227 Pages
English
Gain access to the library to view online
Learn more

Conformations and fragmentation of biologically relevant molecules and their binary complexes with water probed by high resolution UV and mass analyzed threshold ionization spectroscopy [Elektronische Ressource] / Rosen Karaminkov

Gain access to the library to view online
Learn more
227 Pages
English

Description

Technische Universitat¨ Munchen¨Physikalische ChemieConformations and Fragmentation of Biologically RelevantMolecules and their Binary Complexes with Water Probed byHigh Resolution UV and Mass Analyzed Threshold IonizationSpectroscopyRosen KaraminkovVollstandiger¨ Abdruck der vonder Fakultat¨ fur¨ Chemieder Technischen Universitat¨ Munchen¨zur Erlangung des akademischen Grades einesDoktors der Naturwissenschaftengenehmigten Dissertation.Vorsitzender: Univ. Prof. Dr. St. J. GlaserPrufer¨ der Dissertation:1. Univ. Prof. Dr. H. J. Neusser2. Univ Dr. Dr. h. c. A. LaubereauDie Dissertation wurde am 09.11.2009 bei der Technischen Universitat¨ Munchen¨ eingereicht unddurch die Fakultat¨ fur¨ Chemie am 07.12.2009 angenommen.2To my lovely daughters,Vanessa and Plamena,and to my beloved wife,Evelina4PrefaceSince the dawn of humanity the permanent desideration for knowledge and understanding ofthe surrounding world has been the driving force that evolved the ancient studies and promotedthe science over the ages to its modern thrive with its complexity and variety of research fields.Nowadays the macroscopic properties and functionality of chemical and biological systems havebeen ultimately related to the microscopic properties of the their constituents thus giving riseto research areas such as spectroscopy, microbiology, nanoscience, etc., with their realms ofinterest, which are closely interwoven.

Subjects

Informations

Published by
Published 01 January 2009
Reads 11
Language English
Document size 14 MB

Exrait

TPhechnischeysikalischeUniversit¨atChemieM¨unchen
ConformationsandFragmentationofBiologicallyRelevant
HighMoleculesResolutionandtheirUVandBinaryMassComplexesAnalyzedwithThrWesholdaterProbedIonizationby
oscopySpectr

orsitzender:V

Dissertation:derufer¨Pr

voKaraminkRosen

Vollst¨andigerAbdruckdervon
derFakult¨atf¨urChemie
derTechnischenUniversit¨atM¨unchen
zurErlangungdesakademischenGradeseines
DoktorsderNaturwissenschaften
Dissertation.genehmigten

Univ.-Prof.Dr.St.J.Glaser

1.Univ.-Prof.Dr.H.J.Neusser
2.Univ.-Prof.Dr.Dr.h.c.A.Laubereau

DieDissertationwurdeam09.11.2009beiderTechnischenUniversit¨atM¨uncheneingereichtund
durchdieFakult¨atf¨urChemieam07.12.2009angenommen.

2

Tomylovelydaughters,

Plamena,andanessaV

and

to

edvbelomy

elinaEv

wife,

4

efacePr

Sincethedawnofhumanitythepermanentdesiderationforknowledgeandunderstandingof
thesurroundingworldhasbeenthedrivingforcethatevolvedtheancientstudiesandpromoted
thescienceovertheagestoitsmodernthrivewithitscomplexityandvarietyofresearchfields.
Nowadaysthemacroscopicpropertiesandfunctionalityofchemicalandbiologicalsystemshave
beenultimatelyrelatedtothemicroscopicpropertiesofthetheirconstituentsthusgivingrise
toresearchareassuchasspectroscopy,microbiology,nanoscience,etc.,withtheirrealmsof
interest,whicharecloselyinterwoven.Aniceexampleofcollaborationbetweentherapidlyde-
velopingbranchesofscienceistheconnectionbetweenstereochemistryandspectroscopy.While
theformerdealswithspatialarrangementsofatomsinmoleculesandtheeffectsofthesearrange-
mentsonthephysicalandchemicalpropertiesofsubstances,spectroscopyprovidesinsightinto
thestructureandbondingofthemostabundantisomersofachemicalcompound.
Theuniqueconditionsobtainedduringthesupersonicexpansionofmoleculesfromhigherpres-
sureintovacuumresultsintheformationofcoldmoleculesandsmallcomplexeswhichsome-
timeshavebeenreferredtoasthoseofafourthstateofmatter.Furthermore,inmolecularbeams,
thespeciesconcernedcanbeinvestigatedunderconditionsthatremovedpressurebroadening
andmuchofDopplerbroadeningfromtheresultingspectra.Whenmoleculesareinjectedinto
thecarriergas,theyattainverylowrotationaltemperatures,highervibrationaltemperaturesof
theorderof100Kandtranslationaltemperaturesdownto0.1K.Therefore,theso-calledfourth
stateofmatterconsistsofmoleculeswhicharegenerallycoldandwhichhavedifferenttrans-
lational,rotationalandvibrationaltemperatures.Theseconditionsareanexcellentenvironment
formolecularspectroscopyandresultinseveralmajoradvantages.First,weaklyboundspecies,
suchasvanderWaalsandhydrogen-bondedcomplexesandclusters,areheldtogetheratthe
typicallylowvibrationaltemperature,whichallowsahighprecisionspectroscopicstudiesand
secondadvantageisthatspectraoflargemoleculescanberotationallyresolved.Thedevel-
opmentoflasersandtherefinementofthemassresolutiontechniqueshavebeenprovedtobe
anextremelyimportanttoolininvestigationofthespectroscopy,conformations,fragmentation
anddynamicsofmoleculesinsupersonicjetsorpulsedbeams.Themassanalyzedthreshold
ionizationandresonanceenhancedtwophotonionizationtechniques,usedinthiswork,have
demonstratedtheabilitytoobtainbeautifulelectronicspectraevenfortheverylowconcentra-

5

6

thetionslasertypicalresolutionofsobiologicalhighthespeciesspectraincoldofevensupersonicquitejets.complexorBecauseganicthemoleculessamplesarebecomesocoldsimpleand
genenoughbondstoandanalyzehinydrogenbondeddetail.Conformerscomplexesstabilizedareoftroughinteresttheparticularlyformationofinintrabiochemicalmolecularhsystemsydro-
andreactions.alsoforItaisvarietysignificantofotherthatinreasons;thepastfore30xample,years,atheydetailedmaybepictureintermediatesofhydrogenininterbondingmolecularhas
mergedfromhigh-resolutionandmassselectivespectroscopyinvestigationsofweaklybounded
es.xcomple

Contents

Preface

oductionIntrI

3

13

15oductionIntr11.1Molecularconformers................................15
1.2Hydratedcomplexes................................17
1.3Photofragmentationofbiomolecules........................18

DetailsExperimentalII

21

23echniquesTExperimental22.1ResonanceEnhancedTwo-PhotonIonization(R2PI)UVLaserSpectroscopy..24
2.2R2PIExperimentalSetup.............................25
2.2.1Molecularbeamapparatus.........................25
2.2.2LaserSystem................................28
2.2.3ControloftheExperiment.DataAcquisitionSystem...........32
2.3GeneticAlgorithm.................................32
2.4Mass-AnalyzedThresholdIonization(MATI)...................34
2.4.1MATIExperimentalSetup.........................35

DiscussionandResultsIII

37

3ConformationalStructuresof2-PhenylethanolanditsHydratedComplex39
3.1Introduction.....................................40
3.2ExperimentandDataProcessing..........................41
3.3ExperimentalResults................................42

7

8

4

5

6

CONTENTS

3.3.1LowResolutionSpectrum.........................42
3.3.2HighResolutionSpectra..........................43
3.3.3MonomerBands..............................43
3.3.4WaterComplexBands...........................46
3.4AbinitioCalculations................................48
3.4.12-PEMonomer...............................48
23.4.22-PE∙HOComplex.............................51
3.5Discussion......................................53
3.5.1Monomerconformers............................53
3.5.22-PE∙HOConformer............................56
23.6SummaryandConclusions.............................59

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol61
4.1Introduction.....................................62
4.2ExperimentandDataProcessing..........................63
4.3ExperimentalResults................................64
4.3.1LowResolutionSpectrum.........................64
4.3.2HighResolutionSpectra..........................65
4.3.3AbinitioCalculations............................69
4.4Discussion......................................70
4.5SummaryandConclusions.............................75

MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol77
5.1Introduction.....................................79
5.2Experimental....................................79
5.3Results........................................80
5.3.1REMPIspectra...............................80
5.3.2MATIspectra................................80
5.3.3Computationalresults...........................83
5.4Discussion......................................86
5.5Summaryandconclusions.............................90

ConformationalProbingof2-ortho-uorophenylethanol93
6.1Introduction.....................................94
6.2Experiment.....................................95
6.3Results........................................96
6.3.1LowResolutionSpectrum.........................96
6.3.2HighResolutionSpecta..........................97
6.3.3AbinitioCalculations............................102

CONTENTS

9

6.4Discussion......................................105
6.5Summaryandconclusions.............................110

7WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol113
7.1Introduction.....................................114
7.2ExperimentandDataProcessing..........................114
7.3ResultsandDiscussion...............................115
7.3.12-pFPE-WaterCluster...........................115
7.3.22-oFPE-WaterCluster...........................124
7.4SummaryandConclusions.............................130

8R2PIMassSpectroscopyofEphedrine133
8.1Introduction.....................................134
8.2Experiment.....................................134
8.3Results........................................135
8.3.1Abinitiocalculations............................135
8.3.2Experimentalresults............................136
8.4Discussion......................................138
018.4.1Benzene-likeS←Sspectrum.......................138
8.4.2Fragmentationpathways..........................139
8.4.3State-selectivefragmentation........................141
8.5Summaryandconclusions.............................143

9Pseudoephedrine:EvidenceforConformerSpecificFragmentation145
9.1Introduction.....................................146
9.2ExperimentandDataProcessing..........................146
9.3Results........................................147
9.3.1AbinitioCalculations............................147
9.3.2Low-ResolutionSpectra..........................147
9.3.3High-ResolutionSpectra..........................148
9.4Discussion......................................155
9.5SummaryandConclusions.............................156

ConcludingIVemarksR

ppendicesA

159

163

167InteractionsolecularMAA.1ChemicalBonds...................................167

10

CONTENTS

A.2WeakMolecularInteractions............................168
A.2.1AttractiveMolecularInteractions.....................168
A.2.2RepulsiveMolecularInteractions.....................170
A.3TheConceptofWeakMolecularBonds......................170
A.4ClassificationofWeakMolecularBonds......................172
A.4.1HydrogenBonds..............................172
A.4.2Quadrupole-QuadrupoleBonds......................175
A.4.3Charge-TransferBonds...........................175
A.4.4Ion-MediatedBonds............................175
A.4.5HydrophobicInteractions..........................175
A.4.6DispersionBonds..............................175
A.5BindingMotifs...................................176

BAbInitioQuantumChemistryCalculations177
B.1SemiempiricalMethods...............................177
B.2AbInitioQuantumChemistryCalculations.....................178
B.2.1TheMøller-PlessetPerturbationTheory..................178
B.2.2TheCoupled-ClusterMethod.......................179
B.3Density-FunctionalTheory.............................180
B.4MolecularMechanics................................182
B.5MolecularGeometry................................182
B.5.1EquilibriumGeometry...........................182
B.5.2PotentialEnergySurface..........................182
B.5.3GeometryOptimizations..........................183
B.5.4ConformationalSearch...........................183
B.6BSSE........................................184
B.7MolecularVibrationalFrequencies.........................186
B.8ProgrammePackagesforTheoreticalMolecularInvestigations..........187

CInteractionbetweenLightandMatter

189

DMolecularQuantumMechanicalCharacteristicsandMolecularStructure193
D.1Born-OppenheimerApproximation.........................193
D.2RotationofMolecules................................194
D.2.1MolecularRotationalEnergy........................195
D.2.2MolecularRotationalEnergyLevels....................196
D.2.3ElectricDipoleTransitionMoment....................199
D.2.4SelectionRules...............................199
D.2.5TransitionIntensity.............................199

CONTENTS

D.2.6DeviationsfromtheBorn-OppenheimerApproximation
D.3RotationalConstantsandStructureofMolecularSpecies....

Appendices

yBibliograph

esFigurofList

ListablesTof

PublicationsofList

PosterandOralPresentations

..

..

..

..

..

..

..

..

11

200.200.

167

204

215

219

223

225

12

CONTENTS

Part

I

oductionIntr

13

1Chapter

oductionIntr

ormersconfMolecular1.1

Thepositionsoftheatomsdefinethemoleculargeometry,whichisanimportantpropertyof
moleculesandisdirectlyconnectedtotheirreactivity.Historically,thisledtotheestablishment
ofthestereochemicalprinciplesconnectingmolecularshapewithfunctionality.Animportant
contributionwasmadebytheFrenchphysicistBiot(1815)whodiscoveredtherotationofthe
planeofpolarizationoflinearlypolarizedlightbycertainorganicmoleculesinsolutionorin
gaseousphase,aphenomenonreferredtoasopticalactivity.Thiseventwasfollowedbyworks
ofLouisPasteur(1848),whoascribedthispropertytothepresenceofsomeasymmetricgroup-
ingofatomsinamolecule,andlateronvantHoffandLeBel(1874)addedathirddimension
tothetwo-dimensionalchemistryofearlierdays.Itwasnottilltheearlynineteen-fiftieswhen
BartonandHaselintroducedtheconceptofconformationalanalysis.Thisconceptaddedan-
otherdimension,atime-dependent(temporal)onetothethree-dimensionalstereochemistryand
extendeditsscopetoreactionprocesses.Thatgivesamorerealisticdescriptionofthemolecules,
whichexhibitintramolecularmotionssuchasdeformationofbondangles,bondstretchingand
contraction,androtationaboutoneormoresinglebonds.Asaconsequence,achemicalcom-
poundexistsinadynamicequilibriumwithanumberofcontinuouslychangingenergy-preferred
conformationswhichdifferfromeachotherinthedegreeofrotationaboutoneormoresingle
bonds(dihedralangles)andinsomecasesalsoinbondanglesandbondlengths.Theyaredis-
tinctmolecularspeciesseparatedbyenergybarriersandarecalledconformersorisomers(from
Greeki¯somere¯s,havingequalshare:i¯so-,iso-+meros,part,share).Althoughincapableofin-
dependentexistenceatroomtemperatures,theyaredetectableatlowtemperaturesandgaseous
phasebycertainphysicalmethods.Dependingontheirrelativepopulationandenergybarriers
separatingthem,theconformersinfluencethephysicalandchemicalpropertiesofacompound.
Theimportanceofthemoleculesconformericstructuresinvolvedinchemicalreactionseither

15

16

oductionIntr

asreactantsorproductshasbeendiscussedinmodernchemistrytextbooks.F.A.Careyand
R.J.Sundbergintheirorganicchemistrytextbook[1]showexplicitlytherelationshipbetween
reactivityandstereochemistryintroducingtermssuchasstereoselectivityandstereoscpecifity.A
chemicalstereoselectivereactionisoneinwhichasinglereactantformstwoormoreconformeric
products.Whentwoormoreconformericreactantseachprovidestereoisomericproducts,there-
actioniscalledstereospecific.AnexampleforstereospecificreactionsistheSN2substitution
thatresultsintheinversionoftheconfiguration.TheR-reactantgivestheS-productandthe
S-reactantgivestheR-product.AnotherexampleistheepoxidationofE-2-butenethatgives
trans-2,3-dimethyloxirane,whereasZ-2-butenegivescis-2,3-dimethyloxirane.Thisandother

Figure1.1:EpoxidationofE-2-butene

examples[1]manifesttheparamountimportanceofthemoleculararchitectureandtheneedofa
reliableassignmentoftheconformericstructures.
Itiswellknownthatneurotransmittersarethechemicalbridgingstructuresthattransferinforma-
tionbetweenneighboringnervecells[2].Theregionthroughwhichtheneurotransmittermust
passfromonecelltotheotheriscalledthesynapticgap.Suitableneurotransmitterspassthis
wayinmillisecondsandtransfertheirinformationbybindingtothehostcell.Thishighlyse-
lectiveprocessisexplainedbythekey-lock-principle[3].Itisunclearwhetheronlycertain
conformersoftheseneurotransmittersareabletobindtothedockingsiteortheconformationis
adaptedduringdocking.Inordertounderstandtheseprocessesinmoredetailitwouldbenec-
essarytoknowtheconformationallandscapeoftheneurotransmitters.Itisplausibletoassume,
thathowever,theconformationalpreferencesareinfluencedfromthemolecularenvironmentand
solvation,butknowingtheconformationallandscapesoftheisolatedneurotransmitterswouldbe
afirststep.Asecondstepwouldthenbetheinvestigationofsmallcomplexeswithsolvent
molecules.Thisinformationcanbeobtainedfrominvestigationsofcold,isolatedmolecules.
Towardsthisend,thepresentworkdescribesnewresultsonmodelbiologicallyrelevantflexible
moleculesandtheirhydratedclusters,obtainedthroughthecombinationofhigh-resolutionUV
spectroscopyandhigh-levelabinitiocalculations.Theseresultselucidatethepreferableconfor-
mationsofisolatedmolecularspeciesinthegasphase,stabilizedbyweakinteractions,andmay

1.2complexesHydrated

serveasaninputinformationforachemicalsynthesisofnovelcompounds.

complexesHydrated1.2

17

Oncethemonomersconformericlandscapeisdetermined,thenextlevelofevaluationofthe
molecularshapeofflexiblemoleculesistoinvestigatetheirbehaviorinarealisticenvironment.
Water,themostabundantnaturessolventhasattractedalotofattentioninthelasttwodecades.
ThestrengthofwaterasasolventliesinitsabilitytoformH-bondednetworksthatstrongly
influencethepropertiesofthesolutes.Whenasolutemoleculepossessestwohydrogenbond-
ingsites:donorandacceptor,thewatermayselectoneofthemorfillinthegapbetweenthem
formingabridgestructure.Intheprocess,theshapeofthehostmoleculecanbevastlymod-
ified.Additionally,aproductionandastabilizationofahigher-energyconformercanbeob-
served[4].Smallclustersofwatermoleculesinthegasphaseareformedusingsupersonicjet
expansionmethod[5]andareprobedbyvariousspectroscopictechniquesinasize-controlled
manner.Whenthepotentialenergysurface(PES)ischaracterizedbyseveralenergypotential
minimaseparatedbyarelativelyhighpotentialbarriersintheconditionsofexpansioncooling,
therespectiveconformationsaretrappedandtheirpopulationsarecooledtogroundstatelev-
els.Duringtheexpansiontheclustersarecollisionallycooledtolowrotationaltemperaturesand
mostofthevibrationalhotbandsareremoved.Thisreducesthecongestionandcomplexityofan
opticalspectrum,aidingtheassignmentoftheobservedspectralfeatures.Detailedinformation
aboutintermolecularbindingstructuresintheseclustersprovidessubstantialfundamentaldata
foramicroscopicviewreflectingH-bondingnetworksincondensedphasesandcanserveasa
models.theoreticalforpointcalibration

Someofthemoleculeswhosecomplexes(c.1652,composedofparts,fromFrenchcom-
plexe,fromLatincomplexussurrounding,encompassing)withwatermoleculeshavebeen
extensivelystudiedarebenzene[6],phenol[7,8],cresol[9],tropolone[7],benzonitrile[10]in-
dole[10–13],3-methylindole[14,15]and2-phenylethanol[16].Todatemanysystemshavebeen
studied,showingacleartrendforshiftingthefocusfromsmallmodelmolecularspeciestoinves-
tigatinglarger,morecomplexandrealisticsystemswithbiologicalrelevance.Therefore,wein-
vestigatedtheformationofsinglyhydratedclustersin2-phenylethanolanditsfluorosubstituted
derivativesbythecombinationoflow-andhigh-resolutionUVresonance-enhancedtwo-photon
spectroscopy,genetic-algorithm-basedrotationalfittingtechniqueandhigh-levelquantumchem-
icalcalculations.Detailedandcomprehensivediscussionsontheobservedspectroscopicfeatures
arepresentedinChapters3and7.

18

biomoleculesofPhotofragmentation1.3

oductionIntr

Inthelastdecades,thefragmentation(c.1531,fromLatinfragmentumafragment,remnant,
fromrootoffrangeretobreak.)ofbiomolecularionshasbecomearapidlydevelopingfield
ofexplorationsinceitisanindispensablepartoftheproteomictechnology.Thereexistalarge
varietyofmethodstoexplorethisfundamentallyandintrinsicallyinterestingphenomenon.The
fragmentationcanbeinducedbycollisions,lightorsomekindofchargeexchangeprocessthat
leadstothecleavageofjustoneortwobondsoutofthenumerouschemicalbondsthatholdthe
biomoleculetogether(forreview,see[17]).Thefinalgoalofsuchexperimentsistheproduc-
tionoffingerprintreferencemassspectrafortheidentificationofsmallorganicmolecules.In
theseapplicationsitisusuallythefragmentationpatternratherthanthenominalmassesthatpro-
videsdefinitivemarkersandleadstocompoundidentification.Aromaticaminoacidsandtheir
derivativesplayakeyroleinthenear-UVabsorptionofproteinswhichcausesaconsiderable
efforttowardstheunderstandingoftheirspectroscopy,structure,photophysicsandexcited-state
dynamics[18].RecentlyRizzoandcoworkersreportedininterestingfragmentationbehavior
ofprotonatedaromaticaminoacidinacold,22-poleiontrap.Inprotonatedtyrosine,different
conformersexcitedwithsimilarenergyexhibitdifferentfragmentationpatterns,whichsuggests
aconformer-specificexcited-statedynamics[19].Thebiomolecularionproductionoftheproto-
natedspeciesisrealizedbyanelectrosprayionsource(ESI).Theso-producedionsaretrapped
inelectrostaticiontrapsandphotofragmentedbyfixedinenergyUVlaserpulses.Asanext
steponecaneitherrecordaninfrared-ultravioletdoubleresonancespectrumofthephotofrag-
ments[19]orprobethedynamicsandthemechanismsofthephotofragmentation[18].Inour
groupweuseUVresonance-enhancedmultiphotonionizationspectroscopyfortheexcitationof
neutralnonprotonatedbiomoleculesandsubsequentgenerationofbiomolecularions.Usually,
thisisagentleandnondestructivetechniqueforexplorationofthespectroscopicfeaturesofthe
studiedmolecularsystemsinthefirstexcited,S1,electronicstate.Themoleculesunderconsid-
eration,consistofachromophorepartandahydroxyloranaminogroups(orboth)situatedin
thesidechain.Thisfacilitateschargetransferprocessalongthebackboneofthemoleculeand
asuccessivebreakingoftheCα—Cαbond[18].Inthiswork,asimilarsituationwasobserved
fortheneurotransmitterephedrine,wherethespecificfragmentationalbehaviorisexplainedas
aconsequenceofvibrationallyinducedphotodissociation.Thisinterestingcaseisdiscussedin
detailinChapter8ofthepresentwork.Anothercaseofaspecificfragmentationalbehavior
isthephotofragmentationofpseudoephedrine,wherewerecordedahigh-resolutionspectraof
thechargedfragmentsandfoundevidencesforaconformer-dependentphotodissociation.The
resultsarepresentedinChapter9.
Inthisthesisanumberofhigh-resolutionandthreshold-ionizationmassselectiveUVexperi-
mentsaredescribedcoveringmostoftheaforementionedtopics.Thestudiedmolecularsystems
aretwotypes:i)biological,suchasephedrineusedinthedrugsynthesisandii)biologicallyrel-

ofPhotofragmentation1.3biomolecules

19

evantcompoundssuchas2-phenylethanolthatmimictheinteractionsinrealneurotransmitters
ylamine).yleth(2-phen

∗∗∗∗∗InChapter3,aspectroscopicstudyof2-phenylethanol(2-PE)anditscomplexwithwaterispre-
sented.Boththevibrationallyandrotationallyresolvedspectraofthefirstelectronicallyexcited
statehavebeenobservedwhenthemassesofthemonomerandthesinglyhydratedclusterare
detected.Allvibronicbandsintheresonanceenhancedtwo-photonionization(R2PI)spectrum
ofthemonomerblue-shiftedupto220cm−1havebeenmeasuredathighresolutionof70MHz.
Comparingthefitsoftheexperimentaldatawiththeresultsfromthehigh-levelabinitiocalcu-
lationsanassignmentofthespectroscopicfeaturesismade,whichproofstheexistenceoftwo
conformericstructuresinthecoldmolecularbeam.Forthewatercluster,anidentificationofone
ofthevibronicbandsismadewhichimpliesthatthewaterisconnectedtotheterminalhydroxyl
groupofthemoststableconformerofthemonomerwithasinglehydrogenbondasaproton
.donorChapter4describestheeffectsoffluorinationontheconformationallandscapeof2-PE.The
stabilityofthenoncovalentOH∙∙∙πbondoftheunperturbedsystemistestedwithasubstitu-
tiononthepara-positonofthebenzenering.Theresultsofthefittedhigh-resolutionspectrain
conjunctionwiththecomputationalfindingsshowthatthepresenceofanelectronegativeatomat
thepara-positondoesnotinfluencetheconformationalpreferencesandtherovibrationalspectral
profilesof2-para-fluorophenylethanol(2-p-FPE)yieldconformerssimilartothoseof2-PE.
Thesameexperimentalapproachcombinedwithevenhigherlevelofquantumchemistryab
initiocalculationshavebeenusedfortheinvestigationofthegeometricalisomersoftheortho-
fluorinatedderivativeof2-PE.Inthiscase,theproximityoftheFatomtotheflexibleethanoltail
favorsaformationofaσ-hydrogenbondandadistortionofthestabilizingOH∙∙∙πbondofthe
mostabundantconformationofthenonfluorinated2-PE,resultinginaformationofnewconfor-
mationsof2-ortho-fluorophenylethanol(2-o-FPE).Indeed,therotationallyresolvedspectrumof
oneofthevibronicbandsoftheintermediateelectronic,S1,statewascompletelydifferentfrom
theprofilesoftheothervibronicfeaturesinthelow-frequencyregionoftheS1←−S0electronic
transition,whichweascribetoapresenceofasecondconformerwithaterminalhydroxylgroup
rotatedtowardsthepositionoftheFatom.Theotherhigh-resolutionspectrawereassignedas
originatingfromasimilarconformerasinthecaseof2-PEand2-p-FPE.Theexperimentalre-
sultssupportedbyahigh-levelabinitiocalculationsareshowninChapter6.
Theanalysisofthecationicstateoftheprototypeflexiblemolecule2-p-FPEbymassanalyzed
thresholdionizationspectroscopy(MATI)andbysupportingabinitiocalculationsbroughtfor-
wardsomeofthemodernaspectsofapplicationoftheMATItechniqueasaconformerdistin-
guishingtoolandaprobemethodfortheexistenceofstabilizinghydrogenbondsintheneutral
species.TheresultsofthisresearcharepresentedinChapter5.

