INVITEDPRESENTATIONS

Gasadsorptionbynanoporousmaterials:Experimentalchallengesandtheroleofcomputationalmodellingandsimulation

DarrenBroom

HidenIsochemaLtd,422EuropaBoulevard,WarringtonWA57TS,UK

There are many existing and prospective applications for gas adsorption by nanoporousmaterials. Activated carbons and zeolites, for example, are widely used to purify andseparateindustrialgases,whilenewermaterials,suchasmetal-organicframeworks(MOFs),covalentorganicframeworks(COFs)andmicroporousorganicpolymers,havethepotentialto store hydrogen andmethane for use as alternative energy carriers [1]. Experimentallycharacterisingtheadsorptionpropertiesofthesematerialsiscrucialtoassessingtheiruseingas separation and storage technology. However, in some cases, experiments can beparticularly challenging. These include the adsorption of high pressure gases andmulticomponent mixtures, both of which are required for practical applications. Forhydrogen storage, for example, this has led to some high profile controversies [2].Molecular simulation, meanwhile, is an invaluable complementary tool, providing therelevant adsorbate-adsorbent systems can bemodelled to sufficient accuracy [3,4]. Asidefrom providing theoretical insights, it can be used, for example, to help corroborate andvalidate experimental data, to interpolate and extrapolate data in different temperatureandpressureregimes,andtoprovideinformationcurrentlyinaccessibletoexperiment.Thispresentationwillcoversomeoftheexperimentalchallengesencounteredwhenmeasuringgasadsorptionbynanoporousmaterialsanddiscusstherolethatcomputationalmodellingandsimulationcanplay,fromanexperimentalist’sperspective.[1] D.P.Broom,K.M.Thomas(2013)Gasadsorptionbynanoporousmaterials:Futureapplicationsand

experimentalchallenges.MRSBull38:412-421[2] D.P.Broom,M.Hirscher(2016)Irreproducibilityinhydrogenstoragematerialresearch.EnergyEnviron

Sci9:3368-3380[3] D.P.Broometal(2016)Outlookandchallengesforhydrogenstorageinnanoporousmaterials.ApplPhys

A122:151[4] Z.Xiangetal(2010)Multiscalesimulationandmodellingofadsorptiveprocessesforenergygasstorage

andcarbondioxidecaptureinporouscoordinationframeworks.EnergyEnvironSci3:1469-1487

Laserinducedmeltingofmetallicfilms:Modellingmeetsexperiment

DorothyM.Duffy

Dept.ofPhysicsandLondonCentreforNanotechnology,UniversityCollegeLondon,LondonOneoftheuniversalchallengesfacedbythemodellingcommunitytheextrememismatchbetween time and length scales of standard simulation methods and experimentalmeasurements.Inrecentyears,however,advancesinbothcomputationalandexperimentalmethodshaveenableddirectcomparisonstobemadeforalimitednumberofscenarios.Inthis presentation, I will discuss one such example, that of laser irradiated thin films. Thedynamical response of such films can be studied on femtosecond timescales by ultrafastelectron diffraction (UED) and the time evolution of the Bragg peak intensity can bemeasured directly1, bringing the experimental measurements into regimes that are veryaccessibletomoleculardynamics(MD)modelling.

MD has a long and successful history in the field of radiation damage, and cascadesimulations have made exceptional contributions to the fundamental understanding ofradiationeffectsinmaterials. Untilrecently,thesecontributionshavebeenlimitedtolowenergy ion irradiation,wherethe impinging ion interactsprimarilywith theatomicnuclei,knockingthemoutoftheequilibriumlatticesitestocreatedefects.Recentdevelopmentshave,however,enabledmodellingoftheeffectsofirradiationthatexciteselectrons,suchaslasersandveryenergeticions.ThishasbeenachievedbycouplingclassicalMDtothetwotemperaturemodel(2TM)that istraditionallyusedtoevaluatethetemperatureevolutionfor systems in which the atoms are out of equilibrium with the electrons. In the 2TM,different temperaturesareassigned to theatomsand theelectronsand the temperatureevolutionsareevaluatedbysolvingtwocoupledthermaldiffusionequations.Inourmodel,we replace the atomic thermal diffusion equation with MD, which enables us to obtainatomisticdetailabouttheresponseofmetalstoveryenergeticionsandlaserirradiation.2In this presentation I will explain the coupled 2T-MD model and its implementation inDL_POLY.IwilldemonstrateitsapplicationtolaserirradiationofgoldnanofilmsandshowhowthecalculatedtimeevolutionoftheBraggpeaksare inexcellentagreementtothosemeasuredusingUEDbyourexperimentalcollaborators.3Iwillshowthatitisnecessarytomodify the interatomic potentials between highly excited atoms to reproduce theexperiments at veryhigh fluence anddemonstratehowwedevelopeda set of electronictemperaturedependent interatomicpotentials for tungsten.4 Finally, Iwill show that it isnecessary toextend themodel inorder to simulatebandgapmaterials anddemonstratethesuccessoftheextendedmodelforlaserirradiationofsiliconnanofilms.

[1] M.Harbetal.“Carrierrelaxationandlatticeheatingdynamicsinsiliconrevealedbyfemtosecondelectrondiffraction”J.Phys.Chem.B110,25308-25313(2006)

[2] D.M.DuffyandA.M.Rutherford,“Includingtheeffectsofelectronicstoppingandelectron-ioninteractionsinradiationdamagesimulations”J.Phys.:Cond.Matt.,19016207(2007)

[3] S.L.Daraszewicz,etal.Structuraldynamicsoflaser-irradiatedgoldnanofilms.Phys.RevB88,184101(2013)

[4] S.T.Murphyetal.“Dynamicalsimulationsofanelectronicallyinducedsolid-solidphasetransformationintungsten”Phys.Rev.B,92134110(2015)

Moleculardynamicsatnanometriclengthscales

F.Kremer

InstituteofExperimentalPhysicsI,UniversityofLeipzig,Linnéstr.5,04103Leipzig,Germany

Thequestiononwhatlength-scalemolecularandespeciallyglassydynamicsofpolymerstakesplace is of fundamental importance and hasmultifold practical implications as well. RecentresultsbasedonBroadbandDielectricSpectroscopy[1] fornanometricthin(≥5nm) layersofpoly(styrene) [2], poly(cis-1,4-isoprene) [3], poly(styrene‑b‑1,4-isoprene) diblock copolymers[4]andevenforisolatedchains[5]andforbrushesofpoly(2-vinylpyridine)inthedried[6]andswollen[7]statewillbepresented,deliveringtheconcurringresultthatdeviationsfromglassydynamics of the bulk never exceedmargins of±3 K independent of the layer thickness, themolecularweightofthepolymerunderstudyandtheunderlyingsubstrate.-Theexperimentslead to the conclusions that glassy dynamics takes place on the length scale a few polymersegments(≤~1nm),whiletheconformationofthechainasawholeisstronglymodifiedbythegeometrical confinement [8-10]. The results will be discussed with respect to the highlycontroversialliterature[8].

[1] Kremer,F.,A.Schönhals(Eds.)BroadbandDielectricSpectroscopy,Springer,Berlin2003[2] Tress,M.,etal,“GlassyDynamicsandGlassTransitioninNanometricThinLayersofPolystyrene“,

Macromolecules,43,9937-9944(2010)[3] Tress,M.,E.U.Mapesa,W.Kossack,W.K.Kipnusu,M.Reiche,F.Kremer,“GlassyDynamicsinCondensed

IsolatedPolymerChains”,Science,341,1371-1374(2013)[4] Mapesa,E.U.,etal.“Segmentalandchaindynamicsinnanometriclayersofpoly(cis-1,4-isoprene)as

studiedbyBroadbandDielectricSpectroscopyandtemperature-modulatedCalorimetry”,SoftMatter,10592-10598(2013)

[5] Kremer,F.(Ed.)“Dynamicsingeometricalconfinement”,AdvancesinDielectrics,Springer2014[6] Kipnusu,W.K.,etal.“StructureandDynamicsofAsymmetricPoly(styrene-b-1,4-isoprene)Diblock

Copolymerunder1Dand2DNanoconfinement”,ACSAppl.Mater.Interfaces,7,12328-12338(2015)[7] Neubauer,N.,etal.,“GlassydynamicsofPoly(2-Vinyl-Pyridine)brusheswithvaryinggraftingdensity”,

SoftMatter,11,3062-3066(2015)[8] Kremer,F.,M.Tress,E.U.Mapesa;“Glassydynamicsandglasstransitioninnanometriclayersandfilms:A

silverliningonthehorizon”,JournalofNon-CrystallineSolids,407,277-283(2015)[9] Neubauer,N.,etal.“MolecularDynamicsofSwollenPoly(2-vinylpyridine)Brushes”Macromolecules,49,

6101-6105(2016)[10] KipnusuW.K.,M.Elsayed,R.Krause-Rehberg,F.Kremer;“Glassydynamicsofpolymethylphenyl-siloxane

inone-andtwo-dimensionalnanometricconfinement-acomparison”,J.Chem.Phys.146,203302(2017)

PolyesterFilms-CanYouTeachAnOldDogNewTricks?

Prof.BillMacDonald

DuPont-TeijinFilms,WiltonResearchCentrePolyesterfilmhasbeenaroundforoverhalfacenturyandiswellestablishedinspecialty,industrial, packaging and advanced magnetic media and photo systems, electrical andelectronics markets. Old markets die and to survive it is critical that DTF continues toinnovateanddevelopnewfilmstructurestocreatenewmarketopportunitiesormeetnewmarket needs. This talk will discuss the challenges of attempting to develop new filmstructuresfornewmarketsthatoftenareoverhypedandhavechanginggoalpostsandwillusetheareasofflexibleelectronicsandnextgenerationPVtoillustratethis.Markettrendsand where this might offer new opportunities and technical challenges will also bediscussed.

Moleculartheoryandsimulationofthedynamicbehavioroffluidsconfinedinmesoporousmaterials

PeterA.Monson

DepartmentofChemicalEngineering,UniversityofMassachusetts,Amherst,USA

Recently we have developed a dynamic mean field theory (DMFT) that describes thedynamics of adsorption and desorption of fluids in mesoporous materials. The theorypredictstheevolutionofthedensitydistributionofthefluidintheporousmaterialafterastep change in the bulk pressure or chemical potential. An important feature ofDMFT isthatit isfullyconsistentwiththemeanfielddensityfunctionaltheory(DFT)descriptionofthe thermodynamics. In addition, the nucleation mechanisms in phase transitions ofconfinedfluidsareemergentfeaturesofthecalculations.Forinstance,thetheorydescribesthenucleationofcapillarycondensationbyliquidbridging.InthispresentationwedescribeapplicationsofDMFTtoseveralsystems.Forporenetworkswestudytheinfluenceofporeconnectivity upon uptake and desorption dynamics. Studies of adsorption/desorptiondynamicsofmixturesshowthenatureoftheequilibrationofthecompositiondistributioninthe system.Wewill alsocompare thepredictions fromthe theorywith results frombothdynamicMonteCarloandnon-equilibriummoleculardynamicssimulations.

CrystallisationProcessDevelopment:Past,PresentandFuture

AmyRobertson

ChemicalDevelopment,PharmaceuticalTechnologyandDevelopment,AstraZeneca,Macclesfield,UKCrystallisation of the active pharmaceutical ingredient (API) is a key step in themanufacturing process of a drug. Traditionally crystallisation was used to optimise thepurity and yield of the material, however, an understanding of a crystallisation processallowsmanyothersolidstatepropertiestobecontrolledandoptimized.Thesepropertiesinclude polymorphic form, morphology and crystal size. The key variables in a coolingcrystallisation include cooling rate, agitation speed, stirring time and seeding.Understandingtheimpactofeachofthesefactorsinacrystallisationallowsareproducibleprocesstobedesignedtoobtainthecorrectsize,morphologyandpolymorphicformwhichissuitableforsecondaryprocessing.Toaidintheunderstandingofthekeyfactorsthataffectthecrystallisationprocess,ProcessAnalytical Technology (PAT) is now routinely employed. A wide range of probes can beinserted into reactors to monitor processes. The data obtained is used to optimise theprocess,either in real-timeor throughsubsequentexperiments. In recentyears this largeincrease of crystallisation process data has led to the development of models allowingpredictive design and optimisation of processes, in turn minimizing the experimentalprogram required, e.g. Population BalanceModels (PBM), Computational Fluid Dynamics(CFD).However,therearestillmanyareaswherethismodellingcapabilityhasnotyetfullydevelopedandtheopportunitiesandchallengeswillbepresented.

CrystalNucleationofSmallOrganicMolecules

JoopH.terHorst

CMACFutureManufacturingResearchHub,StrathclydeInstituteofPharmacyandBiomedicalSciences(SIPBS),TechnologyandInnovationCentre,UniversityofStrathclyde,U.K.

The trend towardsmore complexmolecules,materials and systemswill proceed into thefuture bringing new crystallization challenges for particulate products such aspharmaceuticals. The solution to these challenges lies in a fundamental understanding ofsolubility, driving force and crystallization kinetics onto which we can build innovative,integrated and intensified continuous crystallization processes. One long-standing andindustrially important issue is the control of crystal nucleation in termsof formand rate.However,duetothevariousnucleationratemeasurementmethods,analyticaltechniques,and simulation methods now available, crystal nucleation research will show substantialprogressinthecomingyears[1,2].Crystalnucleationisastochasticprocessinwhichbuildingunitsattachtoanddetachfromacluster forminganucleus,usuallyonaheterogeneous surface.Variations inexperimentaloutcomesperformedunderequalconditionsthuscapturenucleationbehaviour.Nucleationrates can therefore be determined bymeasuring induction time distributions [3,4]. Suchnucleationratemeasurementsindicatethateitherheterogeneousparticleconcentrationorbuilding unit attachment frequency to the nucleus is orders of magnitude lower thanestimatesfromClassicalNucleationTheory[5].In principle nucleation can be controlled by building units, effective templates andalternativeenergies.Thebuildingunit,forinstance,canbeinfluencedthroughthesolvent’seffect on the self-association of the solutes in solution [6]. The impact of the solvent isreflected in the polymorphic form crystallized. The template can in turn enhance thenucleationofspecificformsbyenhancingtheclusteringofaspecificself-associate[7].[1]R.J. Davey, S.L.M. Schroeder, J.H. ter Horst, Nucleation of Organic Crystals – A Molecular Perspective,

Angew.Chem.Int.Ed.52(2013)2166.[2] H.Yang,J.H.terHorst,CrystalNucleationofSmallOrganicMolecules,pp317-337,In:NewPerspectiveson

Mineral Nucleation and Growth, Editors: A.E.S. Van Driessche, M. Kellermeier, L.G. Benning, DenisGebauer(2017),SpringerInternationalPublishing,DOI:10.1007/978-3-319-45669-0_16.

[3] S.Jiang,J.H.terHorst,CrystalNucleationRatesfromProbabilityDistributionsofInductionTimes,CrystalGrowthDesign11(2011)256-261.

[4] S.A. Kulkarni, S.S. Kadam, H. Meekes, A.I. Stankiewicz, J.H. ter Horst, Crystal Nucleation Kinetics fromInductionTimesandMetastableZoneWidths,CrystalGrowthDesign13(6)(2013)2435-2440.

[5] J.H.terHorst,C.Brandel,MeasuringInductionTimesandCrystalNucleationRates,FaradayDiscuss.179(2015)199

[6] S.A.Kulkarni,E.S.McGarrity,H.Meekes,J.H.terHorst,Isonicotinamideself-association:thelinkbetweensolventandpolymorphnucleation,ChemCommun.48(2012)4983-4985.

[7] S.A.Kulkarni,C.C.Weber,A.S.Myerson, J.H. terHorst,Self-associationduringheterogeneousnucleationontowell-definedtemplates,Langmuir,30(41)(2014)12368-12375.

