1

SupportingInformation

UltrafastBidirectionalChargeTransportandElectronDecoherenceAt

Molecule/SurfacesInterfaces:AComparisonofGold,Grapheneand

GrapheneNanoribbonSurfaces

OlgunAdak1,GregorKladnik2,3,GregorBavdek4,AlbanoCossaro5,AlbertoMorgante*3,5,Dean

Cvetko2,5*,LathaVenkataraman1,6*

1DepartmentofAppliedPhysicsandAppliedMathematics,ColumbiaUniversity,NewYork,NY

2FacultyofMathematicsandPhysics,UniversityofLjubljana,Ljubljana,Slovenia

3DepartmentofPhysics,UniversityofTrieste,Trieste,Italy

4FacultyofEducationUniversityofLjubljana,Ljubljana,Slovenia

5CNR-IOMLaboratorioNazionaleTASC,BasovizzaSS-14,km163.5,I-34012Trieste,Italy

6DepartmentofChemistry,ColumbiaUniversity

email:cvetko@iom.cnr.it,lv2117@columbia.edu,morgante@iom.cnr.it

Contents:

1. MeasurementDetailsandSamplePreparation

2. XPSandNEXAFSMeasurements

3. DFTCalculationsforNEXAFSspectra

4. TheRPESMeasurements

5. TheRPESCore-Hole-ClockMethodwithChargeTransferfromtheSubstrate

6. References

2

1.MeasurementDetailsandSamplePreparation

TheexperimentswereconductedattheALOISAbeamlineoftheElettraSynchrotroninTrieste.

Themeasurementchamberwasmaintainedatanultrahigh-vacuumatpressuresof10−10−10−11

mbar.1-3The4,4’-bipyridine (BP)moleculeswerepurchased fromSigma-Aldrich (purity>99%)

andusedwithoutfurtherpurification.ThemoleculeswereplacedinaPyrexcellandconnected

tothepre-chamberthroughaleakvalve.

TheAu(111)surfacewaspreparedbycyclesofAr+sputteringandannealingat400K.X-

rayphotoemissionspectroscopy(XPS)wasusedtocheckforanychemicalimpurities(O,N,and

C). The BP cellwas heated to several tens of degrees above room temperature and BPwas

leakedintoapre-chambertomaintainapartialBPpressureof10−7mbarfor4-5minutes.The

samplewascooledto-65°Cforthemultilayerphase,anditwaskeptatroomtemperaturefor

themonolayerphase.

TheepitaxialgrapheneonNi(111)surfacewaspreparedbyrepeatedlysputtering(Ar+,

2keV)andannealing(900K)theNi(111)surface.XPSandHeliumatomscatteringwereusedto

check for the presence of any chemical impurities and to probe the surface order. Epitaxial

graphenewaspreparedvia catalyticdissociationofethyleneonNi(111) surface.Anethylene

partialpressureof10-6mbarwasmaintainedfor60minuteswhilekeepingthesurfaceat850K.4

ThegraphenefilmwasprobedusingXPSandultravioletphotoemissionspectroscopywithHeII

lineat40.8eV.Examinationofthegrapheneπ-bandbottom,closetoΓ,wasusedtoidentifythe

graphenelayerasepitaxialgraphene.4Theepitaxialgraphene/Ni(111)samplewasexposedto

ambientconditionsbeforebeingtransferredtotheALOISAchamber.Thesamplewasannealed

3

in the ALOISA chamber at 500 K to recover the pristine graphene. XPS, valance band

photoemission spectroscopy and near edge X-ray absorption fine structure (NEXAFS)

spectroscopy measurements were further performed to characterize the epitaxial

graphene/Ni(111)sample.TocreatetheBPself-limitingmonolayerfilm,apartialBPpressureof

10−7mbarwasmaintainedinthepre-chamberfor3minutes,whilethesamplewascooledto-

50°C.

Graphene nanoribbons (GNR) were formed depositing 10,10’-dibromo-9,9’-biantryl

(AOKBIO,98+%purity)onthecleanedAu(111)samplemaintainedat200Ctoensuresaturating

theAu(111)surfacewiththeGNRprecursor.Thissurfacewasthenannealedto400Ctocreate

GNRsasdescribedpreviously.5XPSandNEXAFSmeasurementswereperformedtocharacterize

theGNRfilmontheAu(111)surface.TocreatetheBPmonolayerfilm,apartialBPpressureof

10−7mbarwasmaintainedinthechamberfor5minuteswhilethesamplewascooledto-45°C.