20

oductionIntr

artP

II

Experimental

21

Details

2Chapter

echniquesTExperimental

Thehancedtwcombinationo-photonofLow-ionizationand(R2PI)High-ResolutionspectroscopyDopplerwith-freemassresonanceselectionen-of
jet-cooled(2-12K)molecularspeciesinthegasphaseisapowerfulexper-
imentaltechniqueforstudyingofisolatedbiologicallyrelevantmolecules,
theirmolecularcomplexes,andfragmentationbehavior.Itallowspartially
rotationallyresolved(FWHM=70MHz)spectraofthevibronicbandsof
theS1←−S0electronictransitionofthestudiedmolecularsystemstobe
ameasuredspeciallyattheirdesignedmasscomputerchannel.-assistedThefittingso-obtainedroutinespectrabasedareonanalyzedgenetical-by
gorithms.Thusaccuraterotationalconstantsintheground,S0andthefirst
excited,S1electronicstate,andthetransitionmomentratiooftheobserved
speciescanbedetermined.TheR2PItechniquetogetherwithmassanalyzed
thresholdionizationspectroscopy(MATI)offlexiblecationicmolecules,can
beusedasasensitivetoolforconformationalassignmentofneutralmolecular
conformersintheS1,electronicstate.

23

24

2.1

echniquesTExperimental

Figure2.1:Resonanttwo-photonionization(R2PI)scheme.

ResonanceEnhancedTwo-PhotonIonization(R2PI)UV

oscopySpectrLaser

TheexcitationschemeforproducingUVR2PIspectraemployspromotionofjet-cooledmolec-
ularspeciesfromtheirground,S0,electronicstatetothefirstexcited,S1,electronicstateanda
subsequentionization.TheenergyschemeispresentedinFig.2.1a.Therecordingofthehighly
resolvedUVR2PIspectrumofarovibronictransitionisprecededbytheidentificationofthe
vibronicbandsintheS1←−S0spectrum.Thelatterisobtainedwhentwophotonswiththe
sameenergy(ν1)fromasinglelasersourceareusedbothfortheexcitationandionizationsteps.
Thesocalledone-colorionizationschemehasthedisadvantagethatwhenscanningabroadre-
gionintheS1electronicstate,thetotalphotonenergy(2ν1)mayexceedthedissociationenergy
intheionandthusleadtoanundesiredfragmentationofthestudiedspecies,especiallyincase

upSetExperimentalR2PI2.2

25

ofweaklyboundcomplexes.Theotherdrawbackisthelowerresolutionofthescanninglaser,
whichcannotresolvetherotationalprofileofthevibronicbands.Thistechniqueisusedmainly
forfindingthepositionofthevibronictransitions.Theaboveshortcomingsareovercameby
usingatwo-colorionizationscheme,inwhichonephotonwithenergyν1isemployedtoexcite
themoleculesfromtheirelectronicgroundstatetothefirstexcitedelectronicstateandasecond
photonwithaconstantenergy(ν2=const)isusedtoionizethealreadyexcitedmolecules.The
energyoftheexcitationphotonisscannedandtheoneoftheionizationphotonisfixedtoavalue
whichexceeds(typicallyby100cm−1)theionizationlimitbutisbelowthedissociationenergy
oftheclusterion(seeFig.2.1a).FortheproductionofhighlyresolvedrovibrationalUVR2PI
spectraascanninglaserwithalinewidthofaround70MHzisused(ν1).Theexcitationandthe
subsequentionizationoccuronlyiftheresonanceconditionisfulfilled(seeEq.2.1).
ΔE=hν1=En−Em(2.1)
where,EnandEmaretheenergiesofthestartingandendinglevelsofthetransition.Asmentioned
above,oneoftheconsequencesofthemultiphotonionizationisthepresenceoffragmentationof
theexcitedspecies.Suchapeculiarcaseisshownfortheflexiblebiologicalmoleculeephedrine
(seeFig.2.1b).Hereevenatverylowenergiesofthesecondphoton[20],thefragmentationof
theparentioncannotbeavoided.ThisinterestingbehaviorisdiscussedindetailinChapter8.

upSetExperimentalR2PI2.2

Theexperimentalsetupconsistsofthreemaincomponents:
•Molecularbeamvacuumapparatuswithpulsedheatablenozzleforproductionofjet-
cooledmolecularspecies[5,21,22]andacoupledtime-of-flight(TOF)withanionde-
tector•Lasersystemforlow-andhigh-resolutionspectroscopy
systemacquisitionData•Onlythemainfeaturesofthesefunctionalunitswillbedescribedhere,togetherwithsomenew
implementations.ForacompletedescriptionseethediplomaworksofHolvan[23],Helm[24],
Siglow[25],Gerthner[26]andthePhDthesesofSussmann[27]andChervenkov[28].

apparatusbeamMolecular2.2.1

Thesupersonicallyjet-cooledmoleculesandmolecularcomplexesareproducedbyadiabatic
expansionthroughapulsednozzleintovacuum.Thevacuumchamberconsistsoftwodiffer-

26

echniquesTExperimental

Figure2.2:Schematicdrawingoftheheatablenozzleforproductionofmolecularbeams.

entiallypumpedsubchambers:aforechamberandaninteractionchamber(seetheenlargement
inFig.2.4).Theforechambercontainsthepulsednozzle(seeFig.2.2),andthecoldmolecular
beamisproducedthere.Theinteractionchamberisconnectedtotheforechamberthrougha
skimmerwithaclearanceof1.5mm,throughwhichthemolecularbeampasses.Theinterac-
tionchamberistheplacewheretheinteractionbetweenthemoleculesandthelaserpulsestakes
place.Itisalsothereceptacleofthetime-of-flightmassspectrometer.Thetwochambersare
evacuatedbytwoturbomolecularpumps(BalzersTPU2200,3200l/sandBalzersTPU510,510
l/s)thatmaintainvacuumof9.10−9mbarinnozzle-nonoperatingregime.Thecoldmolecular
beam[5,21,22]isproducedbyexpansionoftheinvestigatedspeciesentrainedinacarriergas
throughaheatable(upto120◦C)nozzlewithanorificediameterof300µm.Thesampleis
containedinasmallcupinsidethenozzle,justbeforetheorifice.Thisallowsforthevaporsof
thesampletomixwiththecarriergaspriortotheexpansion.Thecontainercanbeheatedandthe
temperaturecanbemaintainedstable(within0.5◦C)byaprogrammableelectronicthermocon-
troller(HengstlerGrado901).Thesampleholderconsistsoftwocompartments,whichallow
theintroductionoftwomolecularspecies.Thesimultaneousheatingprovidesbettervapormix-
ingconditions,henceandanincreasedclusterformationbeforetheinteractionwiththebuffer
gas.ThecarriergasisusuallyArsinceitfavorsthecollisionalcoolingoflargemoleculesdue
toitslargeatomicmass.Necanalsobeused,thoughthecoolingprocessinthiscaseisnotvery
efficient.Thestagnationpressureofthecarriergasinsidethenozzleistypicallyintherange2-4
bar.Themolecularbeamislacedthroughaconicalskimmer(seetheenlargementinFig.2.4).

2.2R2PIExperimentalSetup27
Theskimmerpositionisfixedbutthenozzlepositionisadjustablesothatthenozzleto-skimmer
distancecanbeoptimizedasatrade-offbetweentheionsignalintensityandtheDopplerbroad-
ening.ThereducedDopplerbroadeningΔνDforatransitionwithFrequencyνisdependson
severalparameters[29]:νBν
ΔνD=cLu∞=cKu∞(2.2)
whereLdenotesthedistancebetweennozzleandskimmer,u∞thethermalvelocityoftheex-
pandinggasandBthediameteroftheskimmer(seeFig.2.2).TheratioB/L=Kisalsoknownas
thecollimationratio.Usually,theDopplerbroadeningisreducedto40-50MHz.Thisisavery
importantparameter,sinceitisincludedinthetotalfrequencyofthespectrum:
Δνtot=ΔνL2+Δν2D(2.3)
whereΔνListhefrequencywidthofthelaserlight.Thisrelationholdsundertheconditionsof
broadenings.lineGaussian-shapedFigure2.3:SchematicdrawingoftheIonoptic.

echniquesTExperimental28Theionsproducedasaresultoftheinteractionbetweenthemolecularbeamandthelaserpulses
aremassselectedwithahomebuilt[30,31]linearWiley-McLaren[32]time-of-flight(TOF)mass
spectrometer(seeFig.2.3).Theionopticsofthespectrometerconsistsofthreeelementsdeflec-
tors,whichcreateanelectricfielddeflectingtheionsfromtheinitialstraighttrajectoryofthe
molecularbeaminadirectionperpendiculartotheplanedefinedbythemolecularbeamandthe
pathofthelaserpulses.Thus,theionsaretargetedatthedetector,whichisontopoftheinterac-
tionchamber.Thedeflectorcomprisestworegions:theionextractionregionandtheionsteering
region.Thelaser-generatedionsareextractedbytheelectricfieldcreatedbetweentherepeller
plate(bottomplateinFig.2.3)withapotentialof+300V,andthediaphragmwithapotential
of+220V.Theupwardtrajectoryoftheextractedionsiscorrectedbytheweakerelectricfield
betweenthediaphragmandthemesh(grounded).Theionspassthroughadriftregionof20
cmandthenaredetectedbymicrochannelplateswhoseamplificationrangesbetween106and
108.Itiswell-knownthatthenonrelativistickineticenergyofchargedparticlesisproportional
tothechargeandthepotentialdifference.Mathematically,thisrelationshipisexpressedbythe
equation:1mv2=eU(2.4)
2SincetheR2PIexperimentsareconcernedwithsinglyionizedmoleculesandmolecularcom-
plexesandtheappliedelectricfieldintheTOFisthesameforallions,theaboveequationcanbe
recasttodemonstratethatthearrivaltimeoftheionsatthedetectorisproportionaltothesquare
mass:theirofroot√(2.5)mt∝Thus,themeasuredarrivaltimecanbeeasilytransformedintoionicmass.Byusingreferent
−molecular2systems(e.g.C6H6),theproportionalityconstanthasbeendeterminedtobe2.70
.s.uµSystemLaser2.2.2Thelaserequipmentconsistsoftwofunctionallaserunits,presentedschematicallyinFig.2.4.
•pulsedbroadbandtunabledyelaser,whichisusedasasourcebothfortheexcitation
andionizationphotonsinone-colorlow-resolutionexperiments,andasasourceforthe
ionizationphotonsintwo-colorhigh-resolutionexperiments.
•pulsedamplifiednarrow-bandsingle-modecontinuouswave(cw)ringdyelaser,which
providestheexcitationphotonsforthetwo-colorhigh-resolutionmeasurements.

SetExperimentalR2PI2.2up

twFigureo-photon2.4:ionizationSchematicexperiments.representationofEnlarthegement:lasersystemSchematicforloview-wandofthehigh-resolutionmolecular-beamresonancesetup.enhanced

LaserIonization

29

TheionizationphotonsareproducedbyfrequencydoublinginaBeta-Barium-Borate(BBO)or
Potassium-Dihydrophosphate(KDP)crystaloftheoutputpulsesofacommercialtunablebroad-
banddyelaser(LambdaPhysikFL2002)withalinewidthof0.4cm−1(FWHM).Thislaser
hasanoscillatorandanamplifier.Thefrequencytuningisperformedbytiltingofadiffraction
grating,whichplaysalsotheroleofahigh-reflectiverearmirroroftheoscillator.Bothoscillator
andamplifierarepumpedtransverselybyaXeClexcimerlaser(LambdaPhysikEMG203MSC)
generatingpulsesat308nmwithanenergyof230mJanddurationofca.20ns(FWHM).The
frequency-doubleddyelaserpulseshaveanenergyoftheorderof1mJ.Theso-producedUV
laserpulsespassthroughatelescopicsystembeforetheyenterthevacuumchamber.Thisallows
ustocontrolthelaserspotsizeattheinteractionpointwiththemolecularbeam.

30

SingleModecwRingDyeLaser

echniquesTExperimental

Thenarrow-bandsingle-modecwlaserisaringdyelaser(CoherentCR-699-21)operatingwith
Coumarin6andCoumarin521inthewavelengthrangebetween510and550nm.Itgeneratesa
cwlaserbeamwithabandwidthof2MHzandanoutputpowerrangingfrom200to400mW,
dependingonthepumplaserpowerandthedyeused.Singlecontinuousscanswithin30GHz(1
cm−1)areroutinelyperformed.ThepumpsourcefortheringlaserisanAr+ionlaser(Coherent
Innova400)operatingatoneofitsstrongestlineat488nm(forCoumarin521).Thepumppower
variesbetween4Wand6W.

AmplificationPulsed

firstPulsedrealizedamplificationbySalourofa[35],cwandsingleappliedmodetolaserwasmolecularinventedspectroscopbyH¨yanschbyandRiedleWetallensteinal.[36].[33,34],The
Theoutputlatteroftheconsistsofsingle-modethreedyeringcells.laserisThepulsedseedingamplifiedcwbeaminapropagthree-stageatesthroughhomebtheuiltthreeamplifiercells.
wherepumpingitisgeometryamplified(seebytheSec.pump2.2.2).pulsesTheofenerthegyeofxcimertheelaserxcimer-laserdescribedpulsesaboveisinnotatransvdistributedersal
pinholesuniformlyandattheinterferencethreecellsfiltersbutindowntheproportionseeder10:20:70beampathstartingtoprevfromentthegenerationfirstone.ofThereamplifiedare
spontaneousemissioninthesecondandthirdamplifiers.Theso-amplifiedhigh-resolutionlaser
Thepulsesarefrequenccoupledy-doubledintoapulsesdoublingarecrystalnearlyF(BBOourierorKDP)for-transform-limitedsecondharmonic[37–39].generTheirationfrequenc(SHG).y
arebandwidththenlacedisca.through70-100aMHz,telescopicpulsesystem,durationsimilar10-13tonstheandoneusedmaximumfortheenergybroadband0.5-1.0lasermJ.Thepulsesy
(seewhereSec.they2.2.2),counterpropagpriortoatebeingwithsteeredtheintopulsesthefromvacuumthechamberbroad-band(seedyeSec.laser2.2.1andoandverlapFig.with2.4)
theminspaceandtimeatthecrossingpointwiththemolecularbeam.

etchingStrPulse

TheamplifiedpulsesarenearlyFourier-transformedlimited.Forsuchpulsesthereexistsare-
lationshipbetweentheirdurationandthespectralwidth,referredtoas(time-bandwidhtprod-
uct)[37].ForGaussian-shapedpulsesinthetimedomainthisrelationisgivenbytheformulae:
ΔtΔν=2ln2=0.441(2.6)
πForrectangularpulsesthisproductiscloseto1[40].Ouramplifiedpulsesareclosetorect-
angularwithapulsedurationofca.20ns.(FWHM).Thiscorrespondstoaspectralwidthof

upSetExperimentalR2PI2.2

Figure2.5:Schematicviewofthelaserpulsedelayline.

31

ca.50MHz(FWHM).Apossibleapproachtodecreasethislinewidthistoelongatethepulse
inthetimedomain.Towardsthisend,weconstructedapulsestretcher,schematicallydepicted
inFig.2.5.Thepumppulsesproducedbytheexcimerlaseraresplitbya50/50beamsplitter.
Thetransmittedcomponentisdirectedtowardstheamplifier,whilethereflectedcomponentof
thepulseisroutedalongadelaylinewithatotallengthof2.5m.

Figure2.6:Graphicalrepresentationoftheelongatedexcimerlaserpulse.Thetimewidthismeasuredat
FWHM.

32

echniquesTExperimental

Thisresultsinadelayofca.8.3ns.Thenthedelayedandthenon-delayedpulsesareconcate-
nated.Thedelaylineisoptimizedinsuchawaythatthetwopulsespartiallyoverlap,thisleading
toalongerpulsewithadurationofca.28ns.(seeFig.2.6).Thispulsedurationcorresponds
toalinewidthofca.36MHz.Theso-producedpulsesareusedtopumpthedyecellsofthe
three-stageamplifier.Theamplifiedpulses,however,haveadurationofca.14ns,and,corre-
spondingly,abandwidthofca.70MHz.Theundesiredbroadeningofthebandwidthisbrought
aboutmostlikelybyfastfluorescenceintheamplifierscells,whichisanintrinsicpropertyof
dyes.lasingtheThuswehavedemonstratedthatthepulsestretcheralonedoesnotleadtothedesiredspectral
widthoftheamplifiedpulses,butindeeditimprovedourlaserresolution.

2.2.3ControloftheExperiment.DataAcquisitionSystem

Forthelaserpulsesandthecooledmoleculestoarrivesimultaneouslyattheinteractionpoint,
theopeningofthenozzleandtheexcimerlaserpulsegenerationmustbesynchronized.This
isrealizedbyamastertriggersendingtriggerpulsestothenozzlecontroller(itcancontrolthe
nozzleopeningoftypically50µsandhence,thequantityoftheejectedmolecularspecies)and
adelaygenerator(StanfordResearchSystemsDG535).Thelatterdeterminesthetimeafter
thenozzleopeningatwhichtheexcimerlasermustbefired.Thisisthetimenecessaryforthe
moleculestodriftthedistancefromthenozzletotheinteractionpointwiththelaserpulses.This
delaycanbemanuallycontrolledandoptimizedforeveryparticularcase.Tominimizetheeffect
ofthetemporaljitteroftheexcimerlaser,itspulsesactivateaphotodiode,whichservesasa
triggersettingtimezerofortheTOFandthedataacquisitionsystem.
Thedataacquisitionsystemconsistsofthreegatedintegrators(StanfordresearchsystemsSR
280)andapersonalcomputer(PC).Thegatedintegratorsintegratetheionsignalwithinthese-
lectedtimegatewhosewidthisusually50nsor100ns.Thesignalisthendigitizedandrecorded
PC.aonThelaserscan(bothforlow-andhighresolutionexperiments),therecordingoftheabsolute
frequencymarkers,therelativecalibrationwithaFabry-Perotinterferometer,andthedataacqui-
sitionsystemarecontrolledbyahomemadesoftwareoperatinginLabVIEWenvironment.

2.3Computer-AidedRotationalFitBasedonGeneticAlgo-
rithms

Thecomplicatedrotationalandstructurecongestedofthanvibronicinrigidbandsinsmallerlarge-sizedflexiblespecies.moleculesThisisandduetotheirthelarclustersgerismassmoreof

Genetic2.3Algorithm

33

themoleculeorcluster,theirlowersymmetry,andtheadditionalinternaldegreesoffreedom.
Thismakestheinterpretationoftherotationalstructuremorecomplicated;anassignmentofthe
largenumberofoverlappingrotationaltransitionsisalengthyprocedure,andinmostcasesin-
dividuallinesarenolongerresolvedsothatanassignmentisimpossible.Forthesereasonsitis
usefultoapplypatternrecognitionmethodsforthesimulationandinterpretationofthecomplex
spectraofthesemolecularspecies.Thishasbecomepossiblebytheuseofmoderncomputersaf-
terthefastdevelopmentofcomputationalpowerandadvancedalgorithms.Inourgroupafitting
procedurewithimplementedgeneticalgorithms(GA)[20,41–46]hasbeendevelopedbyJ.E.
Braun.Thedetailsofthealgorithmofthefittingprocedurehavebeendescribedelsewhere.[28]
Hereonlythebasicfeaturesofthedataanalysisaredemonstrated.
Asaninitialstep,theprogramgeneratesseveraltrialsolutionsbased,inmostcases,onarbitrary
valuesoftheparticipatingparameters.Anytrialsolutionisreferredtoasanindividual.Anyset
oftrialsolutionsistermedpopulation.Thesizeofthepopulation,i.e.,thenumberofsolutionsin
agivenset,isafreeparameteranditcanbeuser-defined.Normally,thisnumberdependsonthe
particularproblemandvariesbetweenseveraltenstoseveralhundreds.Afterapopulationhas
beencreated,therecomesthesecondstep,theevaluation.Thequalitycanberankedandaspe-
cialnumbercalledfitnessisusedasanindicatorforthequality.Usually,thefitnessisnormalized
andittakesonvaluesintherangebetweenzeroandone.Asafigureofmeritforthequalityofthe
trialsolutions,acomparisonofanyofthesimulatedspectrawiththeexperimentaloneisused.
Thecomparisonisbased,usually,eitheroncrosscorrelationoronleastsquaresfit.Experience
shows,however,thatcrosscorrelationyieldsbetterresults.Theexperimentalspectrumiscross
correlatedwithdifferentcalculatedspectraandthemaximumoftheso-obtainedcrosscorrelation
isacriterionforthequalityofthesimulation.Thecalculationincludesasimulationofthespec-
trumandconvolutionofthesimulatedstickspectrumwithaGaussianlineshapefunction.The
crosscorrelationcanbetakenasavalueforareaoverlapbetweentheoreticalandexperimental
spectraasafunctionoftheirrelativeposition.Whenthespectraarecompletelydifferent,thefit-
nesstendstozero,andviceversa,whenthespectraareverysimilar,thefitnessconvergestoone.
Whenthequalityestimationisovertherecomestheselectionstep,inwhichofallinitiallygener-
atedspectraonlythefittestonesareselectedandsubjectedtoafurthertreatment.Thenumberof
selectedandpreservedspectraisafreeparameter.Theselectedspectraareallowedtobringforth
anewgenerationthroughtheprocessesofreproduction[41].Thisprocessgeneratestwonew
solutions;eachofthemhavingfeaturesfromeitheroftheparentones.Somemutationsusing
randomnumbergeneratorareappliedtothenewgeneration.Thisoperationisnecessarysinceit
exploitsthewholemultidimensionalspaceandpreventsclingingofthesolutiontoalocalmaxi-
mum.Whenthedescribedprocedureisrepeatediteratively,itultimatelyyieldsthebestsolution.
Veryoften,itisnecessarythatthecontributionofthepeaksinPandRbranches,respectively,be
emphasized.Forthispurpose,theexperimentalspectrumismultipliedbyaspeciallydesigned
weightingfunctionpriortobeingcrosscorrelatedwithanyofthesynthesizedspectra.Thisap-
proachhasturnedouttobesuccessfulinmanycases.Toaugmenttheefficiencyofthefitting

34

echniquesTExperimental

routine,theexploredmultidimensionalvolumecanberestrictedbyimposingsomeconstraints
tothepossiblevaluesofthefittedparameters(rotationalconstants,transitionmoment(TM)ra-
tio,rotationaltemperature,etc.).Theprocedurehasbeensuccessfullyappliedtotheanalysisof
highlyresolvedspectrawithdenselyspacedpeaksorwithalowsignal-to-noiseratioofflexible
biologicallyrelevantmoleculesandtheircomplexes[20,46].

2.4Mass-AnalyzedThresholdIonization(MATI)

Theexcitationschemeofthepulsed-fieldthresholdionizationcombinedwiththeresonance-
enhancedtwo-photonexcitationisshowninFig.2.7.Thefirstlaserpulsewithafixedwavelength
andenergyhν1promotesthemoleculestoavibronicbandintheS1electronicstate.Thesecond
laserwithphotonenergyhν2isscannedacrosstheionizationthresholdsofionicvibrationalstates
abovetheadiabaticionizationenergy(AIE).

Figure2.7:Mass-analyzedthresholdionization(MATI)scheme.

2.4Mass-AnalyzedThresholdIonization(MATI)

35

Recordingthetotalioncurrentyieldsaphotoionizationcurvewithstepsattheindividuallowest
thresholds(seeleftsideofFig.2.7).Togetherwiththepromptions,moleculesinlong-lived
high(n,l)-Rydbergstates[47–50]areexcited,whichareseparatedfromthepromptionsbya
suitableseparationprocess.Theyareionizedbyadelayedpulsedfield(t≈10µsandU≈+1000
V),andtheresultingthresholdionsaredetectedsolelyaftertheMATIseparationprocess,de-
scribedindetailselsewhere[51–53].Asaresult,sharppeakswithnobackgroundareobserved
attheindividualthresholds.Thisleadstoavibrationalspectrumofthemolecularionsasshown
schematicallyontheleftsideofFig.2.7.Theionizationthresholdsareslightlyloweredbythe
presenceofaweakseparationfieldnecessaryfortheseparationprocess(U≈-0.3V).Itcausesa
shiftoftheindividualthresholds(i.e.,ofthepeaksintheionspectrum)towardslowerenergies
ofabout10cm−1,butdoesnotaffecttheirrelativeenergydistance,yieldingacharacteristicvi-
brationalfrequencypattern,weareinterestedin.ThistechniquewasfirstdevelopedbyZhu&
Johnson[51]asacomplementarymethodofthezerokineticenergyphotoelectronspectroscopy
(ZEKE)[54],wheretheweaklyboundRydbergelectronsaredetectedandmonitoredasafunc-
tionoftheionizationlaserwavelength.TheZEKEspectroscopyhasasomewhathigherresolu-
tionandsensitivitythantheMATIspectroscopybecauseduringtheseparationandpulsedfield
ionizationprocessinMATIexperimentsmanyofthehighestRydbergstatesthatcontributeto
mostofthesignalinZEKEexperimentsaredestroyed.Ithasbeenshownbyourgroup[47]
thatRydbergstatesintheregionof40≤n≤110canberesolvedinhigh-resolutionopticaldou-
bleresonanceexperiments.TheadvantageoftheMATItechniqueisitsmassselectivity,which
becomesessentialwhenclusterdissociationexperimentsareperformed.Inthiswork,highRyd-
bergstatesareexcitedbythetwo-photonexcitationschemeviavariousintermediatevibronicS1
statesoriginatingfromdifferentconformers.