ORALPRESENTATIONS

MechanicalPropertiesofPristineandOxidizedEpitaxialGrapheneFilms:Experimentsandcalculations

AngeloBongiorno1,2,3

1DepartmentofChemistry,CollegeofStatenIsland,CityUniversityofNewYork,2800VictoryBoulevard,StatenIsland,NewYork10314,USA

2CUNYGraduateCenter,Ph.D.PrograminPhysics,NewYork,NY100163CUNYGraduateCenter,Ph.D.PrograminChemistry,NewYork,NY10016

Multilayer epitaxial graphene and its oxide exhibit strong in-plane bonding and weakinteractions between the layers [1]. In this work, indentation experiments and densityfunctional theory (DFT) calculations are combined to elucidate the perpendicular-to-the-planesmechanicalpropertiesofgrapheneandgrapheneoxidefilms.Afteran introductiontomaterialsandmethods,wewillfocusonthestructuralpropertiesofgrapheneoxideandthe effect of moisture on their mechanical properties. Then, we will focus on pristinegraphene,andwewilldiscusstheunusualstiffnessandhardnessexhibitedbyatwo-layergraphenefilmonSiC.DFTcalculationsshowthattheobservedstiffeningeffectsmaylieinaphase transformation, occurring at room temperature during the indentation process, ofthetwo-layergraphenefilmintoastiffdiamond-likefilm.[1]Y.Gao,S.Kim,S.Zhou,H.-C.Chiu,D.Nélias,C.Berger,W.deHeer,R.Sordan,L.Polloni,A.Bongiorno,andE.Riedo,Elasticcouplingbetweenlayersintwo-dimensionalmaterials,NatureMaterials14,714(2015).

2DMolecularDynamicsStudyofSubstrate-DependentGrowthofGrapheneonCopper

PeterBrommer,ChenweiZhu,GwilymT.Enstone,GavinR.Bell,DavidQuigley

SchoolofEngineering,UniversityofWarwick,UK

Chemicalvapourdeposition(CVD)growthofgrapheneoncopperhasbeenidentifiedasthemostpromisingrouteforscalablesynthesisofgraphene.It isimportanttounderstandthefundamentalsurfacegrowthmechanismsinordertooptimisematerialquality,butdetailedin situ imaging of growth in CVD is extremely challenging. Comparison of post-growthexperimentaldatawithdynamicsimulationsofgrowthcanovercomethisproblem.Experimental observations of partial-coverage graphene grown on copper foilswithwell-defined crystallographicmicrostructure, e.g. (111), (100) and (210) grains, indicate clearlythat the surface orientation affects the growth rate, island size distribution (ISD) andorientationofgrapheneislands.Ab initosimulationcannotreach lengthscalesrelevanttothese phenomena, and it would be prohibitive to use such methods even to addressinteractionofgrowinggraphenewiththelocalsymmetryofthecrystallinesubstrate.Wepresentatwo-dimensionalmoleculardynamicssimulationmodelforgraphenegrowthoncopperwithaweakvan-der-Waalsepitaxial interaction.Thisminimaloff-latticemodelcapturesvariations ingrowthrate, islandorientationand ISDbetweendifferent low indexcopper faces.We also show effects of patterned substrates, resulting e.g. from a coppersurfacefaceting.Theresultsmaysuggestwhichmorphologicalandcrystallographicfeaturesof a low-cost copper substrate could be optimised for large-grain single crystal graphenegrowth.

Supportedlipidfilmsunderavaryingelectricfield:modellingavoltammetrymolecularsensor

AndreyBrukhno

ComputationalChemistryGroup,SCD/STFC,Sci-TechDaresbury,UK

The behaviour of phospholipid layers onmercury (Hg) electrodes in an electric field hasexcitedconsiderableinterestoverthepastthirtyyears.Partofthereasonforthis interesthasbeentheoccurrenceofverysharpphasetransitionsasafunctionofappliedpotential,which are associated with sudden structural changes in the phospholipid layer. Thesesystemshavebeenstudiedextensivelymainlyusingelectrochemicalmethodsbut still thenatureofthestructuralchangesunderlyingthephasetransitionshasremainedelusive.Forthefirsttime,amolecularsimulationstudyof lipid layerssupportedonaflatandsmoothelectrode is presented, where voltammetry measurements are closely mimicked. Inparticular,oursimulationsreproduceanon-trivialcapacitance-vs-voltagedependence,C(V),recorded in experiment [1,2]. The Monte Carlo simulations reveal the followingrearrangementsinhydrophobicallyadsorbedphospholipidfilms,beingdrivenbythechargeredistributionwithintheinterface:[3]

1. Displacementofthelipidmonolayerfromtheelectrodebyitscounterions(cations),leading to formation of an electric double layer and, consequently, completemonolayerdesorption;

2. Transformationofthemonolayerintoabilayeruponitsdesorption;3. Zwitterionic bilayer readsorption to the electrode through the polar-group

interactionwiththeelectricdoublelayer.The voltammetry peaks reflect a stepwise formation of layers of alternating charge: (a)electric double layer upon transient film desorption, (b) triple or multi-layer upon filmreadsoption. The simulation evidence suggests that the first peak is due to the cationbreakthroughtotheelectroderesulting inthemonolayerdesorption,whereasthesecondpeakrepresentsthefilmrearrangementtoformself-organisedbilayerstructures.

1. F.A.M.LeermakersandA.Nelson,J.Electroanal.Chem.278,53(1990)2. D.BizzottoandA.Nelson,Langmuir14,6269(1998)3. A.V.Brukhno,A.Akinshina,Z.Coldrick,A.NelsonandS.Auer,SoftMatter7,1006(2011)

MolecularDynamicsofPolymerNanocomposites

José-JavierBurgos-MármolandAlessandroPatti

SchoolofChemicalEngineeringandAnalyticalScience,TheUniversityofManchester

Polymer nanocomposites (PNCs) are polymermaterials incorporating nano-sized particles(NPs). They have tremendous potential in industrial formulations and technologicalapplications, including protective coatings and food packaging. Adding NPs to a polymermatrixmodifies the conformation andmobility of the polymer chains at the NP-polymerinterfaceand canpotentially providematerialswithenhancedproperties as compared topristinepolymers.ByperformingMolecularDynamics (MD) simulations,we investigate theeffectofNP sizedispersity and degree of chain stiffness on a variety of properties that characterize theresponse of PNCs at the nano and macro scales. In particular, at the nano scale, weinvestigate the space distribution of small and large polydisperseNPs and their ability todiffuse throughadense isotropic distributionof polymer chains.We find very interestingscaling laws relating the average size and polydispersity index ofNPswith their diffusioncoefficients,generallyunderestimatedbyexisting theories.Toassess theeffectofNPsizedispersityonthemacroscopicresponseofourmodelPNC,weevaluatetwokeytransportproperties,shearviscosityandthermalconductivity,whicharefoundtodisplayanintriguinguniversal behaviourwhen plotted against the polymer/NP specific interface area and theinverseoftheNP'smass,respectively.Atincreasingchainstiffness,weobserveasounddependenceoftheNPs’diffusivityonthelong-range order of the polymer melt, which undergoes an isotropic-to-nematic phasetransition.Thisphasetransitioninducesadynamicalanisotropyinthenematicphase,withthe NPs preferentially diffusing along the nematic director rather than in the directionsperpendiculartoit.NotonlydoesthistendencydeterminetheNPs’mobilityanddegreeofdispersioninthepolymermatrix,butitalsoinfluencestheresistancetoflowofPNCswhenashearisapplied.

HydrotropicSolubilizationofHydrophobicCompoundsinAqueousSolutionsofIonicLiquids

TâniaSintra,K.Shimizu,J.N.CanongiaLopes,SóniaP.M.Ventura,JoãoA.P.Coutinho

CICECO-AveiroInstituteofMaterials,DepartmentofChemistry,UniversityofAveiro,3810-193Aveiro,Portugal

Theabilityofaqueoussolutionsofionicliquids(ILs)tosolubilizehydrophobiccompoundsisdemonstratedinthisworkandamechanismtoexplainthisenhancedsolubilityisproposed.Vanillinandgallicacid, twonaturalantioxidants,and ibuprofenandnaproxen, twodrugs,and technical lignins (kraft and organosolv) are here studied. Their solubility in aqueoussolutionsofILswasmeasuredinthewholecompositionrange,frompurewatertopureILs,and an increase in the solubility of up to 1000-fold was observed. The effects of the ILchemical structures, their concentration and temperature on the solubility of thehydrophobic model compounds were evaluated and compared with the performance ofconventional hydrotropes. Using Dynamic Light Scattering, NMR andmolecular dynamics(MD)simulations,itwaspossibletoinferthattheenhancedsolubilityofbiomoleculesinILaqueoussolutionsisrelatedtotheformationofIL-biomoleculeaggregates.AtreatmentoftheexperimentaldatabasedontheKirkwood-Bufferapproachrecentlyproposed[1]helpsunderstandingthestructureof theseaggregates.Theresultsgathered inthisworkhaveasignificant impact on the understanding of ILs aqueous solutions as novel solvents withenhancedperformance.[1]S.Shimizu,N.Matubayasi,PCCP,2016,18,25621-8Acknowledgments: This work was developed within the scope of the project CICECO-Aveiro Institute ofMaterials (Ref. FCT UID /CTM /50011/2013), financed by national funds through the FCT/MEC and whenapplicable co-financed by FEDER under the PT2020 Partnership Agreement. FCT is also acknowledged forfundingtheprojectDeepBiorefinery(PTDC/AGR-TEC/1191/2014).

PredictionofbinarySALR+SAfluidphasebehaviourbymeansofDensityFunctionalMicelleTheoryCalculations

CarlosA.Ferreiro-RangelandMartinB.Sweatman

SchoolofEngineering,UniversityofEdinburgh,Edinburgh,UKEH93FJ

Simple fluids with competing short-range attractive and long-range repulsive (SALR)interactionsarerelevantinthestudyofmanycomplexbiologicalandsoftmattersystems.Our particular interest for this EPSRC funded project are ‘Synbio’ systems, where largestable clusters of a specific SALR-like solute are used as templates in the ‘greenmanufacture’ofsolidnanoparticlesofsilicaandalsomagnetite.Earlierwork has already elucidated the low density equilibrium phase behaviour of SALRdispersions(M.B.Sweatmanet.al.,JCP140,124508(2014)),whereclusterfluidandclustersolid phases were found above a ‘critical cluster concentration’ (CCC). Furthermore, thethermodynamicmodelpredictedafirstorderclustervapourtoclusterliquidorsolidphase.Thepresentworkextendsthemicelle-DFTthermodynamicmodelpreviouslyappliedtopureSALR fluids in order to study the phase behavior of binary mixtures where both fluidcomponentsinteractthroughahardcorewithshort-rangeattractions(SA),butonlyoneofthe components exhibits an additional long-range repulsion (SALR), since this type ofmixtures are thought to be relevant in the Synbio manufacture process of interest. Theattractivestrengthsofbothspeciesaswellas theattractiveSA/SALRcross-interactionarekeytodeterminethephasebehaviourofthemixture.Interestingly,althoughaweakcross-interactionleadstoabehaviorthatisacompositeoftheindividualcomponents,whenthecross-interaction is strong enough equimolar clusters appear, regardless of whether thepureSAfluidissupercriticalorthepureSALRfluidwouldformclustersornot.These results help to better understand the non-classical nucleation mechanism thatexperimentallymayexhibitstable,giant-clusters.Keywords: Liquidmixtures,clustering,competinginteractions,mesostructure, non-classicalnucleation

TheInfluenceofAminoAcidsonAmorphousCalciumCarbonateStructureandFormation

ColinLFreeman[1],AaronRFinney[1],RiccardoInnocentiMalini[1],P.MarkRodger[2],JohnHHarding[1]

1DepartmentofMaterialsScienceandEngineering,UniversityofSheffield,SirRobertHadfieldBuilding,MappinStreet,Sheffield,S13JD,UK

2CentreforScientificComputingandDepartmentofChemistry,UniversityofWarwick,Coventry,CV47AL,UK

Amorphous calcium carbonate (ACC) often forms prior to the precipitation of calciumcarbonate (CaCO3) from solution but our understanding of the formation of ACC and itsconversion remains poor. Experiments have suggested crystallisation occurs on the ACCsurface[1]butthis interfaceremainsundefined. Inboththenaturalworldandsynthetic,molecularadditivesareusedtomodifythenucleationandgrowthprocess.Acommonideais thatproteins, peptides and similar analogues stabiliseordestabiliseACCand thereforecan stimulate or inhibit the formation of crystalline polymorphs. But understanding thisrequiresmodelsofthebindingmechanismsbetweenthesemoleculesandACCmaterial.

Wepresentacomputational studyonACCand itsprecursors interactingwithavarietyofaminoacidmolecules. Weparticularly focusonhowthedifferentR-groupsof theaminoacidswillinducevariedinteractionswiththeACC.Analysisofourresultssuggeststhatthehydration levelof theACCsignificantlyaffects thechemistryof the interactionsuggestingtheinfluencemayalterwiththechemistryofthesystem.Usingtheseresultsweareabletodiscusswhichaminoacidsmaybemore importantandeffective in stabilisingACC,whichmay influence the dehydration of ACC and which may be significant in stimulating theconversiontoacrystallinephase.

[1] Nielsen, M.H.; Aloni, S.; De Yoreo, J.J. Science 2014, 345 (6201), 1158-1162

MolecularDynamicsSimulationsofZn–AlLayeredDoubleHydroxideswithDifferentIntercalatedIons

aGermánPérez-Sánchez,bTiagoL.P.Galvão,bJoãoTedim,aJoséR.B.Gomes

aCICECODepartmentofChemistry,UniversityofAveiro,PortugalbCICECODepartmentofMaterialsandCeramicEngineering,UniversityofAveiro,Portugal

Layereddouble hydroxides (LDHs) are claymaterials [1] formedby a pile of cationicmixedmetallayers with intercalated molecules, ions and water. This configuration makes them a relevantmaterialindifferentfields,namelycatalysis[2],medicine[3]andinadsorption[4,5]orpurificationprocesses [6]. LDHsmaterialshave theabilityof ionexchangebetween the intercalated ionswiththosethatare insolution. Inthissubject,ourgroupisdevotedonmakingthesematerialsanidealmacromolecularcontainertobeusedasadditives incoatingsandnanostructuredconversionfilmsforcorrosionprotection.RecentlyTedimetal. [7]haveexperimentallydemonstratedtheabilityofZinc-Aluminium LDHs with intercalated nitrates (Zn2Al-NO3) to exchange them by chlorides in acorrosive solution through concentration adjustment used as a trigger response mechanism.However, the studyof themetal hydroxide structure interactionswith the intercalatedmoleculesandtheion-exchangeprocessarequitedifficultfromanexperimentalpointofview.Inthisregard,differentexperimentaltechniqueshavebeenwidelyusedtocharacterizeLDHs,nevertheless,mostofthemprovidedlimitedresolution.Forinstance,defectsproducedduringthesynthesiscandisturbtheX-Raydiffractionpattern,imposingalimittoaccesstotheatomisticmolecularinteraction.Computationalmodellingtechniquescanovercomesomeexperimentallimitationsandenhanceouracquaintance of the LDH structure. In thiswork, an atomistic framework has been developed formolecular dynamics (MD) simulations with Gromacs, using the ClayFF force field, to study themechanism associated with the ion-exchange equilibrium in Zinc-Aluminium LDHs. Thereby, MDsimulations have been carried out to analyse the interaction between the LDH structure and theintercalated ions togetherwith the LDH expansion/contraction caused by the nitrate-chloride ionexchange. Our results have been compared with previous theoretical and experimental studiesdevelopedinourgroup[7,8].TheMDmodelpresentedinthisworkopensthedoortostudyindetailthestructureofLDHsandthecomplexmetallayer-ionsinteractions.[1]Greenwell,H.C.,Jones,W.,Coveney,P.V.,Stackhouse,S.,J.Mater.Chem.,16(2006)708.[2]Lu,Z.,Qian,L.,Tian,Y.,Li,Y.,Sun,X.,Duan,X.,Chem.Commun.,52(2016)908.[3]Li,L.,Gu,Z.,Gu,W.,Liu,J.,Xu,Z.P.,J.ColloidInterfaceSci.,470(2016)47.[4]Lu,Y.,Jiang,B.,Fang,L.,Ling,F.,Gao,J.,Wu,F.,Zhang,X.,Chemosphere,152(2016)415.[5]Yan,Z.,Zhu,B.,Yu,J.,Xu,Z.,RSCAdv.,6(2016)50128.[6] Yokoi, T. Hara,M., Seki, T., Terasaka, S., Kamitakahara,M.,Matsubara. Cryst. Eng. Comm.18(2016)1207.[7]Tedim,J.,Kuznetsova,A.,Salak,A.N.,Montemor,F.,Snihirova,D.,Pilz,M.,Zheludkevich,M.L.,Ferreira,M.G.S.,Corros.Sci.,55(2012)1.[8]Galvão,T. L.P.,Neves,C. S.,Caetano,A.P. F.,Maia, F.,Mata,D.,Malheiro,E., Ferreira,M. J.,Bastos,A.C.,Salak,A.N.,Gomes, J.R.B.,Tedim,J.,Ferreira,M.G.S.,J.Colloid InterfaceSci.,468(2016)86.

BenchmarkingofGCMCSimulationProgramsinApplicationtoGasAdsorption

RichardJ.Gowers,AmirH.Farmahini,DanielFriedrich,LevSarkisov

InstituteforMaterialsandProcesses,SchoolofEngineering,TheUniversityofEdinburgh,UK

GCMC simulation can serve as a powerful tool in the computational screening of newmaterials for problems such as carbon capture and hydrogen storage. There are a widevarietyoffreelyavailableprogramsforperformingthesesimulations,howevertodatetherehasbeennocomprehensivecomparisonof theiraccuracyandperformance. Indeed,evendefining the rate at which sampling is occurring in MC simulation is not immediatelyobvious.In this work we evaluate the performance of several popular Monte Carlo simulationprograms;Cassandra,DLMonte,Music,RaspaandTowhee;inmodellinggasadsorptionincrystallinematerials.Firstly,wesetoutareferencecaseofCO2adsorptioninIRMOF-1andcompare the accuracy of different programs inmodelling this system.We then set out amethodologyfordefiningMCsimulationlengthintermsofthestatisticalinefficiencyoftheresults. Using thismetric, we can then quantify and directly compare the computationalperformanceofthedifferentprograms.OurresultsshowthattherearelargedifferencesinthestatisticalvalueofanindividualMCstepacrossdifferentprograms.ThisisaccompaniedbyareversetrendintherateatwhichMCstepsareproducedbyprograms,whichwouldmakeadirectcomparisonoftheratesatwhich MC steps are performed very misleading. The reference cases and performancemetricsdevelopedinthisworkareintendedtoserveasaresourcetothecommunitybothasastartingpointforvalidationanddevelopmentofnewandexistingprograms,aswellastemplatesforbestpracticesinperformingGCMCsimulation.

SynthesisandCharacterizationofCrystallinePhasesofSmall-MoleculeSemiconductorsandInSilicoEstimationofTheirChargeCarrierMobility

OlgaGuskova,1,2,*DeyanRaychev,1,2YevhenKarpov,1AntonKiriy,1JörgGrenzer,3GotthardSeifert2,4andJens-UweSommer1,2,5

1LeibnizInstituteofPolymerResearchDresden(IPF),01069Dresden,Germany2DresdenCenterforComputationalMaterialsScience(DCMS),TechnischeUniversitätDresden,01062

Dresden,Germany3Helmholtz-ZentrumDresden-Rossendorfe.V.,01328Dresden,Germany

4TheoreticalChemistry,TechnischeUniversitätDresden,01062Dresden,Germany5InstituteofTheoreticalPhysics,TechnischeUniversitätDresden,01069Dresden,Germany

Small-molecule semiconductors represent an important class of materials for organicelectronic devices. The motivation toward investigation of oligomers is caused by morefacilecontroloftheirtransportcharacteristicsowingtomoleculardesignandtheabilityofsmall molecules to pack into well-organised polycrystalline thin films. The coupling ofelectron-deficient fused ring cores, e.g. benzothiadiazole (BTZ) [1,2] anddiketopyrrolopyrrole [3,4] with electron-rich heterocyclic flanks, e.g. thiophene (Th) andfuran(Fu),leadstodonor-acceptor-donormolecularstructuresthatarethemostprevalentactivelayercomponentsinplasticelectronics.Since electrical transport measurements of materials under operating conditions aredifficult to perform, in silico estimation of hopping parameters has become a newmethodology inmaterials design. It requires the hierarchical coupling of time and lengthscales inacombinationof several computational techniques (DFT,MC/MMandMD),andpre-existing knowledge of experimental crystal structures. The aim of this work is toinvestigate the effect of single-atom substitution in the heterocycle, here sulphur foroxygen, on the structural and electronic properties,molecular packing and electron/holemobilityincrystallinephasesofbenzothiadiazole-coredmolecules.Forthefirstcompound(Th-BTZ-Th)withknowncrystalstructure[5],boththeMarcus-Hushtransferrateandthechargecarriermobilityareestimatedforallpossiblehoppingpathwaystakingintoaccountthermaldisorder.It isfoundthattheelectronhoppingproceedsfasterfor both cofacial and herringbone molecular arrangements. The lowest predicted holemobilityof0.3±0.2cm2/V·sis largerascomparedtotheexperimentallymeasuredaverage0.038cm2/V·sandmaximal0.06cm2/V·sholemobility [5].For thesecondobject (Fu-BTZ-Fu), themolecularpacking inthesolidstate isamissing linkfor insilicoestimationofthecharge transport. Here we utilize a joint theoretical packing structure prediction andexperimentalpowderX-raydiffractionanalysis, SEM/TEMandDSCcharacterizationof Fu-BTZ-Fucrystalstodescribethemorphologyandtorefinethecrystalunitcellparametersforfurthercalculations.ThisworkisfundedbytheEuropeanUnion(ERDF)andtheFreeStateofSaxonyviatheESFproject100231947(YoungInvestigatorsGroup"ComputerSimulationsforMaterialsDesign–CoSiMa").TheauthorsaregratefultoDr.PetrFormanek(IPF)fortheSEM/TEMmeasurements.

[1]RaychevD.,etal.Phys.Chem.Chem.Phys.2017,19,8330[2]RaychevD.,etal.Comp.Mater.Sci.2017,126,287[3]MakarovaM.V.,etal.Int.J.QuantumChem.2016,116,1459[4]KarpovY.,etal.Macromolecules2017,50,914.[5]PatiP.B.,etal.ACSAppl.Mater.Interfaces2013,5,12460.

QuantifyingEnergeticBarrierstotheDiffusionofC8AromaticsinPorousOrganicCages

EdwardJacksonandKimE.Jelfs

DepartmentofChemistry,ImperialCollegeLondon,SW72AZ

Separationprocessesaccountfor15%oftheworld’stotalenergyconsumption,1sothereisanactiveresearchinterestinthedevelopmentoflessenergyintensiveseparationmethods,suchas theuseofmembraneseparations.Onesuchseparation is theseparationofpara-xylenefromitsstructuralisomers(Figure1),andthisprojectisexploringthepotentialuseof porous organic cages (Figure 2) to carry out this separation. Porous organic cagemembranes already exist as a proof of concept,2 and the similar sizes of thewindows ofmanyporousorganiccageswithrespecttothekineticdiametersofC8aromaticmoleculescould allow for a shape-selective separation.3 Various computationalmethods have beenemployed toexplore the feasibilityof sucha separation, includingmoleculardynamics toanalyse dynamic changes in window diameters and metadynamics to analyse the freeenergy barriers, both in single molecules and in the cages’ crystalline and amorphousstructures. Results show that a C8 aromatic molecule’s kinetic diameter is roughlyproportional to the energybarrier for crossingbetween cages, indicating that diffusion isconsiderably slower for the larger isomers. This suggests that such a separationmay bepossible,bytrappingthelargerisomersinthecageswhilstthesmalleronesdiffusethroughwithrelativelylittledifficulty.

Figure1:Para-xyleneanditsstructuralisomers

Figure2:TheporousorganiccageCC3a)asasinglemolecule(green=C,blue=N,white=H)b)showingthecrystalpackingbehaviour(red=cyclohexane,blue=aromatic/imine

linker,cyan=porenetwork)1 D.S.ShollandR.P.Lively,Nature,2016,532,435–437.2 Q.Song,S.Jiang,T.Hasell,M.Liu,S.Sun,A.K.Cheetham,E.SivaniahandA.I.Cooper,Adv.Mater.,2016,28,2629–2637.3 T.Mitra,K.E.Jelfs,M.Schmidtmann,A.Ahmed,S.Y.Chong,D.J.AdamsandA.I.Cooper,Nat.Chem.,2013,5,276–281.

a) b)

NegativelyChargedProteinAdsorptiontoNegativelyChargedSurface.MolecularDynamicsCaseStudyofBSAAdsorptiononSilica

KarolinaTokarczyk1,KarinaKubiak-Ossowska2

1InstituteofCatalysisandSurfaceChemistry,PolishAcademyofSciences,2DepartmentofPhysics,UniversityofStrathclyde

How proteins adsorb to inorganic material surfaces is critically important for thedevelopmentofnewbiotechnologies.Sincetheorientationandstructureof theadsorbedproteins impacts their functionality. While it is known that many negatively chargedproteins readily adsorb to negatively charged oxide surfaces, a detailed understanding ofhowthisprocessoccursislacking.UsingfullyatomisticMolecularDynamic(MD)simulationswehave studied the adsorptionofBSA, an important transport protein that is negativelycharged under physiological conditions, to a model silica surface that is also negativelycharged.MDsimulationsprovidedanewinsightintotheprocessesallowingforadsorptionofnegativelychargedproteinonnegativelychargedsurface.Surprisinglytheadsorption isstill driven by electrostatic forces while other forces only regulate the final proteinorientationon the surface.Now thedetails of thenoncovalent interactions thatbind theBSAtothesilicasurfaceareunderstoodandtheroleoflong-rangeelectrostaticandothershort-rangeforcesisfullyelucidated.

SimulationofComplexationofRareEarthElementsinAqueousChlorideSolutions.

SebastienLectez1,AaronR.Finney2,StephenStackhouse1,ColinL.Freeman2andJohnH.Harding2

1SchoolofEarthandEnvironment,Math/EarthandEnvironmentBuilding,TheUniversityofLeeds,Leeds,LS29JT,UnitedKingdom.

2DepartmentofMaterialsScience&Engineering,SirRobertHadfieldBuilding,TheUniversityofSheffield,Sheffield,S13JD,UnitedKingdom.

Rare earth elements (REE) are found in many modern technologies, in particular, in theautomotive, renewableenergyanddefencesectors.There isconcern inEuropeaboutthesecurityofsupplyofthesecriticalmaterials,whichiscurrentlydominatedbyChina.Inordertoaddressthisissue,newsourcesandmoreefficientandenvironmentallyfriendlywaystoextractandrecoverREEarebeinginvestigated.The importance of aqueous solutions in the formation of ore deposits and theirconcentrationof REE, at lowandhigh temperatures has been known formany years [1].Indeed, REEmobility ismainly controlled by the availability and degree of complexation,with ligands such as Cl-, F-, PO43- and SO42-, the pH of fluids and redox conditions[2].However,thestructure,stabilityandmobilityofthesecomplexesatrelevanttemperaturesisnotwellknown.Here, we investigate the structure, stability and mobility of Nd3+, Gd3+ and Er3+ chloridecomplexes in aqueous solution by density functional theory (DFT) and classicalmoleculardynamics.Aforcefieldwasdevelopedandligandexchangereactionssuchas,[REE(H2O)x]3++Cl–=[REE(Cl)(H2O)y]2++(x-y)H2O,were studiedusing adaptivebias and advanced sampling simulations,which also enabledthecalculationofrelativefreeenergies.Complexstabilityconstantsweredeterminedandcomparedwithexperimentaldata,andmechanismsforligandexchangewereelucidated.Inaddition, we report the subtle effects of temperature and salt concentrations on theaforementionedmechanismsandassociatedfreeenergychanges.Thisworkisanimportantfirst step todesigningmoreefficientandenvironmentally friendlymethods to fractionateREEfromsolution.[1]B.W.D.Yardley,EconomicGeology,100,613(2005)[2]R.A.Mayanovicetal.,ChemicalGeology,259,30-38(2009)

InvestigatingtheInterfacialPropertiesandAggregationofBileSalts

OliviaPabois1,CecileA.Dreiss1&ChristianD.Lorenz2

1InstituteofPharmaceuticalSciences,King’sCollegeLondon,LondonSE19NH2DepartmentofPhysics,King’sCollegeLondon,LondonWC2R2LS,UK

The intake of dietary fat and its effects on health has become amajor focus ofmodernsocieties.Therateoffatdigestionandsubsequentabsorptioniscontrolledbytheabilityoflipaseanditsco-enzymeco-lipasetobindtofatdropletinterfaces,aprocesscontrolledandfacilitated by bile salts (BS), which are bio-surfactants with an unusual planar structurepresentinthegastrointestinaltract. Thisprocessisnotonlyrelevanttotheabsorptionofdietary fatand fat-solublenutrientsbutalso thatofpoorlysolubledrugs. A fundamentalunderstandingofthecolloidalandinterfacialprocessesgoverninglipiddigestionanduptakeis needed to guide rational design strategies. As a first step toward obtaining thisunderstanding,weareusingclassicalmoleculardynamicssimulationsincombinationwitharangeof experimental techniques to characterise the interfacial behaviourof bile salts ofvarious chemical composition at the air/water interface and at the interface of a lipidmonolayer. Also, we are using a similar combination of simulation and experimentalmethods to investigate the interactions which govern the self-assembly of bile saltsthemselves as well as in the presence of lipidmolecules. In this talk, I will present theresultsofthesesimulations,indoingsohighlightinghowchangingthechemistryofthebilesalts effects the interfacial properties and how they play a key role in the aggregationmechanisms. Finally, I will put the various simulation results into context of the variousexperimentalresultsthatwehaveobtained.

UnderstandingSelfAssemblyofSilicaPrecursorsUsingMonteCarloSimulations

InderdipShereandAteequeMalani

DepartmentofChemicalEngineering,IITBombay,India.

Molecular precursor of silica is used to synthesis several type of fascinating engineeringmaterials. These materials exhibit superior properties like, excellent thermal insulation,higherporosity,biocompatibilityetc.Severalmolecularmechanismsoccurduringsynthesis,and the knowledge of such eventswould be useful in synthesizing desired products. Thegoal of this work is to understand the molecular events occurring in the polymerizationprocess. Our study involves the polymerization of model silica system using reactionensembleMonteCarlotechniques.Wehavedevelopedalgorithmstocapturethekineticsofmolecular events (i.e. translation, rotation and reactions of particles) occurring duringsynthesis.Thealgorithmswerevalidatedbycomparingtheevolutionofsilicaspecieswithexperimental evolution. The developed algorithms were robust enough to studypolymerization for silica with different functionally.Wehaveanalyzedthesynthesisprocessbytrackingtheevolutionofclustersizeandshape.Thefinalresultingclusterattheendofreactionwasfoundtobeellipsoid(prolate).Wealsostudied the effect of dilution on the polymerization and observed that the degree ofcondensation (DOC)decreaseswithdilution.TheDOC for thesimulationwere lesser thanthetheoreticallimit,whichisduetodelayedkineticsoftheprocess.Wehavealsostudiedthe effect of functionally on polymerization and found that the resulting size, shape andporositydependsstronglyontheinitialprecursormixture.