2. X-ray Photoemission Spectroscopy and Near Edge X-ray Absorption Fine Structure

SpectroscopyMeasurements

X-ray photoemission spectroscopy (XPS) measurements were performed at the ALOISA

beamlinewiththeX-raybeamatgrazing-incidence(4°)tothesample.Photoelectronsfromthe

samplewere collected at the normal to the surface using a hemispherical electron analyzer

withanacceptanceangleof2°,andanoverallenergyresolutionof~0.2eV.Theenergyscale

forXPSspectrawascalibratedbyaligningtheAu4f7/2peaktoabindingenergyof84.0eVfor

theAu(111) andGNRonAu(111)measurements. For the epitaxial grapheneonNi(111), XPS

spectrawerealignedtotheFermilevel.

4

NearedgeX-rayabsorptionfinestructure(NEXAFS)measurementswereperformedon

thenitrogenK-edgebysweepingtheincidentphotonenergyfrom396eVto420eVinstepsof

0.1eV.Thephoton incidenceanglewasset to6°.Spectrawereacquiredusingachanneltron

detectorwithawideacceptanceangleinthepartialelectronyieldmodeandahighpassfilter

withcutoffenergyset to370eV.Thephotonfluxwasmonitoredonthe lastopticalelement

alongthebeampath.Thesamplenormalwasorientedeitherparallel(p-pol)orperpendicular

(s-pol)tothelightpolarization(s-pol).

The relative intensity of the NEXAFS signal in s-pol and p-pol for the N1s to LUMO

transitionwasusedtoobtaintheorientationofthearomaticringrelativetothesurface.The

angleθof the ring to thesurface isdeterminedas tan ! = 2!!/!!where Is and Ip are the

intensitiesoftheLUMONEXAFSpeakfromthes-polandp-polspectrarespectively.

3.DFTCalculationsforNEXAFSSpectra

We calculate the nitrogen K-edge spectra using GPAW, a grid-based real-space projector-

augmented-wave (PAW) code, with the BLYP exchange-correlation functional6, 7. For these

simulations, isolatedmoleculeswere first relaxed to theiroptimizedgeometries.Default grid

spacingsandconvergencethresholdswereemployed.AllNEXAFScalculationswereperformed

using the half-core-hole approximation8. The absolute energy scale was determined by

performing a delta Kohn-Sham calculation and shifting the calculated spectrum using the

calculatedtotalenergydifferencebetweenthegroundstateandthefirstcoreexcitedstate.

5

SI Figure 1. The measured magic-angle N K-edge NEXAFS spectrum (blue) for bipyridine onAu(111) compared with the calculated absorption spectra (dashed red) in the transitionpotential approximation. The calculated peaks with the highest transition probabilities areindicatedandassignedaccordingly.

4.TheRPESMeasurements

TheRPESspectraatthenitrogenK-edgewereobtainedbytakingXPSscansspanning0-60eV

bindingenergyrangewiththephotonenergytunedbetween394eVand423eVinstepsof0.2

eVand0.5eVforallsurfacesstudied.RPESmeasurementsforchargetransfertimecalculations

wereperformedinmagicangleconditions,withthelightpolarizationandtheelectronanalyzer

at54.7°withrespecttothesurfacenormal.ThisyieldedaRPESsignalthatwasindependentof

themolecular orientation on the sample. For the angle-dependent RPESmeasurements, the

anglebetweenthephotonpolarizationandsurfacenormalsettothevaluesstatedinthemain

textwhiletheelectronanalyzerwasalwayscollinearwiththephotonpolarization.TheseRPES

datawerethennormalizedandanalyzedfollowingpreviouslypublishedprocedures.9,10Briefly,

6

thephoton fluxon the lastopticalelementalong thebeampathwasmeasuredandused to

normalizeandcalibratetheRPESscansforanyfluctuationsinphotonintensityandenergy.The

energy scale for XPS spectrum was calibrated as detailed above. The non-resonant

photoemissionspectrumwasobtained fromtheXPSscanat395eVandsubtracted fromthe

entire RPES spectra. The RPES line scanswere then further normalized by the overall Auger

intensitywhichscaleswiththeabsorptioncross-sectionatagivenphotonenergy.

SIFigure2.NitrogenK-edgeRPESmapsof(A)BPmultilayeronAu(111)(B)BPmonolayeronAu(111),(C)BPmonolayeronepitaxialgrapheneand(D)BPmonolayeronGNR.TheLUMO*resonancescansaretakenalongthedashedblacklines.Thedashedorangelinesindicatetheenergyusedforthescansabovetheionizationedge.