2.4.1MATIExperimentalSetup

Theexperimentalsetupcomprisesthreefunctionalunits:vacuumequipmentwithpulsedheat-
ablenozzleforproductionofjet-cooledmolecularspecies[5,21,22]andacoupledlinearre-
flectingtime-of-flightmassspectrometer(RETOF)withaniondetector[55],alasersystemfor
one-colorresonance-enhancedmultiphotonionization(REMPI)andMATI,andadataacquisi-
tionsystem.TheyaredescribedindetailinthePhDthesisofBraun[56].Herejustthemain
presented.bewillfeaturesTheMATIexperimentconsistsoftwodyelasersyielding≈10nslightpulseswithabandwidth
of≈0.3cm−1(LambdaPhysik,FL3002andLambdaPhysik,LPD3000),whichareusedfor
theexcitationofmoleculestohighRydberglevelsorfortheproductionofionsinaresonantly
enhancedone-andtwo-colortwo-photonionization(seeFig.2.8).Bothdyelasersarepumped
synchronouslybyanXeClexcimerlaser(LambdaPhysik,EMG1003i).Thetwocounterpropa-
gatinglaserbeamsintersectaskimmedsupersonicmolecularbeamperpendicularly15cmdown-

36

echniquesTExperimental

streamthenozzleorifice.Thecoldmolecularbeamisproducedbythesupersonicexpansionof
agasmixtureintovacuum.Themixtureofthesamplevaporswiththebuffergasargonat3bar
wasexpandedintothevacuumthroughapulsed-operatedheatablenozzle(GeneralValve)with
anorificediameterof500µm.Thelightpulsesoverlapintimeandspaceinthecenterofthefirst
stageofadouble-stageaccelerationconfiguration[53]formedbythreemetaldiskswithholes.
Thisresultsinaproductionofthresholdions,whichareacceleratedintoalinearreflectingTOF
[55].spectrometermass

Figure2.8:Schematicrepresentationofthelasersystemformass-analyzedthresholdionization(MATI)
xperiments.e

TheAfterdetectorreflection,signalionsiswithdirecordedfferentwithmassesgatedreachinteagrators(multichannelStanfordplateresearchdetectoratsystemsdifferentSR280times.),
thendigitizedandprocessedinamicrocomputersystemwithahomemadesoftwareoperatingin
vironment.enLabVIEW

Results

artP

and

37

III

Discussion

3Chapter

ormationalConfofIdentificationStructuresof2-Phenylethanolandits
MassbyComplexHydratedSinglySelectiveHigh-ResolutionSpectroscopy
Calculationsinitioaband

R.Karaminkov,S.Chervenkov,andH.J.Neusser,J.Phys.Chem.A.112,
(2008).839Theflexibleprototypemolecule,2-phenylethanol(2-PE),anditssinglyhy-
dratedcomplexhavebeeninvestigatedinacoldsupersonicbeambycombina-
tionofhigh-resolutiontwo-colorR2PIspectroscopyandquantumchemistry
abinitiocalculations.Theexistenceoftwomonomerstructuresseparatedby
ahighpotentialenergybarrier,gaucheandantiones,wasproven.Higher-
energyconformersaresupposedtorelaxtotheobservedonesduringthejet
expansionprocess.Wehaveidentifiedtheconformationalstructureofthe
complexbetween2-PEandwater,whichcorrespondstowaterbindingtothe
moststablegaucheconformer.Nodetectablestructuralchangesofthehost2-
PEmoleculehavebeenobserveduponattachmentofasinglewatermolecule.
Aconformationalrelaxationmechanismissuggestedalsoforthe2-PE∙H2O
x.comple

39

40

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

oductionIntr3.1

Manybiologicalprocessesareregulatedbyrelativelysmall,flexiblemolecules.Atypicalex-
ampleforsuchmoleculesareneurotransmitters,[57,58]whichplayanimportantroleinthe
humanbody.Thepropertiesandfunctionalityofneurotransmittersdependsignificantlyon
theirconformationalstructure,whichisdeterminedbyasubtleinterplaybetweenweakin-
tramolecularinteractionsandintermolecularinteractionswiththesolvent.Theimportanceof
theseinteractionshasbeenrealized,andtheirdetailedexplorationhasattractedalotofsci-
entificinterestoverthepastdecade.Inrecentinvestigationthecombinationofvariousspec-
troscopictechniquesinthegasphasewithadvancedquantumchemistrycomputationshasal-
lowedfirstretrievalofstructuralinformationonanumberofneurotransmitters,including2-
phenylethylamine,[59–64]phenylalanine,[65]ephedrine,[20,66–68]norephedrine,[69]pseu-
doephedrine,[67]adrenaline.[69]2-phenylethanol(2-PE)(seeSchemewithatomlabels)isthe
hydroxyanalogoftheneurotransmittermolecule,2-phenyl-ethylamine.Ithasbeenextensively
studiedoverthelastyears[16,46,61,63,70–73]asaneurotransmitterprototypemolecule.The
gauchestructureofthemoststableconformerofthe2-PEmonomerhasbeenconclusivelyes-
tablishedbymicrowave[70]andourhigh-resolutionresonanttwo-photonionization(R2PI)[46]
experimentscombinedwithquantumchemistryabinitiocalculations.Yet,however,thereexist
severalunclearissuesandcontroversiesontheconformationalstructuresofthebaremolecule
andthestructureofitssinglyhydratedcomplex.Thepresentworkpresentsadetailedanalysis
oftheexistingconformersofthe2-PEmonomeranddiscussespossibleprocessesleadingtoin-
terconversionofconformersinthejetexpansion.Theexperimentalfindingsfrommassselective
high-resolutionUVtwo-photonresonanceenhancedionizationspectroscopyofallpronounced
bandsinthevicinityofthestrongestoriginbandarereconciledwiththetheoreticalpredictions
onrotationalconstants,transitionmomentratios,andvibrationalfrequenciesfromhigh-levelab
initioquantumchemistrycalculations.Throughaconfidentassignmentofalltheobservedvi-
bronicbandsinthevicinityofthestrongest000originbandinthespectrumofthe2-PEmonomer,
wehaveconcludedthattwoconformers,gaucheandantionesareobservedundertheconditions
ofadiabaticexpansioninacoldsupersonicbeam.Theanalysisofthetheoreticalpotentialenergy
surfacecorroboratestheassumptionthatthehigher-energyconformersthatpresumablyexistin
thepreexpansionregionrelaxtotheobservedones,thelatterbeingseparatedbyahighpotential
barrier.Wealsoextendedthestudyof2-PEtoitssinglyhydratedcomplex.Thisinvestigation
rendersitselfasanaturalextensionofourpreviousstudyonthe2-PE∙Arcomplex,[46]where
thestructureisdeterminedbytheinterplaybetweentheintramolecularhydrogenbondandthe
intermoleculardispersioninteractions.Thisistoourknowledgethefirstapplicationofmassre-
solvedhigh-resolutionUVspectroscopytostudyinghydratedcomplexesofflexiblebiologically
relevantmolecules,whichareasubjectofintenseresearchoverrecentyears.[4,73–78]The
importantissuesaddressedarewhetherandhowsolventmoleculesaffecttheconformational

3.2ExperimentandDataProcessing

Figure3.1:Atomlabelsofthe2-PE∙H2Ocomplex

41

structureoftheflexiblesolutemoleculeandhowtheflexiblemoleculeadaptsitsstructureto
thelocalenvironmentunderthecombinedactionofintra-andintermolecularhydrogenbonds.
Anotherinterestingaspectiswhetherwaterstabilizesparticularconformersthusactingasan
efficientconformationalselector.Thecomplexbetween2-PEandwaterisaprototypesystem
formicrosolvationandisexpectedtorevealthebindingpatternandtheinfluenceofwateronthe
conformationalshapeandstabilityofthehostflexiblemolecule.

3.2ExperimentandDataProcessing

Theexperimentalsetupusedtomeasuretheone-andtwo-colorresonanceenhancedtwo-photon
ionizationspectraof2-PEhasbeendescribedinChapter2.2andfollowingpublications[10,
46,79,80].Briefly,the2-PEvaporismixedwithAratapressureof3.0barandexpandedinto
vacuumthroughapulsed-operatedheatablenozzlewithorificediameterof300µm.Toincrease
thevaporconcentrationof2-PEinthenozzlethesubstancewasheatedto96◦C.Toreducethe
Dopplerbroadeningaskimmerwithanorificeof1mmindiameterisused.Themolecularbeam
isintersectedperpendicularlybytwocounterpropagatinglaserpulses.
Theanalysisofthehighlyresolvedspectrahasbeenperformedusingourhomemaderotational
fitcomputerroutinebasedongeneticalgorithms.[20,44,46,81]Crosscorrelationwasemployed
asafitnessfunction.Thetypicalnumberofgenerationswas200with200individualsina
generation.

42

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

Figure3.2:One-colorR2PIspectraoftheS1←S0electronictransitionofthe2-PEmonomermeasured
atthemonomer(m/z=122)masschannelunderwater-free(a),water-present(b)conditions,andofthe
2-PE∙H2Ocomplex(c)recordedatitsparent(m/z=140)masschannel.

ResultsExperimental3.3

wLo3.3.1SpectrumResolution

Wehaverecordedlow-resolutionone-colorresonanttwo-photonionization(R2PI)spectramon-
itoringthreedifferentmasschannelsinthevicinityoftheS1←S0electronictransitionof2-PE
undertwodifferentexperimentalconditions.Inthefirstcase,thespectrumwasmeasuredunder
water-freeconditionsbyselectingtheionsignalattheparentmasschannelatm/z=122.Thespec-
trumispresentedinFig.3.2(a).The000vibronicbandandthesmallredshiftedbandat30cm−1
havebeendiscussedinourpreviouspaper.[46]Herewediscusstheblueshiftedvibronicfeatures
inthespectrumat+42,+48,+58,+136,and+213cm−1.Thesecondsetofspectrawasrecorded
underwater-inconditionsmonitoringbothparentmasschannelatm/z=122andthemasschannel
ofthesinglyhydrated2-PEatm/z=140.ThetworesultingspectraareshowninFig.3.2(b,c).

ResultsExperimental3.3

43

Thethreemostintensevibronicbandsmeasuredatthemasschannelofthe2PE-watercluster
arelabeledwith1w1,2w1and3w1,respectively.Ithasbeenfoundthatthepositionsofpeak
+48cm−1inthemonomermasschannelandpeak3w1inthe2PE-waterclustermasschannel
coincide.Therefore,peak+48cm−1islikelytooriginatefromfragmentationofthe2-PE-water
cluster.However,addingwaterdoesnotchangetherelativeintensityofthisband.Thisleadsus
totheconclusionthat3w1isamixtureoftwocloselyingbands:oneofthe2-PEmonomer,and
theotheroneofthe2-PE∙H2Ocomplex.Aclearevidenceforthenatureofthisbandisprovided
onlybyhigh-resolutionspectroscopy,andispresentedbelow.Thereisastrongerfragmentation
behaviorofband2w1incomparisonwiththeotherbands(1w1and3w1)observedatthismass
channel.Thismaybetentativelyexplainedbytheexistenceof2-PE∙H2Oclusterconformersof
differentstability.TheremainingsmallervibronicbandsinthespectrumofFig.3.2(c)measured
atmasschannelm/z=140donotcorrespondtoanyofthebandsinFig.3.2(b)measuredatthe
masschannelofthemonomer.Asanextstep,toelucidatetheconformationalstructuresgiving
risetotheabove-describedvibronicbandsof2-PEanditssinglyhydratedcluster,wemeasured
thebandsunderdiscussionwithamuchhigherresolution.

pectraSResolutionHigh3.3.2

BandsMonomer3.3.3

ThehighlyresolvedUVspectraofallblueshiftedbandsrecordedatthemonomermasschan-
nelunderwater-freeconditionsareshowninFigs.3.3-3.5.Inallbands,awell-pronounced
rotationalstructurecoveringarangefrom3to5cm−1isresolved.Therotationalstructureof
thebandsat+42,+136and+213cm−1issimilartothethatofthepublished000band.[46]Itis
characterizedbyacentraldip,asmallQbranchontheredsideofthedip,andwelldefinedPand
Rbranches.Thespectraofthesethreebandsmanifestahybrida-,b-,andc-typecharacterwith
prominentbcontribution.Theydonotconsistofsinglerotationallinesbutweobservepeaks
producedbyaggregationofseveralrotationallinesresultinginaminimumvalueforthepeak
FWHM.MHz250ofwidth

44

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

Figure3.3:High-resolutiontwo-colorUVR2PIspectrumofvibronicband+42cm−1inFigure3.2,
recordedatm/z=122.Thebandisassignedasaprogressionofthegaucheconformer2ofthe2-
PEmonomerwithitsrotationlesstransitioncenteredat37663.697(7)cm−1.Uppertrace:experimental
spectrum.Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable3.1(for
xt).teseedetails,

Ttheofindtransitiontherotationalmomentratio,constantstheforrotationaltheground,temperature,S0,andandthethefirsteoriginxcited,position,S1,υ0,electronicweusedstates,the
computer-assistedrotationalfitprocedurebasedongeneticalgorithmdescribedearlier.[20,44,
81]calculations(seealsoofSec.the2.3)Asgroundstatestartingvrotationalaluesforconstantsourfits,forwetheusedantitheandvgaucaluesheresultingconformers.fromTabomin-inito
imizethenumberofsimultaneouslyfittedparameters,weusedastepwiseapproachtodetermine
thetionalexperimentalconstantsvwithinaluesof0.5%alloftheirparameters.initialAsvaaluefirstandstep,letwethesearchconstrainedspacetheforthegroundexcitedstaterota-state
valuesrotationalobtainedconstantsintheandfirstforsteptheandotherfixedtheparameterstransitionrelativelymomentbroad.ratiosInbautsecondlettingthestep,weotherusedparam-the
etersfree.TheproducedtheoreticalstickspectrumwasconvolutedusingaGaussianlineshape
withFigs.a3.3-FWHM3.5.ofThe150simulatedMHz.Thespectrabest-fit(theloresultswerforinvtheertedbandspart)areagreeshownwellinintheinbothvertedpeakspectrapositionsin
andbetweenpeaktheeintensitiesxperimentalwiththeandesimulatedxperimentalspectraonesisas(upperhighastrace).95%.TheTheachieevedxperimentallycrosscorrelationobtained
valuesoftherotationalconstants,thetransitionmomentratios,andtherotationaltemperatures

3.3

Experimental

Results

1

Figure3.4:High-resolutiontwo-colorUVR2PIspectraofvibronicbands+48cm−1and+58cm−1in
Figure3.2,recordedatm/z=122.Thebandsareassignedasavibrationalprogressionsoftheanti
conformer5ofthe2-PEmonomerwiththerotationlesstransitionscenteredat37669.601(2)cm−1and
37679.516(7)cm−1,respectively.Uppertraces:experimentalspectra.Lowerinvertedtraces:thebest-fit
simulatedspectrayieldingtheparametersinTable3.1(fordetails,seetext).

45

46

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

aresummarizedinTable3.1.Comparingtheexperimentalvaluesfortherotationalconstants
withtheresultsfromtheabinitiocalculationswecanclearlyassignthebandsat+42cm−1,
+136and+213cm−1tothemoststablegauchestructure.However,differencesareobserved
forthemeasuredtransitionmomentratioofthe+42cm−1bandfromthe+136and+213cm−1
bandsandfromtheory.Theexperimentalvaluesforthe+136and+213cm−1bands(18:74:8)
arethesameastheonesofthe000band[46]butdifferfromthetheoreticallypredictedoneat
theMP2/cc-pVDZleveloftheory(1:96:3).Themeasuredmixedaandb-typespectrumwith
transitionmomentratio(42:50:8)ofthe+42cm−1bandyieldsatransitionmomentorientation
whichisnotpredominantlyalongthebprincipalaxisofinertiaasistheoreticallyexpected.On
theotherhand,theexperimentalrotationalconstantsforbothgroundandexcitedstateareclose
totheonesofconformer2.Weusedtheabove-describedproceduretoassignthespectraofthe
+48and+58cm−1vibronicbandsinFig.3.4fittingwithadifferingrotationalbandstructure
characterizedbyaminorcentraldip,averysmallQbranchontheredsideofthedip,andwell
resolvedPandRbranches.Again,thereisagoodmatchinpeakpositionsandpeakintensitiesof
measuredandsimulatedspectrawithachievedcrosscorrelationof96%and95%,respectively.
ThereisagoodagreementbetweenthevaluesofrotationalconstantsBandCinTable3.1and
thosefoundbyourabinitiocalculationsoftheanticonformer5fortheground,S0,andthefirst
excited,S1,electronicstates(cf.Table3.3).Onlytheexperimentalandcalculatedvaluesforthe
groundrotationalconstantAaresomewhatdifferent.Inaddition,theexperimentalvaluesforthe
transitionmomentratiosofthetwobands(4:96:0and8:92:0,respectively)areveryclosetothe
onecalculated(0:100:0)forconformer5.

ComplexaterW3.3.4Bands

Twoofthevibrationalbands,1w1and3w1,measuredattheparentmasschannel(m/z=140)of
thesinglyhydratedcomplexof2-PE(seeFig.3.2)havealsobeenmeasuredunderhighresolu-
tion.Thelowintensityofband2w1didnotallowthisbandtoberotationallyresolved.Band3w1
yieldsrelativelyhighintensitywhenmeasuredbytwo-photontwo-colorionizationandspans
overafairlybroad(ca.4cm−1)range.Thebandissuperimposedonahighbackgroundand
itsoverallprofileischaracterizedbytwopronouncedPandRbranchesandashallowcentral
dipwithaweakQbranch.Itisworthpointingoutthatthoughitshighintensityband3w1does
notfeatureresolvedrotationalmicrostructure.Thisprecludeditsrotationalfitanalysis.Theex-
perimentalhighlyresolvedspectrumofband1w1ispresentedinFig.3.6(uppertrace).The
spectrumcoversarelativelynarrowfrequencyrangeofca.1.5cm−1.Ithasawell-resolved
structurebuiltupofseparatepeakswithalinewidthof250MHzresultingfromaggregationsof
closelyspacedsinglerotationallines.Thespectrumliesonasmallbackground,andfeaturesa
prominentQbranchinthecenterandlessintensePandRbrancheswithirregularbutresolved
structure.Wefittedthespectrumbythecomputer-aidedroutinebasedongeneticalgorithms

3.3

Experimental

Results

Figure3.5:High-resolutiontwo-colorUVR2PIspectraofvibronicbands+136cm−1and+213cm−1in
Figure3.2,recordedatm/z=122.Thebandsareassignedasavibrationalprogressionsofthegauche
conformer2ofthe2-PEmonomerwiththerotationlesstransitionscenteredat37756.894(4)cm−1and
37833.949(1)cm−1,respectively.Uppertraces:experimentalspectra.Lowerinvertedtraces:thebest-fit
simulatedspectrayieldingtheparametersinTable3.1(fordetails,seetext).

47

48

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

BandarameterP+42cm−1+48cm−1+58cm−1+136cm−1+213cm−1
A´´0.1116(11)0.14765(19)0.14785(30)0.1113(22)0.1118(25)
BC´´´´0.03605(32)0.3211(28)0.02841(9)0.02527(8)0.02857(34)0.02558(41)0.03620(63)0.3203(35)0.03611(63)0.03192(33)
A´0.1083(14)0.14118(14)0.14122(27)0.1080(32)0.1086(26)
CB´´0.03538(33)0.03128(45)0.02803(7)0.02513(9)0.02818(39)0.02513(42)0.03560(49)0.03110(47)0.03550(56)0.03109(34)
µa2:µb2:µc242:50:88:92:04:96:019:74:919:74:9
ν0,cm−137663.697(7)37669.601(2)37679.516(7)37756.894(4)37833.949(1)
T(K)8.5(2)8.6(3)9.9(2)11.3(5)11.2(1)
Best-fitcrosscorrelation(%)94.595.695.294.795.6

Table3.1:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,thebandoriginfrequency,ν0,the
rotationaltemperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofbands+42,
+48,+58,+136,and+213cm−1,showninFigures3.3-3.5.Thenumbersinparenthesesrepresentone
standarddeviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevaluesof
µa2,µb2,andµc2inthetransitionmomentratiodoesnotexceed5(%).

outlinedaboveandderivedthemolecularparametersofthespeciesproducingit.Thesimulated
stickspectrumwasconvolutedbyaGaussianprofilewithalinewidthof150MHz.Theresulting
best-fitsimulatedspectrumisdepictedinFig.3.6(lowerinvertedtrace).Thesimulationrepro-
ducesfairlywellbothoverallprofileandpeakpositions.Theso-obtainedrotationalconstants
fortheground,S0,andthefirstexcited,S1,excitedstate,thetransitionmomentratio,theband
originposition,υ0,andtherotationaltemperature,T,aresummarizedinTable3.3.4.

CalculationsinitioAb3.4

Abinitioquantumchemistrycalculationsforvariousconformationsofthe2-PEmonomerand
itssinglyhydratedcomplexhavebeenperformedusingGAUSSIAN03programpackage.[82]
Fivedifferentconformationsofthe2-PEmonomerandsixdifferentconformationalstructuresof
the2-PE∙H2Ocomplexhavebeenconsidered.

Monomer2-PE3.4.1

Tofindthestableconformationsofthe2-PEmonomerweperformedapotential-energygrid
searchattheMP2/cc-pVDZleveloftheory.Thegridsearchwasrealizedbyscanningthe

CalculationsinitioAb3.4

Figure3.6:High-resolutiontwo-colorUVR2PIspectrumofvibronicband1w1inFigure3.2,recordedat
m/z=140.Thebandisassignedasthe000originbandofconformerconformerBofthe2-PE∙H2Ocomplex
1−invwithertedthetrace:rotationlessthebest-fittransitionsimulatedcenteredatspectrum37641.19(2)yieldingcmthe.parametersUppertrace:inTeable3.5xperimental(fordetails,spectrum.seeLotext).wer

49

C2C7C8O9andC7C8O9H19dihedralanglesofthe2-PEmolecule.Theso-obtainedpotential-
energysurfaceisdepictedinFig.3.7(forthesakeofclearvisualizationthepotential-energy
surfaceisinverted).Fivepotential-energyminimahavebeenpredicted.Thedeepestone(global
minimum)correspondstothegaucheconformer(conformer2)inwhichtheterminalOHgroup
ofthesidechainpointstotheπelectronsofthearomaticring.Othertwohigher-energystructures
(conformer1andconformer3)separatedfromeachother,andfromthemainminimum,bylow
potentialbarrierscorrespondalsotogaucheconformationsbutwithdifferentorientationsofthe
OHgroup(differentC7C8O9H19dihedralangles).Theconformationswiththesidechainpoint-
ingawayfromthebenzenering(conformer4andconformer5)arereferredtoasanticonformers,
andtheygiverisetotheothertwopotentialenergyminima.Thetwostructuresinquestionare
separatedbyalowpotentialbarrierfromeachother,andbyahighbarrieralongtheC2C7C8O9
anglecoordinatefromthegaucheconformers.Theconformationalstructurescorrespondingto
thepotential-energyminimahavebeensubjectedtoafurtherfullstructuraloptimizationandcal-
culationoftheirenergeticsattheMP2/cc-pVDZleveloftheoryfortheground,S0,electronic
state,andattheCIS/cc-pVDZleveloftheoryforthefirst,S1,electronicstate.Inaddition,afre-
quencyanalysisforbothground,S0,andfirstexcited,S1,electronicstatehasbeenperformedat
thesameleveloftheory,respectively.Thepresenceofonlypositivefrequenciesisaverification

50

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

1wBandaramterP10.0671(9)´´A0.0270(3)´´B0.0224(10)´´C0.0654(14)´A0.0268(3)´B0.0221(1)´Cµa2:µb2:µc269:31:0
ν0,cm−137641.19(2)
1.4(1)(K)T91(%)correlationcrossbest-fitTable3.2:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´
,andforthefirstexcited,S1(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,
thebandoriginfrequency,ν0,therotationaltemperature,T,andthebest-fitcrosscorrelationob-
tainedfromtherotationalfitofthe000originbandoftheS1←S0electronictransitionofconformer
Bofthe2-PE∙H2Ocomplexshowninfigure3.6.Thenumbersinparenthesesrepresentonestan-
darddeviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevalues
ofµa2,µb2,andµc2inthetransitionmomentratiodoesis9(%).

thattheoptimizedstructuresindeedcorrespondtopotential-energyminima.Therotationalcon-
stantsofthediscussedconformationsforbothgroundandexcitedelectronicstatesalongwiththe
energieswithoutandwiththezero-pointvibrationalenergycorrectionsforthegroundstate,and
thetransitionmomentratiosforthefirstexcitedelectronicstate,aswellassometypicalinter-
atomicdistances,planar,anddihedralanglesaresummarizedinTable3.3.Thefullyoptimized
conformationsarealsoshowninFig.3.7atopoftheircorrespondingpointonthepotential-
energysurface.Thevibrationalfrequenciesforthefirstexcited,S1,electronicstatearedetailed
inTable3.4.Inadditiontotheabove-describedcalculations,structuraloptimizations,energy
calculations,andvibrationalanalysisofconformers4and5attheMP2/aug-cc-pVTZlevelof
theoryhavebeenperformedaswell.Theuseofthisveryextendedbasissetsignificantlyac-
countingfortheelectroncorrelationwasexpectedtoresolvetheexistingcontroversy[59,63,70]
ontheenergyorderingofconformer4andconformer5,andtotheoreticallysubstantiateour
experimentalresults.Nonoticeablestructuralchangesoccuruponextendingthebasissetfrom
cc-pVDZtoaug-cc-pVTZbutattheaugmentedbasissetconformer5ismorestablethancon-
former4by53cm−1.ThisresultdoesnotincludetheZPVEsincethevibrationalanalysisof
thetwoconformationsatthisbasissetiscomputationallyveryexpensive.Theinclusionofthe
ZPVE,however,isnotexpectedtoaltertherelativeenergyorderingofthetwoconformers.

CalculationsinitioAb3.4

Figure3.7:Potentialenergysurface(invertedforthesakeofabettervisualization)asafunctionofthe
C2C7C8O9andC7C8O9H19dihedralanglesofthe2-PEmonomer.Thegaucheandanticonformations
areseparatedbyahighpotentialbarrierintheC7C8coordinate.Thefiveenergeticallymostfavorable
conformationsofthe2-PEmonomerhavebeenfullygeometricallyoptimizedintheirground,S0,elec-
tronicstateattheMP2/ccpVDZleveloftheory.Thefullyoptimizedstructuresaredepictedontheir
correspondingminimaonthepotentialenergysurface.

51

3.4.22-PE∙H2OComplex
Asaninitialstepweperformedstructuraloptimizationsfortheground,S0,andthefirstexcited,
S1,electronicstateattheMP2/6-31G*andCIS/6-31G*leveloftheory,respectively,ofvarious
conformationsofthe2-PE∙H2Ocomplex.Thestartinggeometriesofthehydratedcomplexes
wereconceivedbyattachingthewatermoietytodifferentplausiblebindingsitesofthethree
lowest-energyconformations(conformers2,4,and1)ofthe2-PEmonomer.Inthisway,6dif-
ferentconformationsofthe2-PE∙H2Ocomplexwereproducedandoptimized(Fig.3.8).The
theoreticallypredictedinertialparametersforthegroundandexcitedstateandthetransitionmo-
mentratiowereusedtosimulatetheoreticalspectra,whichwerevisuallycomparedwiththe
experimentalhighlyresolvedspectrumofthe2-PE∙H2Ocomplex(seeFig.3.6,uppertrace).
Theapparentdisagreementbetweentheexperimentalspectrumandthesimulatedspectracorre-
spondingtothetwohydratedcomplexeswheninvolvingtheanticonformerofthe2-PEmonomer
(conformer4)wasaclearindicationthattheexperimentalspectrumdoesnotoriginatefromany

52

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

Figure3.8:Electronicgroundstate,S0,structuresofthe2-PE∙H2OcomplexoptimizedattheMP2/cc-
pVDZ(conformersA,B,C,andD)andMP2/6-31G*leveloftheory,respectively.Typicalangles,bond
lengths,inertialparameters,andbindingenergiesforthesestructuresarelistedinTable3.5

ofthesetwocomplexes.Forthisreasonthelatterwereruledoutfromafurtherconsideration.
Theremainingfourstructuresofthe2-PE∙H2Ocomplexweresubjectedtofurtherfullgeomet-
ricaloptimizationsandvibrationalanalysisforbothground,S0,andfirstexcited,S1,electronic
statesattheMP2/cc-pVDZandCIS/cc-pVDZleveloftheory,respectively,thusimprovingthe
electroncorrelation.Inthiswayvariousbindingpatternshavebeenthoroughlytheoreticallyex-
plored,inwhichwaterplayseithertheroleofaprotondonororofaprotonacceptorformingσ
orπweakintermolecularhydrogenbonds.Thefoundvibrationalfrequenciesofalloptimized
conformersarepositive,whichisaverificationthatindeedthosestructurescorrespondtominima
onthepotentialenergysurface.Wealsocalculatedthebindingenergieswithoutandincluding
thebasissetsuperpositionerror(BSSE)andzero-pointvibrationalenergy(ZPVE).Theabsolute
bindingenergiesdependessentiallyontheinclusionofthecorrectionsaccountingfortheBSSE
andZPVEbuttheenergyorderingoftheconformationsremainsalwaysthesame.Therotational
constantsforbothgroundandfirstexcitedelectronicstate,bindingenergiesforthegroundstate,
andthetransitionelectronicdipolemomentofallconformationsaresummarizedinTable3.5.
TheoptimizedconformationalshapesaredepictedinFig.3.8,whereconformersA,B,C,and
DhavebeenoptimizedattheMP2/cc-pVDZleveloftheory,andconformersEandF(presented
forcompleteness)havebeenoptimizedattheMP2/6-31G*leveloftheoryThemoststablecon-

Discussion3.5

53

former(theonewiththehighestbindingenergy)isconformerAinwhichthewatermolecule
insertsbetweenthebenzeneringandthesidechainofthemoststable2-PEmonomer(conformer
2).Inthisbridgingstructurethewatermoleculeactsbothasaprotondonorandaprotonacceptor
formingaweakπhydrogenbondwiththebenzenering,andσbondwiththehydrogenatomof
theterminalhydroxylgroupof2-PE.Thesecond-in-energyconformerisconformerBinwhich
thewatermoietybindssidewaysto2-PEthroughtheformationofasingleσbondwhereinthe
watermoleculedonatesaprotontotheoxygenatomoftheOHgroupof2-PE.Thehigher-energy
conformersofthe2-PE∙H2Ocomplexinvolveconformer1ofthe2-PEmonomer.