Keywords:Poroussilica,polymerization,Selfassembly,MonteCarloSimulations

SimulationsToUnderstandReflectivity:HowCoarseCanWeGo?

AndrewR.McCluskey1,2,JamesGrant1,RobertD.Barker3,StephenC.Parker1,KarenJ.Edler1

1. UniversityofBath,ClavertonDown,Bath,UK2. DiamondLightSource,DiamondHouse,Rutherford-AppletonLaboratory,Harwell-Oxford,UK

3. SchoolofScienceandEngineering,UniversityofDundee,Dundee,UK

As thecomplexityof thesystemsstudiedwithneutronandX-ray reflectivity increases, somust the methodology applied to the analysis. The current analytical procedure that iswidelyusedfortheinterpretingreflectivitymeasurementsinvolvestheuseoflayermodels,which are unsuitable for highly complex systems. For this reason, all-atom and coarse-grainedmoleculardynamicssimulationshavebeeninvestigatedtoaidinthecalculationofreflectivity profiles [1-3]. These have involved the use of research group-specific code toslicesnapshotsofthesimulationboxintolayersandapplytheAbelèsmethod;generatingareflectivityprofile.Inthiswork,wehaveproducedtheopen-accesssoftware,falass[4]whichhasbeenusedtodetermine reflectivity profile froma lipid systemat various levels of coarse-graining. ThisinvolvedthesimulationofaDSPCmonolayeratawater-airinterface,usingaseriesofforcefields;includingall-atomandMartinicoarsegrained.UsingthiswehavebeenabletoassesstheeffectivenessofeachforcefieldtoreproducethereflectivityfromtheDSPCmonolayers.This knowledge will inform the level of simulation resolution required to ensure theaccurate modelling of systems of higher complexity, such as tethered lipid bilayers, orprotein-lipidinteractions.

Figure3.ReflectivityfromaMDtrajectory.

[1] Dabkowska,A.P.,etal.(2014).Langmuir,30(29),8803–8811.[2] Darré,L.,etal.(2015).J.Chem.Theor.Comput.,11(10),4875–4884.[3] Koutsioubas,A.(2016).J.Phys.Chem.B,120(44),11474–11483.[4] McCluskey,A.R.(2017).Falass(v0.1).[computerprogram].Availableat:

<http://people.bath.ac.uk/chske/research/software/falass/index.html>[Accessed30Jan2017].

logQ

(Å-1

)

probing radiation

TransferableForceFieldParameterisationofOrganosilicatesUsingMultiscaleModelling

AndrewMilne

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,G11XJ,Glasgow,UK

Silicateandorganosilicatemoleculesareextremely important inawidevarietyofnaturalandindustrialprocesses,includinggeochemistry,biosilicification,bonetissuegrowth,glassandceramicsproduction,polymermanufacture(e.g.,PDMS),andsynthesisofnanoporousmaterials. Although silica is normally encountered in the form of crystalline solids (e.g,quartz, zeolites), their synthesis almost invariablyoccurs in the liquid stateor in aqueoussolution from organo(silicates). With the exception of a few studies devoted to specificmolecules(e.g.,silicicacid)[1,2],thereisnogeneralmolecularmodelthatcanbeappliedtoorganosilicates–infact,thesiliconatomisconspicuouslyabsentfrommostgeneralliquid-state force fields, such asOPLS [3].Hence, it is ourbelief that better force fieldsof suchsystemswouldhelpusbetterunderstand the complex synthesisof suchmaterials and toachieve this we need to integrate information from several levels of theory on differentlengthscales.In this work, we focus on parameterising a new transferable atomistic force field fororganosilicate molecules suitable for the liquid state and aqueous solutions of thesemoleculesusingacombinationofmoleculardynamics (MD)anddensity functional theory(DFT) calculations. Thebasis for our new force field is theUnitedAtom (UA) formof theTransferable Potentials for Phase Equilibria (TraPPE) force field [4], and we extend it toincludeparametersforsiliconandadjacentoxygen,hydrogenandcarbonatoms.Wehaveused quantum chemistry (QM) calculations to calibrate parameters for the torsionalpotentials.TheresultsofQMwerealsofittedtoobtainpointchargesonthemoleculesofinterest.Withtheseparametersinplace,weranmoleculardynamicssimulationsandfittedthe Lennard-Jones parameters to match experimental data for density and heat ofvaporisationof selectedorganosilicatemolecules. Theparameters thatwehaveobtainedfrom this first approach are reasonable but show some systematic discrepancies for theheats of vaporisation, which we hypothesise is due to neglecting polarisation effects. Assuch,specialattentionhasbeenpaidtoproducingelectrostaticchargeswhichcorrespondaccurately to the liquid- or solution-like environment of the molecules, thus hoping tocapture polarisation effects in an implicit way within a non-polarisable model. Theparametersofthenewmodelarevalidatedagainstarangeofthermodynamicproperties,including free energies of solvation. We expect this new model to find applicability inprocesses such as those mentioned above, which rely on interactions involving silicatemoleculesinsolution.[1]J.C.G.Pereira,C.R.A.Catlow,G.D.Price;J.Phys.Chem.A.,2002,106,130-148[2]Miguel Jorge, JoséR.B.Gomes,M.NatáliaD.S.Cordeiro,NigelA.Seaton; J.Phys.Chem.B.;2009,113,708-718[3]W.L.Jorgensen,D.S.Maxwell,J.Tirado-Rives;1996,J.Am.Chem.Soc.118,11225-11236[4]MarcusG.Martin,andJ.IljaSiepmann;J.Phys.Chem.B,1998,102,2569-2577

MultiscaleModellingofPVDFBulkandSurface-InterfacePhysicalProperties

FrancescoPelizza

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,G11XJ,Glasgow,UK

Ferroelectric polymers, such as poly(vinylidene difluoride) (PVDF), have many potentialapplications in flexible electronic devices. PVDF has six experimentally observedpolymorphs,threeofwhichareferroelectric.Inthisworkweusedensityfunctionaltheorytoinvestigatetheenergetics,vibrationalfrequenciesofallPVDFpolymorphs.Furthermore,we studied the crystal structures of γ, ε phases, hence, the polarisation of the stableferroelectricγphase.Then,atheoreticalcharacterisationofthenonferroelectricanalogueof the well know ferroelectric β phase is here presented. The results from a variety ofexchangecorrelationfunctionalswerecomparedanditwasfoundthatvanderWaals(vdW)interactions have an important effect on the calculated crystal structures and energetics,with the vdW-DF functional giving the best agreement with experimental latticeparameters. While the relative phase energies were not strongly dependent on thefunctional, the cohesive energies were significantly affected. The inclusion of vdWinteractionsis,therefore,importanttoobtainthecorrectlatticestructures,polarisationandenergetics of PVDF polymorphs. Furthermore, the investigation of PVDF with classicalMechanicstheoryusinglammpshasbeenperformedaswellinordertofurtherunderstandPVDFchainsorientationonsurface.Bycross-chekingchainsequilibriumconformationatthebulkweproceededtoinvestigatethesurfaceconfiguration.F. Pelizza, B.R. Smith, and K. Johnston. A van der waals density functional theory study of poly(vinylidenedifluoride) crystalline phases. The European Physical Journal Special Topics, pages 1–10, 2016.

Electron-beaminducedchemistryoforganicmolecules

StephenT.Skowron,1ThomasW.Chamberlain,1JohannesBiskupek,2UteKaiser,2AndreiN.Khlobystov,1ElenaBesley1

1SchoolofChemistry,UniversityofNottingham,Nottingham,NG72RD,UK2CentralFacilityofElectronMicroscopy,UniversityofUlm,89081Ulm,Germany

Transmission electron microscopy (TEM) is traditionally used as a tool to characterise materials,providing atomic resolution imaging of low dimensional nanostructures such as graphene andcarbonnanotubes. In this context,damage tomaterials imagedbyTEM (causedbycollisionswiththe highly energetic electrons) is generally considered as something to be avoided or limited.However, with detailed understanding of the effects of the electron beam (e-beam), the energytransmittedfromitcanbeusedtodrivechemicalreactionsthatwouldbeotherwiseunfeasible.A mechanistic understanding of beam-driven chemical reactions can be achieved with thecomparison of experimental TEM images to the results of modelling. The dynamic response ofnanotube-encapsulated organic species to the stimulus of the e-beam has been simulated usingdensity functional theory (DFT)molecular dynamics. By combining these resultswith an accurateanalytical model of the interaction of relativistic electrons and atomic nuclei, the experimentallyobservedbehaviourofthesesystemsunderthee-beamhasbeenquantitativelycharacterised.Theelementaldependenceofthetransferofenergyfromthee-beamwasshowntoplayakeyrolein determining reaction products, and is responsible for the very high susceptibility of carbon-hydrogen bonds to irradiation damage.1 Deuteration is an effective remedy for overcoming thislimitation, increasing lifetimes of organic molecules under electron irradiation and thereforeenhancing the accuracy of structural analysis by TEM.2 A close iterative collaboration betweentheory and microscopy was used to establish TEM as an effective tool for chemical reactiondiscovery and the characterisation of previously unknown reactionmechanisms.3 This has initiallybeendemonstratedwithtwoexamplereactions,inwhichorganicmoleculeprecursorsareactivatedbythee-beam,eventuallyformingnovelone-dimensionalmaterials(Figure1).4

Figure1.Theinitialstagesofe-beaminduced1Dpolymerisationofoctathio[8]circulenemolecules.

ExperimentalTEMimages,top;DFTandmultisliceimagesimulations,bottom.

1. S.T.Skowronetal.,Nanoscale,2013,5,6677–6692.2. T.W.Chamberlainetal.,Small,2015,11,622–629.3. S.T.Skowronetal.,Acc.Chem.Res.,2017,acs.accounts.7b00078.4. T.W.Chamberlainetal.,ACSNano,2017,11,2509–2520.

Phasediagramsofmolecularsolidsfromlattice-switchMonteCarloinDL_MONTE

TomL.Underwooda,NigelWildinga,AndreyBrukhnob,JohnPurtonb,JamesGrantc,SteveParkerc

aDepartmentofPhysics,UniversityofBath,UKbSTFCDaresburyLaboratory,UK

cDepartmentofChemistry,UniversityofBath,UK

Onekeyapplicationofmolecularsimulationistodeterminethephasebehaviourofagivensubstance under given conditions, e.g., temperature and pressure. This problem can berecast in termsof free energies: calculating the free energy differencebetween a pair ofcandidate phases will reveal which of the two is stable at the conditions underconsideration.Lattice-switchMonte Carlo (LSMC) is an efficient method for evaluating the free energydifference between two solid phases. Since its inception LSMC has predominantly beenappliedto`simple'systemssuchastheLennard-Jonesandhard-spheresolids.Howevertherecent implementationof LSMC in the general-purposeMonteCarloprogramDL_MONTEhas opened the possibility of applying the method to molecular crystals modelled by`realistic' force fields. This is an interesting prospect because LSMC is ostensibly moreefficient than other methods for evaluating solid-solid free energy differences. ThusDL_MONTEcouldproveavaluabletoolforstudyingphasebehaviourinmolecularcrystals.Here we present the results of our incipient applications of LSMC in DL_MONTE tomolecular crystals. The key system we have considered is ice. Crucially, we have gonebeyond the problem of using LSMC to determine which ice phase is stable at a singletemperatureandpressure:wehaveusedLSMC toefficiently calculate coexistence curvesforvariousicephases.

IonPermeationinGrapheneOxideMembranes:aMolecularSimulationPerspective

ChristopherD.Williams,JamesDix,PaolaCarboneandFlorR.Siperstein

SchoolofChemicalEngineeringandAnalyticalScience,UniversityofManchester,Manchester,M139PL,UK.

When graphene oxide (GO) membranes are immersed in water the interlayer spacingbetweenGOflakesincreasesandthemembraneswells.Inthisswollenstate,waterandionsare thought to permeate through the membrane via a network of interconnectedunoxidised2Dpores.1Careful controlof the swellinghasnowbeen realised,enabling ionpermeation selectivity through the membrane.2 This breakthrough means that GOmembranescouldbeusedforthewaterpurificationbynanofiltrationinthefuture.In this work, the ion-rejection properties of GO membranes were investigated usingmolecular dynamics simulations and the umbrella sampling technique.2,3 The simulationshave demonstrated, using simple 2D pore models, that relative permeation rates aredeterminedbythefreeenergyassociatedwithiondehydrationuponenteringthepore.Theimplicationsofthisfindingarediscussedinthecontextofdesalinationandtheremovalofproblematicradioactivecontaminants,suchas99Tc.In order to resolve quantitative differences in permeation energy barriers betweenexperiment and simulation, we propose alternatives to the widely assumed permeationpathwayofionsthroughGOmembranes.

1. RKJoshi,PCarbone,FCWang,VGKravets,YSu,IVGrigorieva,HAWu,AKGeimandRRNair,PreciseandUltrafastMolecularSievingThroughGrapheneOxideMembranes,Science,2014,343,752-754.

2. J Abraham, KS Vasu, CD Williams, K Gopinadhan, Y Su, C Cherian, J Dix, E Prestat, SJ Haigh, IVGrigorieva, P Carbone, AK Geim and RR Nair, Tunable Sieving of Ions Using Graphene OxideMembranes,Nat.Nanotechnol.,acceptedforpublication.

3. CDWilliams and P Carbone, Selective Removal of Technetium fromWater Using Graphene OxideMembranes,Environ.Sci.Technol.,2016,50,3875-3881.

LithiumDiffusioninComplexPhosphidosilicateMaterials

StephenR.Yeandel(a)*,DavidO.Scanlon(b),PoojaM.Panchmatia(a)

(a)LoughboroughUniversity,DepartmentofChemistry,EpinalWay,Loughborough,Leicestershire,LE113TU,UK.

(b)UniversityCollegeLondon,DepartmentofChemistry,20GordonStreet,London,WC1H0AJ,UK.

Recent work has demonstrated new materials within the Li-Si-P system [1, 2]. Thesematerials contain a complex framework of phosphidosilicate “super-tetrahedra”, amongwhich lithium atoms are accommodated. Experimentally, the energy barrier to lithiumdiffusion in thesematerials is foundtobe low(<0.1eV),making theman interestingnewsolidstateelectrolytematerialclass.However,issuesexistwithairsensitivityandquestionsremainoverpotentialdopingstrategies.InthisreportpossibledopingstrategieswithintheLi-Si-P systems are evaluated using DFT calculations. The most favourable dopingmechanism is thenused inab initiomoleculardynamicssimulationsandcomparedtothepuresystemtoelucidateanyimpactonthediffusionoflithiumions.AcknowledgementsThiswork recognizes theuseofboth the ‘Hydra’HighPerformanceSystematLoughboroughUniversityandtheARCHERUKNationalSupercomputingService.[1]A.Haffner,T.Bräuniger,D.Johrendt,Angew.Chem.Int.Ed.,55.(2016)13585-13588.[2]L.Toffoletti,H.Kirchhain,J.Landesfeind,W.Klein,L.vanWüllen,H.A.Gasteiger,T.F.Fässler,Chem.Eur.J.,22.(2016)17635-17645.

POSTERPRESENTATIONS

DiffusionandReactionPathwaysofWaternearFullyHydratedTio2SurfacesfromAb-InitioMolecularDynamics

LorenzoAgosta

DepartmentofMaterialsandEnvironmentalChemistry,StockholmUniversity,Sweden

Ab initio molecular dynamics simulations are reported for water-embedded TiO2surfaces to determine the diffusion and reactive behavior at full hydration. A three-domain model is developed for the six surfaces [rutile (110), (100), and (001), andanatase (101), (100), and (001)] that describes surface-bound waters in terms of“hard” (irreversibly bound), “soft” (with reduced mobility but orientation freedom)hydration layers, and bulk water. The model helps interpreting previous experimentaldata and gives a detailed and consistent picture of water diffusion near TiO2 surfaces.Water reactivity is discussed in detail, and in addition to direct water splitting, graphtheoretic analysis reveals a number of water reaction pathways on TiO2 that occur atfull hydration. Hydronium (H3O+) is identified to be the key intermediate state andfacilitates water dissociation by proton hopping between (surface-adsorbed) intactand dissociated waters. These discoveries significantly improve the understanding ofnanoscalewaterdynamicsandreactivitynearTiO2interfacesunderambientconditions.