7

InSIFigure2,weshownitrogenK-edgeRPESmapsofaBPmultilayeronAu(111)anda

BPmonolayer onAu(111), epitaxial graphene andGNR.We see theN1s to LUMO transition

signaturearoundphotonenergyof399eV.Fromthesemaps,weobtainthelinescansthatare

showninFigure3ofthemaintext.Thecore-holeclockmethoddescribedbelowisthenusedto

determinethechargetransfertimeandthefractionoftheLUMO*thatdropsbelowtheFermi

levelinthemonolayersystems.

5.TheRPESCore-Hole-ClockMethodwithChargeTransferfromtheSubstrate

Inthissection,wepresenttheextensionofthestandardcore-holeclockmethod11toobtainthe

bidirectionalchargetransfer timeandthe fractionof theLUMO*thatdropsbelowtheFermi

levelfromtheRPESmeasurements.

Inthecore-holeclockmethod,acoreelectron isphotoexcitedtotheLUMO, leavinga

core-holeonthemolecule;andtheenergydistributionofthesubsequentelectronemissionis

measured.Theprocessrelevanttothecore-holeclockmethodisthefillingofthecore-holeand

the subsequent emission of an electron, leaving the LUMO empty; this is denoted as the

participator decay. In a coupled system, the participator decay gets quenched if the charge

transfer to the substrate occurs within the core-hole life-time. Therefore, by comparing the

participatordecayintensityintheisolatedandthecoupledsystem,thechargetransfertimein

the coupled system can be determined as has been done before.11 In the measurements

presentedhere,whenacore-holeispresent,theLUMO*ispartiallybelowtheFermileveland

canthusgetoccupiedduetoachargetransferfromthesubstrate.

8

Todeterminethebi-directionalchargetransfertime,oneneedstoconsidertwotypesof

decay process that will contribute to the participator decay intensity. First, there is a

participatordecay thatoccurs fromtheLUMO*electron.Additionally, there isanAuger type

participatordecay (fromthe fractionof theLUMO*that isbelowFermi) thatcanoccurafter

chargetransferfromtheLUMO*tothesubstrate.Denotingtheparticipatordecayintensityas

IointheisolatedsystemandasIcinthecoupledsystem,wecangetarelationbetweenIoandIc

as:

CT CHc o o

CH CT CH CT

I I I xτ ττ τ τ τ

= ++ +

(1)

where τCH is the core-hole lifetime, τCT is the charge transfer time, x is the fraction of the

LUMO*belowtheFermilevel.ThefirstterminEq.1isthestandardexpressionthathasbeen

usedbeforeandisattributedtothephotoemissiontypeparticipatordecay.11Thesecondterm

isdue to theAuger typeparticipatordecay thatcanoccuronly if theLUMO* ispartlybelow

FermiandchargetransferfromtheLUMO*tothesubstratehasoccurred.Theprocessgiving

risetothesecondterminEq.1isidenticaltotheAugertypeparticipatordecayoccurringafter

the charge transfer from the substrate to LUMO* following an excitation of a core electron

above the ionization edge. We denote this process as the super-participator decay and its

intensity(Is)isgivenby

CHs o

CH BT

I I x ττ τ

=+

. (2)

Here,τBTisthebi-directionalchargetransfertime.

9

Sincethechargetransfertimeinbothdirectionsisequal,Eq.1and2canbesolvedto

findxandτCTwhichyields

(1 )1 (1 )CT CHff

βτ τ

β−

=− −

. (3)

1 (1 )

fx

fββ

=− −

(4)

wherefisIc/IoandβisIs/Ic.

Theequationsabovearevalidwhenthelightpolarizationisperpendiculartothenodal

planeofthenodalplaneofthemolecularπ-system.Therefore,weneedtoaccountforthefact

that the molecular π-system makes a small angle with the substrate, as determined from

NEXAFS measurements presented in the main text. We also need to consider the light

polarizationwhencomparingtheAugertypesuper-participatorintensitieswithphotoemission

typeparticipatorintensitysincethereisaclearangulardependenceinthelatter.