Discussion3.5

ormersconfMonomer3.5.1

Inourabinitiocalculationsofthe2-PEmonomerattheMP2/cc-pVDZleveloftheory,wehave
consideredfiveofthemoststable2-PEmonomerstructureswhichhavebeendiscussedinprevi-
ousworks.[16,61,70,72,73]Themoststableconformeristheonewiththehydroxylhydrogen
positionedabovethebenzenering,anddesignatedhereinafterasconformer2.Thenexttwocon-
formationsaretheanticonformer4,andthegaucheconformer1,separatedfromthemoststable
speciesby586cm−1and685.86cm−1,respectively.Theenergygapbetweenthem,however,is
verysmall,andthereforetheaboveorderingmaybechangeduponimprovingtheusedcompu-
tationalmethodandbasisset.Themostunstableconformationsarethegaucheconformer3,and
theanticonformer5,distancedby783.30cm−1and788.35cm−1awayfromthelowest-energy
conformer2.Weunambiguouslyassignbands+136and+213cm−1toconformer2wherethe
agreementofthetheoreticalandexperimentalvaluesfortherotationalconstantsoftheground,
S0,andthefirstexcited,S1,electronicstatesandthetransitionmomentratioisthebest.This
istheconformerthatproducesalsothestrongestband(seeFig.3.2).[46]Inthecaseofband
+42cm−1,wherethetransitionmomentratiodiffersformtheonepredictedforconformer2,the
assignmentisbasedontherotationalconstantsmatchonly.Asimilartoour(42:50:8)changeof
thetransitionmomentratiohasbeenpredicted(42:36:22)byMonsetal.[16]Wesupporttheir
explanationfortheoriginofthisphenomenon:the+42cm−1bandmaygainintensitythrough
vibroniccouplingtoanotherexcitedelectronicstate,whichleadstoastrongdependencyofthe
transitionmomentratioonthenatureofthevibration.Anadditionalargumentthatconformer2
producesthebandsat+42cm−1,+136cm−1,and+213cm−1isthegoodagreementbetweenthe
bandpositionsandthetheoreticallypredictedfundamentalvibrationalfrequenciesforthiscon-
former(seeTable3.4).Onthebasisoftheabovevibrationalanalysis,wetentativelyconclude
thattheveryweakbandat+90cm−1alsooriginatesfromconformer2ofthe2-PEmonomer.The
otherunassignedweakbandat+75cm−1isverylikelytostemfromfragmentationofawater

54

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

S0120.5-61.62.54060.5-90.0105.60.03176112.30.03677110.80.110230ConformerGaucheS-120.665.32.78-64.81:96:390.612109.00.03047113.00.03524113.80.11200

µEEµCBA3τ2τ1τααααdConformer
:−1µc2SZPVE),(incl.cmb2−1−1−1
cm−11theofstateelectronic,Theoreticaldistances,interatomic
gaucheS11andd,antiplanarS0angles,ConformerGaucErelheSMP2fromobtainedbeenevhagyenervibrationalzero-pointtheofinclusionwithandgies,enerevRelati.monomer2-PEtheofgeometryoptimisedratio(TM)momenttransitiondipoleelectricThe.elyvrespecti,theoryofxcited,e3.3:ableTelrelr,a:2cm,cm,cm,067.063.3663.166.565.3-61.6-170.7-168.6gde(C7C8O9H19),0177.3176.771.667.464.860.570.69.36gde(C2C7C8O9),88.590.389.4140.8109.990.690.0105.110.4gde(C3C2C7C8),4106.7109.3106.1109.2105.9109.0105.6109.5106.9gde(C8O9H15),3107.5112.0112.8114.1113.1113.0112.3108.9107.7gde(C7C8O9),2111.0113.3111.2116.8113.3113.8110.8114.7112.7gde(C2C7C8),1120.7120.4120.5122.1121.1120.6120.5120.9120.8gde(C1C2C7),---3.533.412.782.543.893.78Å(H15-C1),
12µa2:µb2MP2theatcalculatedmonomer2-PEtheofconformers
:S-18:77:5--0:100:0-23:77:0-1:96:30SConformer
0Conformerhas586.00-783.30-0-685.86788.35-769.00-649.69-703.59-0-652.780.025160.025410.025080.025330.027400.029210.030470.031760.029090.029750.028190.028400.028120.028310.033740.035260.035240.036770.034810.035770.145420.145700.144570.144490.125530.117070.112000.110230.117160.11622
beenS13obtainedwithoutmonomer2-PEtheofstructuresconformationalstablemostevfitheofαangles,dihedral,
.theoryofelvlecc-pVDZ/cµ2τground,theforconstantsrotationaland,
fromCIS/CISandcc-pVDZ/0SConformer
cc-pVDZS14/SAnticalculationSelvlecc-pVDZ0theand,first
0ofConformerthe

S105.966.50.02921703.593.41113.167.40.03526-113.3109.90.11707783.30121.10ConformerS63.10.02740-3.53114.10.0337471.623:77:0116.8140.80.125531-122.13109.2

6S586.00120.5-649.69176.70-89.4106.10.02533112.80.02831111.20.1444963.3ConformerS0.028120.14457120.467.0--177.310:100:090.34109.30.02508112.0113.3-AntiS0.025410.028400.14570120.70-769.0000-88.5106.7107.5111.0788.35ConformerS120.6050.025160.028190.14542109.4107.5113.1--010:100:089.3-

Discussion3.5

55

Conformer1Conformer2Conformer3Conformer4Conformer5
ν,cm−1Modeν,cm−1Modeν,cm−1Modeν,cm−1Modeν,cm−1Mode
31torsion45torsion26torsion42torsion40torsion
102stretching89stretching120bending88bending89bending
129torsion135torsion148torsion99torsion96torsion
226bending230stretching232bending230bending232bending

Table3.4:Theoreticallowestvibrationalfrequenciesforthefirstexcited,S1,electronicstateofthegauche
andanticonformationsofthe2-PEmonomer,calculatedattheCIS/cc-pVDZleveloftheory.Theapplied
0.94.isactorfscale

complexof2-PE.Thecalculatedinertialparametersofconformers4and5areverysimilarwith
exceptionofrotationalconstantA.Thisisnotsurprisingnotingthatthesetwostructuresdiffer
fromeachotheronlybytheorientationoftheterminalOHgroupofthesidechain.Notwithstand-
ingthesimilarityofthetheoreticallypredictedrotationalconstantsofthediscussedconformers,
therotationalfitofourhigh-resolutionspectraunambiguouslydemonstratesthatbothvibrational
bandsat+48and+58cm−1originatefromstructure5.Thisassignmentisinagreementwiththe
findingofMonsetal.[16]basedonUV-UVholeburningandinfraredspectroscopy.Theassign-
mentofband+48asoriginatingfromconformer5iscorroboratedalsobydispersedfluorescence
experiments.[73]Bothclose-lyingbandswithsimilarrotationalstructures,+48and+58,may
resultfromtransitionsfromconformer5intheelectronicground,S0,statetoconformers4and
5inthefirstexcited,S1,electronicstateduetothetorsionalpotentialforthehydroxylhydrogen
intheelectronicallyexcitedstatefavoringthesetransitions.[16]Thisisaplausibleexplanation
sinceintheanticonformers4and5,thebehavioroftheOHgroupisalmostnegligiblyperturbed
bytherestofthemoleculeand,forthisreason,itspropertiesareexpectedtobearresemblanceto
theonesoftheOHgroupinethanol,andhencetheOHpotentialenergymodelforethanol[83]
canbeappliedtotheanti2-phenylethanolaswell.Alikephenomenonhasbeenobservedalso
[84]p-tyrosol.forAllvibrationalbandsmeasuredunderhighresolutionhavebeensuccessfullyassignedtocon-
formers2and5,respectively.Noevidencesfortheexistenceoftheotherthreeconformershave
beenfound.Thisputsforwardthefollowingissues:i)whyonlytwooutoffivetheoreticallypre-
dictedconformationalstructureshavebeenexperimentallyobserved;ii)whythehighest-energy
conformer5isobservedinthecoldmolecularbeamratherthananyofthelow-lyingstructures.
Theexplanationofthefirstquestionrequirescognizanceofthepotential-energysurfaceanddis-
cussionoftherelaxationdynamicsduringtheprocessoftheadiabaticexpansionpresumingthat
inthepreexpansionregionofthenozzlethereexistseveraldifferentconformationalstructures
atroomtemperaturethatrelaxduringtheprocessoftheadiabaticcooling.Asseenfromthe
shapeofthepotentialenergysurfacedepictedinFig.3.7,thegaucheconformersareseparated
fromtheanticonformersbyahighpotentialbarrier,whereasthepotentialbarriersbetweenthe

56

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

gaucheandbetweentheanticonformers,respectively,areverylow.Thereforewemayconclude
thatthehighpotentialbarrierprecludestheinterconversionbetweengaucheandanticonform-
ersononehand,andontheotherhandthelowpotentialbarriersbetweenthegaucheandanti
conformers,respectively,favortherelaxationalintraconversionwithinthetwogroupsofspecies
duringtheexpansion.Thus,itisplausibletoassumethatundersuchcircumstancesthegauche
conformersrelaxtothemoststablegaucheconformer,whichisthemoststableconformeras
well,andtheanticonformersrelaxtothelowest-energyanticonformer.Itisnotclear,however,
whytheexperimentallyobservedanticonformerisconformer5,whichislessstableby202.35
cm−1thanconformer4aspredictedattheMP2/cc-pVDZleveloftheory.Experimentalevidences
fortheexistenceofconformer5havebeenpresentedalsobymicrowavespectroscopy.[70]To
clarifythisquestion,weperformedadditionalabinitioquantumchemistrycalculationswitha
veryextendedbasissetforthesetwoconformers.Augmentingthebasisset,wehavefoundthat
attheMP2/aug-cc-pVTZleveloftheorywithanimprovedelectroncorrelationconformer5is
energeticallymorestablethanconformer4by45cm−1.Thisfactcombinedwiththepresumed
lowpotentialbarrierbetweenthetwoconformersalsoattheMP2/aug-cc-pVTZleveloftheory
favorsafastandefficientrelaxationofconformer4toconformer5,andhence,theexperimental
observationofthelatter.

3.5.22-PE∙H2OConformer
Comparingtherotationalconstantsfortheground,S0,andthefirstexcited,S1,electronicstate
andthetransitionmomentratioobtainedfromthefitofthehighlyresolvedtwo-photontwo-
colorspectrummeasuredattheparent(m/z=140)channelofthesinglyhydratedcomplexof
2-PEwiththerespectivetheoreticallypredictedinertialparametersandtransitionmomentratios
oftheoptimizedstructuresofthecomplex,wehaveconcludedthatband1w1originatesfromthe
2-PE∙H2OconformerBwherewaterbindssidewaystothemoststable(conformer2)conformer
of2-PEthroughtheformationofastrongsingle3c3hydrogenbond(bondlength1.9Å)wherein
waterdonatesaprotontotheoxygenatomoftheside-chainhydroxylgroupof2-PE.Thethe-
oreticallypredicteddistancebetweentheOatomofthewatermoietyandthenearesthydrogen
atomfromthebenzeneringis2.44Å,whichislikelytosuggesttheexistenceofanadditional
weakσhydrogenbondinwhichthewatermoleculeisaprotonacceptor.Thisassumptionis
somewhatjustifiedbythewellresolvedstructureofthisband,whichistypicalforrigidoral-
mostrigidcomplexeswherethepositionoftheattachedspeciesisfixedandtherelativemotions
ofthetwomoietiesarestronglyconstrained.Sincethetwohydrogenbondsarealmostperpen-
diculartoeachother,theexistenceoftheabove-mentionedadditionalweakσhydrogenbondis
supposedtohinderthetumblingmotionofthewatermoietyaboutthestrongσhydrogenbond
thuspreventingcomplicationoftherotationalstructureofthespectrum.Therelativelysmall
theoreticallypredictedenergygapof94cm−1fromthemoststableconformer(conformerA)can

Discussion3.5

)11S(ConformerDConformer0S)11S(ConformerCConformer0SS014:2:84-85:10:5-71:29:0-24:1:75--391-463-216-0628-725-460-0--131-240-94-0
0.059070.064370.061980.066730.064380.067480.060380.063830.029920.031200.026660.026910.026600.027630.031180.032720.026530.028700.020770.021810.021250.022560.027170.02898
)21S(Conformer4.113.954.194.022.862.523.172.93Å(H15-C1),120.7120.6120.7120.4120.4120.2120.5120.4gde(C1C2C7),(C2C7C8),114.4112.0114.4112.1113.6110.6114.1111.0gde108.8107.4104.2107.9112.9112.1112.9112.1gde(C7C8O9),109.2106.5109.6107.2108.9105.3109.3105.8gde(C8O9H15),90.993.282.988.885.287.584.388.7gde(C3C2C7C8),66.063.867.063.663.258.363.758.9gde(C2C7C8O9),177.0176.9-172.9-177.8-65.5-57.7-87.0-88.2gde(C7C8O9H19),2.182.042.021.892.051.912.011.87Å,(O9-H21)146.2150.6170.8167.3159.3162.5109.4101.9gde,59.167.0174.3-160.261.362.9-26.8-0.6gde,:cm,
BConformer0S)21S(ConformerA1−cmConformerConformerdα1α2α3α4τ1τ2τ1dα5τ4cm,Acm,Bcm,C2µaEelrEelrEelr
ZPVE),1−cm+c21µ(BSSE),−(BSSE:1−1−1−2b
µ

cc-pVDZ/xcited,efirsttheand,S0
/The.elyvrespecti,theoryofelvlecc-pVDZ
theofgeometryoptimizedtheofcalculationcc-pVDZ/
CISandground,theforconstantsrotationaland,angles,dihedral,α
cc-pVDZ/compleydratedhsinglytheofstructuresconformationalstablemostfourtheofgiesenerevrelatiThex.compleOandwithout2-PEofx2
3.5:ableTMP2theatcalculated2-PEofesxcompleydratedhsinglytheofstateelectronic,S1(TM),momenttransitiondipoleelectric2-PEMP2fromobtainedbeenevhagyenervibrationalzero-pointandcorrection(BSSE)errorsuperpositionsetbasistheofinclusionwithvrespectithetorisevinggi2-PEofconformationsparentThecalculations.theofheadertheinparenthesesinreportedareesxcompleydratedhext.teseedetails,orFtable.
2CISfromobtainedbeenhasratio,µc
,angles,planard2:µb
:a2µdistances,interatomicTheoreticalH∙

57

58

ConformationalStructuresof2-PhenylethanolanditsHydratedComplex

alsobeattributedtotheassumedadditionalhydrogenbond.Comparingthegeometricalparam-
eters(interatomicdistances,planaranddihedralangles)ofthe2-PEmolecule(conformer2)in
barestate(seeTable3.3)andinthecomplexwithwater(seeTable3.5),onecaninferthatno
appreciablestructuralchangesordeformationsofthe2-PEspeciestakeplaceuponcomplexation
withwater.Thisdemonstratesthatthebackboneofthehostgauche2-PEconformerisstabilized
byanintramolecularhydrogenbondbetweentheterminalOHgroupandtheπelectronsofthe
benzenering,whichisstrongerthantheintermolecularhydrogenbondsresponsibleforthewa-
terattachment.Aconspicuousdifference,however,isobservedbetweenthetransitionmoment
ratiosof2-PEmonomer(conformer2)anditssinglyhydratedcomplex(conformerB).Sincein
thelattertheπelectronsofthearomaticchromophorearenotinvolvedintheinteractionwiththe
watermoietyitisveryunlikelythatthereisanalterationofthetransitionmomentvectorrelative
tothebenzenering.Weattributetheobserveddrasticchangeofthetransitionmomentratioasa
puremasseffectdescendingfromreorientationoftheprincipalaxesofinertiaduetheattachment
ofawatermolecule.Itisimportanttonotethatafterexcitationofband1w1,the2-PE∙H2Odoes
notfragmentunderthewater-inconditionsoftwo-photonone-colorexperimentsasnotraceof
asignalhasbeenobservedattheparentmasschannelofthemonomer(m/z=122),thispointing
x.complestableatoAninterestingissueisthesearchforthetheoreticallypredictedmoststableconformerofthe
2-PE∙H2Ocomplex(conformerA).SinceconformerAispredictedtobethemoststableone,
thissuggeststhatitmustyieldahighintensityband.Thisis,however,notthecaseforband2w1,
andwetentativelyassignband3w1asoriginatingfromconformerA.Thehighintensityofthis
bandisinlinewiththeabovediscussion.Theunresolvedrotationalstructureofthisbandcan
beexplainedbyatumblingmotionoftheintercalatedwatermoietyabouttheπintermolecular
hydrogenbond.UnlikethecaseofconformerBwherethetwointermolecularbondsarealmost
perpendicular,inthiscase,thetwointermolecularbonds,theOH∙∙∙πandtheH-Oσbondslie
almostonastraightline,andtherotationofthewatermoleculeabouttheπhydrogenbondisnot
hindered,whichleadstoabroadeningandsmearingoftherotationalstructureofthespectrum.
OurassignmentisinlinewiththeoneofHockridgeetal.,[64]andwealsoobservefragmenta-
tionofband3w1intothemasschannelofthe2-PEmonomer(m/z=122).Hockridgeetal.[64]
haveassignedband2w1toconformerB,andband1w1doesnotappearatthemasschannelof
thesinglyhydratedcomplexof2-PE(m/z=122).Wehaveperformedaone-colorexperiment
monitoringthemasschannelofthedoublyhydratedcomplexof2-PEbutwedidnotobserve
anypeaks.Itisveryunlikelythatononehandtheabundanceofdoublyhydratedcomplexesof
2-PEinthemolecularbeambehigh,andontheotherhandtheyfragmentcompletelytosingly
hydratedcomplexessoastoyieldanintensitycommensuratetotheintensityofthemoststable
conformerA(band3w1).

ConclusionsandSummary3.6

Summary3.6Conclusionsand

59

2-PE,thehydroxyanalogofthesimplestaromaticneurotransmitter,2-phenylethylamine,and
itssinglyhydratedcomplexhavebeeninvestigatedbycombinationofmassselectivehigh-
resolutiontwo-colorR2PIspectroscopycombinedwithquantumchemistryabinitiocalcula-
tions.Allprominentblueshiftedvibronicbandsofthe2-PEmonomerupto220cm−1have
beenmeasuredunderhigh-resolutionof70MHz.Onthebasisofcomparisonoftheinertial
parametersobtainedfromtheanalysisofthehighlyresolvedspectraandthoseresultingfrom
thestructuraloptimizationsofthelowest-energyconformationsof2-PE,wewereabletoas-
signall5bandsasresultingfromtwoconformers,thelowest-energygaucheconformerandthe
second-in-energyanticonformerof2-PE.Thisstructuralassignmentiscorroboratedbythefair
agreementbetweenthetheoreticallypredictedvibrationalfrequenciesforthegaucheandanti
conformersandthevibrationalbandpositions.Theobservationofonlytwooutof5theoret-
icallypredictedconformationsundertheconditionsofmolecularjetexpansioninconjunction
withadetailedtheoreticalstudyofthepotentialenergysurfaceof2-PEhasputforwardtheissue
ofconformationalinterconversion.Ourcalculationsdemonstratethatdifferentmembersinthe
gauchemanifoldandintheantimanifoldareseparatedfromoneanotherbyonlylowpotential
barriers,whilethegaucheconformersareseparatedfromtheanticonformersbyasignificant
potentialbarrier.Onthisbasisweconcludethatintheprocessofadiabaticjetexpansionall
gaucheconformersrelaxtothemoststablegaucheconformers,whichisalsothelowest-energy
conformersof2-PE,andtheanticonformersrelaxtothemoststableanticonformer.Thehigh
potentialenergybarrierbetweenthem,however,precludestheanticonformersfromafurther
relaxationtothemoststablegauchegeometryofconformer2.Thiscorollaryimpliesthatenergy
considerationsalonearenotsufficientforanadequateexplanationoftheobservedstructures.It
isimportantalsotherelaxationalkineticsintheexpansionprocesstobeconsidered.
Itisknownthatthepropertiesofmanybiologicallyrelevantmoleculesdependsubstantiallyon
theirenvironment.Toelucidatethisissue,asafirststep,weinvestigatedthecomplexationof2-
PEwithasinglewatermolecule.Comparingtheresultsfromthehigh-resolutionexperimentwith
thosefromabinitiocalculations,wehavefoundthatthemoststableconformersofthesingly
hydratedcomplexof2-PEbelongtothemoststablestructureofthemonomer.Therelatively
scarcevibronicspectrumofthe2-PE∙H2Ocomplex(onlythreebands)impliesthatthemanydif-
ferentconformationsofthesinglyhydratedcomplexof2-PEwhicharetheoreticallysupposed
toexistinthepreexpansionregioneitherdissociateorrelaxtotwostableconformationpresum-
ablyseparatedbyahighpotentialbarrierwhichprecludestheinterconversionofthehigher-lying
conformertothemoststableone.Wedidnotobservesignificantstructuralchangesofthe2-PE
monomeruponitscomplexationwithasinglewatermoleculewithinourexperimentalresolu-
tion,whichisanindicationthatinthecaseofthegaucheconformerof2-PEtheintramolecular
hydrogenbondstilldominatesovertheintermolecularhydrogenbondsresponsibleforthewater

Complex

Hydrated

its

and

2-Phenylethanol

esStructurormationalConfof

60

releantv

species

ofinfluencethe

Fromthiswecanconcludethatamorecompletesolvationshellisnecessarytobring

attachment.

is

of

entsolv

molecules

on

the

structure

of

xiblefle

biologically

andasystematicinvestigationof

aboutstructuralchangesofthehostmolecule,

numberincreasing

intended.

4Chapter

FluorineSubstitutionandNonconventional
OH∙∙∙πIntramolecularBond:
andoscopySpectrUVHigh-ResolutionofCalculationsinitioabophenyl)ethanol-uorp2-(

RosenKaraminkov,SotirChervenkovandHansJ.Neusser,Phys.Chem.
CThehem.Phys.,para-fluorinated10,2852fle(2008)xibleneurotransmitteranalog2-phenylethanolhas
beeninvestigatedbyhighlyresolvedresonance-enhancedtwo-photonion-
izationtwo-colorUVlaserspectroscopywithmassresolutionandabinitio
structuraloptimizationsandenergycalculations.Twostableconformations,
gaucheandanti,separatedbyahighpotentialbarrierhavebeenidentified
inthecoldmolecularbeambyrotationalanalysisofthevibronicbandstruc-
tures.Thetheoreticallypredictedhigher-lyingconformationsmostlikelyre-
laxtothesetwostructuresduringtheadiabaticexpansion.Thelowest-energy
gaucheconformerisstabilizedbyanintramolecularnonconventionalOH∙∙∙π-
typehydrogenbondbetweentheterminalOHgroupofthesidechainandthe
πandelectronstheoreticaloftheresultsphenylring.demonstratesThegoodthatevenagreementthesubstitutionbetweenthewitheaxperimentalstrongly
ableelectroneeffectgationvetheatomstrengthof2-phenandylethanolorientationatofthetheparOHa∙∙∙πpositionbond.hasnonotice-

61

62

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

oductionIntr4.1

Molecularshapeandchargedistributionplayanimportantroleintheselectivityandfunctional-
ityofbiologicallyrelevantmolecules(forreview,seeref.[3]andref.4423).Ithasbeenfound
thattheconformationalstructuresofflexiblemoleculesarestabilizedbyasubtlebalancebe-
tweenvariousnonconventionalweakinteractionsinvolvingneighboringfunctionalgroupsand
localizedcharges.[3,60,85–89]OH∙∙∙πbondsconstituteaparticularclassofnoncovalentweak
hydrogenbonds.TheimportanceofOH∙∙∙πbondshasbeendemonstratedinalargenumber
ofpublicationsonmodelsystemsoverrecentyears.[46,59–63,65,73,90]Aninterestingis-
sueishowmolecularsubstituentsinfluencethestrengthofOH∙∙∙πbondsandhencetheshape
andstabilityoftherespectivemolecularconformations.Fluorinesubstitutionbringsabouta
significantrearrangementoftheelectrondensityofthehostmoleculeand,forthisreason,fluo-
rinesubstitutedmoleculescanbeusedasconvenientmodelsystemstoinvestigatetheeffectof
electronegativegroupsandatomsontheconformationalpreferencesofthestudiedspecies.To
shedlightonthedelicateinterplaybetweenasubstitutedelectronegativeatomandanoncon-
ventionalOH∙∙∙πbondandtheresultingconformationalstructure,wehaveinvestigated2-(p-
fluorophenyl)ethanol(2-pFPE)(Scheme1),whichisthepara-fluoroderivativeoftheextensively
studied[16,46,61,63,70,72,73,84]neurotransmitteranalogmolecule2-phenylethanol(2-PE).
Chapter)viousprealso(see

Figure4.1:Atomlabelsof2-pFPE

Themoleculesskeletoniscomposedofaflexiblesidechainwithahydroxylgroupattheend
andanaromaticringwithapara-fluorosubstitution.Theparentmolecule,2-phenylethanol,has
andattractedtheapresencegreatofinterestainthenonclassicalrecentπ-hyearsydrogenduetoitsintramolecularconformationalbond.preferencesThereforeinittheisgasimportantphase

4.2ExperimentandDataProcessing

63

toidentifythemoststablemoleculargeometriesand,inparticular,tounderstandtheeffectof
electronegativesubstituentsonthegeometricalstructureoftheflexiblemolecule.Thefirstex-
perimentalresultsfromdispersedemissionspectroscopyandtheoreticaldataon2-pFPEhave
beenpublishedrecentlybyChakrabortyandco-workers.[91]Inthisworkwepresentwhatis,
tothebestofourknowledge,thefirstinvestigationbyhigh-resolutionUVspectroscopyonthe
effectoftheelectron-attractingfluorinesubstitutiononthestabilityofthemostfavorablecon-
formersofthisflexiblemolecule.Inchapter3,wehaveshownthatthenonfluorinatedmost
stablegauchegeometryof2-phenylethanolisstabilizedbyaπ-hydrogenbondingofthealco-
holichydrogenofthesidechainwiththephenylringbutthehigher-energyantistructureisalso
presentundermolecular-beamconditions.[90]Here,weusedthehigh-resolutionmass-selective
spectroscopytodetermineunambiguouslyanddirectlytheoriginofmostoftherovibronicbands
intheS1←−S0spectrumof2-pFPE.

