How Nitrogen Doping A ff ects Hydrogen Spillover on Carbon- Supported Pd Nanoparticles New Insights from DFT - Warczinski - 2021 - Unkno

《How Nitrogen Doping A ff ects Hydrogen Spillover on Carbon- Supported Pd Nanoparticles New Insights from DFT - Warczinski - 2021 - Unkno》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

pubs.acs.org/JPCCArticleHowNitrogenDopingAffectsHydrogenSpilloveronCarbon-SupportedPdNanoparticles:NewInsightsfromDFTLisaWarczinski*andChristofHättigCiteThis:J.Phys.Chem.C2021,125,9020−9031ReadOnlineACCESSMetrics&MoreArticleRecommendations*sıSupportingInformationABSTRACT:Theprocessofhydrogenspilloveronmetalnanoparticledecoratedcarbonsurfaceshasbeendiscussedformanyyearsduetoitsimportanceforheterogeneouscatalysisandhydrogenstorage.Thepresentdensityfunctionaltheory(DFT)studysupportsrecentexperimentalobservationsofthehydrogenspilloverprocess.Byuseofmodelsystemsofcarbon-supportedpalladiumnanoparticles,thedetailsofhydrogenspilloverareinvestigatedandtheeffectsofnitrogendopingonsuchprocessesareelucidated.Itisfoundthatgraphiticnitrogenatomsandwatermoleculessignificantlyfacilitatethehydrogenspilloverreactionandserveasanchoringsitesforthespilloverhydrogenatoms.■INTRODUCTIONadsorbedasprotonsonthecarbonsupportmaterial,whichmightbefurtherstabilizedbywatermolecules.Inahydrogenspilloverprocess,activatedhydrogenspeciesIngeneral,itisverydifficulttodetectspilloverhydrogenadsorbedorformedonafirstsurfacearemovedtoasecond1,11,12speciesinexperiments.Severalexperimentalresultsonsurfaceonwhich,underthesameconditions,theseactivatedhydrogenspeciescouldnotadsorborform.1Forthefirsttime,thehydrogenspillovereffectwerelaterproventobe2thehydrogenspillovereffectwasreportedin1964byerroneous.Therefore,thepurposeofthepresentstudyistoKhoobiar,whoobservedatransportofhydrogenatomsfromusequantumchemistrytosupporttherecentexperimental2observations8−10andtoimprovetheunderstandingofthePttoWO3.Inthepastdecade,hydrogenspilloverfrommetalnanoparticlesontocarbonmaterialshasattractedtremendoushydrogenspilloverprocessandofthepropertiesofspilloverattentionasthespillovereffectsignificantlyenhancesthehydrogen.Weparticularlyaimtoshedlightontheeffectof3−7carbonmaterial’scapabilityofstoringhydrogen.Thisopensdifferentnitrogenfunctionalgroupsonthepropertiesofupthepossibilityofusingmetalnanoparticledecoratedcarbonspilloverhydrogenandtoelucidatehownitrogendopingcouldmaterialsashydrogenstorageinfuturefuelcellsystems.Alsoenhancethehydrogenspilloveroncarbon-supportedmetalDownloadedviaUNIVFEDDEMINASGERAISonMay14,2021at20:25:30(UTC).theusageofreversiblystoredspilloverhydrogenfornanoparticles.Furthermore,westudytheinteractionofheterogeneouscatalyzedreactions,e.g.,forhydrogenationspilloverhydrogenatomswithwatermolecules,whichontheSeehttps://pubs.acs.org/sharingguidelinesforoptionsonhowtolegitimatelysharepublishedarticles.reactions,isofhighinterest.8realcatalystsaredifficulttoavoidandmightplayacrucialroleUptothepresent,severalexperimentalstudieshavetriedtoforspilloverhydrogen.Forourstudy,weapplydensityshedlightonthedetailsofthehydrogenspillovermechanism,functionaltheory(DFT)tomodelsofmesoporouscarbonwhichturnedouttobequitedifficultduetothecomplicatedsupportedpalladiumnanoparticles,inwhichaPd21clusterischaracterizationofthecarbonsupportandthehydrogen2910supportedonahydrogenterminatedgraphenelayer.Aswespilloverspecies.Inrecentyears,Lietal.andWangetal.haveshownthatgraphiticnitrogen(6%relativeabundanceinobservedthatsurfacefunctionalgroupsandresidualmoisturetheNMCsupportmaterial)andpyridinicnitrogenspeciessignificantlyenhancethespilloverprocess.Inacomprehensive(45%relativeabundanceintheNMCsupportmaterial)studyonhydrogenspilloveronplatinum/nitrogen-dopedsignificantlytunethepropertiesoftheanchoredPdnano-mesoporouscarbon(NMC)compositesfromtheyear2018,8Yangetal.providednovelinsightsintothehydrogenspillovereffect.TheywereabletogenerateandtogiveexperimentalReceived:December23,2020evidenceofspilloverhydrogenonthenitrogen-dopedcarbonRevised:March21,2021supportmaterial.Furthermore,theycharacterizedspilloverPublished:April21,2021hydrogenasamobileandreversiblystoredhydrogenspecies,presentinachemicallyinactivestate.Onthebasisoftheseobservations,theysuggestedthatspilloverhydrogenis©2021TheAuthors.PublishedbyAmericanChemicalSocietyhttps://doi.org/10.1021/acs.jpcc.0c114129020J.Phys.Chem.C2021,125,9020−9031

1TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticle13particlesinapreviousstudy,thenitrogendopingismodeledbyincorporatedgraphiticandpyridinicnitrogenatoms.Similarapproacheshavebeenusedinseveralstudiesandhaveprovidedvaluableinsightsintothepropertiesofcarbon-supportedPdnanoparticles.However,mostofthesestudies14−16usedsinglePdatomsorverysmallPdclusters.Asan17example,Rangeletal.investigatedthehydrogenspillover18processonPd4clusters.OneofourpreviousstudieshasdemonstratedthatitisessentialtoalsoconsiderlargerclustersizeslikePd21,whichareclosertoPdnanoparticlesizespresentinexperimentalsystems.Severalaspectsofhydrogenadsorption,dissociation,anddiffusiononcleananddopedgraphiticsurfaceshavebeenstudiedbeforeboththeoreticallyandexperimentally;e.g.,seerefs19and20andtherecent21reviewbyGerberetal.InthecurrentworkwestudytheeffectsofdopingagraphiticsurfacewithchemicallydifferentnitrogensonthespilloverfromPdnanoparticlestogetherwithbindinganddiffusionofthespilloverhydrogenandprovideadditionalinsightsintotheelectronicstructureofhydrogenatthedifferentbindingsites.Theexperimentallyobservedhydrogenspilloverisacomplexmultistepprocess,influencedbyseveralaspectsasFigure1.Modelsystemforthepuremesoporouscarbonsupporthydrogenpressure,temperature,impurities,andcatalystmaterial(CMC).morphology.Theexperimentalcatalystsystems,Pdnano-particlesonnitrogen-freemesoporouscarbon,Pd/CMC,andPd/NMC,includesurfacedeformations,furtherdefectslikesuccessfullyusedtostudyhydrogenadsorptiononcarbon-five-andseven-memberedrings,orsurfactantsandresidualsupportedPdnanoparticles,18waschosenbecausetheuppermoisture.AsstudyingalltheseaspectsatthesametimewouldfaceatomsreflectaPd{111}facetandpossessestwoadjacentintroducetoomanydegreesoffreedom,thepresentstudyfacecenteredcubic(fcc)sitesatthetop.ItsinitialgeometryutilizesmodelsystemsanddealswiththedifferentstepsofthewastakenfromShamsievetal.23hydrogenspilloverprocessseparately.Asaconsequence,weAgain,forthenitrogen-dopedsystems(Pd/NMC)oneorfocusonqualitativeobservationsanddonotcompareourthreecarbonatomswerereplacedbynitrogenatomstomodel,resultstospecificexperimentalvalues.Elucidatingtherespectively,graphiticorpyridinicN-sites.TheconfigurationsenergeticsoftheindividualstepsconstitutesanimportantofPd21onthesupportswithgraphiticandpyridinicNforthesteptowardabetterunderstandingofhydrogenspilloverandpresentstudyareshowninFigure4.ThePd21clusterhasbeenoftheeffectofnitrogendopingonsuchprocesses.validatedinpreviousinvestigations.InastudyonhydrogenFurthermore,thepresentstudycangivesomecomputationalabsorptiononcarbon-supportedPdnanoparticles18itturnedevidenceifhydrogencouldactuallybereversiblystoredonPd/outasasuitablemodelforthemetal−supportinteractionNMC.betweenPdnanoparticlesandthecarbonsupportanditsinfluenceonhydrogenadsorption.Themodelsforthe■COMPUTATIONALDETAILSnitrogen-dopedsupportwereevaluatedinacombinedModelSystems.Tomodelthenitrogen-freemesoporoustheoreticalandexperimentalstudyontheanchoringofPd13carbonsupportmaterial(CMC),wechoseahydrogennanoparticlesatnitrogendefectsinNMCandlaterappliedterminatedgraphenelayercomprising96carbonatoms(seeinastudyofthebifunctionalcatalyticroleofPd/NMCforthe22Figure1).Innitrogen-dopedmesoporouscarbonthenitrogenreductionof5-hydroxymethylfurfural.Forfurtherdetailsweatomscanbepresentinfourforms:graphiticN,pyridinicN,referthereaderstorefs13,18,and22andtheSupportingpyrrolicN,andN−Olikespecies.13Aspyrrole-likeandN−O-Informationprovidedwiththesestudies.likenitrogenatomsareeitherlocatedontheedgesofthesupportmaterialorincludelargesurfacedeformations,bothof■METHODOLOGYwhichwouldintroduceundesiredsideeffectsintheelectronicThepresentdensityfunctionaltheory(DFT)studywas24structurecalculationsofthereactions,thesesystemshavenotperformedusingtheTURBOMOLEprogrampackagewith2526beenconsidered.Accordingly,tomodeltheNMCsupporttheTPSSfunctionalandthedef2-SVPbasisset.Stuttgart−27material,one(graphiticN,seeFigure2a)orthree(pyridinicCologne28-electroneffectivecorepotentials(def2-ecp)N,seeFigure2b)carbonatomsofthegraphenelayerwerewereusedforthePdatoms.Allcalculationsincludedthe28replacedbynitrogenatoms.multipoleacceleratedresolutionofidentity(MARI-J)29Tomodelthecarbon-supportedpalladiumnanoparticlesapproximationwithoptimizedauxiliarybasissetsand30(Pd/CMC),wechoseaPd21clustersupportedbyahydrogenGrimme’sD3correctionforLondondispersioninteractions.terminatedgraphenelayerof150carbonatoms(seeFigure3).ThemoststablestructuresofourmodelsystemsaswellastheThePd21clusterpossessesadiameterofaround0.85nm,minimaofhydrogenonthesesystemsweredeterminedbycorrespondingtosizesatthelowerboundaryofexperimentallyground-stategeometryoptimizations.ToobtainaccuratedeterminedparticlesizedistributionsforPd/CMCandPd/relativeenergies,singlepointenergieswerecombinedwithNMC,whichshownanoparticlesof1−8nmaswellassomezero-pointvibrationalenergies(ZPVEs)fromvibrational22sub-1nmPdparticles.Thiscluster,whichhasbeforebeenanalysis.9021https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

2TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticleFigure2.Modelsystemsforthenitrogen-dopedmesoporouscarbonsupportmaterial(NMC):(a)graphiticNsupport;(b)pyridinicNsupport.TPSS/def2-SVPgeometries,whichdocumentthegoodagreementofenergeticscalculatedwithdifferentfunctionalsandbasissetsforoursystems,canbefoundintheSupportingInformation(seeTablesS1−S3).Reactionpathwayswerepredictedusingthechain-of-statemethodimplementedinthewoelflingmoduleofTURBO-33MOLE,andthefoundinitialtransitionstatestructureswerethenfurtheroptimizedwiththetrust-regionimageminimiza-34tion(TRIM)algorithm.Alltransitionstateswerefinally35validatedbyintrinsicreactioncoordinate(IRC)calculations.Toallowforaqualitativecomparisonofthecalculatedresultstoexperimentalstudies,Gibbsfreereactionenthalpiesandfreeactivationenthalpiesweredeterminedwithintheharmonicoscillatorandrigidrotorapproximationunderreactionconditionsof473.15Kand0.1MPa(chosenanalogouslytotheexperimentalreactionconditionsinref8)withascaling36factorof1.0190fortheTPSSfunctional.Favorableadsorptionenergiesanddissociationenergiesareprovidedwithanegativesign,andenergybarrierslikeactivationbarriersFigure3.ModelsystemforPdonthemesoporouscarbonsupportareprovidedwithapositivesign.Toarriveatadeepermaterial(Pd/CMC).Adaptedfromref13withpermissionfromtheunderstandingoftherelevantmetal−support,hydrogen−PCCPOwnerSocieties.support,andmetal−hydrogeninteractions,westudiedintrinsic37bondorbitals(IBOs),intrinsicatomicorbital(IAO)charge37analyses,andelectrondensitydifferences.Astheπ-systemofAtthislevel,resultsfromDFTwithdifferentfunctionalsandthedifferentsupportmaterialscanbedescribedwellbyHückelbasissetscandifferintherangeofsomekJmol−1perbond.Intheory(seeFigureS1foravalidationofHückelvsDFT),we18apreviousstudywehaveshownthatmovingtoalargerbasishadacloserlookatHückelorbitalenergiesandpopulationsetortootherDFTfunctionalsdoesaffecttheDFTenergeticsanalysis.Tomodelthemolecularsystemafterspilloverofofthecatalystsystems,butcalculatedenergydifferencesarehydrogentothesupportmaterialswithintheHückeltheory,similarandtheobservedtrendsoftheenergydifferencesarethecarbonatomattachedtothespilloverhydrogenwasessentiallythesame.Duetothelimitedcomparabilityofourdeletedfromtheπ-system.idealizedmodelsystemstoexperimentalcatalysts,weanywaylimitourconclusionstoqualitativeobservationsandomit■comparisonstospecificexperimentalvalues.OurchoiceoftheRESULTSANDDISCUSSIONmeta-GGATPSSfunctionalandthedef2-SVPbasissethasElectronicPropertiesofSpilloverHydrogen.Asafirstbeenshowntobereasonableinseveralpreviousstudies,asitstep,weinvestigatedtheelectronicpropertiesofspillovershowsaverygoodcost/performanceratioforthesystemshydrogen.Whenahydrogenatomspillsovertothecarbon13,18,22,31,32underconsideration.Acomparisonofsomesupportmaterial,itbindstoonecarbonatom.InthisexemplaryTPSS/def2-SVPenergydifferenceswiththoseconfigurationthecorrespondingcarbonatomcomesoutoffromPBE/def2-SVPandTPSS/def2-TZVPonthebasisofthesupportplaneandchangesitshybridizationpartiallyfrom9022https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

3TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticleFigure4.ModelsystemsforPdonthenitrogen-dopedmesoporouscarbonsupportmaterial(Pd/NMC):(a)PdongraphiticNsupport;(b)PdonpyridinicNsupport.Adaptedfromref13withpermissionfromthePCCPOwnerSocieties.Figure5.Spilloverhydrogenonthegraphenesupportmaterial:(a)geometry;(b)IBOfortheC−HbondcomposedofaCsp3hybridandaHsorbital.Isosurfacevalueis±0.09.Figure6.Electrondensitydifferenceplotforspilloverhydrogenonthegraphenesupportmaterial:(a)topview;(b)sideview;(c)sideviewrepresentedwithameshtorevealinsidechanges.Densitygainiscoloredingreen;densitylossiscoloredingray.Isosurfacevalueis±0.0025.9023https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

4TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticlesp2tosp3(seeFigure5a).Consequently,theintrinsicbondtheπsystem.Thegraphiticnitrogenatompossessesanorbital(IBO)analysisshowsansp3hybridC−Hbondingadditional(incomparisontoC)electronintheπ-system,orbital(seeFigure5b).whichisunpairedandfacilitatesthebindingofahydrogenBesidesthelocalchangeofthecarbonsupportgeometry,radical.ThiscanbevalidatedwiththeHückelorbitalenergies,alsotheelectronicstructureofthesupportmaterialisaffectedwhichshowahighestoccupiedmolecularorbital(HOMO)asthecarbonatomattachedtothespilloverhydrogendoesnotdoublyoccupiedandsignificantlybelowtheFermi-level(seecontributetotheπ-systemanymore.Accordingly,thespilloverFigureS2b).Furthermore,intheelectrondensitydifferenceithydrogendisturbstheπ-systemofthesupportmaterial,whichcanbeobservedthatthegraphiticnitrogenatomleadstoaleadstoaveryunfavorableH2dissociationfreeenthalpy(H2+delocalizationofshiftedelectrondensityfromtheHsorbital2CMC→2H-CMC)ofΔG=+362.0kJmol−1.disstotheπ-systemofthesupportmaterial(seeFigure7andTheperturbingeffectofspilloverhydrogenontheπ-systemFigureS3forelectrondensitydifferenceplotsforsystemswithcanfurtherbevalidatedwithHückelmolecularorbitaltheorytwoandsixgraphiticNatoms).whichallows,viaacomparisonwithDFTresults,separationofqualitativeeffectsoriginatingfromtheresonanceenergyoftheπ-systemfromthoseresultingfromothersourcesase.g.geometricstrain.TheHückelorbitalenergiesshowthatthespilloverhydrogenshiftsthehighestoccupiedmolecularorbital(HOMO),containinganunpairedelectron,uptotheFermi-level(seeFigureS2a).Togetfurtherinsightintothepropertiesofspilloverhydrogenitself,wecalculatedtheelectrondensitydifferenceforspilloverhydrogenonthegraphenesupportmaterial(seeFigure6).LiketheIBOanalysis,theelectrondensitydifferenceplotalsoproposesthatachemicalbondbetweenthespilloverhydrogenandonecarbonatomisformedasanelectrondensityshiftfromtheπ-bondsoftheboundcarbonatomtoansp3hybridC−Hbondingorbitalcanbeobserved.Moreover,theplotshowsthatelectrondensityisshiftedfromtheHsorbitaltotheπorbitalsofthesurroundingCatoms.Thissuggeststhatspilloverhydrogenispositivelycharged.OntheFigure7.ElectrondensitydifferenceplotforspilloverhydrogenonbasisofIAOatomicchargeanalyses,wefoundthatthepartialthegraphiticNsupportmaterial.Densitygainiscoloredingreen;chargeofspilloverhydrogen(q=+0.16)ismorepositivethandensitylossiscoloredingray.Isosurfacevalueis±0.0025.thepartialchargeofhydrogenatomslocatedattheedgeofthesupportmaterialorofthehydrogenatomsinthemethaneItisimportanttonotethattheextentoftheeffectofthemolecule(bothq=+0.13).graphiticnitrogenatomontheπ-systemdependsonits8IncontrasttothesuggestionofYangetal.,wedonotfindpositionwithrespecttothespilloverhydrogen(seeFigure8thespilloverhydrogeninthestateofa“real”proton,astheelectronofthespilloverhydrogenatomislocalizedinansp3hybridC−HbondingorbitalwiththeHnucleuslocatedatapositionwheretheelectrondensityishigherthanforanisolatedHatom(seeFigure6c).However,spilloverhydrogenatomspossessarelativelyhighpositivepartialchargeduetoelectrondensityshiftedfromtheHsorbitaltotheπ-system.Furthermore,westudiedtheeffectofnitrogendopingonthepropertiesofspilloverhydrogen.Dopingwithpyridine-likenitrogenatomsslightlyincreasesthepositiveGibbsfreeenthalpyfortheH2dissociation(foraspilloverhydrogennexttothepyridinicN)toΔG=+364.8kJmol−1.AsthedisselectronegativecharacterofthepyridinicNleadstoapositivechargeattheneighboringcarbonatoms,onewouldexpectthatdopingwithpyridine-likenitrogenatomswouldmoresignificantlyhinderthebindingofahydrogenradical.However