1964), p. 508–514.

[10] W. Bruning and A. Holtzer, The effect of urea on hydrophobicbonds: The critical micelle concentration of n-dodecyltrimethylam-monium bromide in aqueous solutions of urea, J. Am. Chem. Soc. 83(1961), p. 4865–4866.

[11] M.J. Stevens and K. Kremer, The nature of flexible linearpolyelectrolytes in salt free solution: A molecular dynamics study,J. Chem. Phys. 103 (1995), p. 1669–1690.

[12] O. Biermann, E. Hadicke, S. Koltzenburg, M. Seufert, andF. Muller-Plathe,Hydration of polyelectrolytes studied by moleculardynamics simulation, arXiv:cond-mat/0101115v1 [cond-mat.mtrl-sci], 9 Jan 2001.

[13] D. Rieth, B. Muller, F. Muller-Plathe, and S. Wiegand, How doesthe chain extension of poly(acrylic acid) scale in aqueous solution?A combined study with light scattering and computer simulation,J. Chem. Phys. 116 (2002), p. 9100–9106.

[14] Q. Liao, A. Dobrynin, and M. Rubinstein, Molecular dynamicssimulations of polyelectrolyte solutions: Osmotic coefficient andcounterion condensation, Macromolecules 36 (2003), p. 3399–3410.

[15] R.G. Winkler, M.O. Steinhauser, and P. Reineker, Complexformation in systems of oppositely charged molecules: A moleculardynamics simulation study, Phys. Rev. E 66 (2002), p. 021802–7.

[16] U. Micka, C. Holm, and K. Kremer, Strongly charged, flexiblepolyelectrolytes in poor solvent: Molecular dynamics simulations,Langmuir 15 (1999), p. 4033–4044.

[17] E.J.W. Wensink, A.C. Hoffmann, P.J. van Maaren, and D. van derSpoel, Dynamic properties of water/alcohol mixtures studied bycomputer simulation, J. Chem. Phys. 119 (2003), p. 7308–7317.

[18] S.U. Noskow, G. Lamoureux, and B. Roux, Molecular dynamicsstudy of hydration in ethanol–water mixtures using a polarizableforce field, J. Phys. Chem. B 109 (2005), p. 6705–6713.

[19] S. Patel, Z. Yang, B.A. Bauer, and J.E. Davis, Interfacial structure,thermodynamics and electrostatics of aqueous methanol solutionsvia molecular dynamics simulations using charge equilibrationmodels, J. Phys. Chem.B 113 (2009), p. 9241–9254.

[20] M.S. Sulatha and U. Natarajan, Origin of the difference in structuralbehaviour of polyacrylic acid and polymethacrylic acid in aqueoussolution discerned by explicit solvent explicit ion MD simulations,Ind. Eng. Chem. Res. 50 (2011), p. 11785–11796.

[21] F. Muller-Plathe and W.F. van Gunsteren, Solvation of poly(vinylalcohol) in water, ethanol and an equimolar water–ethanol mixture:Structure and dynamics studied by molecular dynamics simulation,Polymer 38 (1997), p. 2259–2268.

[22] C. Oostenbrink, A. Villa, A.E. Mark, and W.F. van Gunsteren,A biomolecular force field based on the free enthalpy of hydrationand solvation: The GROMOS force field parameter sets 53a5 and53a6, J. Comput. Chem. 25 (2004), p. 1656–1676.

[23] M.L. Miller, K. O’Donnell, and J. Skogman, Crystalline polyacrylicacid, J. Colloid Sci. 17 (1962), p. 649–659.

[24] K. Kogej, Association and structure formation in oppositely chargedpolyelectrolyte–surfactant mixtures, Adv. Colloid Interf. Sci. 158(2010), p. 68–83.

[25] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, andJ. Hermans, Intermolecular forces, in B. Pullman, ed., Reidel Pub.Co., Dordrecht, 1981, pp. 331–342.

[26] S. Miyamoto and P.A. Kollman, Settle: An analytical version of theSHAKE and RATTLE algorithm for rigid water models, J. Comput.Chem. 13 (1992), p. 952–962.

[27] H. Hess, C. Kutzner, D.V. Spoel, and E. Lindahl, GROMACS 4:Algorithms for highly efficient, load balanced and scalablemolecular simulation, J. Chem. Theory. Comput. 4 (2008), p. 435–447.

[28] D. Van der Spoel, P.J. Van Maaren, and H.J.C. Berendsen, Asystematic study of water models for molecular simulation:Derivation of water models optimized for use with a reaction fielddielectric, J. Chem. Phys. 108 (1998), p. 10220–10230.

[29] J.P. Ryckaert, G. Ciccotti, and H.J.C. Berendsen, Numericalintegration of the equations of motion of a system with constraints:Molecular dynamics of n-alkanes, J. Comput. Phys. 23 (1977),p. 327–341.

[30] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. DiNola,and J.R. Haak, Molecular dynamics with coupling to an externalbath, J. Chem. Phys. 81 (1984), p. 3684–3690.

[31] A.R. Khokhlov, S.G. Starodubtzev, and V.V. Vasilevskaya,Conformational transitions in polymer gels: Theory and experiment,Adv. Polym. Sci. 109 (1993), p. 123–171.

[32] O.E. Philippova, N.L. Stinikova, G.B. Demidovich, andA.R. Khokhlov, Mixed polyelectrolyte/ionomer behavior ofpoly(methacrylic acid) gel upon titration, Macromolecules 29(1996), p. 4642–4645.

[33] A.R. Khokhlov and E.Y. Kramarenko, Weakly charged polyelec-trolytes: Collapse induced by extra ionization, Macromolecules 29(1996), p. 681–685.

[34] M.T.M. Klooster, F.V. Touw, and M. Mandel, Solvent effects inpolyelectrolyte solutions: 1. Potentiometric and viscometrictitration of poly(acrylic acid) in methanol and counterion specificity,Macromolecules 17 (1984), p. 2070–2078.

[35] S.K. Pearsall, M.M. Green, and H. Morawetz, Titration ofpolycarboxylic acids in methanol solution: Polymer chain extension,ionization equilibria and conformational mobility, Macromolecules37 (2004), p. 8773–8777.

[36] Y. Nishiyama and M. Satoh, Swelling behavior of poly(acrylic acid)gels in aqueous ethanol – Effects of counterion species and ionicstrength, Macromol. Rapid Commun. 21 (2000), p. 174–177.

[37] Y. Nishiyama and M. Satoh, Solvent- and counterion-specificswelling behavior of pol(yacrylic acid) gels, J. Polym. Sci. Part BPolym. Phys. 38 (2000), p. 2791–2800.

[38] N. Wu and J. Parris, Interaction of water soluble acrylic polymerswith alcohols in aqueous solution, Coll. Surf. A: Physicochem. Eng.Aspects 167 (2000), p. 179–187.

Molecular Simulation 153

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

06:

49 0

2 Se

ptem

ber

2013