4.2ExperimentandDataProcessing

Thesetupusedformass-selectedresonanttwo-photonionization(R2PI)inthepresentwork
hasbeendescribedindetailinSec.2.2.Liquid2-pFPE(97%)obtainedfromAldrichwasused
withoutfurtherpurification,andvaporizedat100◦Cinagasstreamofargonatapressureof
about3.5bars.Thegasmixturewasexpandedinafree-jetexpansionintovacuumthrougha
pulsed-operatedheatablenozzlewithorificediameterof300µm.ToreducetheDopplerbroad-
eningthejetwasskimmedbeforeenteringtheinteractionregionwiththecounterpropagating
laserbeamsthatwereperpendicularlyfocusedontothemolecularbeam130mmdownstreamof
thenozzle.Thelow-resolutionspectraof2-pFPEaswellasitshigh-resolutionspectrumwere
measuredbyR2PIwithmassselectivityinasupersonicmolecularbeam.Thespectralresolu-
tionofabout70MHzofthenarrowbandUVlaserpulsesisnotsufficientforsinglerotational
lineresolutioninthecongestedspectraofamoleculeofthissize.Forthisreason,weuseda
computer-basedmethodforfittingoftherotationalstructureoftheexperimentalhighlyresolved
spectra.Themethodisbasedongeneticalgorithms,whichhasproventobeadecisivetoolin
thedeterminationofthemolecularparameters(seeSec.2.3).Crosscorrelationisemployedasa
figureofmeritformatchbetweenthesimulatedandtheexperimentalspectra.Fortheproduction
ofthesyntheticspectraofthe2-pFPEmonomer,werantheprogramusing300generationswith
500individualsinageneration.[20,46]

64

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

ResultsExperimental4.3

SpectrumResolutionwLo4.3.1

ThevibronicstructureoftheS1←−S0spectrumof2-pFPEwasinvestigatedbylow-resolution
mass-selectedone-colourR2PI.TheresultingspectrumisdepictedinFig.4.2.Itwasmea-
suredbymonitoringthemonomerparentmass(140u)andthemassofthemostprominent
fragment(109u)appearingintheR2PImassspectrum.Thereisadominating00bandandsev-
eralblueshiftedvibronicbands,thestrongestonesbeingat+42,+80and+120cm0−1.Onthered
sideofthestrongestband,thereareonlytwoweakbands,whichareprobablyduetohotbands,
fragmentingdimersof2-pFPE,orcomplexesof2-pFPEandAr.

Figure4.2:One-colorR2PIspectraoftheS1←S0electronictransitionofthe2-pFPEmonomermeasured
atthemonomerC8H9OF(m/z=140)masschannel(a)andthefragmentmasschannelC7H6Frecordedat
m/z=109masschannel(b).

Wchaineofidentifythethemolecularfragmentatcation.109uMostaslikelyoriginatingthisoccursfromtheafterthedetachmentabsorptionofCHof2aOHfurtherfromthephotonside
+intheunchanged2-pFPElocalizationcation.ofThistheresultscharge.intheThisisdiformationfferentoffromthethechargedsituation,fragmente.g.in(C7H6ephedrine,F)withwherean

ResultsExperimental4.3

65

achargetransferoccursbeforefragmentationofthecationbecauseofthepresenceofanitrogen
atomwithaloneelectronpairinthesidechain.[68]Ananalysisofallvibronicbandsinthe
low-resolutionone-colorR2PIspectrashowsaslightlydifferentfragmentationratiofortheband
at+80cm−1observedatthemonomerandthefragmentmasschannels.Apossibleexplanation
forthiswillbegivenbyrotational-bandstructureanalysisofthisbandinthenextsection.To
estimateanupperlimitoftheionization0energy(IE),weperformedtwo-colorR2PIexperiments
offixingthethesecondexcitationphotonphotonstepwisetotothethe00lobandwestinthepossibleS1←−limitS0(35spectrum026cmand−1)sodecreasingthatthetheparenteneriongy
signalstillcouldbeobserved.ThusweobtainedanupperlimitfortheIEof72100cm−1,which
issomewhathigherthantheonefor2-phenylethanol[71]of71500cm−1.Nofragmentationwas
observedinthecaseofthe2-phenylethanolcation.[46]Mostlikelythisisduetoalessefficient
photonabsorptioninthecationafterladderswitchingofthemultiphotonabsorptionfromthe
neutralmoleculetothecation.[92]Torevealpossibleconformationsof2-pFPE,weperformed
high-resolutionexperimentsforresolutionoftherotationalbandstructuretogetherwithabinitio
calculationsemployingextendedbasissets.

pectraSResolutionHigh4.3.2

Thehigh-resolutionUVspectrumofthestrongestmonomerband000andthespectraofthenext
twopronouncedblueshiftedvibronicfeatures(+42and+80cm−1)arepresentedinFig.4.3-4.5
(uppertraces).Allspectrawererecordedatthemonomermasschannel140u.The000andthe
+42cm−1bands(Fig.4.3and4.4)haveasimilarappearanceoftheirrotationalstructurethus
indicatingthattheyoriginatefromthesameconformer.Theywillbediscussedfirst.

66

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

0bandFigurise4.3:assignedasHigh-resolutiontheorigintwoftheo-colorgaucUVheR2PIconformerspectrum1(seeoftheFig.004.6)bandoftherecorded2-patFPEm/z=monomer140.withThe
1−itsinvertedrotationlesstrace:best-fittransitionsimulatedcenteredatspectrum37067.6161(27)yieldingthecm.parametersUpperinTtrace:ablee4.1xperimental(fordetails,spectrum.seetext).Lower

Therotationalconstantsandthetransition-momentratiosweredeterminedwiththecomputer-
cedureassistedandroutinetoshortendescribedbothabosearchveandtimeinandmorespace,detailweinusedSectionas2.3startingTofvaluesacilitatethetheonesfittingreceivpro-ed
fromlow-level(MP26-311++G**)abinitiocalculations.Theresultingbestfitsimulatedspec-
bytraconarevalsoolutingshowntheinFig.theoretical4.3andstick4.4spectra(lowerwithinavertedGaussiantrace).lineThewidthbestof210agreementMHzwasFWHM.obtainedThe
simulatedspectraagreeverywellbothinpeakpositionsandinintensitieswiththeexperimen-
0talthe+ones,42cmwith−1anvibronicachievedband.crossTheecorrelationxperimentallyof97.5%foundforvthealues00forthevibronicrotationalbandandconstants,95.5%thefor
demonstratetransition-momentthatbothratioseandxperimentaltherotationalrotationalspectratemperaturesconsistareofsummarizedprominentinPTandableR4.1.branchesThefitsin
asthea−,wingsb−,andandac-typeminorhybridscontribwithutionafromsignificanttheQbbranch-typeinthecontribcenterution..TheThetwyodospectranotearexhibitassignedsingle
rotationallinesbutregularlyspacedpeakswithaFWHMofapproximately250MHzformed
bytationalaggregationconstantsofsewithveraltheresultsrotationalfromlines.theAabinitiocomparisonofcomputationstheexperimentaldeterminesvthealuestwoforobservthero-ed

Experimental4.3Results

Figure4.4:High-resolutiontwo-colorUVR2PIspectrumofvibronicbandat+42cm−1recordedat
m/z=140.Thebandisassignedtothelowest-frequencyfundamentaltorsionalmodeofthegauchecon-
former1withitsrotationlesstransitionat37109.575(24)cm−1.Uppertrace:experimentalspectrum.
Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable4.1(fordetails,see
xt).te

BandarameterP000cm−1+42cm−1+80cm−1
A´´0.10571(74)0.1059(23)0.14984(45)
B´´0.02409(36)0.02414(38)0.01986(41)
C´´0.02218(36)0.02186(56)0.01809(37)
0.14215(38)0.1020(18)0.10165(77)´A0.01939(43)0.02400(19)0.02409(39)´B0.01853(40)0.02147(36)0.02188(31)´Cµ2a:µb2:µc228:56:1611:58:319:71:20
ν0,cm−137067.6161(27)37109.575(25)37147.621(4)
24(4)8.9(1)8.5(5)(K)TBest-fitcrosscorrelation(%)97.595.492
Table4.1:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,thebandoriginfrequency,ν0,
therotationaltemperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofbands
000,+42and+80cm−1,showninFigures4.3-4.5.Thenumbersinparenthesesrepresentonestandard
deviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevaluesofµa2,µb2,and
µc2inthetransitionmomentratiodoesnotexceed5(%).

67

68

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

1−TheFigureband4.5:isassignedHigh-resolutionasthetworigino-coloroftheUVantiR2PIconformerspectrum4of(seethe+Fig.80cm5)ofbandthe2-precordedFPEatmonomerm/z=with140.
1−itsinvertedrotationlesstrace:best-fittransitionsimulatedcenteredatspectrum37147.621(4)yieldingthecm.parametersUpperintrace:Tablee4.1xperimental(fordetails,spectrum.seetext).Lower

vibronicfeaturesasbelongingtothemoststablegaucheconformer.Theonlyobserveddiffer-
encebetweentheexperimentalandtheoreticalresultsoriginatesfromtheexchangedvaluesof
thea-andc-contributionsinthetransition-momentratiosofthebands(seeTable4.1).The
structureofthehighlyresolvedspectrumofthebandat+80cm−1iscompletelydifferentfrom
thatoftheotherdiscussedbands.ThespectrumfeaturesnicelystructuredP-andR-branches
withregularlyspacedsharppeaks.Fromitsappearance,thespectrumcanbeassignedaspre-
dominantlyb-typewithsmallcontributionsalsofrombotha-andc-types.Apeculiartraitofthe
spectrumistheconvergingseriesofpeaksintherightsideterminatingwithanabruptcut-offat
37149.7cm−1.Toobtaintherotationalconstantsfortheground,S0,andthefirstexcited,S1,
electronicstates,thetransition-momentratioandtherotationaltemperature,weemployedthe
above-mentionedcomputer-aidedfitroutine.Becauseofitsdifferentappearance,presumably
themolecularparametersgivingrisetothisspectrumhavetobedifferentfromtheonesofthe
otherbands.Thereforeasinitialvaluesfortherotationalconstants,weemployedthetheoreti-
callypredictedvaluescorrespondingtotheanticonformers,whichdonotdiffermuchfromeach
otherascanbeseenfromTable4.2.Theresultingbest-fitsimulationisshowninFig.4.5(lower
invertedtrace)andthemolecularparametersbringingforththisfitaresummarizedinTable4.1.
ThefitwasabletoreproduceexcellentlyboththepeakpositionsandpeakintensitiesintheP-
andR-branchesand,inparticular,theconvergingbehaviorandthefalloffoftherightsideofthe
spectrum.Somediscrepanciesbetweentheexperimentalandthesimulatedspectraareobserved,
however,inthecentralpart.Thisisanindicationthatsomephenomenanotaccountedforby
thetheoreticalmodelofarigidasymmetric-topHamiltonianemployedforthesimulationofthe

ResultsExperimental4.3

Figure4.6:Potential-energysurface(invertedforthesakeofabettervisualization)asafunctionofthe
C2C7C8O9andC7C8O9H19dihedralanglesofthe2-pFPEmonomer.Thegaucheandanticonforma-
tionsareseparatedbyahighpotentialbarrierintheC7C8coordinate.Thefiveenergeticallymostfavor-
ableconformationsofthe2-pFPEmonomerhavebeenfullygeometricallyoptimizedintheirground,S0,
electronicstateattheMP2/ccpVDZleveloftheory.Thefullyoptimizedstructuresaredepictedontheir
correspondingminima(maximaintheFigure)onthepotential-energysurface.

69

spectrumareinvolved,whichleadtospectralshiftsorbroadeningofthepeaks.Amoreinvolved
commentonthisissueispresentedinthediscussionsection.

CalculationsinitioAb4.3.3

Abinitioquantum-chemistrycalculationsusingtheGaussian03programspackage[82]have
beenperformedtostudytheoreticallytheconformationalstructure,vibrationalmodes,anden-
ergeticsofthe2-pFPEmonomer.Toidentifythestableconformationsofthe2-pFPEmonomer,
weperformedagridsearchexploringtheenergycorrespondingtothevariousshapesoftheside
chainof2-pFPE.ThegridsearchwascarriedoutattheMP2/cc-pVDZleveloftheory,andin-
cludedascanoftheC2C7C8O9andC7C8O9H19dihedralangles.TheC3C2C7C8dihedral
anglewaskeptfixedat90°.Thethree-dimensionalpotential-energysurface(PES)isdepictedin
Fig.4.6(forthesakeofabettervisualization,thePESisinverted).Itfeaturessixpotentialmin-
imapointingtosixstableconformationsofthe2-pFPEmonomercorrespondingtothreegauche
andthreeantistructures.Thedeepest(global)minimumbelongstothegaucheconformer(con-
former1),inwhichtheterminalOHgroupofthesidechainfacestheπelectronsofthebenzene
ring.Thetwoenergyminimaseparatedfromeachotherandfromtheglobalminimumbylow

70

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

potentialbarrierscorrespondalsotogauchestructuresbutwithadifferentorientation(different
dihedralangleC7C8C9H19)oftheOHgroupofthesidechain.Theseconformationsaredes-
ignatedasconformer3andconformer5,respectively.Inallantigeometries,thesidechainis
extendedawayfromthebenzenering(dihedralangleC2C7C8O9=180°),thedifferencebe-
tweenthembeingonlytheorientationoftheOHgroup(dihedralangleC7C8C9H19).Two
oftheantistructures,however,withdihedralanglesC7C8C9H19=63.91°andC7C8C9H19=-
63.91°respectively,arecompletelysymmetricalwithrespecttotheC2C7C8O9symmetryplane
andgiverisetotwoequivalentpotentialenergyminima.Thatiswhyhereinafteronlyoneof
thesestructures,theconformerwithdihedralangleC7C8O9H19=-63.91°,willbeconsidered.
Itisreferredtoasconformer4.Theotherdifferingantistructureisconformer2withafully
extendedsidechaininwhichthedihedralangleC7C8O9H19=180°.Itisinterestingtopoint
outthattheanticonformationsareseparatedbylowpotentialbarriersfromeachotherandby
ahighbarrieralongtheC2C7C8O9coordinatefromthegauchestructures.Theso-identified
fiveconformationshavebeenfurthersubjectedtoafullstructuraloptimisation,frequencyanal-
ysisandcalculationoftheirenergeticsintheground,S0,electronicstateattheMP2levelof
theoryemployingthecc-pVDZbasisset.Fullstructuraloptimisationandfrequencyanalysis
havealsobeenperformedforthefirstexcited,S1,electronicstateattheCIS/cc-pVDZlevelof
theory.Thevaluesofthetypicalinteratomicdistances,planaranddihedralanglesdefiningthe
conformationalshape,therotationalconstants,thetransition-momentratioandtheenergieswith
andwithoutinclusionofthezero-pointvibrationalenergy(ZPVE)ofthe5stableconformations
calculatedusingthecc-pVDZbasissetaresummarizedinTable4.2.Inthecalculationofthe
zero-pointvibrationalenergies,pertinentscalefactorshavebeenused.[93,94]Allcalculated
vibrationalfrequenciesarepositive,thisbeingevidencethattheoptimizedstructuresindeedcor-
respondtopotentialminima.ThefullyoptimizedconformationalstructuresaredepictedinFig.
4.6atoptheircorrespondingpointonthepotential-energysurface.

Discussion4.4

Themoststablecalculatedgeometryoftheconformers,withoutandwiththeinclusionofthe
ZPVE,correspondstothegaucheconformer1withitsterminalOHgroupofthesidechain
pointingtotheπelectronsofthebenzenering.Thisstructureisassumedtobestabilizedby
anintramolecularOH∙∙∙πhydrogenbond,asisalsothecaseinthe2-phenylethanolmonomer.
[16,46,70]Thenext-in-energystructurelies446cm−1above(seeTable4.2)themoststable
geometryandbelongsalsotoagaucheconformer3,whichdiffersfromconformer1onlybythe
orientationoftheOHgroup,inthiscasepointingupwards.Thethird-in-stabilityconformation
correspondstotheantistructurewithitsterminalOHgrouppointingsideways(conformer4)
whichisdistancedby556cm−1fromthelowest-energyconformation,andby110cm−1from

Discussion4.4

firsttheand,0elvlecc-pVDZmonomerFPE
hegauc1S/p
Scc-pVDZ/theofcalculation
SCIS5andConformer0cc-pVDZCIS/fromforconstantsrotationaland,ground,theτ2µc
ConformerS0.theoryofelvlecc-pVDZ/
1antiS4obtainedS02-theofstructuresconformationalstablemostevfitheof
beencalculatedmonomerFPEMP2theat
hashe:1b2Sgaucµangles,S00:89:11-0:100:0-5:91:4-0:100:0-5:94:1--594.2-614.8-517.2-712.2-0-572.0-556.0-446.2-617.4-0
:30.108730.138170.140110.105100.104140.115390.111070.137150.138670.109380.023020.023300.019680.019470.023560.023770.019710.019510.023830.024290.019600.020120.018220.018100.020610.020800.018250.018140.021460.02198
a2µConformerpdihedral2-elrtheE,α120.0120.9120.0120.6gde(C1C2C7),121.4121.4119.8120.7120.3120.9114.1114.1113.2111.4114.6112.8113.0111.1113.7111.0gde(C2C7C8),113.2113.2111.7112.8109.0107.6107.3107.5113.1112.3gde(C7C8O9),105.9105.9109.4106.2109.6107.0109.4106.8109.2105.7gde(C8O9H19),119.3119.390.689.3107.0106.789.188.592.590.8gde(C3C2C7C8),69.469.4177.2176.672.169.0-180.0-180.066.360.4gde(C2C7C8O9),66.166.1-68.4-63.9-170.8-169.4-180.0-179.9-66.7-63.3gde(H19O9C8C7),cmcmcmµ:,4.2:ableTxcited,eratio(TM)momenttransitiondipoleelectricThe.elyvrespecti,theoryof2-theofgeometryoptimizedMP2fromobtainedbeenevhagyenervibrationalzero-pointtheofinclusionwithandwithout
ofConformerarameterP3.453.45--3.873.77--2.762.55Å(H19C1),r1.351.351.321.351.321.351.321.351.321.35Å(C5F13),rα1α2α3α4τ1τ2τ3,A,BC2µaEE
1antiSconformers2ConformerS0angles,planar,danti
1gies,enerevRelati.monomerFPE
andhehegauc1Sgaucpcmdistances,interatomicTheoreticaltheofstateelectronic,
1−1SZPVE),c21µ−:(incl.1−1−1−2bcm
elelrr,

71

72

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

theprecedingstructure.Conformer4isfollowedbythegaucheconformer5,theenergygap
betweenthem,however,beingverysmallatonly16cm−1.Becauseoftheverysmallenergy
difference,itisverylikelythatthelatterenergyorderingmayalteruponchangingtheemployed
modeltheoryandbasisset.Thehighest-energyconformationat+617cm−1istheanticonformer
2withacompletelyextendedsidechainawayfromthebenzenering.Weassignthestrongest
peaktothe000originbandofthemoststablegaucheconformer1.Thesameconformergivesrise
alsotoband+42cm−1.Theassignmenthasbeenmadeonthebasisoftheverygoodagreement
ofthetheoreticallypredictedrotationalconstantsforthisconformer(seeTable4.2)withthe
onesobtainedfromthefitofthehighlyresolvedspectrabothfortheground,S0,andthefirst
excited,S1,electronicstates(seeTable4.1).Thevibrationalanalysisperformedforthefirst
excitedelectronicstate,S1,ofconformer1showsthatthelowest-frequencyvibrationalmodeis
at47cm−1andcorrespondstoatorsionofthesidechainabouttheC2C7bond.Thisresultisin
agoodaccordwiththepositionofthebandat+42cm−1whichprovidesaconfidentargumentto
assignthisbandtothefundamentaltorsionalmodeofconformer1.Somediscrepancybetween
theoryandexperiment,however,isobservedinthetransition-momentratioforbothbands.This
mightbeduetothelimitedaccuracyoftheexcitedstatecalculationattheCIS/cc-pVDZlevelof
theory.Therotationalstructureofthebandat+80cm−1differssignificantlyfromthestructure
ofthestrongestband(000)andthebandat+42cm−1.Comparingtherotationalconstantsfor
thegroundandthefirstexcitedelectronicstatesobtainedfromthefitofthespectrumwiththe
onestheoreticallypredicted,wehaveconvincinglyestablishedthatthisbanddescendsfroman
anticonformer.Thisimplicationisinaccordwiththeresultsfromthedispersedfluorescence
experimentsofPanjaetal.,[91]demonstratingthatthisbanddoesnotoriginatefromthecon-
formerbringingforththemainbandandthebandat+42cm−1.Itisworthpointingout,however,
thatthetheoreticallypredictedvaluesoftherotationalconstantsfortheanticonformers2and
4areveryclose,whichisnotsurprisinginviewoftheverysimilarstructuresofthetwocon-
formersdifferingonlybytheorientationoftheterminalOHgroup,whichpresumablydoesnot
significantlyaffecttheprincipalmomentsofinertiaandhencetherotationalconstantsofthetwo
conformers.TheexperimentalvalueofconstantAdifferssomewhatfromthecalculatedvalues
forthisconstantforbothanticonformers,2and4.Constrainingtherotationalfitinthevicinity
ofthetheoreticallypredictedvalueofconstantAdidnotyieldacceptableresults.Goodfitswere
producedonlyafterliftingtheconstraintsonconstantAwhereuponitsvaluesdriftedawayfrom
thetheoreticallypredictedones.Comparingtherotationalconstantsforband+80cm−1alone
inthiscaseallowsfortheassignmentofthebandtoanantistructure,butprecludesadecisive
assignmentofband+80cm−1tooneofthetwoanticonformers2or4.Energyconsideration
providesatentativefavourforconformer4asitistheoreticallypredictedtobemorestableby
45cm−1comparedtoconformer2.Theseresults,however,shouldnotberegardedasarig-
orousargumentsincetheenergygapisverysmalland,aswehavedemonstratedforthe2-PE
monomer[46](seepreviousChapter)theenergyorderingmaychangeuponchangingthelevel
oftheory.Ithasbeennotedabovethatdespitetheexcellentagreementbetweentheexperimental

Discussion4.4

73

andthesimulatedspectrainFig.4.5inbothpeakpositionsandpeakintensitiesinthewings
(P-andR-branches),thecentralpartofthespectrumcouldnotbefittedsatisfactorily:twoof
thepeaksinthesimulatedspectrum(peaksBandCinFig.4.5)areshiftedwhileanotherone
(peakAintheexperimentalspectruminFig.4.5)isverybroadenedandspreadout.Duetothe
increasedbackgroundinthisregion,itisverylikelythatsomespectralcontaminationispresent
thatmayderivefromfragmentationofhydratedcomplexes.Thisisaplausibleassumptionin
viewoftheobservedatthispositionbandofthesinglyhydratedcomplexof2-pFPEmeasured
atitsmasschannelunderlow-resolutionconditions.[95]TheobservedspectralshiftofpeakC
by0.07cm−1fromthetheoreticalpositionisanindicationthatalocalizedperturbationispresent
inthecentralpartofthespectrum.Weperformedananalysisoftherotationalstructureofthe
centralregionofthehighlyresolvedspectrumofband+80cm−1.Usingtheparameterspro-
ducingthefitshowninFig.4.5,wegeneratedastickspectrumusingthePGOPHERprogram
package[96]andintheso-producedspectrum,wemadeaFortratanalysisandanassignmentof
therotationaltransitionsassociatedwiththesharppeakCappearingtotherightofthecentraldip
justbeforetheonsetoftheR-branch.Allrotationallinesundertheenvelopeofthispeakturned
outtooriginatefromtransitionstosmallJ´(J´<10)-valuerotationallevelswithidenticalvalueof
Ka´(Ka´=3)intheexcitedelectronicstate.Thisfindingpointstoacouplingofrotationallevels
withsmallquantumnumberJandwiththesamevalueofKa´=3asother,dark,levels,[97]
whiletheotherrotationallevelswithhighervaluesofJanddifferentvaluesofKa(givingrise
tothemajorityofthespectrallines)remainunaffected,whichleadstoagoodfitinthewings.
Sincethisbandisassignedtothevibrationless000bandoftheanticonformer,thequestionarises
astowhattheidentityofthecoupledstatesis.TheterminalOHgroupinbothanticonformers,
2and4,doesnotformhydrogenbonds(asisthecaseforthemoststablegaucheconformer1)
and,forthisreason,thetwoconformationsareseparatedbyasmallpotentialbarrier(estimated
fromthePESdepictedinFig.4.6tobeca.570cm−1)thusfavoringtheinternalrotationof
theOHgroupabouttheC8O9bond.Inthiscase,itisverylikelythatacouplingbetweenthe
internalrotationoftheOHgroupandtheoverallrotationofthemoleculeoccursforcertainrota-
tionalenergylevelswhichareinresonancewitheachother.Thiscouplingmayleadtoshifting
andpushingaparttheindividualrotationallines,thelatterleadingtoaslightbroadeningofthe
observedpeaks.Anotherplausibletypeofcouplingleadingtosimilarspectralshiftsofonly
0.07cm−1andbroadeningmaydescendalsofromtheinteractionbetweentheabove-discussed
rotationalenergylevelsoftheanticonformerwithhigherrovibroniclevelsofthelowest-energy
gaucheconformer1.Thoughinthiscasetheexpectedcouplingisweak,thisdoesnotcontradict
theobservedspectralshiftofonly0.07cm−1oftheresultingeigenstates.Aninterestingissuethat
meritsdiscussioniswhyonlytwoconformershavebeenobservedoutof5theoreticallypre-
dictedstablestructureswhichdifferlittleinenergy.Theanswertothisquestioncanbefound
intheshapeofthepotential-energysurfaceandtheresultingrelaxationdynamicsofthecooling
process.ThePESinFig.4.6showsthatthegaucheconformationsareseparatedfromeach
otherbyonlysmallpotentialenergybarriersoflessthan600cm−1.Thesameistruealsofor