Multi-scalematerialgrowthinextremeenvironments

AndrewBellandPaulMulheran

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,UK

We aim to develop a kinetic Monte Carlo (kMC) model of surface evolution in extremeenvironments when a crystalline material is bombarded by high energy atoms causingsputtering,growthandsurfaceroughening.WeexploitresultsofMolecularDynamics(MD)simulationstodevelopthekMC,whichhasthepotential tocapturethe importantgrowthphysicswhilebeingordersofmagnitudefaster,allowingustosimulateontimeandlengthscales that would be unfeasible for MD due to the computational cost. Here we use alattice-basedmodelwithdeposition,sputteringandsurfacediffusionrules inspiredbytheMDsimulations.Whilewecansuccessfullycapturemanyfeaturesofthesurfaceevolution,challenges remain to fully replicate the movement of ad-clusters and vacancy islands.Neverthelessthetoolallowsustoconsiderfinitesizeeffectsandthescale-uptolongtimeand length scales. Ultimately, we wish to explore the feasibility of coupling the KMC tocomputational fluiddynamics (CFD)methodologies suchasdirect simulationMonteCarlo(DSMC)applicabletorarefiedgasenvironments.

EndeavoursofDL_MONTE-2–projectprogress&recentupdates

AndreyBrukhno*,TomUnderwood†,JamesGrant‡,JohnPurton*,NigelWilding†andSteveParker‡

*STFC,DaresburyLaboratory,UK,†Physics&‡Chemistry,UniversityofBath,UK

DL_MONTE-2 is thecurrentCCP5 flagshipproject thathasbeen instigatedbyand fundedthroughanEPSRC“SoftwarefortheFuture”programme.ThedevelopmentofDL_MONTE-2is aimed at providing the academic research community with a state-of-the-art generalpurpose software engine forMonte Carlo (MC) simulation of condensed/softmatter andother complex materials. Amongst the objectives of DL_MONTE-2 is to develop novelfunctionalitieswithin theMC simulationmethodology, as a complementary alternative tomoleculardynamicssimulations,e.g.fortheusersofDL_POLYpackage(s).The poster reports on the advances in development of DL_MONTE-2 software – recentupdates,releases,aswellashighlightthemostprominentnewfeaturessuchasfreeenergydifference(FED)methodologies,lattice/phaseswitch(PSMC)implementations,planarnano-pore(slit)confinementandcurrentprogressintacklingorganic-inorganicinterfacesandbio-chemicalapplications.

Mesoscalemodellingofdimethacrylate-basedbiomaterials

LauritzT.Bußfeld,PeterBehrens,AndreasM.SchneiderInstituteforInorganicChemistry,LeibnizUniversityHannover

Thedevelopmentofadvanceddentalmaterialsbasedonpolymershasattractedgrowinginterestintherecentdecade.Particularlyresinscurablebyphoto-inducedradicalpolymerizationareadvantageousoverconventionallyusedimplantslikeceramicsandnoblemetalalloys,regardingprocessabilityandaestheticproperties. Additional incorporation of inorganic filler materials leads to durable dental composites,ubiquitouslyused forpartial remodellingand fillingof cavities inhumanbonesand teeth.[1]However,the decrease in volume of polymer-based materials with increasing double bond conversion duringpolymerizationisamajordrawback.Besideutilizationoffillermaterialsforthecompensationoftheso-calledpolymerizationshrinkage,theselectionofwell-suitedmonomersleadstosignificantimprovement.Thetargetofourworkistoelucidatetheunderlyingprocessesandpredictthemechanicalpropertiesofthecomposites.Whilethestructureofpolymerscannoteasilybeaccessedfromanexperimentalpointofview,computationalmethodsfacilitatedetailedinvestigations.

InvestigationviaCoarseGrainedmodelsForpolymers,computerbasedsimulationmethodsofferapowerfultoolforinvestigationandpredictionof mechanical properties. Due to the high radii of gyration of macromolecules, sufficiently largeensembles of atoms and consequently large simulation cells are necessary, leading to long simulationtimeson theatomistic scale.Oneway tomanage the simulationof cellswith such spatialextent isbyutilizing Coarse Grained (CG) or mesoscale models, which lead to a significant acceleration of thecalculations.ThemainadvantageofCGmodelsliesintherepresentationofmonomersorunitsoflargermoleculesby single superatoms,often calledbeads.Using thismethodeliminates the time-consumingsimulation of e.g. bond vibrations or atomic pair interaction, which are then described by effectivepotentials.Ontheotherhand,informationofindividualcovalentorhydrogenbondsislost.Toavoidthisdisadvantage,asimulationondifferentsizescalesmaybeapplied,knownasmultiscalesimulation.Intheaforementionedmesoscalemodelspotentialparametersareemployedwhichcanbederivedfromreferencemodelsorexperimentaldata.Inpracticeitiscommontouseatomisticmodelsasareference.Inthisapproach,parameterizationoftheCGforcefieldisbasedonstructuralobservablesandisachievedbyIterativeBoltzmannInversion.[2,3]Inthiswork,weuseaurethanedimethacrylate,whichiscommonlyused indental applicationnowadays, asamodel system.This systemshowsa suitable viscosityof themonomermeltwhileexhibitingonlyamoderatepolymerizationshrinkage.Themodelconsistsofthreedifferentbead typespermonomer (Figure1). The terminalmethacrylate groupsare the reactiveunitswith regard to the polymerization and therefore are described as an individual bead type. The otherbeads were defined in such a way that all beads obtain approximately the same size. Most of theparameters for the mesoscale force field are derived from atomistic monomer melt simulations. Foroligomers and polymers only one additional bead type is introduced to describe the backbone units,therefore only two additional bond and angle interaction potentials have to be generated. The latterhavetobederivedfromatomisticsimulationsofanoligomer,whichresult inthemajorcomputationaleffortofourparameterizationapproach.

Figure1:Stickmodeloftheurethanedimethacrylatemonomerunderinvestigation.Transparent

ellipsoidsindicatethelocationofthebeadsdefinedinthemesoscalemodel.[1] X.Wangetal.,MaterialsScienceandEngineeringC,2016,59,464-470.[2] D.Reithetal.,J.Comput.Chem.,2003,24,1624-1636.[3] V.Agrawaletal.,Macromolecules,2014,47,3378-3389.

Anaccurateandtransferablemolecularmodeltopredictbinaryadsorptioninopenmetalsitecontainingmetalorganicframeworks

ChristopherCampbell1,MiguelJorge1,JoséR.B.Gomes2andMichaelFischer3

(1)DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,Glasgow,UK,(2)CICECO,DepartmentofChemistry,UniversityofAveiro,Portugal,(3)UniversityofBremen,KlagenfurterStraße,

Bremen,Germany

Thiswork focuses on developing accurate computationalmodels to predict adsorption inMetal-OrganicFrameworks(MOFs)withcoordinativelyunsaturatedmetalsites(CUS).MOFsareadsorbentmaterialsofgreatinterestwithintheresearchcommunityduetotheirmanydesirable characteristics such as large surface areas, high porosity and high degree oftailorability. In particular, MOFs containing CUS have demonstrated highly selectiveadsorptionbyformingstrongcoordinationbondswithspecificadsorbates,whichgivesthemgreat potential for challenging gas separations. The caveat ofMOF variability is that thisintroduces an overly deep pool of potential MOFs to be assessed through purelyexperimentalmeans.Furthermore,experimentallymeasuringcompetitivebinaryadsorptionis very difficult, and it is therefore often estimated on the basis of single-componentadsorptionthroughIdealAdsorbedSolutionTheory(IAST).However,IASThasbeenshowntobeinaccurateforsystemsinvolvingspecificgas-solidinteractions,asisthecaseinMOFswithCUS[1].Assuch, there ispotential forcomputationalmodellingtoplayapivotal role inadsorbentmaterial design, e.g., through high-throughput screening of MOFs. The caveat is thataccurate predictions of adsorption in new MOFs by computer simulation require thedevelopment of realistic molecular models, which is especially difficult for MOFs thatcontainCUS.Ithasrecentlybeendemonstratedthatconventional“off-the-shelf”molecularmodels are unable to correctly describe adsorption in CUS-containingMOFs [2]. Herewereport the latest developments of a new approach that combines accurate quantummechanical (QM) calculations with classical Monte Carlo simulations to enable accuratepredictions of adsorption for these complex materials [3,4]. The new model was firstvalidated against ethylene adsorption in Cu-BTC and showed good agreement withexperiment. Crucially, we demonstrate that the QM-derived potential parameters aretransferabletodifferentadsorbatesofthesametype(e.g.,ethylenetopropylene)aswellasbetween different MOFs with the same type of open metal site (e.g., Cu-paddlewheelMOFs). This has allowed us to predict multi-component adsorption isotherms, andcorresponding selectivities, in the context of several challenging gas separations (e.g.,ethylene/ethane; propylene/propane; CO/nitrogen) using MOFs with open metal sites.Overall, our new model provides detailed insight into the molecular level adsorptionmechanismsonMOFswithCUS,andconstitutesauseful tool todesignnewmaterials forchallengingseparations. [1]NaomiF.Cessford,NigelA.Seaton,andTinaDüren,Ind.Eng.Chem.Res.,51,4911-4921,2012[2]M.Fischer,J.R.B.Gomes,M.Jorge,Mol.Simul.,40,537-556,2014.[3]M.Fischer,J.R.B.Gomes,M.Froba,M.Jorge,Langmuir.28,8537-8549,2012[4]ChristopherCampbell, CarlosA. Ferreiro-Rangel,Michael Fischer, JoséR.B.Gomes, andMiguel Jorge, J.Phys.Chem.C,121,441-458,2017

ContinuousintegrationforDL_POLY_4:Abird’seyeview

AlinMElenaandIlianTTodorov

DaresburyLaboratory,ScientificComputingDepartment,Cheshire,UK

Whencomestoscientificsoftwaredevelopmentmethodologiesweseemtohaveaspecialone for each project. This approach hinders massively sustainability of the softwaredeveloped.Inthisposterwegothroughdifferentstagesofcontinuousintegrationforareallife,wellestablishedscientificcode,DLPOLY_4,CCP5flagship.Theexperiencedrawsfrominvolvement in few other collaborative scientific projects, lone projects and open-sourcecommunity. One is exposed to the various and the very different aspects needed for CI,ranging from hardware/software infrastructure, testing and tools to policy creation andenforcement.

DL_MONTE2:Developmentsandteachingmaterial

JamesGrant,AndreyBrukhno,TomUnderwood,SteveParker,JohnPurton,NigelWilding

ChemistryDepartment,UniversityofBath,UK

WewillreportdevelopmentswithintheDL_MONTEEPSRCfundedCCP5flagshipSoftwarefortheFutureproject.RecentworkhasincludedmolecularLatticeSwitchMonteCarloandcapability for 2D geometries, including interactionswith interfaces. Wewill also discussinstructive material developed from a workshop held at Bath, introducing Monte Carlomethodsandthefunctionalitythathasbeenintroducedtothecode.

Expandingthescaleofcondensedphasemodels

JamesDenhomandBenHourahine

DepartmentofPhysics,UniversityofStrathclyde,JohnAndersonBuilding,107RottenrowEast,GlasgowG40NG

The size of systems that can be simulated is limited by both the computational cost ofmodellingtheirbehaviourandtherequirementstonumericallyrepresentthem.Forsuitablecases, renormalization groupmethods can be used to exponentially expand the availablesize of the system that can be studied. However this is usually used, for example, whenexaminingscalefreeprocessessuchassecondorderphasetransitions.Hereweinsteadaimtouserenormalizationgrouprescalingtechniquestoallowthesameparameterspacetobeusedtosimulatesystemsonalargerscale.We apply these ideas to rescale both rates and length scales for Kinetic Monte Carlosimulationsoflatticesystemsundergoingcoarseningprocesses.TheIsingmodel(seeFig.1)captures theessentialbehavioursof systems includingmagneticorderingandalloyphaseseparations.Asafirstareaofapplication,westudythelowtemperatureGlauberdynamicsof the d=2 model on a square lattice. Our approach involves coarse graining the modelthroughblockspintransformations,categorizingtheeventsoccurringonthecoarsegrainedmodel and determining their associated kinetic rates. Our initial investigations arepresentedandpotentialproblemsarediscussed.

Figure1:Showssnapshotsofcoarseninginthenearest-neighbourzerotemperatureGlauber

dynamicsofthekineticIsingmodelat(a)0,(b)66,(c)150,(d)440and(e)1156×105spinflipevents.

Phasebehaviourofself-assembledmonolayerscontrolledbytuningphysisorbedandchemisorbedstates

SaraFortuna1,RenginPeköz2,DavidL.Cheung3,DavideDonadio4,andKarenJohnston5

1CenterforBiomedicalSciencesandEngineering,UniversityofNovaGorica,Slovenia;SISSA,Trieste,Italy;2

PhysicsGroup,FacultyofEngineering,AtilimUniversity,Turkey;3SchoolofChemistry,NUIGalway,Ireland;4

DepartmentofChemistry,UniversityofCaliforniaDavis,U.S.A.;5DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,U.K.