,thehigherstructuralflexibilityatthedefectallowsforgeometricaldeformationsofthesupportand,therefore,Figure8.SketchofspilloverhydrogenatompositionsongraphiticN-allowsforaH−Cbondwithastrongersp3character,whichdopedgraphenesupport.compensatesfortheunfavorableelectroniceffectofthepyridinicN.Incontrasttodopingwithpyridinicnitrogen,dopingwithagraphiticnitrogenatomsignificantlydecreasesthefreeforasketchofthedifferentpositions).ThiseffectisalmostenthalpyforthedissociationofH2toΔGdiss=+150.4kJquantitativelyduetotheresonanceenergiesoftheπ-sytemasmol−1iftheHatomisboundatthemostfavorablesite,whichthecomparisonoftheenergydifferencesobtainedattheDFTistheCatominorthopositiontothenitrogen.ThislevelwiththeresonanceenergiesfromHückeltheoryobservationisagainexplainedbytheresonanceenergiesforstabilizationenergiesforthedifferentpositionsinTable19024https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

5TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticleshows.Positiverelativevaluesindicatealowerstabilizationenergyand,therefore,asmallereffectofthegraphiticN.Table1.HückelandDFTEnergyDifferencesfortheDifferentSpilloverHydrogenPositionsSketchedinFigure8withRespecttotheMostStableOrtho-Position,forHückelinUnitsoftheResonanceParameterβandforDFTinkJmol−1hydrogenpositionHückel,|β|DFT,kJmol−1o0.000.0m0.5160.4p0.2519.8a0.51101.4b0.4261.5c0.4980.4Figure10.Geometryofcarbon-supportedPd21loadedwith26d0.4890.6hydrogenatoms.e0.5085.7atoms(furtherdetailscanbefoundinref18).When,Theextentoftheeffectisverysimilartoarenesubstitutionsubsequently,severalhydrogenmoleculesarechemisorbedpatternsandcanbesummarizedasfollows:onthecluster,hydrogenatomsareattachedtothesideofthe•withinfirstshell:o

6TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticlegrapheneandsupportdopedwithpyridinicN.Table2showstheactivationenergies(ΔEact)andthefreeactivationTable2.ActivationEnergiesandFreeActivationEnthalpiesfortheHydrogenSpilloverfromPd21ontoDifferentSupportMaterialsΔGact,modelsystemΔE,kJmol−1kJmol−1actPd21ongraphenesupport;trianglesite130.8140.8Pd21ongraphiticNsupport;trianglesite94.897.9Pd21ongraphiticNsupport;bridgesite46.850.4Pd21onpyridinicNsupport;trianglesite193.8200.3enthalpies(ΔGact)forthehydrogenspilloverfromPd21onthethreedifferentsupportmaterials.ForthegraphiticNsupport,alsotheactivationenergyandfreeactivationenthalpyfromthelessstablereactantarelisted.Table3showsthecorrespondingreactionenergies(ΔEreact)andfreereactionenthalpies(ΔGreact).Figure11.HydrogenspilloverpathwayfromPd21tothegraphiticNsupportmaterialfromtwodifferentinitialsites.Table3.ReactionEnergiesandFreeReactionEnthalpiesfortheHydrogenSpilloverfromPd21ontoDifferentSupportMaterialsΔEreact,ΔGreact,modelsystemkJmol−1kJmol−1Pd21ongraphenesupport;trianglesite129.2143.4Pd21ongraphiticNsupport;trianglesite86.094.4Pd21ongraphiticNsupport;bridgesite38.047.0Pd21onpyridinicNsupport;trianglesite187.9187.4Onthepuregraphenesupportmaterialthehydrogenspilloverreactionstartingfromthetrianglesiteshowsquiteahighactivationenergyandfreeactivationenthalpy.EvenifoneassumesthatstartingfromthelessstablebridgepositionwilldecreasetheGibbsfreereactionenthalpybyaround45kJmol−1(approximatedfromtheresultforthegraphiticNsystem),itstillmountsuptoaround100kJmol−1.Thisindicatesthatthehydrogenspillovertopurecarbonsupportmaterialsiskineticallydifficult,whichisalsoinagreementwithFigure12.HydrogenspilloverpathwayfromPd21tothepurepreviousabinitiomoleculardynamics(AIMD)simulations,graphenesupportmaterial.whichhavenotfoundasinglecaseofhydrogenspilloverfrom39Pd6orPd13clusterstopuregraphene.DopingwithgraphiticNatomssignificantlydecreasestheactivationenergyandfreeactivationenthalpyofthespilloverreaction(forthetrianglesitetoaround95kJmol−1;forthebridgesitetoaround50kJmol−1).ThiseffectcanbeexplainedonthebasisofHammond’spostulate:Ingeneral,thehydrogenspilloverreactionisanendothermicreaction(seeTable3).Thetransitionstateisverysimilartotheproductstate,whichiscalled“late”transitionstate.AsthedopingwithgraphiticNatomsdecreasesthepositivereactionenergyand,therefore,stabilizestheproductstate(duetoitsadditionalelectron;seeprevioussection),thedopingalsostabilizesthetransitionstate.AnalogoustotheexplanationforthedecreaseoftheactivationenergyforthegraphiticNsystem,alsotheincreaseoftheactivationenergyforthepyridinicNsystemcanbeexplainedusingHammond’spostulate.Asdescribedintheprevioussection,thepositivelychargedcarbonatomsnexttothepyridinicNsignificantlyhinderthebindingofthespilloverhydrogen.Incomparisonwiththefreesupportmaterial,theFigure13.HydrogenspilloverpathwayfromPd21tothepyridinicNbindingofthePdclusteronthesurfacedefecthindersasupportmaterial.changeofthesupportgeometryuponhydrogenspillover.9026https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

7TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticleAccordingly,dopingwithpyridinicnitrogendestabilizesthehydrogenatomisnotstronglyacidic;i.e.itdoesnotprotonateproductofthehydrogenspillover(seeTable3)and,therefore,HOtoformHO+,whichisinagreementwithourprevious23alsoincreasestheactivationenergy.findings,showingthatspilloverhydrogenispositivelychargedIngeneral,ourcalculationsrevealthatthehydrogenspilloverbutisnotinthestateofacidicprotons.issignificantlyfacilitatedbygraphiticNatoms.ThefreeToelucidatethedetailsoftheC−H−Ointeraction,weactivationenthalpyforthespilloverfromthelessstablebridgecalculatedtheelectrondensitydifferenceofthespilloversiteisonlyaround50kJmol−1.AccordingtotheArrheniuslawhydrogen−watersystem(seeFigure15a)andcomparedittoÄÅÅÉÑÑtheelectrondensitydifferenceofthewaterdimer(seeFigureν=νexpÅÅÅÅ−ΔEactÑÑÑÑ15b).Theelectrondensitydifferencesofthesetwosystemsare0ÅÅkTÑÑnearlyequivalent.ThisindicatesthatthepreviouslyfoundÅÅÇBÑÑÖ(1)positivepartialchargeofspilloverhydrogenishighenoughsowiththeactivationenergyΔE=46.8kJmol−1andaactthatitcanformhydrogenbondswithwatermolecules.Thecommonestimateofthepre-exponentialvalueofν=kTB=1−1energyofsuchahydrogenbondisEhbond=−37.0kJmolata0h×107s−1,17,40,41onecanexpectahydrogenmigrationrateofνhydrogenbondlengthofrb=2.24Å,comparedtoEhbond=−1∼7×107s−1at473.15K.Thissufficientlyhighrategives−28.7kJmolandrb=1.90Åforthewaterdimer.Thisisin8someevidencethatdopingwithgraphiticnitrogenatomscouldagreementwiththeexperimentalstudiesofYangetal.,Liet910actuallyenablethehydrogenspilloverprocessonPd/NMC,al.,andWangetal.,whichproposethatspilloverhydrogenwhichisalsoinagreementwiththeexperimentalobservationisstabilizedbywatermolecules.ofhydrogenspilloveronPt/NMCbyYangetal.8DiffusionofSpilloverHydrogen.ToinvestigateifHydrogenBondsFormedbySpilloverHydrogenandspilloverhydrogencanplayaroleinhydrogenationreactionsWaterMolecules.Togetaninsightintotheinteractionofandifitlendsitselfforthepurposeofhydrogenstorage,itisspilloverhydrogenwithresidualmoisture,weoptimizedtheimportanttoalsostudythediffusionofspilloverhydrogenongeometryofawatermoleculeclosetospilloverhydrogen.Thethesupportmaterial.Table4liststheactivationenergiesequilibriumgeometryindicatesthatthespilloverhydrogenatomsomehowinteractswiththeoxygenatomofthewaterTable4.ActivationEnergiesandFreeActivationEnthalpiesmolecule(seeFigure14).However,theintrinsicbondorbitalforSpilloverHydrogenDiffusiononDifferentSupportMaterialsdiffusionpathwayΔE,kJmol−1ΔG,kJmol−1actacta:onpuregraphenesupportmaterial105.0107.6b:ongraphiticNsupportmaterial121.5123.9c:onpyridinicNsupportmaterial45.146.9(ΔEact)andthefreeactivationenthalpies(ΔGact)forthediffusionpathwaysonthethreedifferentsupportmaterials,whicharesketchedinFigure16.Table5showsthecorrespondingreactionenergies(ΔEreact)andfreereactionenthalpies(ΔGreact).Inallthesepathwaysthespilloverhydrogendiffusesinachemisorbedstatefromonetotheneighboringcarbonatom,surpassingatransitionstateinwhichtheHatombridgesacarbon−carbonbond.Figure14.Equilibriumgeometryofwateronaspilloverhydrogen.Onthepuregraphenesupportmaterialthechemisorbeddiffusionpathway(pathwaya)showsaquitehighactivationenergyandGibbsfreeactivationenthalpy.AnIBOanalysisand(IBO)analysisdoesnotgiveanybondingorbitalsbetweentheanIAOpartialchargecalculationrevealthatthespilloverwaterOandthespilloverHatom.Accordingly,thespilloverhydrogeninthetransitionstateisboundtotwocarbonatomsFigure15.Electrondensitydifferenceplots:(a)interactionofwaterwithaspilloverhydrogen;(b)interactionofthewaterdimer.Densitygainiscoloredingreen;densitylossiscoloredingray.Isosurfacevalueis±0.0025.9027https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

8TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticlehydrogenfirstdesorbsviaafirsttransitionstatetoaphysisorptionminimum,diffusesinthiswell,andthenadsorbsatthenextbutoneneighboringcarbonatom,surpassingasecondtransitionstate(seeFigure17).AccordingtoIBOFigure16.Sketchofspilloverhydrogendiffusionpathways:(a)diffusiononpuregraphenesupportmaterial;(b)diffusionongraphiticNsupportmaterial;(c)diffusiononpyridinicNsupportmaterial.Table5.ReactionEnergiesandFreeReactionEnthalpiesforSpilloverHydrogenDiffusiononDifferentSupportMaterialsΔEreact,ΔGreact,diffusionpathwaykJmol−1kJmol−1Figure17.Spilloverhydrogendiffusionpathwayviaaphysisorptiona:onpuregraphenesupportmaterial∼0.0∼0.0minimum.b:ongraphiticNsupportmaterial−62.6−62.3c:onpyridinicNsupportmaterial−53.4−32.5analysesandIAOchargecalculations,inbothtransitionstatesthespilloverhydrogenisphysisorbedasahydrogenatomandandissignificantlystrongerpositivelychargedcomparedtonotboundinachemisorbedstate.Inthephysisorptionspilloverhydrogenintheequilibriumgeometry(q=+0.25minimumthehydrogenatomisinadoubletspinstate(thecomparedtoq=+0.16).DuetothebridgingpositionofthecorrespondingspindensityisshowninFigureS4).spilloverhydrogen,thevalenceelectronsoftwocarbonatomsItisimportanttonotethattheactivationenergyandfreearepartlytakenoutoftheπ-system,whichexplainstheactivationenthalpyforthechangeofspilloverhydrogenfromunfavorableenergyofthetransitionstate.However,thethechemisorbedstatetothephysisorptionminimumareactivationenergyandfreeactivationenthalpyofthesignificantlylowercomparedtotheactivationenergyandfreechemisorbeddiffusionpathwaycanbesignificantlyreducedactivationenthalpyofthechemisorbeddiffusion(ΔEact=68.5byahydrogenbondedwatermoleculeviaaGrotthus-likekJmol−1andΔG=70.2kJmol−1comparedtoΔE=105.0actactmechanism(ΔG=77.7kJmol−1;seeFigureS5andTablekJmol−1andΔG=107.6kJmol−1).Thisobservationisinact1actS4forthereactionpathway).agreementwithpreviousstudiesreportingonbarrierless40,42Thesmallestactivationenergyandfreeactivationenthalpydiffusionofphysisorbedspilloverhydrogenatoms.canbeobservedforpathwayc,inwhichthespilloverhydrogenComparedtothechemisorbeddiffusion,thehydrogenatomatomislocatedclosetoapyridinicnitrogenatom.ThefoundinthetransitionstateofthechangetothephysisorbedstateisGibbsfreeactivationenthalpyisonly46.9kJmol−1,whichnotchemisorbedattwocarbonatomsbutonlyphysisorbedatcorrespondstoanexpectedhydrogenmigrationrateofν∼1×onecarbonatom,leadingtoasmallerperturbationoftheπ-108s−1at473.15K(seeeq1).AccordingtotheIBOanalysis,systemand,therefore,toasignificantlyreducedactivationthespilloverhydrogeninthetransitionstateissolelyboundtoenergyandfreeactivationenthalpy.Accordingly,onthepurethecarbonatomoftheproductstate.Incontrasttothepuregraphenesupportmaterial,hydrogendiffusionshouldoccurviagraphenesupportmaterial,itisnotsymmetricallyboundtotheaphysisorptionpathway.Thefreeactivationenthalpyforthetwocarbonatomsofreactantandproduct.Thiseffectcanbechangeofthehydrogenatomfromthechemisorbedtotheexplainedbythedifferentchargesofthecarbonatomsphysisorbedstateis70.2kJmol−1,whichcorrespondstoan(reactant,positivelycharged;product,negativelycharged),expectedrateofν∼3.5×105s−1at473.15K(seeeq1).inducedbythepyridinicnitrogendopant.Asaconsequence,Onthebasisoftheseobservations,wealsostudiedthetheperturbationoftheπ-system,andthereforealsothechangefromthechemisorbedtothephysisorbedstatefortheactivationenergyandfreeactivationenthalpy,issignificantlyinitialconfigurationsofthenitrogen-dopedsystemsinlowered.pathwaysbandc.ThecorrespondingactivationenergiesandClosetographiticNdopantsthecalculatedactivationfreeactivationenthalpiesforallsystemsaresummarizedinenergyandGibbsfreeactivationenthalpyforthespilloverTable6.hydrogendiffusionareveryhigh,whichisaconsequenceofitsClosetoapyridinicNatomtheactivationenergyandfreesignificantspilloverhydrogenstabilization.ThisresultindicatesactivationenthalpyareslightlyhighercomparedtothepurethatspilloverhydrogenwillaccumulateclosetographiticNgraphenesystem,whichcanbeexplainedwiththeslightlyatomsinthesupportmaterial.morestablereactantinthepyridinicsystem.AsthetransitionWhenwestudiedthediffusionofhydrogenspillovertothestateofthechemisorbeddiffusionpathwayissignificantlynextbutoneneighboringcarbonatomonthepuregraphenefavored(HatomsolelyboundtotheCatomoftheproductsupportmaterial,weobservedthatthechemisorbedspilloverstate),spilloverhydrogendiffusion,closetopyridinicNatoms,9028https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

9TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticleTable6.ActivationEnergiesandFreeActivationEnthalpiesshouldbeconsidered.ItisimportanttonotethatthechangefortheChangefromtheChemisorbedtothePhysisorbedfromthechemisorbedtothephysisorbedstateissignificantlyStatehinderedbyaneighboringhydrogenatomandtheactivationenergyisincreasedupto163.2kJmol−1forthegraphenemodelsystemΔE,kJmol−1ΔG,kJmol−1actactsupportsystem.Thiscanbeexplainedbythefactthatthepuregraphenesupport68.570.2reactantofachemisorbedspilloverhydrogenisstabilizedbyapyridinicNsupport80.076.3neighboringchemisorbedspilloverhydrogenatom,asthegraphiticNsupport107.6107.4secondhydrogenatomcancelsouttheradicalstateofthesystem.Accordingly,alsothehydrogendesorptionviaashouldoccurviaachemisorbeddiffusiononceHwasboundatphysisorbedintermediateisunfavorableandtheformationofsuchaposition.spilloverhydrogenatomdomainsisverylikely.ClosetoagraphiticNatomtheactivationenergyandfreeOfcourse,twophysisorbedspilloverhydrogenatomsactivationenthalpyaresignificantlyhighercomparedtothecomingclosetogethercouldeasilydesorbasH.However,2puregraphenesystem.Thiscanalsobeexplainedwiththehighasthefreeactivationenthalpyforthediffusionfromthestabilityofthereactant.EventhoughthephysisorbeddiffusionsupportmaterialbacktothePdclusterisverysmall(e.g.,3.5pathwayisnoticeablyfacilitatedcomparedtothechemisorbedkJmol−1forthespilloveronthegraphiticNsupportmaterial),diffusionpathway(HatomphysisorbedatoneCatomspilloverhydrogenwillratherdiffusebacktothePdclustercomparedtochemisorptionattwoCatomsinthetransitionthandesorb.Theseresultssuggestthatifhydrogenspilloveronstate),thefreeactivationenthalpyforthechangefromthePd/NMCoccurs,thespilloverhydrogenwouldbereversiblyphysisorbedtothechemisorbedstatestillmountsupto107.4storedonthecarbonsupportmaterial.ThisisinagreementkJmol−1.8withthestudyofYangetal.,whichgivesexperimentalTheexpectedratesofν∼1×108s−1forthehydrogenatomevidenceofreversiblystoredspilloverhydrogenspeciesonPt/migrationinthevicinityofpyridinicnitrogenatomsandofν∼NMC.3.5×105s−1forthechangefromthechemisorbedtothephysisorbedstateonthepuregraphenesupportmaterial■CONCLUSIONSindicatethatspilloverhydrogenatomsshouldbemobileontheUsingquantumchemistry,wewereabletoprovideanin-depthcarbonsurface.ThiswasalsoexperimentallyproposedbyYang8insightintothehydrogenspilloverprocessoncarbon-etal.Moreover,spilloverhydrogenwillaccumulateclosetosupportedpalladiumnanoparticlesandintotheeffectofgraphiticNatoms.dopingwithgraphiticandpyridinicnitrogenonsuchprocesses.DesorptionofSpilloverHydrogen.TofurtherclarifyWefoundthatthehydrogenspilloverreactionissignificantlythepossibleusageofspilloverhydrogeninthecontextoffacilitatedbygraphiticnitrogenatoms,leadingtoarelativelyhydrogenstorage,alsothedesorptionoftwoneighboring−1lowGibbsfreeactivationenthalpyofaround50kJmolandaspilloverhydrogenstoH2hastobeconsidered.Tables7and87−1feasiblemigrationrateofaround7×10sforthespilloverofahydrogenatomfromaPd−PdbridgepositionatreactionTable7.ActivationEnergiesandFreeActivationEnthalpiesconditionsof473.15Kand0.1MPa.ThisobservationcanbeforSpilloverHydrogenDesorptiononDifferentSupportexplainedonthebasisofHammond’spostulateasthegraphiticMaterialsnitrogenstabilizestheproductstate.DiffusionofspilloversupportmaterialΔE,kJmol−1ΔG,kJmol−1hydrogenatomstakesplacewithfeasiblemigrationratesatactactgraphenesupport232.0231.2puregraphenesupportviaaphysisorbedpathwayandinthegraphiticNsupport128.4128.3vicinityofpyridinicnitrogenatomsviaachemisorbedpathway.pyridinicNsupport176.8169.2Closetographiticnitrogenatomsnodiffusionshouldbeobserved,asthegraphiticnitrogenstabilizesthespilloverTable8.ReactionEnergiesandFreeReactionEnthalpiesforhydrogenbindingduetoitsadditionalelectrondistributedtoSpilloverHydrogenDesorptiononDifferentSupporttheπ-system.SpilloverhydrogendesorptioniskineticallyMaterialshinderedonallsupportmaterials.OnthebasisoftheseresultsandtheresultsofourprevioussupportmaterialΔE,kJmol−1ΔG,kJmol−113reactreactstudy,thefollowingcanbeconcluded:Hydrogenspillovergraphenesupport−199.9−370.2fromPdtopurecarbonsupportmaterials(asinPd/CMC)isgraphiticNsupport−138.4−194.3kineticallydifficult.InPd/NMCthepalladiumnanoparticlespyridinicNsupport−183.8−243.8arepreferentiallyanchoredonpyridinicnitrogenatoms.Graphiticnitrogenatoms,whichmakeuparound6%oftheshowtheactivationenergies(ΔEact)andfreeactivationNfunctionalgroupsoftheNMCsurfaceusedinref13,enthalpies(ΔGact)aswellasthereactionenergies(ΔEreact)significantlyreducethebarriersofthehydrogenspilloverandfreereactionenthalpies(ΔGreact)ofthedesorptionprocess.Thisgivessomeevidencethatgraphiticnitrogenpathwaysonthedifferentsupportmaterials(seeFiguresS6,atomscouldactuallyenablethehydrogenspilloverprocessonS7,andS8forthereactionpathways).Pd/NMC.ThecalculatedpositiveGibbsfreereactionItisobservedthatthespilloverhydrogendesorptionisenthalpiesforallspilloverreactionpathwaysareduetothehighlykineticallyhinderedonallsupportmaterials.Onthefactthatourcalculationsjustincludeonehydrogenatomonbasisofthepreviousobservationthatthechangefromthethecluster.InthechemicalequilibriumofahighhydrogenchemisorbedtothephysisorbedstateforspilloverhydrogenoncoverageonthesupportedPdnanoparticle,asitispresentthegraphenesupportmaterialshowsanactivationenergyofwhenthehydrogenspilloveroccurs,ΔGofthehydrogenonly68.5kJmol−1,alsothehydrogendesorptionviaanchemisorptiononthePdclustershouldbezero.ThecalculatedintermediateinwhichonehydrogenatomisphysisorbedfreechemisorptionenthalpyofH2onourmodelsystemison9029https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

10TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticletheorderof−60kJmol−1,whichwillcompensateforthe(3)Lueking,A.;Yang,R.T.HydrogenSpilloverfromaMetalOxideremainingpositivefreereactionenthalpyofthespilloverCatalystontoCarbonNanotubesImplicationsforHydrogenprocess.Moreover,inthechemicalequilibriumthesupportedStorage.J.Catal.2002,206,165−168.Pdclusterismostprobably(afterhydrogenspillover)(4)Lachawiec,A.J.;Qi,G.;Yang,R.T.HydrogenStorageincontinuouslyreloadedwithhydrogenfromthegasphase,NanostructuredCarbonsbySpillover:Bridge-BuildingEnhancement.Langmuir2005,21,11418−11424.whichchangestheoverallreactionequation.(5)Contescu,C.I.;Brown,C.M.;Liu,Y.;Bhat,V.V.;Gallego,N.Besidestheireffectonthespilloverprocess,graphiticandC.DetectionofHydrogenSpilloverinPalladium-ModifiedActivatedpyridinicnitrogenatomscanserveasanchoringsiteswhereCarbonFibersduringHydrogenAdsorption.J.Phys.Chem.C2009,spilloverhydrogenspecieswillaccumulate.Whileatgraphitic113,5886−5890.nitrogenswefindaratherhighbarrierforthemigrationtoor(6)Blackburn,J.L.;Engtrakul,C.;Bult,J.B.;Hurst,K.;Zhao,Y.;awayfromthecarbonsinorthopositiontotheNatoms,theXu,Q.;Parilla,P.A.;Simpson,L.J.;Rocha,J.-D.R.;Hudson,M.R.;respectivebarriersaremuchlowerclosetopyridinicNatoms.Brown,C.M.;Gennett,T.SpectroscopicIdentificationofHydrogenWhenthehydrogenspilloverprocessactuallyoccursonPd/SpilloverSpeciesinRuthenium-ModifiedHighSurfaceAreaCarbonsNMC,thespilloverhydrogenwillbereversiblystoredonthebyDiffuseReflectanceInfraredFourierTransformSpectroscopy.J.carbonsupportmaterialasthebarriersformigratingbackfromPhys.Chem.C2012,116,26744−26755.thecarbonsupporttothePdnanoparticlearequitesmalland(7)Liu,X.M.;Tang,Y.;Xu,E.S.;Fitzgibbons,T.C.;Larsen,G.S.;spilloverhydrogendesorptionisunlikely.Gutierrez,H.R.;Tseng,H.-H.;Yu,M.-S.;Tsao,C.-S.;Badding,J.V.;Ingeneral,wedonotfindthatspilloverhydrogenisintheCrespi,V.H.;Lueking,A.D.EvidenceforAmbient-Temperaturestateof“real”protons.However,thespilloverhydrogenisReversibleCatalyticHydrogenationinPt-dopedCarbons.NanoLett.2013,13,137−141.positivelychargedandisstabilizedsignificantlybywater(8)Yang,F.;Hu,B.;Xia,W.;Peng,B.;Shen,J.;Muhler,M.Onthemoleculesofresidualmoistureviahydrogenbonds.Thesenatureofspilloverhydrogenspeciesonplatinum/nitrogen-dopedwatermoleculesalsoreducethebarrierformigrationofmesoporouscarboncomposites:Atemperature-programmednitro-hydrogenatomsonthecarbonsupport.ViaaGrotthus-likebenzenedesorptionstudy.J.Catal.2018,365,55−62.mechanismtheymightenhanceconsiderablythediffusionof(9)Li,Q.;Lueking,A.D.EffectofSurfaceOxygenGroupsandspilloverhydrogenonCMCandNMCsupports.WateronHydrogenSpilloverinPt-DopedActivatedCarbon.J.Phys.Chem.C2011,115,4273−4282.■ASSOCIATEDCONTENT(10)Wang,L.;Yang,F.H.;Yang,R.T.;Miller,M.A.Effectof*SurfaceOxygenGroupsinCarbonsonHydrogenStoragebysıSupportingInformationSpillover.Ind.Eng.Chem.Res.2009,48,2920−2926.TheSupportingInformationisavailablefreeofchargeat(11)Cheng,H.;Chen,L.;Cooper,A.C.;Sha,X.;Pez,G.P.https://pubs.acs.org/doi/10.1021/acs.jpcc.0c11412.Hydrogenspilloverinthecontextofhydrogenstorageusingsolid-Hückelcalculations,stabilitywithrespecttoDFTstatematerials.EnergyEnviron.Sci.2008,1,338−354.functionalandbasisset,electrondensitydifferences,(12)Wang,L.;Yang,R.T.HydrogenStorageonCarbon-Basedspindensity,additionalreactionpathways,anddifferentAdsorbentsandStorageatAmbientTemperaturebyHydrogenSpillover.Catal.Rev.:Sci.Eng.2010,52,411−461.energycontributions(PDF)(13)Warczinski,L.;Hu,B.;Eckhard,T.;Peng,B.;Muhler,M.;Hättig,C.AnchoringofpalladiumnanoparticlesonN-doped■mesoporouscarbon.Phys.Chem.Chem.Phys.2020,22,21317−AUTHORINFORMATION21325.CorrespondingAuthor(14)Granja-DelRío,A.;Alonso,J.A.;López,M.J.StericandLisaWarczinski−Ruhr-UniversityBochum,44780Bochum,chemicaleffectsonthehydrogenadsorptionanddissociationonfreeGermany;orcid.org/0000-0002-1884-9251;andgraphene-supportedpalladiumclusters.Comput.Theor.Chem.Email:lisa.warczinski@rub.de2017,1107,23−29.