74

High-ResolutionUVSpectroscopyof2-para-uorophenylethanol

theanticonformers.Ontheotherhand,however,thegaucheconformersareseparatedfromthe
anticonformersbyahighpotentialbarrierofca.2000cm−1.Evenifinthepre-expansionregion
allconformersarepopulated,uponadiabaticexpansionthehigherlyingconformersrelaxtothe
lowest-energyoneswithinthegaucheandtheantimanifolds,respectively,becauseofthelow
potentialbarriersbetweenthegaucheandtheantispecies,respectively.Thusthehigher-energy
gaucheconformersrelaxtothemoststablegaucheconformer1,theprocessofrelaxationbeing
favoredbythelowpotentialbarriersbetweenthegauchespecies.Anidenticalmechanismcan
explaintherelaxationoftheanticonformerstothemoststableantistructure.Theinterconver-
sionbetweentheantiandthegauchestructures,however,ishinderedbythehighbarrieralong
theC7C8coordinate(seeFig.4.6).Thisprecludesafurtherrelaxationofthelowest-energy
anticonformertothemoststablegaucheconformer1priortophotoexcitationandthusthefor-
merremainstrappedinitslocalminimumatthelowtemperatureofthemolecularbeamandwe
observeonlytwoconformers:onegaucheandoneanti.Ontheotherhand,theobservedlocal
perturbationofband+80cm−1,whichcorrespondstoananticonformation,demonstratesthat
afterphotoexcitation,anisomerizationprocessmayoccur.Atthelowexcessenergyofthe000
band,however,theobservedlocalperturbationofthisbandleadingtoasmallshift(lessthan0.1
cm−1)ofspecialrotationallineswouldresultinaslowisomerizationoccurringintherangeof1
nsinthestatisticallimit.Athigherexcessenergies,fasterisomerizationcanbeexpected.Ifthis
occursbeforetheabsorptionoftheionizationphoton,thismayleadtoadifferentisomerization
pathwayandtoaselectivedifferentfragmentation,similartotheoneofourrecentobservationsin
ephedrine.[68]Furthermore,itissurprisingtoobservethattherotationaltemperatureoftheanti
conformer(band+80cm−1)resultingfromthefitistwiceashighasthatofthegaucheconformer
(themainbandandthebandat+42cm−1)(seeTable4.1).Aninterestingaspectthatdeserves
specialattentionisthecomparisonbetween2-phenylethanol(2-PE)[16,46,61,70](seeprevious
Chapter)and2-pFPE.Inbothspeciesthemoststableconformeristhegaucheonewhosecon-
formationalstructureisstabilizedbyanintramolecularnonconventionalOH∙∙∙πbondbetween
theterminalOHgroupandthebenzenering.Thepotential-energysurfaceisverysimilarfor
bothspecies,thoughtheirorderingaccordingtotheenergyisdifferent:in2-PEtheorderingis
gauche,anti,gauche,gauche,anti,whereasfor2-pFPEtheorderingisgauche,gauche,anti,
anti,gauche.Theenergygapbetweenthemoststablegauchestructureandthenext-in-energy
oneisalsoslightlydifferentthoughofthesameorderofmagnitude:586cm−1for2-PEvs.446
cm−1for2-pFPE.Thedifferentorderingoftheconformationsmaybearesultoftheredistribu-
tionoftheπelectrondensityofthebenzeneringcausedbythefluorineatom.Thecovalently
boundfluorineatombringsaboutasignificantbreakingofthesymmetryoftheπelectronden-
sityofthebenzenering,whichismorepronouncedthantheonecausedbythesidechain,and
leadstoanexcessofπelectrondensityinthevicinityofthefluorineatomandadepletionofthis
densityintheareaclosetothesidechain.Thisdifferentdistributionoftheπelectrondensityis
notexpectedtosignificantlyinfluencethebindingstrengthandthestabilityofthemoststable
gaucheconformerinwhichthesidechainisbenttothearomaticringandstabilizedbyanon-

ConclusionsandSummary4.5

75

conventionalOH∙∙∙πbond.Butontheotherhand,thedepletedπelectrondensityclosetothe
sidechainaffectsthestabilityofthesidechainitselfandhencetheenergyorderingoftheother
conformationsinwhichtheOHgrouppointsawayformthering.Inthiscasethereducedπelec-
trondensityofthearomaticringnexttothesidechaininducesaredistributionoftheelectronic
densitywithinthesidechainofthelatterandhenceaffectsitsconformationalstabilityandthe
conformations.theofordering

ConclusionsandSummary4.5

Thepara-fluorinatedflexiblemolecule2-phenylethanol,thehydroxyanalogueofthesimplest
neurotransmitter,2-phenylethylamine,hasbeeninvestigatedbycombinationofhigh-resolution
R2PItwo-colormass-selectivelaserspectroscopyandquantum-chemistryabinitiocalculations
onitsstructureandenergetics.Inthiswork,allprominentvibronicbandsinthelow-resolution
spectrumofthe2-pFPEmonomerupto120cm−1havebeenmeasuredunderhighresolutionof
70MHz.Comparingthevaluesoftherotationalconstantsobtainedfromthefitsofthehighly
resolvedspectrawiththetheoreticallypredictedonesforthe5lowest-energyconformations,we
wereabletoconvincinglyassignallofthesevibronicbandstotwoconformationalgeometries:
thelowest-energygauchestructureandahigherlyingantistructureof2-pFPE.Theobserva-
tionofonlytwooutoffivetheoreticallypredictedconformationsrequiresacloseinspection
ofthepotentialenergysurface.Thelatterclearlydemonstratesthatthegaucheconformersare
separatedfromeachotherbylowpotentialbarriersandthesameholdsalsofortheanticon-
formers.Thegaucheconformations,however,areseparatedfromtheantistructuresbyahigh
potentialbarrier.Thisfavorstherelaxationofalloriginallyproducedgaucheconformerstothe
moststableconformerinthegauchemanifoldandtherelaxationofallinitiallyproducedanti
conformerstothemoststablestructurewithintheantimanifold.Thehighbarrierbetweenthe
moststablegaucheconformerandthehigher-energyanticonformation,however,precludesthe
furtherrelaxationoftheantistructuretothemoststablegaucheone.Wehaveobservedlocal
perturbationsintherotationalstructureofthehigher-lyinganticonformer.Thispointstoacou-
plingofspecialrotationallevelsoftheanticonformertorovibronicstatesofalower-energy
conformationinthefirstexcitedelectronicstate,S1.AthigherexcessenergyintheS1state,
thismayresultinisomerizationprocessesinthestatisticallimit.Fromourinvestigation,we
havefoundevidencethatthefluorinationattheparapositionof2-phenylethanoldoesnotalter
theshapeofthegaucheconformerstabilizedbyanintramolecularOH∙∙∙πbondandthatthis
conformerremainsthelowest-energyone.Thisimpliesthatevenupontheconsiderablechange
oftheπ-electron-densitydistributioninthearomaticringbroughtaboutbytheattachmentofa
stronglyelectronegativeatom,theintramolecularnonconventionalOH∙∙∙πbondisnotweakened
appreciably.Thisresultdemonstratesthesignificanceandrobustnessofthistypeofbondforthe

76

stabilization

of

the

High-Resolution

conformational

structures

UV

of

oscopySpectr

xiblefle

of

para2-

molecules.

ophenylethanol-uor

77

78MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

5Chapter

IonizationesholdThrAnalyzedMassFlexibleofoscopySpectr2-para-uorophenylethanolConformers
IntramolecularanwithoutandwithOH∙∙∙πBond

R.Karaminkov,S.ChervenkovandH.J.Neusser,Phys.Chem.Chem.
Phys.11,2249(2009).
Thecationicstateoftheprototypeflexiblemolecule2-para-
fluorophenylethanolhasbeeninvestigatedbycombinationofmass-analyzed
threshold-ionization(MATI)spectroscopyandquantumchemistryabinitio
densityfunctionaltheory(DFT)calculationsemployingtwodifferent
functionals:theB3LYPfunctionalandthenewhybridfunctionalM05.The
MATIspectrameasuredviavibronicbandsintheS1intermediatestatebe-
longingtothemoststablegaucheconformerstabilizedbyanintramolecular
OH∙∙∙πhydrogenbondarestructureless,whilethespectrarecordedviabands
oftheanticonformerfeaturewell-resolvedpeaks.Thisresultisinagood
accordwithourtheoreticalpredictionsshowingthatuponionization,theanti
conformerretainsitsstructure,whilethelowest-energygaucheconformer
undergoesasignificantstructuralchangeresultinginabreakoftheOH∙∙∙π
bond.Thisandthegoodagreementbetweenthemeasuredbandpositions
andthetheoreticallypredictedfrequenciesforthecationicanticonformer
confirmtheconformationalassignment.Theresultforthecationprovides
clearevidencefortheexistenceofanonclassicalintramolecularOH∙∙∙π
hydrogenbondiftheelectrondensityinthearomaticringissufficientlyhigh
asisthecaseonlyfortheneutralmolecule.

oductionIntr5.1

oductionIntr5.1

79

Inrecentyears,avarietyofspectroscopictechniques[3,20,70,85,87,98–100]havebeenused
todeducetheconformationallandscapeofflexibleprototypemolecularsystemsofbiological
relevance.Animportantmodelsystemis2-phenylethanol(2-PE),thehydroxylanalogueofthe
neurotransmitter2-phenylethylamine,whosemoststableconformerintheground,S0,andfirst
excited,S1,electronicstatesisstabilizedbyanon-conventionalOH∙∙∙πbond.[46,90]Inor-
dertoexplorethestabilityandthestrengthofthisnoncovalentweakhydrogeninteraction,we
introducedperturbationoftheπ-electronsystemofthebenzeneringbyfluorinesubstitution
attheparaposition,andanalyzeditseffectontheconformationalstructures.(SeeChapter4.)
Spectroscopicdataon2-(para-fluorophenyl)ethanol(2-pFPE)(seeScheme1)werepublishedby
Chakrabortyetal.[91]andcomplementedbyastudyfromourgroup(seeChapter4)witha
clearassignmentofthegaucheandantistructuresbyhighresolutionUVspectroscopy.Thetwo
conformersareseparatedbyarelativelyhighpotentialbarrierofabout2000cm−1intheground
neutralstate,(seeFig.4.6)whichundermolecularbeamconditionsprecludestherelaxationof
thehigherlyinganticonformer(ca.570cm−1)tothelowestenergygaucheconformer,stabi-
lizedbyaπ-hydrogenbondingoftheterminalhydroxylgroupofthesidechainwiththephenyl
ring.Anevenstrongerperturbationofthep-electrondensityiscausedbytheionizationofthe
molecule.Dessentetal.havestudiedtheionization-inducedconformationalchangesofaromatic
moleculeswithasidechainbythresholdionizationtechniques.[101]Inthiswork,forthefirst
timewehaveappliedthemass-analyzedthreshold-ionization(MATI)spectroscopytoinvestigate
thefateofanOH∙∙∙πbondingthatexistsintheneutralmolecule,aftertheionizationprocess
takingaπelectronoutofthering.Furthermore,existenceandstabilityofgaucheandanticon-
formersafterionizationareinvestigated.TheobtainedMATIspectraprovideinformationon
theconformer-specificionization-inducedgeometrychangesandyieldadditionaldataforthe
conformers.theseofidentificationspectroscopic

Experimental5.2

Thetwo-photonexcitationschemeinaMATIexperimentemployspromotionofthemolecules
andfromtheirsubsequentground,eS0xcitation,statetotovaeryhighselectedRydbervibronicgstatesstateinwhichthearefirstefinallyxcited,Sionized1,byelectronicadelayedstate
detailpulsedinSec.electric2.4.1field.andour[51–53]preTheviousexperimentalpublications.setup[102,of103]theBrieflyMATI,eliquidxperiment2-pFPEis(puritydescribed97%)in
waspurchasedfromSigma-Aldrich,usedwithoutfurtherpurification,andheatedto110.A
mixtureofthesamplevaporwiththebuffergasargonat3barwasexpandedintovacuumthrough
apulsed-operatedGeneralValvenozzlewithanorificediameterof500µm.Theso-produced

80MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

molecularbeamwasskimmedbeforeenteringtheinteractionregionwithtwocounterpropagating
laserbeamsthatwereperpendicularlyfocusedontothemolecularbeam150mmdownstreamof
thenozzle.TheexcitedRydbergmoleculesareseparatedfromthedirectlyproducedphotoions
inthefirststageoftheaccelerationregionofalinearreflectingtimeofflightmassspectrometer
(RETOF)byaweakelectricfield(ca.0.180.22Vcm−1)andionizedbyastrongelectricfield
of1000Vcm−1inthesecondstage.Theresultingthresholdionsareacceleratedandinjected
bythesameelectricfieldintothedriftregionofthelinearRETOFmassspectrometer.[55]The
massresolvedsignalsaredetectedatdifferentmasschannelsandrecordedwithgatedintegrators,
digitized,andfinallyprocessedinapersonalcomputerunderLabVIEWenvironment.

Results5.3

spectraREMPI5.3.1

Theone-colorresonanceenhancedmulti-photonionization(REMPI)spectrumoftheexcited,
S1,electronicstateof2-pFPEisshowninFig.5.1.Itcomplementsthealreadyshownspectrum
inthepreviouschapterwithanextendedscantotheblueside.Inadditiontothevibrationsinthe
neighborhoodofthestrongbandat37067.6cm−1,astrongbandshiftedby553cm−1totheblue
isseen,whichcorrespondstothefalseoriginν6vibronicbandinbenzene.[97]Inthepreceding
chapterwehaveidentifiedthepronouncedbandat37067.6cm−1astheoriginofthegauche
conformer,whichisstabilizedbyanOH∙∙∙πhydrogenbond.Therotationalstructureofthe
weakbandat80cm−1totheblueofthestrongestbandwasfoundtobecompletelydifferentand
assignedtotheoriginoftheanticonformerbyafitoftherotationalstructure.Usingthetypical
experimentalvibrationalfrequenciesfoundforthegaucheconformer,wetentativelyassignthe
weakbandsat+147cm−1and+632cm−1toprogressionbandsoftheanticonformer(seebelow).
Toshedlightontheidentityofthesebandsandtoexaminethebehaviorofthegaucheandtheanti
conformersafterionization,weperformedMATIscansaftertwo-photonexcitationviavarious
intermediatevibronicstatesinthefirstexcited,S1,electronicstate.

spectraTIMA5.3.2

Asafirststep,inFig.5.2thetotalioncurrentisshownasafunctionofthetwo-photonexcitation
energymeasuredbyfixingthefirstphotonenergytoeitheroriginintheS1intermediatestateat
37067.6cm−1or37147.6cm−1oftherecentlyidentifiedtwoconformersunderconsideration,
andscanningthesecondphotonenergy.ThetotalioncurrentinFig.5.2b,measuredviathe000
bandoftheanticonformerinS1intermediatestateshowsastep-likebehaviorwithtwosteps

Results5.3

Figure5.1:One-colourS1←S0REMPIspectrumof2-pFPE,recordedatthemonomermasschannelm/z
=140.Conformerassignmentsandrelativepeakpositionsareincludedinthespectrum.Bandsoriginating
fromgaucheandanticonformersaredesignatedby(g)and(a),respectively.

81

separatedbyonly40cm−1,whilethetotalioncurrentspectrumviathestrongpeakat37067.6
cm−1(the000bandofthegaucheconformer)doesnotshowastep-likebehaviourbutagradu-
allyincreasingtotalioncurrent(Fig.5.2a).Next,usingthetwo-colorexcitationscheme,only
thresholdionsaredetected.TheresultingMATIspectrumofthegaucheconformerof2-pFPE
measuredviatheS1,000originat37067.6cm−1asafunctionofthetwo-photonenergyisshown
inFig.5.3a.Thespectrumisfeaturelessandnoisywithasmoothonsetataround71200cm−1
andanincreasingintensityforhigherexcitationenergies.InFig.5.3b,theMATIspectrumis
shownwhenmeasuredviatheS1,000bandat37147.6cm−1,whichhasbeenassignedtothe
originoftheanticonformerof2-pFPE.[104]Itdisplaysawell-resolvedvibronicstructure.No
peakisobservedontheredsideofthestrongpeakat71238cm−1.Thisallowsustodetermine
theadiabaticionizationenergy(AIE)as71238±5cm−1(ionizationenergiesaregivenwithout
fieldcorrection[49]).InadditiontotheMATIspectraviatheS1originsoftheobservedtwo
conformersof2-pFPE,wehaverecordedMATIspectraviatwoothervibronicbandsintheS1

82

MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

Figure5.2:Totalioncurrentspectraof2-pFPEmeasuredviagauche(a)andanti(b)originsofthetwo
conformersintheS1electronicstate.

intermediatestate.Theresultingspectraviathebandsat+632cm−1and+147cm−1intheS1
intermediatestateareshowninFig.5.4aandbandthespectrummeasuredviatheoriginofthe
anticonformerintheS1intermediatestateisshowninFig.5.4cwithanextensionupto864cm−1
excessenergy.EachofthetwospectrainFig.5.4aandbconsistsofseveralwell-pronounced
peaks,whichareshiftedtohigherenergyby574cm−1and66cm−1,respectively,fromtheAIEat
71238cm−1.TheseshiftsaresimilartotherespectiveblueshiftofthevibronicS1intermediate
states.ThecorrespondingpositionsofthelowfrequencyvibronicMATIpeaksmeasuredvia
differentintermediatestatesaremarkedwithdottedlinesinFig.5.4bandc.Thepresenceofa
similarvibronicstructureintheD0cationicstatemeansthatthespectraoriginatefromtheanti
conformerofthe2-pFPEmoleculeandassignsthecorrespondingintermediatestatebandstothe
.conformeranti

Results5.3

Figure5.3:MATIspectraof2-pFPEmeasuredviagauche(a)andanti(b)originsofthetwoconformers
intheS1electronicstate.Relativebandpositionsareincludedinthelowerspectrum.

5.3.3esultsrComputational

83

Theoreticalquantumchemistrycalculationofthestructure,energetics,electrondensitydistribu-
tionandthevibrationalfrequencieswithoutandwithinclusionofanharmonicityofthe2-pFPE
cationicconformershavebeenperformedusingtheGaussian03suiteofprograms.[105]Wecon-
sideredthemoststablestructuresofthe2-pFPEmonomerintheneutralground,S0,electronic
state(seeChapter4)asstartinggeometriesfortheoptimizationsinthecationicstate.Den-
sityfunctionaltheorywiththecommonlyusedB3LYPfunctional,therecentlydevelopedhybrid
M05functional[106]andaug-cc-pVTZbasissetwereemployedforthestructuraloptimizations,
energetics,electrondensitydistributionintheHOMO,andcalculationsofthenormal-modefre-
quencies.B3LYP/cc-pVDZandM05/cc-pVDZleveloftheorywasusedforthecomputationof
theanharmonicfrequencies.Forthemoststablegaucheconformerintheneutralwehaveap-
pliedalsothecomputationallymoredemandingUMP2leveloftheorywithtwodifferentbasis

84

MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

Figure5.4:MATIspectraof2-pFPEviathe+632cm−1(a),the+147cm−1(b),andtheanti000(c)bands
intheS1intermediateelectronicstate.Thenumbersaretherelativebandpositionstothe000bandat71
1.−cm238

sets,6-31/G(p,d)andcc-pVDZ,respectively,tochecktheabovecalculations.[107]

Results5.3

Figure5.5:Conformationsofthe2-pFPEcationobtainedfromabinitioquantumchemistryfullstructural
optimisationattheDFTM05/aug-cc-pVTZleveloftheory.

85

Allcalculationspredictedthesamestructuralchangeinthecation.Itisinterestingtopointout
thattheDFTcomputationswithB3LYPaswellasM05functionalsrevealedthattwoofthe
threetheoreticallyfoundgaucheconformersandthetwoanticonformersintheneutralground,
S0,electronicstateconvergetoonegaucheandoneantistructureinthegroundcationic,D0,
state,respectively,duringtheoptimizationprocess.Onlyoneofthegauchestructuresretainsits
geometryuponionization.Thus,thetotalnumberofthetheoreticallyoptimizedconformations
ofthe2-p-FPEisdecreasingfrom5to3inthecationicstate:twogaucheandoneantistructure.
Itisworthnotingthatnoneofthetwogaucheconformersinthecationfeaturesanonclassical
OH∙∙∙πhydrogenbondbetweentheterminalOHgroupofthesidechainandtheπelectrons
ofthebenzenering.AllthreeoptimizedconformationalstructuresattheM05/aug-cc-pVTZ
leveloftheoryofthe2-pFPEcationaredepictedinFig.5.5.Theoptimizedstructuresatthe
B3LYP/aug-cc-pVTZleveloftheoryarethesameastheonesshowninFig.5.5withonly
marginaldifferencesinsomeoftheplanaranddihedralangles.Theenergyorderingobtained
fromtheB3LYPcalculationsisasfollows:ConformerC(0cm−1),ConformerA(+18cm−1),and
ConformerB(+517cm−1).TheenergyorderingresultingfromtheM05calculationsassertsthat
themoststableconformerisconformerA(0cm−1),followedbyconformersC(+44cm−1)and
B(+735cm−1).Fig.5.6showsacomparisonbetweentheHOMOelectrondensitydistribution
intheneutralandthecationicstateforthemoststableconformercalculatedattheM05/aug-
cc-pVTZleveloftheory.Thenormal-modefrequenciesupto1100cm−1forconformersA,B,
andCaswellasthecorrespondingfrequencieswithinclusionofanharmoniccorrectionsfor
conformerBcalculatedwithbothB3LYPandM05functionalsarelistedinTable5.1.

86

MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

Figure5.6:HOMOelectrondensitydistributionof2-pFPEintheneutralground,S0,electronicstatefor
themoststablegaucheconformer(a)andinthecationicground,D0,electronicstateforthemoststable
gaucheconformerA(b)calculatedattheDFTM05/aug-cc-pVTZleveloftheory.

Discussion5.4

TheMATIspectraofthe2-pFPEmonomermeasuredviatheoriginbandsofthegaucheand
theanticonformersinthefirstexcited,S1,electronicstateshowsignificantlydifferentpattern
(seeFig.5.3):thereisawell-resolvedMATIspectrumonlyfortheanticonformation,andan
unstructuredsignalforthegaucheconformer.TheresponsibleFranck-Condonfactorscanbe
vieweduponasaprobeforthestructuralchangesoftherespectivespecies,takingplaceonion-
ization.UnfavorableFranck-Condonfactorsapplytothegaucheconformer,and,forthisreason,
weexpectthatthestructureofthisspeciesinthecationshouldbesubstantiallydifferentfrom
thatintheneutralstate.Onthecontrary,theanticonformerproducesawell-resolvedMATI
spectrum,whichisanindicationforasmallornegligiblestructuralchangeuponionization.The
resultsfromourquantumchemistryabinitiocalculationsprovidefurtherargumentsforthese
conclusions.Thestructuraloptimizationsinthecationclearlydemonstratethattheanticon-
formerretainsitsstructure(seeFig.5.5,ConformerB),whereastheprobedmoststablegauche
conformationstabilizedbyanonconventionalOH∙∙∙πbondintheneutralstatedoesnolonger
existinthecation,andtheoptimisationresultsinanothergeometrywithadifferentarrangement
ofthesidechain(seeFig.5.5,ConformerA).ConformerAisthemoststableconformerinthe
cationaccordingtotheM05calculations.Theenergygapbetweenthisandthenext-in-energy
conformerC,however,isonly44cm−1,whichsuggeststhatareorderingisnotveryunlikelyto

Discussion5.4

Harm.1070578501015562100946399744298242290437288333484029481320779419176312171974CM05ConformerHarm.69104859343101756010124591001440975419896377879331834301797267777188747113704YPB3L

Anharm.71173109758027979545952499960418102268882366835341780335798303814177747103M05Anharm.6636610815936197853797849097742697237587437483434280332678430277818172690BYPB3LConformerHarm.

Harm.73166110658357100954810054981002428994376898347849340812314809305791182766100M05Harm.6886810956006610175481014497100243099938689735884534581933180130679718474787YPB3L

AHarm.M05Harm.YPB3L4641747113811219416925220129228932332336937142141743343447247056756557158371570976876579679284083386084489688692089599299410121013101710141030102510851047
Conformer

FPEpbeen87

1stablethreethefor−2-theofconformations
haactorsfscaleNo.theoryofelsvlecc-pVDZ/correcttoappliedbeenev
theoryofelsvleaug-cc-pVTZ/evhacorrectionsanharmonicincludingFrequencies.
1cm1100toup)−
M05andcc-pVDZM05/aug-cc-pVTZ,vibrationalnormal-modepredictedTheoretically(cmfrequencies
DFTtheatYPB3LB
/5.1:ableTYPB3LDFTtheatcalculatedcationconformerforonlycalculatedfrequencies.normal-modethe

88MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

occuruponchangingthemodeltheoryandthebasissetasalreadydemonstratedbytheB3LYP
calculationwheretheorderingofAandCisinverted(seeabove).Insuchacaseconformer
Cbeingthemoststableconformerinthecationcannotbeproducedbydirectionizationfrom
itsneutralcounterpartsincethelatterisahigh-energyconformation,andisnotpresentinthe
coldmolecularbeam,butitcanbeproducedasaresultfromarelaxationfromthehigher-energy
gauchecationconformerA.Nevertheless,whicheverofthetwogaucheconformations,AorC,
isfinallyproduceduponionizationofthemoststablegaucheconformerintheneutralstate,no
resolvedMATIspectrumisanticipatedsincebothendingstructuresdifferconsiderablyfromthe
initialoneintheS1state.Withrespecttotheabovestatement,wehavemeasuredtotalioncur-
rentandMATIspectraviathestrongpeakintheintermediate,S1,electronicstateat+553cm−1,
whichwerewithunresolvedpeakstructure.Onthisbasis,weassignedthisintermediatebandas
originatingfromthegaucheconformerof2-pFPE.(seeFig.4.6)Bycontrast,theMATIspectra
obtainedviaexcitationofthevibronicbandsat+147cm−1and+632cm−1inthefirstexcited,
S1,electronicstatefeaturesimilarpeakpatternwiththeMATIspectrumoftheanticonformer
accessedviaitsoriginband(seeFig.5.4aandb)butwithdifferentintensityprofileduetothe
differentFranck-CondonfactorscorrespondingtothetwointermediatevibronicstatesintheS1
electronicstate.Thisresultprovidesaclear-cutevidencethatthevibronicbandsat+147cm−1
and+632cm−1originatefromthesameanticonformer.TheAIEoftheanticonformationof
71238cm−1foundfromtheMATIspectrumofFig.5.4acanbecomparedwiththeAIEofthe
parentmolecule2-PE(71501±16cm−1)measuredbyWeinkaufetal.,[71]25andthusisfound
tobeshiftedtotheredby263cm−1.Theblue-shiftedbandsinFig.5.4shouldcorrespondto
vibrationalexcitationsinthecationasshowninTable5.2bycomparisonbetweentheexperi-
mentalandthetheoreticalfrequenciescalculatedattheDFTB3LYP/cc-pVDZleveloftheory
withincludedanharmoniccorrection.AsseenfromTable5.1thenormal-modefrequencies
obtainedfromtheDFTM05/cc-pVDZleveloftheorywithoutandwithinclusionoftheanhar-
moniccorrectioncanexplaintheobservedbandpositions,especiallyforthelowest-frequency
regionwherelarge-amplitudetorsionalmotionispresent.Wewereabletoassignalloftheex-
perimentallyobservedlow-frequencybandsupto365cm−1,whicharesensitivetothestructure
ofthespecies,eithertonormalmodesortocombinationmodesoftheanticonformer.Next,we
comparethecomputedionizationenergywiththemeasuredonefortheobservedanticonformer.
ThecalculatedAIEfortheanticonformerof2-pFPEis68260cm−1attheB3LYP/aug-cc-pVTZ
leveloftheoryand66686cm−1attheM05/aug-cc-pVTZleveloftheory.Thesevaluesarelower
thantheexperimentalionizationenergyof71238cm−1butyettheyareagoodestimate:the
ratiosoftheexperimentalandthetheoreticalionizationenergiesare1.044(B3LYP)and1.068
(M05),respectively.ThecalculatedAIEforthegaucheconformerofthe2-pFPEmonomerat
thesameDFTB3LYP/aug-cc-pVTZleveloftheoryis67936cm−1,and66201cm−1atthe
M05/aug-cc-pVTZleveloftheory.Assumingthattheaboveratiosholdalsoforthiscase,we
expectanAIEof70925cm−1(deducedfromB3LYP)and70719cm−1(deducedfromM05)for
thegaucheconformerwhosedirectexperimentalmeasurementisprecludedbytheunfavourable

Discussion5.4

Exp.Freq.Theor.Freq.Assignment
3961τ1torsionabouttheC2C7bond
5866β1bendingsidechainvs.ring
8590τ2torsionabouttheC7C8bond
99-Combinationmodeτ1+β1
134-Combinationmodeτ2+β1
169181Butterflymodeβ2orovertoneofτ2
212-Combinationmodeτ1+β2
302302σstretchingalongthelongaxis
modeaggingW342337torsiongroupOH374365Tbasisableof5.2:theAssignmenttheoreticallyofthepredictedobservedfrequenciesvibronic(cm−1bands)withintheanharmonicMATIspectracorrectionsshownforintheFig.anti5.3bconformeronthe
ofthe2-pFPEcation(conformerBinFig.5.5)calculatedattheDFTB3LYP/cc-pVDZleveloftheory.