Thephotoemissiontypeparticipatordecayintensityisproportionalto(ε·n)2whereεis

thepolarizationunitvectorandn isthevectornormaltothenodalplaneofthemolecularπ-

system.Wecalculate the (ε·n)2 as a functionofε andn assuming that there isnoazimuthal

dependenceonthemolecularorientationonthesurface.Weuse

sin( ) cos( )x zε θ θ∧ ∧

= + (5)

n = sinα cos(φ −π / 2) x∧

+ sinα sin(φ −π / 2) y∧

+ cosα z∧

(6)

10

whereθisthepolaranglebetweenthesurfacenormalandthepolarizationunitvector,αisthe

polaranglebetweenthesurfacenormalandthenormaltothemolecularπ-system,andφisthe

azimuthalangleofthemolecularπ-system.

Averaging(ε·n)2overφ,andusingEq.5andEq.6weget:

22 2

2 2sin sin( ) co· s cos

2n θ α

θε α= + (7)

Since measurements are made at the magic angle (θ =54.7°), cos2θ=sin2θ/2, and thus the

average(ε·n)2isindependentofα.Weobtain:

Icm = Io

m τCTτCH +τCT

+Iom

cos2θx

τCHτCH +τCT

(8)

Ism =

Iom

cos2θx

τCHτCH +τ BT

(9)

where mcI , m

sI and moI arethemeasuredintensitiesatthemagicangle.SolvingEq.8andEq.9

yields,

(1 )

1 (1 )CT CHff

βτ τ

β−

=− −

(10)

x = cos2θ f β1− f (1−β)

. (11)

wherefis mcI / m

oI andβis msI / m

cI .

11

6.References

1. Floreano,L.;Cossaro,A.;Gotter,R.;Verdini,A.;Bavdek,G.;Evangelista,F.;Ruocco,A.;Morgante,A.;Cvetko,D.J.Phys.Chem.C,2008,112,(29),10794-10802.2. Cvetko,D.;Lausi,A.;Morgante,A.;Tommasini,F.;Prince,K.C.;Sastry,M.MeasurementScience&Technology,1992,3,(10),997-1000.3. Vilmercati,P.;Cvetko,D.;Cossaro,A.;Morgante,A.Surf.Sci.,2009,603,(10-12),1542-1556.4. Patera,L.L.;Africh,C.;Weatherup,R.S.;Blume,R.;Bhardwaj,S.;Castellarin-Cudia,C.;Knop-Gericke,A.;Schloegl,R.;Comelli,G.;Hofmann,S.;Cepek,C.AcsNano,2013,7,(9),7901-7912.5. Batra,A.;Cvetko,D.;Kladnik,G.;Adak,O.;Cardoso,C.;Ferretti,A.;Prezzi,D.;Molinari,E.;Morgante,A.;Venkataraman,L.ChemicalScience,2014,5,4419.6. Enkovaara,J.;Rostgaard,C.;Mortensen,J.J.;Chen,J.;Dulak,M.;Ferrighi,L.;Gavnholt,J.;Glinsvad,C.;Haikola,V.;Hansen,H.A.;Kristoffersen,H.H.;Kuisma,M.;Larsen,A.H.;Lehtovaara,L.;Ljungberg,M.;Lopez-Acevedo,O.;Moses,P.G.;Ojanen,J.;Olsen,T.;Petzold,V.;Romero,N.A.;Stausholm-Moller,J.;Strange,M.;Tritsaris,G.A.;Vanin,M.;Walter,M.;Hammer,B.;Hakkinen,H.;Madsen,G.K.H.;Nieminen,R.M.;Norskov,J.;Puska,M.;Rantala,T.T.;Schiotz,J.;Thygesen,K.S.;Jacobsen,K.W.J.Phys.:Cond.Mat.,2010,22,(25).7. Ljungberg,M.P.;Mortensen,J.J.;Pettersson,L.G.M.JournalofElectronSpectroscopyandRelatedPhenomena,2011,184,(8-10),427-439.8. Cavalleri,M.;Odelius,M.;Nordlund,D.;Nilsson,A.;Pettersson,L.G.M.Phys.Chem.Chem.Phys.,2005,7,(15),2854-2858.9. Batra,A.;Kladnik,G.;Vazquez,H.;Meisner,J.S.;Floreano,L.;Nuckolls,C.;Cvetko,D.;Morgante,A.;Venkataraman,L.Nat.Commun.,2012,3,1086.10. Kladnik,G.;Cvetko,D.;Batra,A.;Dell'Angela,M.;Cossaro,A.;Kamenetska,M.;Venkataraman,L.;Morgante,A.J.Phys.Chem.C,2013,117,(32),16477-16482.11. Bruhwiler,P.A.;Karis,O.;Mårtensson,N.Rev.Mod.Phys.,2002,74,(3),703-740.