Theself-assemblyofmoleculesonsurfacesinto2Dstructuresisimportantforthebottom-upfabricationoffunctionalnanomaterials,andtheself-assembledstructuredependsontheinterplay between molecule-molecule interactions and molecule-surface interactions.Halogenated benzene derivatives on platinum have been shown to have two distinctadsorptionstates:aphysisorbedstateandachemisorbedstate,andtheinterplaybetweenthe two can be expected to have a profound effect on the self-assembly and phasebehaviourofthesesystems.Wedevelopedalatticemodel[1,2]thatexplicitlyincludesbothadsorption states,with representative interactionsparameterisedusingdensity functionaltheorycalculations [3,4,5].Thismodelwasused inMonteCarlosimulationsto investigatepattern formation of hexahalogenated benzene molecules on the platinum surface.Molecules that prefer the physisorbed state were found to self-assemble with ease,dependingontheinteractionsbetweenphysisorbedmolecules.Incontrast,moleculesthatpreferentiallychemisorbtendtogetarrestedindisorderedphases.However,changingtheinteractionsbetweenchemisorbedandphysisorbedmoleculesaffectsthephasebehaviour.We propose functionalising molecules in order to tune their adsorption states, as aninnovativewaytocontrolmonolayerstructure, leadingtoapromisingavenuefordirectedassemblyofnovel2Dstructures.[1]Fortunaetal.,J.Phys.Chem.B,1141849(2010) [2]Fortunaetal.,J.Chem.Phys.144134707(2016) [3]Peközetal.,J.Phys.Chem.C1186235(2014) [4]Peközetal.,J.Phys.Chem.C11620409(2012) [5]Johnstonetal.,Surf.Sci.644113(2016)

Multi-ScaleModelingoftheSynthesisofSurfactant-TemplatedMesoporousSilicaMaterials

GermánPérez-Sánchez1,Szu-ChiaChien2,ScottM.Auerbach2,3,PeterA.Monson2,MariaNatáliaD.S.Cordeiro4,JoséR.B.Gomes1,MiguelJorge5

1CICECO–AveiroInstituteofMaterials,DepartmentofChemistry,UniversityofAveiro,CampusUniversitáriodeSantiago,3810-193Aveiro,Portugal;2DepartmentofChemicalEngineering,UniversityofMassachusetts,686NorthPleasantStreet,Amherst,MA01003-9303,USA;3DepartmentofChemistry,UniversityofMassachusetts,686NorthPleasantStreet,Amherst,MA01003-9303,USA;4REQUIMTE,

DepartmentofChemistryandBiochemistry,FacultyofSciences,UniversityofPorto,RuadoCampoAlegre687,4169-007Porto,Portugal;5DepartmentofChemicalandProcessEngineering,Universityof

Strathclyde,75MontroseStreet,GlasgowG11XJ,UnitedKingdom

Nanoporous silica materials are produced by templated synthesis, whereby a surfactantliquid-crystalactsastemplate,andthusallowcontroloverporesizebychangingsynthesisconditions [1]. They are thus excellent candidates for computational design,whichwouldhave tremendous benefits in areas like gas separation, catalysis, and drug delivery.Accomplishingthisgoal,however,requiresmodelsthatcanaccuratelypredicthowsilicateprecursor and surfactant templates interact in aqueous solutions, giving rise to orderedmesostructures. This is a very challenging task, given thewide range of time and lengthscales thatarerelevant to theself-assemblyof thesestructures. In thisstudy,wepresentresultsofanewmulti-scalesimulationapproachtounderstandporoussilicasynthesisatthemolecularandmesoscale levels,basedonconstructingachainofmodels that range fromthe quantum-mechanical level, through classical atomistic simulations, up to amesoscalecoarse-grained(CG)descriptionofthesystem[2].OurCGmodelwasvalidatedbyaccuratelyreproducing the phase diagram and average micelle aggregation number of cationicsurfactantsolutions,inremarkableagreementwithexperimentalmeasurements.Using this newlydevelopedCGmodel,wehaveperformedextensivemolecular dynamics(MD) simulations in themicrosecond time scale to probe in detail the phase diagram ofsilica/surfactant aqueous solutions, changing several key synthesis conditions such astemperature,pHanddegreeof silicapolymerisation [3,4].Our resultsprovideanswers toseveralhithertocontentiouspointsregardingthesynthesisofnanoporoussilicamaterials:i)silicate monomers replace bromide counterions and strongly adsorb on the surface ofsurfactant micelles; ii) monomeric silica adsorption leads to micelle size increase andeventually to a sphere-to-rod transition; iii) silicate oligomers are necessary to promoteformation of hexagonal liquid crystals, by forming “charge bridges” linking differentsurfactant micelles; iv) nanoporous silicas form through a co-operative templatingmechanism controlled by charge-matching interactions between silicates and surfactants;v)thepostulatedliquid-crystaltemplatingmechanism,wheresilicatesaremerespectators,is not viable for this class of materials under typical synthesis conditions. Ourmodellingapproach thus provides unprecedented insight into how thesematerials actually form insolution,andpavesthewayforcomputationalmaterialdesignefforts.[1] Beck,J.S.;Vartuli,J.C.;Roth,W.J.;Leonowicz,M.E.;Kresge,C.T.;Schmitt,K.D.;Chu,C.T.W.;Olson,D.H.;Sheppard,E.W.;McCullen,S.B.;Higgins,J.B.;Schlenker,J.L.,J.Am.Chem.Soc.,114,10834(1992).[2] Pérez-Sánchez,G.;Gomes,J.R.B.;Jorge,M.,Langmuir,29,2387(2013).[3] Pérez-Sánchez,G.;Chien,S.-C.;Gomes,J.R.B.;Cordeiro,M.N.D.S.;Auerbach,S.M.;Monson,P.A.;Jorge,M.,Chem.Mater.28,2715(2016).[4] Chien, S.-C.; Pérez-Sánchez, G.; Gomes, J. R. B.; Cordeiro,M. N. D. S.; Jorge,M.; Auerbach, S.M.;Monson,P.A.,J.Phys.Chem.C,121,4564(2017).

MonteCarloforUndergraduateswithDL_MONTE

ChrisKing,JamesGrant,SteveParker,JohnPurton,NigelWilding

ChemistryDepartment,UniversityofBath,UK

ACCP5fundedbursaryhasdevelopedteachingmaterial forundergraduatecomputationalcourses. Wewillpresentanoverviewof the topics covered, including theexample casesandassociated teachingmaterial. These tutorialsdescribe theuseofDL_MONTEcode inChemistry and Physics computational labs to provide clear and instructive problems toillustrateMonteCarlomethodsandtheirroleinmolecularsimulation.

Light-induceddeformationofazobenzene-containingcolloidalspheres

MarkusKoch1,2,*,OlgaGuskova,1,3NinoLomadze,4SarahLoebner,4SvetlanaSanter4

(1)Leibniz-InstituteofPolymerResearchDresden,01069Dresden,(2)InstituteofTheoreticalPhysics,TechnischeUniversitätDresden,01069Dresden,(3)DresdenCenterforComputationalMaterialsScience,TechnischeUniversitätDresden,01062Dresden,(4)InstituteofPhysicsandAstronomy,Universityof

Potsdam,14476Potsdam,Germany

Materials containingazobenzene (azo)havebeenattracting large scientific interest in thepastyearsduetotheirphotophysicalproperties.Thekey-featureofazoisafullyreversiblephotoisomerizationreactionwhichconverts the trans into thecis stateandback. Ifazo isembedded inpolymeric systems, this leads toconformationalchangesofmacromoleculesand induces stresses which yield macroscopic deformations in such multi-componentsystems[1].Therefore,azobenzeneshavepromisingprospectsinfunctionalmaterials.Here, we present the results of a joint theoretical-experimental study of how the light-triggered isomerization of azo on molecular scale causes observable deformations of apolymersample.Forthis,weinvestigatephotoresponsivecolloidalspheresofpoly[(methylmethacrylate)-co-(Disperse Red 1 acrylate)], which have been prepared throughhydrophobicaggregationofthepolymerchainsintetrahydrofuran/watermixture[2].InAFMexperiments it is foundthatunder illuminationwithpolarized light thesecolloidalspheresstronglyelongateinthedirectionoftheE-fieldvectoroftheirradiationwithatotaldeformationofupto3-foldofitsinitialdiameter.

# Figure1:AFMmeasurementsofthesameareabeforea),c)andafterb)d)irradiationwith

singlebeampolarizedlightof4minuteswithwavelengthof491nm,intensityof100mW/cm².

Forthemodelingpart,weuseall-atomisticmoleculardynamics(MD)simulationstoimitatetheexperimental samples in closechemicaldetail.Wemodelpoly[(methylmethacrylate)-co-(DisperseRed1acrylate)]chainwithinPCFF(PolymerConsistentForceField)[3]andthesimulationpackageLammps[4].Toimplementthebehaviorofazounderlightillumination,weapplyaneffectiveorientationpotentialonthetransisomersasintroducedinRef.5.Inthismanner,wemodeltheeffectthattheazogroupswillorientperpendicularlytotheE-field vector of the incoming light. This orientation behavior is thought to be the drivingmechanismthatgeneratesdeformationsinazo-polymermaterials.Wegratefullyacknowledgesupport fromtheGermanResearchFoundation(DFG), theprojectsGU1510/3-1andSA1657/13-1.[1]Yadavalli,N.S.,Saphiannikova,M.andSanter,S.,Appl.Phys.Lett.105,051601(2014).[2]Li,Y.,He,Y.,Tong,X.,andWang,X.,Langmuir22(5),2288-2291(2006).[3]Hwang,M.J.,Stockfisch,T.P.,andHagler,A.T.,J.Am.Chem.Soc.,116(6),2515–2525(1994).[4]Plimpton,S.,J.Comput.Phys.,1(117),1–19(1995).[5]Toshchevikov,V.,Saphiannikova,M.,andHeinrich,G.,J.Phys.Chem.B,113(15),5032-5045(2009).

d) b) c) a)

Uncoveringthemechanismsofislandformationinonedimensionfromthegapandcapturezonedistributions

HrvojkaKrcelic

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,UK

Understanding of the mechanisms of island nucleation and growth during thin filmdeposition from the observed statistical properties of the island patterns is an importantproblem. To be able to predict and control the evolution of nanostructures, we need asystematic way to deduce the critical island size for nucleation, the laws for nucleation,whethertheislandsaremobile,etc.OneestablishedapproachistoperformkineticMonteCarlosimulationswithpre-definedrules,andcomparetheresultingislandpatternstoselectthesetofrulesthatgivethebestfitforthedata.However,thisisnotentirelysatisfactory.Inthisworkweexploreanotherpossibility,thatistosolvetheinverseproblemusingnumericaltechniquesdesignedforthistask.We focus on one-dimensional substrates, which mimics the behaviour of monomersdiffusingalongastepedge,andlimitourstudytothescale–invariantaggregationregime.Here, the nucleation process translates to the fragmentation of a line into inter – islandgaps.Uponnucleationwedescribenewlycreatedgapsbytrackingthefragmentationoftheold, larger one. This novel, retrospective approach gives an analytical description of thenucleation process through Distributional Fixed Point Equations which leads to integralequationsforgapandcapturezonedistributions.TheseequationsbelongtothebroadclassofFredholmintegralequationsofthefirstkind,andtheyincorporateinformationaboutthecritical island size and the nucleation mechanism (diffusion or impingement drivennucleation). This offers a possibility of analysing statistical data about the nucleatedstructures,eitherexperimentalorobtainedfromasimulation.WegeneratethisdatainkineticMonteCarlosimulations,inwhichmonomersarerandomlydepositedonto a substratewhere theydiffuse andundergo reversible attachment to thesurfaceand/orothermonomers,resultinginislandformation.Islandsgroweitherthroughthe capture ofmonomers that diffuse in their corresponding capture zones or by havingmonomersdepositeddirectlyonthem(impingement).Weuseapointislandmodelwhereislandsoccupyasinglespaceonthelatticeandrecordtheirsizethroughoutthesimulation,andwekeepthemimmobile.Toexplorethemechanismsdrivingtheprocessesofislandformationwesolvetheinvertedmodel integral equation for theobtainedMonteCarlodata. This leadsus toa caseof aninverse problem that is ill-posed and we use regularisation techniques to solve it. Theobtainedsolutionsreflectdifferentcriticalislandsizesandfitfairlywellwithintheexpectedbounds,howevertheyalsocontainahighlevelofuncertaintyduetotheill-posednessoftheproblemandthenoisepresentintheinputdata,sofurtherrefinementoftheregularisationiscurrentlyunderway.

MolecularDynamicsSimulationsofPolymersatSurfaces

DavidMcKechnieandKarenJohnston

ChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,Glasgow,G11XJ

Polymer carbon fibre composites have a wide range of industrial uses due to their highstrength-to-weight ratio and electrical conductivity. The properties of the polymercompositedependonnotjusttheindividualpropertiesofthepolymerandincludedcarbonfibres,butalsoontheinteractionsbetweenthemandtheirimpactonthefinalstructure.Thisprojectusesmoleculardynamicssimulationstoinvestigatetheimpactsofasurfaceonpolymerproperties.Toprovideinsightintocarbonfibrecomposites,weinvestigateamodelsystem of polyethylene on a graphene surface. Three systems were simulated: i) bulkpolyethylene, ii)apolyethylene filmwithgrapheneononesideandvacuumon theotherand iii) polyethylene sandwiched between two graphene surfaces. The effect of thegraphene surfaces on structural properties, including radius of gyration (Rg), end-to-enddistance (Re) and conformation tensors were investigated. The polymers were found toextendlaterallynearthesurfaceleadingtoanincreaseinbothRgandRewhilereturningtotheirmeltvalues inabulk-likeregionwithin3Rgfromthesurface.ThisshowsqualitativeagreementtopreviousworkbyDaoulasetal.1andDeVirgilisetal.2onpolymermelt/solidinterfaces.Theglasstransitiontemperatureandhowitisaffectedbythesurfaceswillalsobepresented.[1] Daoulas, Kostas Ch, Vagelis A. Harmandaris, and Vlasis G.Marantzas. “Detailed atomistic simulation ofpolymer melt/solid interface: structure, density, and conformation of a thin film of polyethylene meltadsorbedongraphite.”Macromolecules38.13(2005):5780-5795.[2]DeVirgilis,A., et al. “Structureanddynamicsof apolymermelt at anattractive surface.” TheEuropeanPhysicalJournalE:SoftMatterandBiologicalPhysics35.9(2012):1-11.

DisjoiningPressureObtainedbyMolecularDynamicsSimulationsandItsApplicationtoMultiscaleModellingofDropletCoalescence

AleksandarY.Mehandzhiyski,BrianA.Grimes

DepartmentofChemicalEngineering,NorwegianUniversityofScienceandtechnology,SemSælandsvei4,NO-7491Trondheim,Norway

Coalescence of water droplets dispersed in a crude oil is an important process in the chemicalengineeringandpetroleumindustry.Despitebeingextensivelystudied,theprocessofcoalescenceisstillnotcompletelyunderstood.Particularly, the intermolecular interactionsbetween thecomplexmolecular structures of adsorbed crude oil components (asphaltenes, resins, etc) have a strongeffect on the disjoining pressure and can be a significant factor that controls the stability of theemulsionandmayhinderthedrop-dropcoalescence.However,whendevelopingprocessmodelsforcrude oil systems, only the bulk properties of the crude oil, such as density, viscosity etc. areaccountedinmostmodelsandthespecificchemicalfootprint(e.g.,asphaltene/resincontent)ofthecrude oil is often neglected. On the other hand, numerous laboratory experiments have beenperformed to study the interfacial phenomena for crude oil-water systems since differentinterfacially active compounds in the crude oil affect the behavior in these complex systems.Therefore, with the current work we aim to close the apparent gap between crude oil chemicalcharacterizationandprocessmodellingofcrudeoil-watersystems.Themainobjectiveistoobtainanewsemi-empiricalexpressionforthecoalescencetimethattakesintoaccountageneraldescriptionofthechemicalfootprintofthecrudeoil.Thefirststepinthistaskistosimulatethedisjoiningpressureatliquid-liquidinterfacesforvariousinterfacialconcentrationsofmodelasphaltenemoleculeswithmoleculardynamics(MD).TheMDsimulationspresentedherecanbeusedtoestimatethedisjoiningpressureinathinliquidfilmbetweentwodropsasafunctionof the thin film thickness for different interfacial concentrations of surface active compoundsindigenoustocrudeoil.Thedisjoiningpressureisoftenproblematictobeobtainedfrommolecularsimulations for liquid-liquid systems. Thus, we investigate different methodologies in order tocalculate the disjoining pressure fromMD simulations for various film thicknesses and interfacialconcentrations of asphaltenemolecules at an oil-water interface. Based on theMD simulations asemi-empiricalrelationofthedisjoiningpressureasafunctionofthe interfacialconcentrationandfilmthicknesswillbeobtained.Theresultingfunctionalbehaviorsforthedisjoiningpressurewillbesubsequentlyemployed incontinuumfilmdrainagemodels tosolve for thecoalescencetimewiththe interfacial concentration ofadsorbed molecules as aparameter.

Figure1Differentscalesaccountedinthedropletcoalescencemodelinwater-in-oilemulsions.