ConformerAgaucheBanti
S1←S0,0003706837148
AIEexp70925∗71238
∗∗71970AIEtheo(B3LYP)6793668260
AIEtheo(M05)6620166686
Table5.3:ElectronicoriginsofS1←S0transitions(cm−1)forconformersAandB.AIEexp:experimental
valuesoftheadiabaticionizationenergies(cm−1)forconformersAandB.AIEtheo:theoreticalvalues
oftheadiabaticionizationenergies(cm−1)forconformersAandBcalculatedattheDFTB3LYPand
M05/aug-cc-pVTZleveloftheories.

89

Franck-Condonfactorfortheionizationstep.TheaboveresultsaresummarizedinTable5.3.
ThepositionoftheonsetoftheunstructuredMATIspectrumofthegaucheconformer(seeFig.
5.3a)isatca.72000cm−1.ThisimpliesthattheFranck-Condonfactorsfavortransitionsonlyto
vibrationalstatesinthecationatmorethan1000cm−1abovetheAIEandcorrespondstomore
than10quantaofthelowest-frequencytorsionalmodeofthesidechaininducingatransition
fromtheOH∙∙∙π-bondedgaucheconformerintheS1statetothelowest-energystructureofthe
cation.ItdemonstratesthebreakingoftheOH∙∙∙πhydrogenbondofthegaucheconformer
followedbyasignificantreshapingofthesidechainuponionization.Wehaveemployedthree
differenttheoreticalmodelstoinvestigatethestructuralchangesintheneutralconformersupon
ionizationandthuscheckingtheirsuitabilitytoaccountforweakintramolecularinteractions,in
particularthenonclassicalOH∙∙∙πbond.WestartedwiththeUMP2leveloftheorywithtwo
basissets(seesection5.3.3)leadingtooptimizedstructuresoftheneutralmoststablegauche

90MassAnalyzedThresholdIonizationSpectroscopyof2-para-uorophenylethanol

conformerof2-pFPEinthecation.Theso-obtainedcationicstructuresclearlydemonstratethat
theneutralconformerisdestroyeduponionizationasaresultofthebreakingofthestabilizing
OH∙∙∙πbond.WedidnotconsidertheseresultsindetailbecausetheUMP2calculationsstarted
withtheneutralsoftheotherfourconformerswerenotsuccessfulduetospincontamination
errorsinthecation.Asanextstep,theDFTcalculationswereperformedwithtwodifferent
functionals.First,weusedthestandardB3LYPfunctionalwhichhasrevealedthethreestable
structuresintheion,asmentionedabove,butthisleveloftheoryisnotexpectedtoconsider
thedispersioninteractionsthatshouldbeimportantindeterminingtheconformationalstructures
ofhydrogen-stabilizedconformers.Forthisreason,weemployedthenewlyelaboratedhybrid
functionalM05,[106]whichisspeciallydesignedtoaccountforsuchweakinteractions.Itwas
interestingtoseethatthetwotheoreticalfunctionalspredictedthesameneutral-to-cationstruc-
turalchanges.UnliketheverysimilarstructuresresultingfromtheB3LYPandM05calculations,
theenergyorderingandtheenergygapsbetweenthepredictedconformersisdifferentaspointed
outabove,theanticonformerBbeingdistancedawayby735cm−1fromthemoststablegauche
conformerAinthecaseoftheM05calculations.Asshownabove,onlythreestableconformers
aretheoreticallypredictedforthecation:oneanti(B)andtwogaucheones(AandC).None
ofthesegeometriesisfavorablefortheformationofanOH∙∙∙πbond,andhencewecancon-
cludethatthisbonddoesnotsurvivetheionizationprocessleadingtothecation.Acomparison
betweentheHOMOelectrondistributionsintheneutral[107]andinthecationcalculatedat
theDFTM05/aug-cc-pVTZleveloftheory(seeFig.5.6)showsasignificantdepletionofthe
π-electrondensityinthebenzeneringinthecationduetotheejectedπelectron,andasignificant
electrondensityrelocationalsointhesidechain.Thereducedelectrondensityinthebenzene
ringthuscanberenderedresponsiblefortheweakeningoftheOH∙∙∙πbondandultimatelyfor
itsbreakingafterionization.Thisisanimportantresultsinceitprovidesexperimentalevidence
forthenatureoftheOH∙∙∙πbondasbeingformedasaresultoftheattractionbetweenthepar-
tialpositivechargelocatedattheterminalhydrogenatomandthepartialdelocalizednegativeπ
electronchargedistributedoverthearomaticringintheneutralmolecule.

conclusionsandSummary5.5

Inthisworkwehaveinvestigatedtheeffectofionizationonthestructureofaprototypeflexible
molecule,2-para-fluorophenyl-ethanol,whichexistsintwoconformericconfigurationsinthe
coldmolecularbeam.Bycombiningmass-analyzedthreshold-ionizationspectroscopywithhigh
levelquantumchemistryabinitiocalculationsperformedattheB3LYP/aug-cc-pVTZandthe
M05/aug-cc-pVTZlevelsoftheory,wehaveassignedtheconformersofthe2-para-fluorophenylethanol
cation.Ourcalculationspredictthreestablestructuresinthecation:twogaucheandoneanti.
TheMATIspectraviavibronicbandsintheS1intermediatestatecorrespondingtogaucheand

5.5conclusionsandSummary

91

anticonformers,respectively,areutterlydifferent:thespectrummeasuredviatheoriginofthe
gaucheconformerincreasesinintensitygraduallybutdoesnotshowresolvedstructure,whereas
thespectraobtainedviabandsoftheanticonformerarecharacterizedbyawell-pronounced
peaks.WeascribethisdifferencetodifferentFranck-Condonfactorsratherthantoafastre-
laxationprocess.Thisisjustifiedbyourcomputations,whichdemonstratethatthemoststable
anticonformerintheneutralundergoesanegligiblestructuralchangeuponionizationandhence
theionizationshouldhavefavorableFranck-Condonfactors.Ontheotherhand,theionization
ofthemoststablegaucheconformerintheneutral,whichisstabilizedbyanOH∙∙∙πhydro-
genbond,leadstoabreakingofthisbondandtoasignificantstructuraldeformationresulting
intheunstructuredMATIspectruminthiscase.Theseresultsprovidefurtherevidenceforthe
existenceofanonclassicalOH∙∙∙πbondinthisflexiblemodelmoleculeandshedlightonits
natureasoriginatingfromattractionbetweenthepositivepartialchargeontheterminalHatom
ofthesidechainandthedelocalizednegativechargeoftheπelectronsoftheneutralbenzene
ring.TheunderstandingofthenatureoftheOH∙∙∙πbondisrelevanttostudyingthestability
ofsecondarystructuressuchaspolypeptidesandproteinsheldtogetherbyweakinteractions.In
conclusion,wehaveshownthatMATIspectroscopyprovidesinformationparticularlyonweak
interactionswiththeπelectronsofaromaticmoleculesbecauseasignificantchangeintheπ
electrondensityoccursduringanionizationprocess.Thisisadvantageousto,e.g.,dispersed
emissionspectroscopywhereduringS1←−S0transitionnolargechangesintheπelectronden-
sityareinvolved.

92

Mass

Analyzed

esholdThr

Ionization

oscopySpectr

of

para2-

ophenylethanol-uor

Chapter6

CompetitionbetweenπandσHydrogen
BondsandConformationalProbingof
2-orthoHigh-Resolution-uorophenylethanolElectronicbySpectroscopy

R.Karaminkov,S.Chervenkov,H.J.Neusser,J.Chem.Phys.130,034301
(2009).Theflexiblemodelmolecule2-ortho-fluorophenylethanolhasbeeninvesti-
gatedbylow-andhigh-resolutionresonance-enhancedtwo-photonionization
spectroscopyincombinationwithhigh-levelabinitioquantumchemistrycal-
culations.Onedominantconformationhasbeenidentifiedinthecoldmolec-
ularbeamcorrespondingtothemoststabletheoreticallypredictedgauche
structurestabilizedbyanintramolecularOH∙∙∙πbond.Atentativeassign-
mentofahigher-lyinggaucheconformerpresentinthemolecularbeamsep-
aratedbyhighpotentialbarriersfromthemoststableonehasbeenmade.
Themissingotherhigher-energytheoreticallypredictedconformationsmost
likelyrelaxtothemoststableonesduringtheprocessoftheadiabaticexpan-
sion.Thegoodagreementbetweentheexperimentalandtheoreticalresults
demonstratesthateveninthecaseofasubstitutionwithanelectronegative
atomattheorthoposition,bringingaboutasignificantredistributionofthe
electrondensityinthebenzeneringandprovidingaconvenientbindingsite
fortheformationofacompetingOH∙∙∙Fσhydrogenbond,thenonclas-
sicalOH∙∙∙πbondremainsthepreferredbindingmotifforthemoststable
.conformer

93

94

oductionIntr6.1

ConformationalProbingof2-ortho-uorophenylethanol

Theconformationalpreferenceofaflexiblemoleculeistheoutcomeofasubtleinterplayofnon-
bondedweakinteractionsbetweendifferentgroupsofthemoleculeandwithitssurroundings
(Forreview,seeRefs.[3,60,89]).Conformationalshapesdeterminethepropertiesandfunc-
tionalityofmanybiologicallyimportantmolecules.Towardthisend,muchdiscussionhasbeen
devotedinthepasttwodecadestoscrutinizingtherelativeinfluenceofpairwiseinteractions
betweenspecificmoleculargroups[86,87]andChapters3,4.Insuchanendeavor,gas-phase
conformationalstudiesofsmallbiologicalmoleculesandtheirconvenientmolecularanalogsun-
dersolvent-freeconditionsareverywellsuitedforinvestigationoftheintrinsicpropertiesof
thesemolecules,andforthisreason,haveattractedthescientificinterestofthespectroscopic
community[20,46,61,67,75,80,85,108–110]andChapters3,4.

Figure6.1:Atomlabelsof2-oFPE

2-ortho-fluorophenylethanol(2-oFPE)isaderivativeof2-phenylethanol(2-PE),whichisthehy-
droxylanalogofthesimplestaromaticamineneurotransmitter,2-phenylethylamine.[16,46,59,
60,64,72,73,111]Microwaveandhigh-resolutionelectronicspectroscopystudieshaveconclu-
sivelyestablishedthatagauchegeometryisthepreferredconformationofthe2-PEmolecule
inthegroundelectronicstate.[46,70]Thehigh-resolutionUVspectroscopyhasfoundalsothat
almostnostructuralchangeinthemoststableconformeroccursuponelectronicexcitationand
thatnofastisomerizationprocessestakeplace.[46]Theprimaryfactorthatstabilizesthisconfor-
mationisanonclassicalhydrogenbondingoftheterminalhydroxylfunctionalgroupoftheside
chainwiththearomaticπ-electrons.[46]Thefindingcorroboratesthepredictionsofquantum
chemistrycalculationsandconclusionsofearlierlow-resolutionlaserspectroscopicmeasure-
mentsandmicrowaveexperiments.[16,72,73]Recentlyweusedthesamespectroscopicmethod
todeterminehowtheconformationalpreferenceisaffectedonbindingwithaforeignmolecular
species,suchasargon[46]andwater[seeChapter3].ToshedlightonthestabilityoftheOH∙∙∙π

Experiment6.2

95

hydrogenbondanditsimportanceforthestabilizationofconformationalshapes,theconforma-
tionsofanumberoffluorinesubstitutedaromaticalcoholswerestudiedinthepastbymeasuring
thelow-resolutionelectronicspectraofthosemoleculesunderthesupersonicjetexpansioncon-
ditionusingresonance-enhancedmultiphotonionization(REMPI)spectroscopicmethod.[86]
Directfluorinesubstitutionatthebenzeneringsitesresultsinadistortionofthesymmetryofthe
aromaticπ-electrondistributionontheringandthechargesontheringcarbonatomsandsub-
stitutions.Imetal.notedthatthevibrationalfeaturesinthespectraofortho-fluorinesubstituted
moleculesareconsiderablydifferentcomparedtootherisomericforms.[86]Theyattributedthe
differencestotheresultsofdirectproximalinteractionbetweentheF-atomandOHgroupof
thechain.Thepresentstudyaimsatunderstandingtherelativeimportanceofthenonclassical
hydrogenbondin2-PEcomparedtoaclassicalhydrogenbondofthesameterminalhydroxyl
functionalgroupofthesidechainwithafluorineatomsubstitutedattheorthopositionofthearo-
maticring,insettlingtothefinalconformationalformsofthemolecule.Thisissueisstillopen
becauserecentstudiesinourgroupwithparafluorinesubstituted2-PEshowthattheconforma-
tionalbehaviorofthemoleculeisnotverydifferentfromitsunsubstitutedspecies.(seeChapter
4)Anobviousreasonisthemuchlargerseparationoftheparafluorineatomfromthehydroxyl
groupprecludingtheformationofadirectOH∙∙∙Fhydrogenbond.Theaboveresultalsoim-
pliesthattheOH∙∙∙πhydrogenbondremainsevenundertheconditionsofadepletedπ-electron
densityresultingformthecovalentattachmentoftheelectronegativeFatomtothearomaticring.
Inthischapter,wehaveusedthelow-andhigh-resolutionmass-selectiveresonancetwo-photon
ionization(R2PI)spectroscopysupportedbyhigh-levelabinitiotheoreticalcalculationsforde-
terminingthemoststableconformationalisomericformsof2-orthofluorophenylethanol(See
Scheme6.1)inthegroundstateandunambiguousdeterminationofthestructuresoftheabove
isomericspecies.Theissuesthatwehaveaddressedarewhetherastableisomericspecieshaving
intramolecularOH∙∙∙Fhydrogenbondcanbeisolated,andhowtheconformationaldistribution
ofthenonfluorinated2-PEisaffectedbecauseofsubstitutionoftheelectronegativefluorineatom
attheorthopositionofthearomaticring.

Experiment6.2

TheexperimentalapparatusandtechniqueusedfortheproductionoftheR2PIlow-andhigh-
resolutionspectrahavebeendescribedindetailsinSection2.2.Ashortdescriptionwillbe
presentedherepointingoutonlythedifferentexperimentalconditions.The2-oFPEsamplewas
purchasedfromAldrich,andheatedto90inahomemadenozzlewithanorificeof500µm
wherethesamplevaporsweremixedwithbuffergasargonatapressureofabout3.5barand
supersonicallyexpandedintovacuum.Askimmerwasusedtocollimatethemolecularbeam
andthustoreducetheDopplerbroadening.Thelightpulsesusedintheone-andtwo-color

96

ConformationalProbingof2-ortho-uorophenylethanol

experimentswereproducedbyfrequencydoublingoftheoutputpulsesofacommercialbroad-
banddyelaser(LambdaPhysikFL2002)operatingwithCoumarin153andthepulsedamplified
outputofacontinuouswavesingle-moderingdyelaser(Coherent,CR699-21)operatingwith
Coumarin334,respectively.Thespectralresolutionachieved(∼70MHz)isnotsufficienttore-
solvesinglerotationallinesinthecongestedrovibronicspectraoflargemolecules.Therefore,a
singlerotational-lineassignmentisnotpossible,andweapplyacomputer-basedmethodforfit-
tingoftheexperimentalhighlyresolvedspectra.Themethodemploysgeneticalgorithm,which
hasbeenpreviouslydemonstratedbyus[20,46]andothergroupstobeadecisivetoolinthe
determinationofthemolecularparameters.[112,113]Crosscorrelationhasbeenemployedasa
qualityfactorforthematchbetweenthesimulatedandtheexperimentalspectra.Fortheproduc-
tionofthesyntheticspectraofthe2-oFPEmonomer,werantheprogramusing300generations
with500individualsinageneration.[20]

Results6.3

6.3.1SpectrumResolutionwLo

ThevibrationalstructureoftheS1←−S0spectrumof2-oFPEwasexaminedbylow-resolution
one-colormassselectedR2PIspectroscopy.InFig.6.2,thespectraof2-oFPEmeasuredatthe
parentmasschannel(bottomtrace)andatthemassoftwofragments(uppertraces)areshown.
Here,justlikeinthecaseof2-parafluorophenylethanol(2-pFPE),(seeChapter4)weobservea
fragmentationofthesidechain(CH2OH)resultinginamassofmz=109.Inaddition,wehave
observedaweaksignalatonemorefragmentatmassmz=121thatstemsfromdetachmentof
thefluorineatom.Incomparisonto2-pFPE,thefragmentationbehaviorofthe2-oFPEmonomer
isnotsopronounced,andtheintensityofthefragmentsignalsisaboutsixtimeslowercom-
paredtothatof2-pFPEforcomparablelaserintensities.Theoriginbandoftheintermediate
electronic,S1,stateisdetectedatallmasschannels,butwhileatmassesmz=140andmz=109
itisaprominentfeature,atmassmz=121thereisanotherbandwithhigherintensitylocatedat
+54cm−1fromthe000transition.Therearethreemorebandsthatareobservedattheparentmass
channel(mz=140)andthefragmentmasschannel(mz=109),namelyat+3,+43,and+86cm−1,
whichcoincidewiththecorrespondingsignalsintheFluorescenceExcitationspectrum.[114]To
shedlightontheconformationalpreferencesof2-oFPE,wemeasuredthehigh-resolutiontwo-
colorspectraoftheabove-mentionedbandsatmasschannelmz=140.Further,wecombinedthe
experimentalresultswiththefindingsfromabinitiocomputationswithlargebasissets.

Results6.3

Figure6.2:One-colorR2PIspectraoftheS1←S0electronictransitionofthe2-oFPEmonomermea-
suredatthemonomerC8H9OF(m/z=140)masschannel(c)andthefragmentmasschannelsrecordedat
m/z=109(b)andm/z=121(a).

SpectaResolutionHigh6.3.2

97

Thehigh-resolutionUVspectraofthestrongestmonomerband000andthespectraofthevibronic
bandsat+3,+43,+86,+109,and+125cm−1arepresentedinFigs.6.3and6.8(uppertraces).
Allofthemhavebeenrecordedatthemonomermass(mz=140)channel.Thehighlyresolved
spectraoftheoriginbandandthebandsat+43,+86,+109,and+125cm−1,featureverysimilar
rotationalstructure,andhence,canbeattributedtooneandthesameconformer.Thesebands
first.discussedbewill

98

ConformationalProbingof2-ortho-uorophenylethanol

0bandFigureis6.3:assignedasHigh-resolutiontheorigintwoftheo-colorgaucUVheR2PIconformerspectrum1(SeeoftheFig.006.9)bandoftherecorded2-oFPEatm/z=monomer140.withThe
1−initsvertedrotationlesstrace:best-fittransitionsimulatedcenteredatspectrum37589.183(7)yieldingthecm.parametersUpperintrace:Tablee6.1xperimental(fordetails,spectrum.seetext).Lower

Figure6.4:High-resolutiontwo-colorUVR2PIspectrumofthebandat+3cm−1of2-oFPErecordedat
140.=z/m

Boththerotationalconstantsintheground,S0,andthefirstexcited,S1,electronicstate,and
thetransitionmomentratioweredeterminedwiththecomputer-aidedroutineoutlinedabove.
Asstartingvaluesforthefittedparameters,weusedtheresultsfromthetheoreticalpredictions

Results6.3

BandRotationalconstant(cm−1)000cm−1+43cm−1+86cm−1+109cm−1+125cm−1
BA´´´´0.03676(79)0.0694(14)0.03756(88)0.0677(13)0.03727(53)0.0680(17)0.03689(78)0.0682(14)0.0356(12)0.0682(17)
C´´0.02749(46)0.02792(75)0.02780(71)0.02810(51)0.02675(57)
BA´´0.03619(93)0.0681(14)0.03694(85)0.0666(14)0.03678(74)0.0664(17)0.03649(92)0.0665(12)0.0348(11)0.0668(15)
C´0.02665(49)0.02699(61)0.02690(69)0.02736(56)0.02582(51)
TMratioµa2:µb2:µc260:30:10(5%)52:29:19(9%)82.4:17.6:0(6%)69.4:30.6:0(4%)74.3:20.7:5.0(6%)
Originν0(cm−1)37589.183(97)37631.951(11)37675.5306(64)37698.9949(74)37713.8964(65)
TBest-fitemperaturecrossT(K)correlation(%)97.0(2)11.0(4)93.0(3)10(2)95.0(5)10.1(7)93.0(6)11.0(7)96.6(9)11.6(6)
Table6.1:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,thebandoriginfrequency,ν0,
therotationaltemperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofbands000,
+43,+86,+109,and+125cm−1,showninFigures6.3-6.8.Thenumbersinparenthesesrepresentone
standarddeviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevaluesof
µa2,µb2,andµc2inthetransitionmomentratioisgiveninparentheses.

99

(MP2/cc-pVDZandCIS/cc-pVDZlevelsoftheory).Theresultingbest-fitspectrawereproduced
byconvolutingthestickspectrawithaGaussianprofilewith210MHzfullwidthathalfmaxi-
mum(FWHM),andareshowninthefiguresdepictingtheexperimentalspectra(lowerinverted
traces).Thesimulatedspectrabeargoodresemblancewiththeexperimentalonesintermsof
peakpositionsandpeakintensities.Theachievedcrosscorrelationsforthefitsrangefrom93%
to96%.Theexperimentalvaluesoftherotationalconstants,transitionmomentratio,rotational
temperature,andbandpositionaresummarizedinTable6.1.Therotationalconstantsforallof
thediscussedconformationsarealmostidentical;therearesmalldifferencesinthetransitionmo-
mentratios,whichareresponsiblefortheslightlydifferentshapesofthehighlyresolvedspectra.
Thefitsclearlyshowthatthespectraconsistofwell-pronouncedPandRbranchesinthewings
andaprominentbroadQbranchinthecenter.Thespectraareassignedashybrida-,b-,and
c-types,withdominantacontribution.Theexperimentalspectradonotdisplaysinglerotational
linesbutaggregationsofrotationallinesresultinginpeakswithaFWHMof∼250MHz.The
comparisonoftheexperimentallyfoundparameterswiththeonesobtainedfromtheabinitiocal-
culationsbringstheconclusionthatthebandsinquestionoriginatefromthemoststablegauche
conformation(conformer1)of2-oFPE(seeFig.6.9).Wepresentinthispaperthespectraofsev-
eralmeasuredbandsindetailtodemonstratethatthesebandscanbeclearlyassignedtothesame
conformer,andruleoutanyinterpretationstatingthatthebandsunderscrutinydescendfrom
differentconformations(seeFigs.6.3and6.5-6.8).Wehavefittedalsothehigh-resolution
spectraoftwootherbandsintheregionabove100cm−1,namely,thebandsat+109and+135
cm−1.Theyareverysimilartotheabove-discussedrotationallyresolvedbands,andweshowthe
bandsat+109and+125cm−1asatypicalrepresentativesforthebandsinthisregioninFigs.
6.7and6.8.Thevibronicbandat+3cm−1inFig.6.4featuresaconspicuouslydifferentrota-

100

ConformationalProbingof2-ortho-uorophenylethanol

Figure6.5:High-resolutiontwo-colorUVR2PIspectrumofthebandat+43cm−1recordedatm/z=140.
Thebandisassignedtoaprogressionoftheoriginofthegaucheconformer1(SeeFig.6.9)ofthe2-oFPE
monomerwithitsrotationlesstransitioncenteredat37631.951(11)cm−1.Uppertrace:experimental
spectrum.Lowerinvertedtrace:best-fitsimulatedspectrumyieldingtheparametersinTable6.1(for
xt).teseedetails,

Figure6.6:High-resolutiontwo-colorUVR2PIspectrumofthebandat+86cm−1recordedatm/z=140.
Thebandisassignedasthesecondovertoneofthebandat+43cm−1thegaucheconformer1(SeeFig.
6.9)ofthe2-oFPEmonomerwithitsrotationlesstransitioncenteredat37675.531(6)cm−1.Uppertrace:
experimentalspectrum.Lowerinvertedtrace:best-fitsimulatedspectrumyieldingtheparametersinTable
6.1(fordetails,seetext).

Results6.3

Figure6.7:High-resolutiontwo-colorUVR2PIspectrumofthebandat+109cm−1recordedatm/z=140.
Thebandisassignedtoacombinationmodeofthegaucheconformer1(SeeFig.6.9)ofthe2-oFPE
monomerwithitsrotationlesstransitioncenteredat37698.995(7)cm-1.Uppertrace:experimental
spectrum.Lowerinvertedtrace:best-fitsimulatedspectrumyieldingtheparametersinTable6.1(for
xt).teseedetails,

Figure6.8:High-resolutiontwo-colorUVR2PIspectrumofthebandat+125cm−1recordedatm/z=140.
Thebandisassignedtoacombinationmodeofthegaucheconformer1(SeeFig.6.9)ofthe2-oFPE
monomerwithitsrotationlesstransitioncenteredat37713.896(7)cm-1.Uppertrace:experimental
spectrum.Lowerinvertedtrace:best-fitsimulatedspectrumyieldingtheparametersinTable6.1(for
xt).teseedetails,

101

102

ConformationalProbingof2-ortho-uorophenylethanol

tionalstructurefromtheoneoftheabove-discussedbands,whichcannotbeattributedsolelyto
adifferenttransitionmomentratio.Thespectrumconsistsofirregularlyspacedpeaksformedby
bunchingofsinglerotationallinessuperimposedonabroadbackground.ItfeaturesstrongPand
RbranchesandarelativelyweakQbranch.Duetoitslowintensity,thesignal-to-noiseratioof
thisspectrumissmallercomparedtotheoneoftheotherhighlyresolvedspectrapresentedhere.
Fromtheprofile,thespectrumcanbeassignedasamixeda-,b-,andc-typewithsignificantb
contributioninvariancewiththeotherbandspresentedabove.Toanalyzethespectrum,weem-
ployedthesamecomputer-assistedfitprocedureweusedalsofortheotherbands.Presumably,
thisbandoriginatesfromadifferentconformationinthecoldmolecularbeamandforthisrea-
son,toaid-inthefit,weusedasinitialvaluesfortherotationalconstantstheonesobtainedfrom
theoreticalpredictionsforsomeofthenext-in-energyconformers(conformers7and9).The
simulatedstickspectrumwasconvolutedwithaGaussianprofilewithaFWHMof210MHz.
Theinsufficientsignal-to-noiseratio,however,precludedtheunambiguousassignmentofthis
bandtooneofthetwoconformationssincetheuseoftheirtheoreticallypredictedparametersas
startingvaluesforthefitsyieldedverysimilarrotationalprofiles.Allattemptedfitsemployingas
startingvaluestheparameterscorrespondingtotheothertheoreticallypredictedconformations
(1,2,3,4,5,6,and8)failedtoreproducetheexperimentalspectrum.Thatiswhy,thebandcan
beattributedtoeitherconformer7orconformer9.