(15)López-Corral,I.;Germán,E.;Juan,A.;Volpe,M.A.;Brizuela,AuthorG.P.DFTStudyofHydrogenAdsorptiononPalladiumDecoratedChristofHättig−Ruhr-UniversityBochum,44780Bochum,Graphene.J.Phys.Chem.C2011,115,4315−4323.Germany;orcid.org/0000-0002-5752-2710(16)Ramos-Castillo,C.M.;Reveles,J.U.;Zope,R.R.;deCoss,R.Completecontactinformationisavailableat:PalladiumClustersSupportedonGrapheneMonovacanciesforhttps://pubs.acs.org/10.1021/acs.jpcc.0c11412HydrogenStorage.J.Phys.Chem.C2015,119,8402−8409.(17)Rangel,E.;Sansores,E.;Vallejo,E.;Hernández-Hernández,A.;López-Pérez,P.A.StudyoftheinterplaybetweenN-graphenedefectsNotesandsmallPdclustersforenhancedhydrogenstorageviaaspill-overTheauthorsdeclarenocompetingfinancialinterest.mechanism.Phys.Chem.Chem.Phys.2016,18,33158−33170.(18)Warczinski,L.;Hättig,C.Aquantumchemicalstudyof■ACKNOWLEDGMENTShydrogenadsorptiononcarbon-supportedpalladiumclusters.Phys.ThisworkhasbeensupportedbytheDeutscheForschungs-Chem.Chem.Phys.2019,21,21577−21587.gemeinschaftwithintheframeworkoftheSino-German(19)Lueking,A.D.;Psofogiannakis,G.;Froudakis,G.E.AtomicHydrogenDiffusiononDopedandChemicallyModifiedGraphene.J.initiative“NovelFunctionalMaterialsforSustainableChem-Phys.Chem.C2013,117,6312−6319.istry”throughGrantHa2588/9-1.(20)Sánchez,M.;Ruette,F.Calculationsofadsorption,coadsorption,diffusion,andreactionbarriersofHatomsintheH2■REFERENCESformationonapositivelychargedcoronene.Chem.Phys.Lett.2020,(1)Conner,W.C.;Falconer,J.L.SpilloverinHeterogeneous738,136913.Catalysis.Chem.Rev.1995,95,759−788.(21)Gerber,I.C.;Serp,P.ATheory/ExperienceDescriptionof(2)Prins,R.HydrogenSpillover.FactsandFiction.Chem.Rev.SupportEffectsinCarbon-SupportedCatalysts.Chem.Rev.2020,120,2012,112,2714−2738.1250−1349.9030https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031

11TheJournalofPhysicalChemistryCpubs.acs.org/JPCCArticle(22)Hu,B.;Warczinski,L.;Li,X.;Lu,M.;Bitzer,J.;Heidelmann,(42)Psofogiannakis,G.M.;Froudakis,G.E.DFTStudyoftheM.;Eckhard,T.;Fu,Q.;Schulwitz,J.;Merko,M.;Li,M.;Kleist,W.;HydrogenSpilloverMechanismonPt-DopedGraphite.J.Phys.Chem.Hättig,C.;Muhler,M.;Peng,B.FormicAcid-AssistedSelectiveC2009,113,14908−14915.Hydrogenolysisof5-Hydroxymethylfurfuralto2,5-DimethylfuranoverBifunctionalPdNanoparticlesSupportedonN-DopedMesoporousCarbon.Angew.Chem.,Int.Ed.2021,60,6807−6815.(23)Shamsiev,R.S.;Danilov,F.O.QuantumchemicalstudyofH2adsorptiononPd21cluster.Russ.Chem.Bull.2017,66,395−400.(24)TURBOMOLEversion7.1,2016;version7.2.12017;andversion7.3,2018;adevelopmentofUniversityofKarlsruheandForschungszentrumKarlsruheGmbH,1989−2007,TURBOMOLEGmbH,since2007.https://www.turbomole.com(2016−2018).(25)Tao,J.;Perdew,J.P.;Staroverov,V.N.;Scuseria,G.E.ClimbingtheDensityFunctionalLadder:NonempiricalMeta−GeneralizedGradientApproximationDesignedforMoleculesandSolids.Phys.Rev.Lett.2003,91,146401.(26)Weigend,F.;Ahlrichs,R.Balancedbasissetsofsplitvalence,triplezetavalenceandquadruplezetavalencequalityforHtoRn:Designandassessmentofaccuracy.Phys.Chem.Chem.Phys.2005,7,3297−3305.(27)Andrae,D.;Häußermann,U.;Dolg,M.;Stoll,H.;Preuß,H.Energy-adjustedabinitiopseudopotentialsforthesecondandthirdrowtransitionelements.Theor.Chim.Acta1990,77,123−141.(28)Sierka,M.;Hogekamp,A.;Ahlrichs,R.FastevaluationoftheCoulombpotentialforelectrondensitiesusingmultipoleacceleratedresolutionofidentityapproximation.J.Chem.Phys.2003,118,9136−9148.(29)Weigend,F.AccurateCoulomb-fittingbasissetsforHtoRn.Phys.Chem.Chem.Phys.2006,8,1057−1065.(30)Grimme,S.;Antony,J.;Ehrlich,S.;Krieg,H.Aconsistentandaccurateabinitioparametrizationofdensityfunctionaldispersioncorrection(DFT-D)forthe94elementsH-Pu.J.Chem.Phys.2010,132,154104.(31)Koßmann,J.;Hättig,C.InvestigationofinterstitialhydrogenandrelateddefectsinZnO.Phys.Chem.Chem.Phys.2012,14,16392−16399.(32)Stodt,D.;Noei,H.;Hättig,C.;Wang,Y.AcombinedexperimentalandcomputationalstudyontheadsorptionandreactionsofNOonrutileTiO2.Phys.Chem.Chem.Phys.2013,15,466−472.(33)Plessow,P.ReactionPathOptimizationwithoutNEBSpringsorInterpolationAlgorithms.J.Chem.TheoryComput.2013,9,1305−1310.(34)Helgaker,T.Transition-stateoptimizationsbytrust-regionimageminimization.Chem.Phys.Lett.1991,182,503−510.(35)Fukui,K.Thepathofchemicalreactions-theIRCapproach.Acc.Chem.Res.1981,14,363−368.(36)Kesharwani,M.K.;Brauer,B.;Martin,J.M.L.FrequencyandZero-PointVibrationalEnergyScaleFactorsforDouble-HybridDensityFunctionals(andOtherSelectedMethods):CanAnharmonicForceFieldsBeAvoided?J.Phys.Chem.A2015,119,1701−1714.(37)Knizia,G.IntrinsicAtomicOrbitals:AnUnbiasedBridgebetweenQuantumTheoryandChemicalConcepts.J.Chem.TheoryComput.2013,9,4834−4843.(38)Jewell,L.L.;Davis,B.H.Reviewofabsorptionandadsorptioninthehydrogen−palladiumsystem.Appl.Catal.,A2006,310,1−15.(39)Blanco-Rey,M.;Juaristi,J.I.;Alducin,M.;López,M.J.;Alonso,J.A.IsSpilloverRelevantforHydrogenAdsorptionandStorageinPorousCarbonsDopedwithPalladiumNanoparticles?J.Phys.Chem.C2016,120,17357−17364.(40)González-Herrero,H.;Cortes-delRío,E.C.;Mallet,P.;Veuillen,J.-Y.;Palacios,J.J.;Gómez-Rodríguez,J.M.;Brihuega,I.;Ynduráin,F.Hydrogenphysisorptionchannelongraphene:ahighwayforatomicHdiffusion.2DMater.2019,6,021004.(41)Gokhale,A.A.;Kandoi,S.;Greeley,J.P.;Mavrikakis,M.;Dumesic,J.A.Molecular-leveldescriptionsofsurfacechemistryinkineticmodelsusingdensityfunctionaltheory.Chem.Eng.Sci.2004,59,4679−4691.9031https://doi.org/10.1021/acs.jpcc.0c11412J.Phys.Chem.C2021,125,9020−9031