AdsorptionofContaminantsonClayMineralSurfaces

L.Ohene-Yeboah1,J.Grant2,J.Skelton2,S.C.Parker2

(1)CentreforDoctoralTraining,CentreforSustainableChemicalTechnologies,UniversityofBath,BA27AYUK,(2)DepartmentofChemistry,UniversityofBath,BA27AYUK

The accumulation and persistence of hazardous compounds (HCs) in surface and groundwater as well as living organisms have emerged as an adverse effect of humananthropogenic behaviour. A variety of HCs, from emerging contaminants found inpharmaceutical residues and personal care products to household chemicals,biocides/pesticidesandmanufacturingwastespersist in theenvironment. Someof thesecompounds have been shown to cause adverse effects in aquatic organism as well aspromoteincreasedriskofdevelopingthyroiddisorders,tumoursanddiabetes.Inthisstudy,weaimtobuildonthefoundationsofourunderstandingofhowHCsinteractwith the environment by applying atomistic simulation methods to determine thephysicochemical factors controlling the distribution of pollutants and theirmetabolites inaqueousandterrestrialenvironments,andthenapplythistoidentifyingsustainablewaysofcontroltheirtransport.Claymineralsareubiquitousintheenvironment,sotheyprovideagoodstartforourstudy.Additionally,ithasbeensuggestedthatclaymineralscouldactasgeosorbants for theremediationoforganicpollutant, soanevaluationof theiradsorptivepropertieswouldbeuseful.Initial studies centred on HCs adsorption on two model clay surfaces; sodiummontmorillonite and pyrophyllite. Simulations were performed using a combination ofdispersion corrected DFT and classicalMDmethods, to calculate the binding free energyand identify favourable sites forHCadsorptiononclays invacuumandwater. Our initialinvestigation focusedon10molecules includingdibenzo-dioxinsand twopoly-chlorinatedderivatives, chloro- and hexachloro-benzene, emerging contaminants amphetamine andtwootherderivativesandMDAanditsanalogueMDMA.Thesecompoundswerechosentodetectthe linksbetweenthechemicalpropertiesoftheseHCsandtheirpossiblefateandeffectsintheenvironment.

Multiscaleapproachtodesigningnovelcolloidaldrugdeliveryvehicles

DemiPink1,M.JayneLawrence2&ChristianD.Lorenz1

(1)DepartmentofPhysics,King’sCollegeLondon,LondonWC2R2LS,UK,(2)InstituteofPharmaceuticalSciences,King’sCollegeLondon,LondonSE19NH

Alargenumberofthesmallmoleculescurrentlyunderdevelopmentasdrugcandidatesarehydrophobic,andasaresultmanyofthesedrugsmaynotmakeittomarket.Therefore,itisessential to incorporate suchdrugs into formulationswhichprotect the drug in thebodyand release it at the target tissue.Phosphaditylcholines are amphiphilic lipids that self-assemble to formmicelles. This, alongwith their biocompatibility,makes them attractivecompoundsto investigatewithregardstotheiruse inthedeliveryofpoorlysolubledrugswhose solubility can be increased by encapsulating within the hydrophobic core of themicelle. The resulting micelles are then degraded by the pancreatic enzyme PLA2whichreleasesthedrug.The length of the phosphaditylcholine carbon tail can greatly impact the properties andstructureoftheresultingmicelles,this isparticularlyevident inshort-tailedlipidsaswhenmoving froma6 carbon tail (diC6PC) toa7 carbon tail (diC7PC) the shapeof themicellechangesfromsphericaltorod-like.AtomisticmoleculardynamicsimulationshavebeenrunonmultiplesystemswithdiC6PCanddiC7PClipidstodeterminehowtheminordifferenceinlipidtaillengthcanimpactthearchitectureofthemicelleandtheresultingeffectthismayhaveonhydrophobicdrugencapsulation.MD simulations have also been used to study systems which contain both the parentphosphaditylcholinemoleculesandtheirproductsafterhydrolysisbyPLA2.Thiswasdonetodeterminetheimpactofdegradationonthenatureofthephospholipidaggregatewhichisanessential factorwhenusingphospholipiddrugdeliveryvehicles.Thissimulationwork iscombined with experimental work, including SANS and SAXS data, in order to provide acomprehensive andmultiscale picture of how the differentmicelle shapes are generatedandhowdegradationaltersthenatureofthesemicelle.

Modellingthegelationofporousnanomaterials

MartinProstredny,PaulMulheran,AshleighFletcher

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,G11XJ,Glasgow,UK

Resorcinol-formaldehyde (RF) gels have been studied extensively, subsequently findingmanyuses.However, despite significant researchefforts, there is still noacceptedmodeldescribing the formation of these structures. This work combines experiments andsimulations in order to better understand the process of creating and growing clustersduring RF gel formation. The investigated areas include the influence of gelationtemperature, and the concentration, and type, of catalyst used. The catalyst studied issodiumcarbonateand,byusingothersourcesofsodiumions,wehaveexploredtheeffectofsodiumions intheseprocesses. Inordertoobtaintexturalpropertiesofthedriedgels,nitrogenadsorptionwasanalysedusingBrunauer-Emmett-Tellertheory[1],forsurfaceareaanalysis, and Barrett-Joyner-Halenda theory [2], for pore size and volume analysis. Theacquireddatasuggestthat,withincreasingresorcinol/catalyst(R/C)ratio,thesurfaceareadecreases while the average pore width increases. This is most likely caused by theformationofalargernumberofsmallerclustersforthelowerR/Cvalues,andfewerlargerclusters for higher R/C ratios. The experimental data form the basis of a model for thegrowthandgelationofRFclusterswithwhichweintendtodeveloppredictivecapabilitiesfor future work. Currently, themodel is in development and uses a kineticMonte Carlomethodintwo-dimensions;extensiontothreedimensionswillfollow.

Comparisonoftheclustergrowthmodel(RHS)withanSEMimageofRFgelclusters(LHS) [1]S.Brunauer,P.H.Emmett,andE.Teller,JournaloftheAmericanChemicalSociety,1938(60)309-319.[2]E.P.Barrett,L.G.Joyner,andP.P.Halenda,JournaloftheAmericanChemicalSociety,1951(73)373-380.

Competitiveadsorptionforclearairapplications

PaulRapp

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,GlasgowG11XJ,UnitedKingdom

Environmental andhealth concerns frompollutionare significant social economicdrivers,whichpushlegislationtowardspollutionpreventionandsustainability.Inadditiontohealthconcerns, the emission of the greenhouse gas carbon dioxide must be reduced to keepclimate change below 2C. Therefore, it is essential to develop and test environmentallyfriendlymaterialsthatarehighlyoptimizedtoremovespecificpollutantspecies.Activatedcarbon isawell-knownaffordablecarbondioxideandpollutantadsorber.However, therearemanytypesofactivatedcarbonsandtheseadsorbsomespeciesmoreeffectivelythanothers. To tailor activated carbons for different pollutant species, it is necessary tounderstand or control their chemical and structural composition. This project aims toidentifywhichcharacteristicsofactivatedcarbonareoptimalforremovalofpollutants,suchascarbondioxide.Asastartingpointwetakegraphiteasamodelsystemandstudytheadsorptionofcarbondioxide using a combined computer simulation and experimental approach. Quantumcalculations found that carbon dioxide adsorbs weakly on the graphite surface viaphysisorption (van der Waals interactions) with almost negligible preference for specificadsorption sites. Grand canonical Monte Carlo isotherms, with force fields based onquantum adsorption datawere performed for graphite slit pores and amorphous carbonstructures.Resultsrevealedmuchhigheradsorptionformicroporousstructurescomparedtomesoporousslitporesatlowpressure.Experimentalresultsfoundgraphiteinpowderedform has a significantly lower isosteric heat of adsorption compared to theoreticalisotherms, which is attributed to a broader pore size distribution and larger pore sizes.Future work is aimed towards a) competitive adsorption of several pollutant species ongraphiteandb)adsorptiononmodifiedgraphitetodeterminethemosteffectiveactivatedcarbons.

Designofmetal-organicframeworksforsensingsmallorganicmoleculeslikeethanolamine

MalteSchäfer,AndreasM.Schneider,PeterBehrensInstituteforInorganicChemistry,LeibnizUniversityHannover,Germany

Sensingdevicesare importantcomponents ineveryday life.Theappliedsensingprinciplesvary from rather simple (e.g. detection inmass change) to quite sophisticatedones as inchemicalprocesses.Therefore,thedevelopmentofsensingmaterialsandtechniquesisanimportant.Forexample,thedetectionofethanolamineisinthefocusofcurrentactivities,sinceitsreleasefromthehumanbodycanindicatecertaindiseases.[1]Here,wepresentthedevelopment of metal-organic frameworks (MOFs) as selective sensing materials for thedetectionofethanolamine.Metal-organicframeworksaremicroporoushybridmaterialswithoutstandingandtuneableproperties leading toavarietyofpotentialapplications.Theyconsistof inorganicbuildingunits (IBUs)andorganic linkermolecules,whichconnectthese IBUstoformaframework.MOFsareusedforgasstorage,gasseparation,incatalysisaswellasinsensingmaterials.[2]In this work a zirconium-based MOF (UiO-66, Figure 1) is investigated. Due to its highthermal and chemical stability, Zr-basedMOFs in general are suitable sensing materials.UiO-66crystalizesinthefcutopology,exhibitingacontinuousporesystemconsistingoftwodifferentporetypes.[2]Theyareformedbyatetrahedralandanoctahedralarrangementofthe surrounding inorganic building units, respectively. Within these pores the candidatemolecules,e.g.ethanolamine,canbeadsorbed.InordertodesignaMOFmaterialshowingspecific interactionswithethanolamine, the linkermoleculesaremodifiedby introductionof various substituents (Figure 2). In order tominimize the experimental effort, possiblecandidatestructureswereinvestigatedbymolecularmodellingmethods.Fortheexamination,asuitableforcefieldwasderivedwhichallowsnotonlyforthecorrectdescriptionofZr-basedMOFs,butofchargeinteractionsandhydrogenbonds;especiallythelatterareimportantasethanolaminecanformsuch(Figure2).

ConstructionprincipleofUiO-66.Organiclinkers(lowerleft)connectinorganicbuildingunits(upperleft).Aporewithoctahedralshapeisindicatedbyayellowsphereandonewitha

tetrahedralshapebyagreensphere.

Once a suitable force field was identified, Molecular Mechanics and Dynamics wereemployed to find the minimum energy structure of different candidate structures. Theresultinginteractionenergies,thepreferredadsorptionsitesaswellastheconformationoftheethanolaminemoleculeareusedtofindaMOFpredictedtopossesshighsensitivityandselectivityfortheadsorptionofethanolamine.

Ballandstickmodelofethanolamine(left)andpossiblypositionsforsubstitutionsatthe

linkermoleculesoftheUiO-66.

[1]A.A.Farooquietal.,Neurochem.lResearch,1997,22,523-527.[2]A.U.Czaja,N.Trukhan,U.Müller,Chem.Soc.Rev.,2009,38,1284-1293.[3]J.H.Cavkaetal.,J.Am.Chem.Soc.,2008,130,13850-13851.

Strain-inducedtuningofsurfaceenergyandreductiondriveinspinelLiMn2O4cathodes

IvanScivetti1andGilbertoTeobaldi2

(1)TheStephensonInstituteofRenewableEnergy,UniversityofLiverpool,Liverpool,UK,(2)DaresburyLaboratory,Sc.Tech.,Warrington,UK

LiMn2O4(LMO)implementationinrechargeableLi-ionbatteries(LIBs)forstationarystorageishamperedbythelimitedlifetimeofthematerialanditsinterfaces,startingfromtheSolidElectrolyte Interphase [1,2]. Recent experiments [2] and Density Functional Theory (DFT)simulations[3]indicatethattheformationandeffectivenessoftheSEIonLMOarerelatedto the surface orientation and reduction drive. In this context, we analyse the role ofgeometrical strain for the relative energy, magnetic ordering and the reduction drive ofseveral LMO surfaces. DFT simulations reveal LMO surfaces to be markedly sensitive togeometrical strain. Strain lower than 10% can induce insulator-metal and ferromagnetic-antiferromagnetictransitions,altertherelativeenergyofLMOsurfaces,andinducechangesas largeas1.0eV in the surfacechemicalpotential, thence reductiondrive.Promptedbyadvances inthesynthesisofmetal-oxidecore-shellnanostructures[4], theuseofstrainedLMOcoatingasSEI-formationagentisputforwardtowardsengineeringoflongerlivedSEIonLMOsubstrates. 1.K.Leung.JCPC2012,116,9852-9861.2.M.Hirayamaet.al.JACS2010,132,15268−15276.3.I.Scivettiet.a.J.Phys.Chem.C2015,119,21358−21368.4.X.Xiaet.al.ACSNano2012,6,5531.

TheEffectofAtomicPointChargesonAdsorptionIsothermsofMetalOrganicFrameworks

KristinaSladekova,ChristopherCampbell,IzabelaCebula,MiguelJorge

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,75MontroseStreet,GlasgowG11XJ,UnitedKingdom

Metal-organic frameworks (MOFs) are crystalline and porous materials consisting ofcoordinationbondsbetweentransition-metalcationsandorganic ligands. [1]The intrinsicability of MOFs to adsorb molecular fluids is one of their most studied and promisingproperties.MOFsexhibitmanycharacteristicphysicalattributesincludinghighporevolume,large surface area and well-defined crystalline structure which make them suitable foradsorption-based applications such as gas storage, purification, sensing devices and drugdelivery.[2]ApromisingavenueforscreeningandpotentiallydesigningMOFsforparticularapplications istousecomputersimulationstounderstandthe interactionsbetweenMOFsand adsorbates and to predict their adsorption performance. In this context, Grand-Canonical Monte Carlo (GCMC) enables one to compare simulated adsorption isothermswithexperimentaldatawhileprovidingadegreeofmolecularleveldetailthatisdifficulttoobtain in experiments. [3] The success of this approach relies on the development ofmolecularmodelsthatareabletoaccuratelypredictadsorptioninMOFs.Inparticular,theassignment of atomic point charges to each atom of the framework is essential formodellingCoulombicinteractionsbetweentheMOFandpolaradsorbatemoleculessuchasCO2,COandwater.MultiplemethodshavebeendevelopedtoassignatomicpointchargesinMOFs.[4]Inthiswork,weperformadetailedandsystematicinvestigationoftheeffectofdifferent point charges on the adsorption isotherm predictions in various frameworksrepresentative of differentMOF families (including CuBTC, IRMOF-1, UIO-66,MIL-47 andMOF-177).We compared point charges obtained by differentmethods including ChelpG,QEQ, Mulliken, DDEC and REPEAT, among others, and based on quantum mechanicalcalculations at different levels of theory and with different approaches (e.g., periodic vscluster).GCMCsimulationsusingthesedifferentchargesetsallowedustoclarifytheeffectof varying atomic point charges on the behaviour of adsorption isotherms and to betterunderstand the impact of electrostatic interactions on adsorption prediction quality. Ourresults show that the impactofpoint chargesvaries significantlyaccording to the typeofMOFstructure.Wealsoobservethatsomemethodsproduceisothermsthatareconsistentwitheachother,whileothermethodsleadtoisothermsthatdeviatesignificantlyfromtherest.Basedonthisstudy,weproposesomeguidelinestoimproveexistingmodelsinordertoincreasetheaccuracyofgasadsorptionpredictioninMOFs.[1] Shekhah,Osama.Layer-By-LayerMethodForTheSynthesisAndGrowthOfSurfaceMountedMetal-OrganicFrameworks(Surmofs).Materials3.2(2010).[2] Coudert, François-Xavier, and Alain H. Fuchs. Computational Characterization And Prediction OfMetal–OrganicFrameworkProperties.CoordinationChemistryReviews307(2016).[3] Chen, Linjiang, Carole A. Morrison, and Tina Düren. Improving Predictions Of Gas Adsorption InMetal–Organic Frameworks With Coordinatively Unsaturated Metal Sites: Model Potentials, Ab InitioParameterization,AndGCMCSimulations.TheJournalofPhysicalChemistryC116.35(2012):18899-18909[4] Nazarian, Dalar, Jeffrey S. Camp, and David S. Sholl. A Comprehensive Set Of High-Quality PointChargesForSimulationsOfMetal–OrganicFrameworks.N.p.,(2017).

Assessingcoarse-grainedmodelsfortheself-assemblyofblockcopolymers

OliviaSobek,MiguelJorge

DepartmentofChemicalandProcessEngineering,UniversityofStrathclyde,Glasgow

Block copolymers find applications in many fields, including adhesives, plastics, drugdelivery and photonics. Several of these rely on the ability of block copolymers to self-assembleintoorderedmesophasesinsolution.Onesuchapplicationofparticularinteresttoourresearchgroupistheiruseastemplatesinthesynthesisofporoussilicamaterials,suchasSBA-15[1].Becauseoftheirhighlyorderedpores,highsurfaceareas,highfunctionalityand low cost,mesoporous silicas have been of great interest for an increasing variety ofapplications and research. Understanding the synthesis mechanism for this class ofmaterials,however,modelsthatcanpredicthowblockcopolymertemplatesself-assembleinaqueoussolution.Thisstudyaimstoproduceanaccuratecoarse-grained(CG)modelofself-assembling block copolymers, including those used in the synthesis of SBA-15mesoporoussilica(i.e.,Pluronicsurfactants).Suchamodelwillenableustoprobethelargetime and length scales that are needed to describe the mesostructure formation fromsolution,thusclarifyingthemechanismsbywhichthesematerialsareformed.Our approach is based on the established Martini CG force field [2],which has beenpreviously applied tomodel these systems [3,4].Wehave found thatexistingmodels areunable to accurately describe micelle aggregation self-assembly of Pluronic surfactants,althoughtheyaredesignedtomatchsingle-chainproperties.Wehavethussystematicallytested the existing MARTINI parameters against experimental Gibbs free energies ofsolvationinbothwaterandhexadecane,leadingtoimprovedmappingschemesforpolymermolecules. In some cases, parameters needed to be adjusted to accurately describe thesolvation free energies in both solvents. In the future, we will test these improvedparameters against the phase diagram of block copolymer surfactants, in order toeffectivelyreplicatethemicelleaggregationandformationprocesses.

[1]J.Thielemann,F.Girgsdies,R.Schlgl,andC.Hess.Porestructureandsurfaceareaofsilicasba-15:influenceofwashingandscale-up.BeilsteinJournalofNanotechnol,11:110–118,2011[2]S.MarrinkandP.Tieleman.Perspectiveonthemartinimodel.Chem.Soc.Rev.,42:6801–6822,2013[3]P.CarboneandS.Nawaz.Coarse-grainingpoly(ethyleneoxide)/poly(propyleneoxide)/poly(ethyleneoxide)(peo/ppo/peo)blockcopolymersusingthemartiniforcefield.J.Phys.Chem.B,118(6):1648-1659,2014[4]G.Rossi,P.F.J.Fuchs,J.Barnoud,andL.Monticelli.Acoarse-grainedmartinimodelofpolyethyleneglycolandofpolyoxyethylenealkylethersurfactants.J.Phys.Chem.B,116(49):14353–14362,2012

Predictivemodellingviaelectroniccoarse-graining:thecuriouscaseofthewaterinterface

VladSokhan

ScientificComputingDepartment,STFC,UK

Most of the classical force fields include many-body interactions only in the mean-fieldapproximation,througheffectivechargesandcondensedphasefittedvanderWaalsterms.As the result, these force fields are parameterised for certain thermodynamic conditionsandhavelowpredictivepoweroutsidethetrainingsets.Polarisableforcefieldsgobeyondpairapproximationbyincludingmany-bodypolarisation,althoughalmostinvariablykeepingdispersionatthepairlevelsincemany-bodydispersionisnon-classicalinnature.HereIwilldescribe a novel approach based on coarse-graining electronic degrees of freedom thatallowsmodellingmany-bodydispersionandpolarisationonequalfootingbyacoupledsetof quantumharmonic oscillators (quantumDrudemodel). Solvednonperturbatively usingpath integral molecular dynamics this approach provides an attractive basis to studynonhomogeneous systems since it contains sufficiently rich physics to capture complexphenomenaofsoftcondensedmatter.Iwillpresentourrecentresultsofhydrogenbondingandmolecularorientationstudyatthefreewaterinterface. 1.V.P.Sokhan,A.P.Jones,F.S.Cipcigan,J.Crain,andG.J.Martyna.PNAS112,6341(2015).2.A.P.Jones,J.Crain,V.P.Sokhan,T.W.Whitfield,andG.J.Martyna.Phys.Rev.B87,144103(2013).3.F.S.Cipcigan,V.P.Sokhan,A.P.Jones,J.Crain,andG.J.Martyna.PCCP17,8660(2015).

GrowthandreproductionofgiantSALRclusters,withimplicationsfortheirchemicalevolution

M.B.Sweatman

SchoolofEngineering,UniversityofEdinburgh,King’sBuildings,MayfieldRoad,Edinburgh,Scotland,UK.EH93FB

Particles with SALR (short-range attraction and long-range repulsion) interactions arecommontomanyphysicalsystems,especiallybiological,colloidalandsoftmatter,yettheirbehaviourisstillnotcompletelyunderstood.UsingMonteCarlosimulations,itisshown[1]thatgiantclusterscangrowandreproduceinthesefluids.Growthandreproductionshouldtherefore be common to a wide range of natural and synthetic systems under suitableconditions.If,inaddition,clusterfitnessselectioncanoccurthenchemicalevolutionofgiantclustersinSALRfluidsmightbeobservedinsuitablesystems.

SnapshotsfromMonteCarlosimulationsillustratinggiantSALRclustergrowthandreproduction,wheretis

thenumberofMonteCarlocyclesperformed. [1]M.B.Sweatman,submitted

CrossingBoundaries–TheEffectofInterfacesonthePropertiesofFluoriteOxides

AdamSymington(1),ChrisKing(1)NicholasWilliams(3)JamesGrant(1),MarcoMolinari(2),StephenCParker(1)

(1)UniversityofBath,ClavertonDown,Bath,BA27AY,(2)UniversityofHuddersfieldQueensgate,Huddersfield,HD13DH,(3)AWEAldermaston,Reading,Berkshire,RG74PRUnitedKingdom

Fluorite structuredoxides have a variety of applications, ranging fromnuclear fuel in thecaseofUraniumDioxide(UO2)tocatalysis,fuelcellsandbio-medicineinthecaseofCeriumDioxide (CeO2). In both examples theirmaterial properties depend on defect processesoccurringattheirinterfaces.In thisprojectweuseacombinationofDFTandpotentialbasedsimulationtechniquestomodelinterfaceprocessesoftwoimportantfluoriteoxides–UO2andCeO2.Interfacesintheformofgrainboundariesandsurfacesareimportantstructuraldefectsthatinfluence the properties of the materials. We have investigated the free energy of lowindex surfaces of CeO2 and UO2 as well as 6 tilt grain boundaries. Here we have usedDL_MONTEtoinvestigatethe{111},{110}and{100}surfacesduetotheirprevalenceintheliterature[1] [2] aswell as the grain boundaries previously studied inUO2byWilliams etal[3].We have also begun building on the work ofWilliams et al by considering the effect oftrivalentdopantsonoxygendiffusionatUO2grainboundaries.HerewehaveusedDL_POLYto determine the effect of a wide range of trivalent dopants across a range ofconcentrationsonthediffusionpropertiesofthese6grainboundaries.[1]M.Molinari,S.C.Parker,D.C.SayleandM.S.Islam,JournalofPhysicalChemistryC2012,116,7073-7082.[2]M.Nolan,S.Grigoleit,D.C.Sayle,S.C.ParkerandG.W.Watson,SurfaceScience2005,576,217-229.[3]N.R.Williams,M.Molinari,S.C.ParkerandM.T.Storr,JournalofNuclearMaterials2015,458,45-55.

AtomisticSimulationoftheinteractionoftheUranylIonwithFe(OH)2.APotentialOptionfortheLongTermStorageofNuclearWaste.

RogerD.ThorntonandDavidJ.Cooke

DepartmentofChemicalSciencesUniversityofHuddersfield

GreenrustisarelativelyrecentlydiscoveredfamilyofFe(II)Fe(III)layereddoublehydroxidewhichformsingroundwaterandduringsteelcorrosion.Themineralisunstableinairanditis this instability that has led to it being suggested as an efficientmeans of reducing thereactive actinyl ions (A(VI)O2)2+ ion into the more stable A(IV)O2 leaving themencapsulatedinamineralform.Onewayofexploringthepropertiesofsuchstructuresisviacomputer simulation at the atomistic level. In this projectwe have begun to explore thefundamentalpropertiesofgreenrust,howitsstructurechangeswithFe(III)concentrationand the content of the interlayers. Our ultimate goal is to describe the mineral’ssusceptibilitytotheadsorptionandabsorptionofactinylionsbutthiswillrequirearobustpotential model capable of modelling all the components of the system and it is thedevelopmentandtestingofpotentialmodelsthatisthecurrentfocusoftheproject.InthisposterwepresentpreliminaryresultsbasedonmodellingthemineralFe(OH)2andtheaqueousuranyl ionUO22+.Wealsopresent resultsofmodelling theUO22+ ionwithvariousanions,namelyFeO22-,SO42-,CO32-,(OH)22-,andCl22-andhowthesecomparetostructuresdetailed in the crystal database. The results from thisworkwill then feed intomodellingtheinteractionoftheuranylionwiththemorecomplexdoublehydroxide.

It’sJustaPhase:AComputationalInvestigationoftheRelativeStabilityofMagnesiumCarbonateMinerals

JoshuaTse1,JamesGrant1,MarcoMolinari2andStephenCParker1

(1)DepartmentofChemistry,UniversityofBath,(2)UniversityofHuddersfield,Queensgate,Huddersfield,HD13DH

WehaveappliedatomisticsimulationtomodelHydratedMagnesiumCarbonateminerals.ThechallengeistofindtheorderofstabilityofMgcontainingmineralsatdifferentH2OandCO2partialpressures.BruciteandHydromagnesiteareofinterestastheypotentiallytrapradionuclidesinnuclearwasterepositories.Computationalmodellingisasafeandeffectivetooltoassessthestructure,stabilityandinteractionofthesemineralswithradionuclides.DFT does not include long range dispersion forces by default. However, it is widelyrecognisedasbeingessentialforthecorrecttreatmentofintermolecularinteractions.Wecompared four commonly used approaches (GrimmeD3, optB88-vdW, optB86b-vdWanduncorrected DFT) which were run using the VASP code. Analysis has shown that theinclusionofvdWforces,improvesthestructuralandenergeticpropertiesoftheseminerals.The best representation was given by the optB86b technique, reproducing experimentalresults,withallvalueswithin±5%.We have shown that the functional optB86b correctly reproduces the experimentalstructuresandenthalpies. Using thismethod,wecalculate the relative stability,of theseminerals, as a function of H2O and CO2 chemical potential. This enables us to generatephase diagrams, which illustrate the most stable phase at different conditions. Atatmosphericconditions(350ppmCO2,32mbarH2Oat298K),weshowthatMagnesiteisthemost thermodynamically stable phase, of those considered. In addition, asmagnesite isslowtoformwehavealsopredictedthelikelymetastablephasesthatwillcrystallisefirst.

EffectsofChlorideTreatmentonGrainBoundariesinPolycrystallineCdTeThin-filmSolarCells

MichaelJ.Watts,PoojaM.Panchmatia,RogerSmith

LoughboroughUniversity,EpinalWay,Loughborough,Leicestershire,LE113TU,UK.

Polycrystallinethin-filmCdTephotovoltaiccellsshowgreatpromise,potentiallybeingbothmoreefficientandlowerincostthansiliconsolarcells.Keytotheirefficiencyistreatmentwithchlorinecompoundsbuttheprecisemechanismoftheperformanceincreaseremainsunknown. The primary candidate is a grain boundary based effect,where the chlorine isknown to strongly segregate upon treatment. Initial results using DFT and classicalpotential-based methods to study chlorine’s behaviour at observed grain boundarystructuresarepresented.AcknowledgementsThisworkrecognizestheuseofboththe‘Hydra’HighPerformanceSystematLoughboroughUniversityandtheARCHERUKNationalSupercomputingService.

UncertaintyQuantificationforClassicalEffectivePotentials

SarahWishart

SchoolofEngineering,UniversityofWarwick,UK

Effectivepotentialsareanessentialingredientintheimplementationofclassicalmoleculardynamics (MD)simulations. However, little is understoodof the errors incurredinrepresenting the complex energy landscape of an atomic configuration by aneffectivepotentialcontainingconsiderablyfewerparameters.Theprobabilisticsloppymodelmethodhas been implemented in thepotfitopen source force matching code. Thisadvancementintroducesuncertainty quantification (UQ) into the interatomic potentialgenerationprocess. Uncertainties in the effective potential are propagatedthrough MDtoobtainuncertaintiesinquantitiesofinterest,i.e.cohesiveenergies,thermaldiffusivityorelasticconstants.Totesttheimplementationtheelasticconstantsfornickelarecalculated,alongwiththecorrespondinguncertaintiesintheparametersofthemorsepotentialusedintheirgeneration.[1]S. L. Frederiksen, K. W.Jacobsen, K. S. Brown, and J. P. Sethna. Phys. Rev. Lett., 93(16), 2004.[2]P.BrommerandF.Gähler.Modell.Simul.Mater.Sci.Eng.,15(295-304),2007.

UnderstandingTheAdsorptionofHeavyMetalCationsatIron(oxyhydro)oxide-WaterInterfaces

W.Zhang1,JD.Baran1,SC.Parker1,M.Molinari2,RL.Zhu3,Q.Zhou3

(1)DepartmentofChemistry,UniversityofBath,ClavertonDown,BathBA27AY,UK,(2)DepartmentofChemistry,UniversityofHuddersfield,Huddersfield,HD13DH,U.K.,(3)KeyLaboratoryofMineralogyand

Metallogeny,GuangzhouInstituteofGeochemistry,CAS;Guangzhou510640,China

Heavy metals are introduced into the environment through a wide variety of humanactivities. Themobility of these heavymetals in sediments is of concern because of thepotential for bioaccumulation. In the environment, these contaminantsmay adsorb ontosurfaces of minerals, either become immobile or drift throughout the ecosystem. Pastresearchhasexperimentallydiscoveredthatsomeofthemorestableiron-bearingminerals(e.g.hematite,goethite)aresuitablesorbentsforheavymetalcationssuchascopper,leadand zinc1 2. Furthermore, co-adsorption of heavy metal cations and oxyanions (e.g.phosphateandsulphate)hasbeenfoundtosignificantly influencetheadsorptionprocess,alongsidewith the effects of pH3. However, themechanisms are unknown. Studying theadsorptionandco-adsorptionmechanismsmayprovide ideas fornewtreatmentmethodsofheavy-metal-containingwastewaters.Our current research has focused on hematite and goethite, two naturally abundant andstableiron-bearingmineralsassorbents,whereweaimtouseacombinationofelectronicstructureandpotentialbasedtechniques.Ihavebegunbymodellingthebulkstructureandelectronicpropertiesofhematiteandgoethite.Theresultsshowthatwecanmodelthebulkreliably andwe are nowutilising these techniques to simulate surfaces of bothminerals.Clearly, in the natural environment the surfaceswill be in contactwithwater and hencehavealsoinvestigatedtheinteractionofthesesurfaceswithwater.

Fig1.1anexampleofhydroxylatedoxygen-terminated(001)surfaceofhematite.

The results show, as observed previously, that hydroxylation of (001) of the oxygenterminated surface of hematite is very stable when water is present, while the ironterminatedsurfaceisthemoststableinultra-highvacuumconditions.Thecurrentplansareto investigate how the interaction of heavy metal cations with the sorbent surfaces ismodifiedbythesurfacestructure,waterandoxyanionconcentration.1. H. J. Shipley, K. E. Engates and V. A. Grover, Environmental Science and Pollution Research, 2013, 20,

1727-1736.2. M.Hua,S.J.Zhang,B.C.Pan,W.M.Zhang,L.LvandQ.X.Zhang,JournalofHazardousMaterials,2012,

211,317-331.3. J.Liu,R.L.Zhu,T.Y.Xu,Y.Xu,F.Ge,Y.F.Xi,J.X.ZhuandH.P.He,Chemosphere,2016,144,1148-1155.