CalculationsinitioAb6.3.3

TheGAUSSIAN03suiteofprograms[82]andtheTURBOMOLEprogrampackageforabinitio
electronicstructurecalculations[115]havebeenemployedtotheoreticallyinvestigatethecon-
formationalstructure,vibrationalmodes,andenergeticsofthe2-oFPEmonomer.Toassignthe
stableconformationsofthespecies,wefirstperformedatwodimensionalgridsearchyielding
thepotential-energysurface(PES)correspondingtovariousshapesofthesidechain.Thegrid
searchwasperformedattheMP2/cc-pVDZleveloftheoryincludingscansoftheC2C7C8O9
andC7C8O9H19dihedralanglesfrom0to360withanincrementof10.TheC3C2C7C8dihe-
dralanglewaskeptfixedat90.ThePESconsistingof1296pointsisdepictedinFig.6.9(the
PESisshowninvertedforthesakeofabettervisualization).Itfeaturesninepotential-energy
minima(maximainFig.6.9)implyingninestableconformationalstructures.Thelattercanbe
subsumedintothreegroups:oppositegauche,anti,andadjacentgauche,eachofthemcontain-
ingthreedifferentconformations.Intheoppositegaucheconformations,thesidechainisina
gauchegeometryontheoppositesideoftheFatomandthedihedralangleC2C7C8O9is∼60
forallofthem(seeinsetsinFig.6.9,conformers1,2,and3).Theanticonformers(seeFig.6.9,
conformers4,5,and6)arecharacterizedbyanextendedsidechainawayfromthebenzenering
withtheirdihedralangleC2C7C8O9being∼180.Inthecaseoftheadjacentgaucheconfor-
mations,thesidechainisinagauchegeometryonthesamesideoftheFatomandthedihedral

Results6.3

Figure6.9:Potential-energysurface(invertedforthesakeofabettervisualization)asafunctionofthe
C2C7C8O9andC7C8O9H19dihedralanglesofthe2-oFPEmonomer.Thegaucheandanticonforma-
tionsareseparatedbyahighpotentialbarrierintheC7C8coordinate.Thenineenergeticallymostfavor-
ableconformationsofthe2-oFPEmonomerhavebeenfullygeometricallyoptimizedintheirground,S0,
electronicstateattheMP2/ccpVDZleveloftheory.Thefullyoptimizedstructuresaredepictedontheir
correspondingminima(maximaintheFigure)onthepotential-energysurface.

103

anglewithineachC2C7C8O9oftheisgroups∼-60diffforerthefromthreeeachspeciesotherbyinthethegrouporientationof(conformersthe7,terminal8,9).OHThegroupspecies(di-
hedralangleC7C8O9H19).Theglobal(deepest)minimumcorrespondstotheoppositegauche
pointsconformertotheπ(conformerelectrons1),ofinthewhichbenzenetheOHring.groupTheofotherthetwsideochainminimaisawayseparatedfromfromtheFeachatomotherand
Itandiswfromorthwhiletheglobalpointingminimumoutthatbylothewbarriersconformationsalsowithincorrespondthetogroupsoppositearegaucseparatedhefromconformers.each
otherbylowpotentialbarriers,whilehighpotentialbarriers(ridgesalongtheC2C7C8O9co-
nineordinate)identifiedseverthestableoppositeconformationsgauche,ofanti,the2-andoFPEadjacentmonomergauchahevespeciesbeenfurtherfromonetheoreticallyanother.eAllx-
ploredbyperformingafullstructuraloptimization,frequencyanalysis,andcalculationoftheir
Fullenergeticsstructuralintheoptimizationground,S0,andelectronicfrequencystateanalysisusingforMP2alllevelspeciesofhatheoryvebeenwithcarriedcc-pVDZoutbasisalsoset.for
thefirstexcited,S1,electronicstateemployingtheCIS/cc-pVDZleveloftheory.Thevalues
ofsometypicalinteratomicdistances,planaranddihedralanglesdefiningtheconformational
structureofthe2-oFPEmonomer,therotationalconstants,thetransitionmomentratio,andthe

104

ConformationalProbingof2-ortho-uorophenylethanol

−−−1c2)(cmZPVE)(including(cmbµ2
STheoretical−1)1theofstateelectronic,
∙S00∙0∙ointeratomic191:2:71hegauc∙∙∙∙∙∙S
distances,and487.4573.20enerevRelati.monomerFPEgies,
danti,planar∙∙∙∙∙∙S2
95:1:41conformersangles,of613.1666.9r,eltheEα∙∙∙0SConformer
2-dihedraloEEaµ2CBA3τ2τ1ταααα1.311.351.321.361.321.361.321.361.321.361.321.361.321.36(A)(C3F11)rr2.772.63(A)H19F11)or(H19C1ConformerarameterP
93:5:21elrelr(cm(cm(cm4321
µMP2fromobtainedbeenevhagyenervibrationalzero-pointtheofinclusionwithandwithout2-theofgeometryoptimised,theoryofratio(TM)momenttransitiondipoleelectricThe.elyvrespectixcited,e6.2:ableT:66.560.3-55.6-50.968.365.4-67.5-64.267.667.7171.6-169.7-66.7-61.9(C7C8O9H19)-65.5-60.9-58.0-53.3176.3177.4-179.7-177.870.562.168.466.262.960.6(C2C7C8O9)109.694.887.383.887.278.286.376.8127.893.0107.597.093.385.1(C3C2C7C8)110.0105.9109.6106.6109.3106.1109.4106.2109.1106.1109.6107.0109.1105.7g)(de(C8O9H19)113.2112.5113.3113.0111.8112.4111.9112.5113.8112.5108.9107.3113.0112.0g)(de(C7C8O9)113.9110.8113.9111.8113.4111.6113.4111.6115.6111.8114.8112.0113.9110.5g)(de(C2C7C8)121.5121.7122.2122.4122.2122.6122.3122.7122.8122.3122.3122.0122.1122.1g)(de(C1C2C7)
a2−µ:0.026680.027850.027280.027950.022210.022400.022110.022410.023160.025670.024250.025440.025460.02676)10.033680.035280.034560.035520.027940.028110.027800.028110.032610.035430.033780.035560.034620.03631)10.084850.071410.066720.069640.067010.068270.06645)0.070730.070780.074170.074000.084280.084390.084081
:∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙SConformer
µangles,b2566.7607.60µτ:∙∙∙SConformer
hasMP2theatcalculatedmonomerFPE∙∙∙∙∙∙S3
∙∙∙∙∙∙S4
94:2:4beenstructuresconformationalstablemostenvsetheof2-theof
obtained466.8503.7∙∙∙S
0.theoryofelvlecc-pVDZ/1
Conformerc2ground,theforconstantsrotationaland,
from95:1:41CIS//CISandcc-pVDZ∙∙∙∙∙∙S6
0cc-pVDZConformerS90:9:11,calculationand257.3252.6∙∙∙2.333.202.702.63S
omonomerFPEelvlecc-pVDZ/0∙∙∙∙∙∙S7
theof382.2413.4∙∙∙S
0firsttheConformer∙∙∙∙∙∙S9
91:0:91

Discussion6.4

105

relativeenergieswithoutandwiththezero-pointvibrationalenergy(ZPVE)aresummarizedin
Table6.2.TheZPVEshavebeencalculatedusingappropriatefortheemployedleveloftheory
scalefactors.[93,94]Acompellingargumentthatthetheoreticallypredictedstructuresindeed
correspondtopotentialenergyminimaisthepresenceofall-positivevibronicfrequenciesinthe
ground,S0,electronicstate.Thenormalmodefrequenciesupto1000cm−1ofalltheoretically
predictedstableconformersarelistedinascendingorderinTable6.3.Toascertainthereli-
abilityofourtheoreticalresultsweperformedsingle-pointenergycalculationsforallalready
optimizednineconformationsattheMP2/aug-cc-pVDZleveloftheory.Theseresultsshowed
thatconformer1isagainthemoststablestructurebutsomereorderingofthenext-in-energy
conformationswaspresent,inparticular,conformer7movedtothefourthposition.Sincewe
areconcernedonlywiththelowest-energyconformationswhosepresenceinthecoldmolecular
beamcanbeexperimentallyverified,weperformedafullstructuraloptimizationofconformers
1,7,and9attheMP2/aug-cc-pVDZleveloftheory.Thesecalculationsyieldedthefollowing
orderingofthediscussedconformationsaccordingtotheirenergy:conformer1,9,and7,the
lattertwobeingspacedfromthemoststableoneby406and465cm−1,respectively.Weper-
formedalsoatestcalculationfortheexcitedstatebyoptimizingthestructureandcalculatingthe
normal-modefrequenciesforthemoststableconformer1attheCC2/aug-cc-pVDZ[116,117]
andMP2/aug-cc-pVDZleveloftheory,respectively.Theso-optimizedstructuredoesnotdiffer
muchfromtheoneoptimizedattheCIS/cc-pVDZleveloftheory.Asmalldiscrepancybe-
tweenthetwolevelsoftheorywasobservedonlyforthelowest-frequencymode,whichcanbe
attributedtotheinherentanharmonicityofthetorsionalmotioncorrespondingtothismode.

Discussion6.4

Duetothereducedsymmetryofthe2-oFPE,itspotential-energylandscape(seeFig.6.9and
Table6.2)ismorecomplicatedcomparedtotheonesof2-PEand2-pFPE,anditfeatures
ninepotential-energyminima.Thetheoreticallypredictedmoststablegeometryofthe2-oFPE
monomerwithoutandwithinclusionoftheZPVEisthatoftheoppositegaucheconformer1,
inwhichtheterminalOHgroupofthesidechainpointstothearomaticring.Thestructure
isassumedtobestabilizedbyanintramolecularnonclassicalhydrogenbondbetweentheOH
groupandthedelocalizedπelectronsofthebenzenering,similarlytothecasesofthenonflu-
orinated2-PE(Refs.[46,70]andChapter3)andthepara-fluorinated2-FPE.(seeChapter4)
Thesecond-in-energyconformationistheadjacentgaucheconformer7lying257cm−1above
thelowest-energygeometry.TheskeworientationoftheterminalOHgroupofthesidechain
suggeststheexistenceofaσhydrogenbondbetweenthisOHgroupandthefluorineatomat
theorthopositionofthearomaticringstabilizingthisconformer.Thenext-in-energystructure
ofthe2-oFPEmonomercorrespondstoconformer9,whichalsobelongstothegroupofthe

106

ConformationalProbingof2-ortho-uorophenylethanol

adjacentgauchespecies.Inthiscase,thestructureresemblesverymuchtheoneofconformer
1and,forthisreason,itisverylikelythatthisgeometryisalsostabilizedbyanonconventional
intramolecularhydrogenbondbetweentheOHgroupandtheπelectronsofthearomaticring.
Thelowest-energyantistructureisconformer6withitssidechainpointingtothesideofthe
fluorineatom.Itisatthefourthpositionandisdistancedawayby467cm−1fromthemost
stableconformer(conformer1),andby∼80cm−1fromtheprecedinggeometry(conformer9).
Conformer6isfollowedbytheoppositegaucheconformer2withtheOHgroupinanupright
position,theenergygapbetweenthembeingonly20cm−1.Thedifferenceintheenergiesis
smallandreorderingofthesetwoconformationsuponalteringofthemodeltheoryandthebasis
setislikelytooccur.Thenexthigher-energyconformationistheanticonformer4withitsOH
grouppointingtotheoppositesideofthefluorineatom,whichlies566cm−1abovethemost
stablestructure.Twocloselyspacedconformationsfollow,conformer5withacompletelyex-
tendedsidechainawayfromthearomaticringandtheoppositegaucheconformer3,lyingat
610and613cm−1,respectively,abovethemoststablestructure.Theenergygapbetweenthese
twospeciesisverysmalland,forthisreason,theorderingoftheconformationsmaydependon
theemployedleveloftheory.ItisinterestingalsotopointoutthattheinclusionoftheZPVE
correctionalsoinfluencestheenergyorderingofthespecies,invariancewiththecaseofthe
lower-lyingconformations.Thehighest-energyconformationistheadjacentgaucheconformer
8withitsterminalOHgroupdirectedupwards.Thisconformerislessstableby848cm−1com-
paredtothelowest-energyone.Apreliminaryidentificationoftheobservedconformationsin
thecoldmolecularbeamcanbemadebycomparisonofthemeasuredvibrationalbandpositions
inthelow-resolutionR2PIspectrum(seeFig.6.2)withthetheoreticallypredictedfrequencies
forthefirstexcited,S1electronicstate(seeTable6.3).Themostprominentbandat37589
cm−1isassignedtothemoststableoppositegaucheconformer1aswillbeconfirmedbythe
rotationalanalysisbelow.Twoblueshiftedvibronicbandsat+43and+86cm−1areobserved
intheR2PIspectra.Thebandat+43cm−1perfectlymatchesthetheoreticallypredictedlow-
estfrequencyof43cm−1correspondingtothetorsionalmodeofthesidechainabouttheC2C7
bondinconformer1.Thebandat+86cm−1canbeconfidentlyassignedtothefirstovertone
oftheabove-describedtorsionalmode.TheR2PIspectruminFig.6.2featuresaweakband
at+80cm−1,whichwetentativelyattributetoOH-πbendingmodeinconformer1whichis
theoreticallypredictedtobeat+83cm−1.Thereareseveralweakbandsat109,125,and135
cm−1observedintheR2PIspectra.Thebandat+135cm−1canbereadilyassignedtothelow-
frequencytorsionalmodeabouttheC7C8bondinconformer1,whichiscalculatedtobeat+134
cm−1.Further,weassumethatthebandsat+109and+125cm−1alsooriginatefromthesame
conformerthoughtheycannotbeassignedtoitsfundamentalnormal-modevibrations(seeTa-
ble6.3).Inthiscase,weputforwardthehypothesisthatthesebandsstemfromanharmonic
couplingsinvolvingovertonesandfrequencymixingofthenormalmodevibronicbandsat+43,
+80,and+135cm−1.Theassumptionthattheabovebandsdescendfromthesameconformer
issubstantiatedbytheanalysisoftheirhighlyresolvedelectronicspectradiscussedbelow.An

Discussion6.4

107

interestingfeatureemergesfromtherightwingofthestrongestband.Thisbandisblueshifted
by3cm−1.Sincethereisnotheoreticallypredictedconformationfeaturingverylow-frequency
vibrations,weascribetheweakbandat+3cm−1astheoriginbandofanotherconformation.
Anotherprominentvibronicbanddetectedinlaser-inducedfluorescenceexcitation(LIF)spec-
trum[114]islocatedat+30cm−1.Thesameband,however,appearsonlyasasmallbumpin
theR2PIspectrum(seeFig.6.2).Thissuggeststhatthisbandmayoriginatefromaconformer
whoserelaxationdynamicsisdifferentforthedifferentbuffergases,HeandAr,usedintheLIF
andR2PIexperiments,respectively.Itisplausible,however,toassumethatthisbandmaystem
alsofromtheconformationbringingforththe+3cm−1bandin[114].Inthiscasetherespective
conformerisexpectedtohavealowestfrequencyof∼27cm−1inthefirstexcited,S1,electronic
state.Inspectingthetablewiththecalculatedvibrationalfrequencies(seeTable6.3),wecan
findthattherearethreeconformershavingfrequenciesinthevicinityabove27cm−1;theseare
conformers2,4,and9,thelatterbeingthemostlikelyonesinceitisoneofthelow-energycon-
formations.Toverifyorruleoutthistentativeassignmentadditionalargumentswillbeadduced
below.Tocorroboratetheabovetentativeassignmentsaswellastocomplementthemissing
informationandtomakereliableimplicationsonthestructureandrelaxationdynamicsofthe
observedconformationsofthe2-oFPEmonomer,thedatafromthedispersedfluorescence(see
ourpaper)andthehighlyresolvedR2PI(seeFigs.6.3-6.8)experimentsmustbeanalyzedand
discussed.Thedispersedfluorescence(DF)spectraobtaineduponprobingthe000,+3,and+30
cm−1bandsareverysimilar[seeourpaper]bothintermsofbandpositionsandbandintensities.
AnassignmentofthesebandsbasedontheirDFspectraandcomparisonwiththetheoretically
predictedfrequenciesofninestableconformersof2-oFPEisgiveninourcombinedpaperwith
V.RamanathanandT.Chakraborty.TheDFspectra[114]producedasaresultoftheexcitation
ofthebandsat000,+43,and+86cm−1featurethesamepeakspacingbutdifferentintensitydis-
tributions.Theinterpretationinthiscaseisratherstraightforward:thethreebandsbelongtothe
sameconformation,andhencehavethesamevibrationalstructureinthegroundelectronicstate.
ThevariationintheintensitydistributiondescendsfromdifferentFranck-Condonfactorsforthe
differentvibronicstatesinthefirstexcitedelectronicstate.Themostrigorousandunambiguous
assignmentisprovidedbyhigh-resolutionexperiments.Wemeasuredunderhighresolutionthe
R2PIspectraofthebandsat000,+3,+43,+86,+109,+125,and+135cm−1.Therotationalstruc-
turesofallbandswiththeexceptionofthebandat+3cm−1havebeensuccessfullyfittedand
therespectivemolecularparametershavebeenobtained.Therotationalconstantsfortheground,
S0,andthefirstexcited,S1,electronicstateobtainedfromthefitsofthebandsunderscrutiny
(seeTable6.1)areverysimilarpointingthatallthesebandsoriginatefromthesameconformer.
Slightdifferencesmightoriginatefromtheadditionalexcitationofdifferentvibrations.Compar-
ingtheexperimentalrotationalconstantsobtainedfromthefitwiththecorrespondingrotational
constantsforthetheoreticallypredictedconformations(seeTable6.2),wefindthatthebest
matchisobservedforthemoststableoppositegaucheconformer1.Thisprovidesacompelling
argumentthatconformer1givesrisetothemostintensebandat37589.18cm−1aswellasto

108

anticonformers−1.ofthe2-andcminfrequenciesvibrationalnormal-modepredictedTheoretical
cminenvgiarefrequenciesCIStheforactorsfscalereliablenosincerescaledbeennotevhastatexcitedetheforfrequenciesground,theforfrequenciesvibrationalhegaucthe6.3:ableT
CISandcc-pVDZ/ground,thefor−1
0SThe32].[Ref.0.9970ofactorfscaleappropriatetheusingwnscaleddobeenevhastateelectronic,oMP2theatcalculatedmonomerFPE
SAllailable.vaaretheoryofelvlecc-pVDZ/cc-pVDZ/.Theelyvrespecti,theoryofelvle0xcited,efirsttheand,
1Sofstateelectronic,

ConformationalProbingof2-ortho-uorophenylethanol

106510381005958938945906926873886782854759837722766671744628689556583531551514528423494410441383417364373308336224276185201134150839243460S
S1

S1047337548836279782715247649422444977479242024366908861434085828511000S1068291526764321021263514731944231423670902188402618871133378550779891

S357550834179763165257619412694997469191984396918841494115838469510230S3465307652110282575167329362444296749091924016178661463705517849710711

S3285467986110212075347759442074737659282074516888641093556038569710520S5367783810592974947179042234196629021823846178421073485678107910823301

S5468086210352875347809442714747659292084496898651063336038560971044317S5357764010602864937199052244176639011823836178551063415668111801072306

S5457966010243115347779452900474765929207450687864109360602856961046321S5357773910603074947189042231417664902181384616841107356566811801079320

S8278197131252576594302745027509262134356848811474195848541091034359553S7656510222974297199311223429660902192403626869128373548777811063340429

S831639742735237659410257501752925205433686887143417583854901045322550S7665010162705277181940222428660904188406622874127371549776711064308529

S831469583185227680945275501749931202435684882143423581857801039363549S3810073165257151929224430665908187401622873129369548781681063341532764

Conformer1

Conformer2

Conformer3

Conformer4

Conformer5

Conformer6

Conformer7

Conformer8

Conformer9

Discussion6.4

109

thebandsat+43,+86,+109,125,and+135cm−1.Somediscrepancybetweentheexperimental
andtheobservedparametersisfoundforthetransitionmomentratio,whichstemsmostlikely
fromtheinsufficientaccuracyofthecalculationsintheexcitedelectronicstate.Noconvincing
fitoftherotationalstructureofthebandat+3cm−1isavailableduetothelowsignal-to-noise
ratiooftheexperimentalspectrum(seeFig.6.4).Allattemptsforfittingthespectrumusingas
startingparameterstheonescorrespondingtothemoststableconformerfailed.Thebestvisual
agreementwasachievedonlywhenemployingasstartingvaluestheparametersofthesecond-in-
energyconformer7andthethird-in-energyconformer9.Fromenergyconsiderations,conformer
7isthemorefavorableone.Itstheoreticalvibronicfrequencies,however,donotmatchtheones
experimentallyobservedintheDFspectrum(fordetails,seeourpaper).Thoughconformation
9isenergeticallylessfavored,itsrotationalconstantsandvibrationalfrequenciesareinagood
accordwiththeexperimentalobservations.Thatiswhywetentativelyassignthebandat+3
cm−1asoriginatingfromconformer9.Inourexperiment,wehaveidentifiedtwoconformations
andwehaveindicationsforathirdoneoutofninetheoreticallypredictedstablestructures.The
intensityratioofthethreeconformersis∼94:5:1.Thisinvokesthequestionwhypreferably
onlythemoststableconformationispresentinthebeam.Aplausibleexplanationcanbefound
intheshapeofthePESandintherelaxationdynamicsduringtheprocessoftheadiabaticexpan-
sion.ThePESischaracterizedbythreetroughs(ridgesinFig.6.9)hostingthethreegroupsof
conformations,theoppositegauche,theanti,andtheadjacentgaucheones.Theconformations
withineachoftheabovegroupsareseparatedbylowpotentialbarriers,whichfacilitatesthe
relaxationfromthehigher-lyingtothelowest-energyconformationintherespectivegroup.The
conformationsinthedifferentgroups,however,areseparatedbyhighpotentialbarriers,which
shouldprecludetheirinterconversion.Thusthismechanismreducesthetotalnumberofpossibly
observedconformationstothree,correspondingtothelowest-energystructureswithineachof
thethreegroups.Next,wehavetoexplainwhythetwodominantconformationspresentinthe
coldmolecularbeamaregauchestructures.Thisresultisinstrikingvariancewiththecasesof
nonfluorinated2-PE(Ref.[46])andparafluorinated2-FPE(Ref.[104])whereonegaucheand
oneanticonformerhavebeenobserved.Aspointedoutabove,duetothereducedsymmetry
ofthespecies,theorthofluorinated2-FPEprovidesalargernumberofdistinctconformational
structurescomparedtothenonfluorinatedandthepara-fluorinatedcounterparts.Takinginto
accounttheenergyordering,wefindthatthelowest-energyantistructureisatthefourthposi-
tionbeingprecededbythreegaucheconformations(seeTable6.2).Theexistenceofseveral
conformationswithlowerenergyimpliesmorerelaxationchannels,whicheventuallyleadsto
adepletionofthepopulationoftheanticonformationinthecoldmolecularbeamclosetothe
detectionlimit.Animportantaspectthatdeservesattentionisthecomparisonbetweenthecon-
sideredconformersandthenonfluorinated2-PEandtheparafluorinated2-FPE.Asitwasshown
inourrecentstudies,themoststablestructuresofboth2-PE(seeChapter3and[46])andthe
2-pFPE(seeChapter4)arestabilizedbyanonclassicalOH∙∙∙πhydrogenbondbetweenthe
terminalOHgroupofthesidechainandtheπelectronsofthearomaticring,andthestabilityof

110

ConformationalProbingof2-ortho-uorophenylethanol

thisbondisalmostunaffectedbytheparasubstitutionwithfluorine.Theorthofluorination,in
additiontotheOH∙∙∙πbond,providesthefeasibilityforaσhydrogenbondbetweentheterminal
OHgroupandtheorthosubstitutedfluorineatom(conformer7).Eveninthiscase,wehave
foundthatthelowest-energyoppositegaucheconformerisstabilizedbyanonclassicalOH∙∙∙π
hydrogenbondratherthanbyaσbond.Theresultclearlydemonstratesthatevenunderthe
conditionsofasignificantlydistortedπelectrondensityinthebenzeneringbroughtaboutbythe
presenceofthestronglyelectronegativefluorineatomandthefavorablegeometryenablingthe
formationofacompetingσbond,thenonclassicalOH∙∙∙πbondprevailsandremainsimportant
forthesustainingoftheconformationalstructureoftheflexiblemolecule.

conclusionsandSummary6.5

Inthischapterwehaveinvestigatedtheinfluenceoffluorinesubstitutionontheweakintramolec-
ularforcesstabilizingcertainconformationalstructuresofflexiblemolecules.Asaprototype
system,wehavechosen2-phenylethanol,whichisthehydroxylanalogofthesimplestneu-
rotransmitter,2-phenylethylamine.Orthofluorinated2-phenylethanolhasbeeninvestigatedby
combinationoflow-andhigh-resolutionresonanceenhancedtwo-photonionizationexperiments
supportedbyhigh-levelquantumchemistryabinitiocalculationsoftheenergetics,structure,and
normal-modevibrationalfrequencies.Allvibrationalbandsobservedinthelow-resolutionR2PI
spectrahavebeenpreliminaryassignedinafirststepbycomparisonoftheexperimentalband
positionswiththetheoreticallypredictedfrequenciesinthefirstexcitedelectronicstate.Further
argumentsareprovidedbythecomparisonoftheobservedvibrationalbandpositionsintheDF
spectrawiththetheoreticalfrequenciesinthegroundelectronicstate.[114]Thefinalassignment
reliesonthecomparisonoftherotationalconstantsobtainedfromthefitsofthehighlyresolved
spectrawiththeonestheoreticallypredictedfortheoptimizedconformationalstructures.We
havefoundthatthemostabundantconformerinthecoldmolecularbeamisthetheoreticallypre-
dictedmoststableonecorrespondingtoagauchestructurestabilizedbyanonclassicalOH∙∙∙π
hydrogenbondbetweentheterminalOHgroupofthesidechainandtheπelectronsoftheben-
zenering.Higher-energygaucheandanticonformershavebeententativelyidentified,although
theyweaklycontributetothetotalconcentrationof2-orthofluorophenylethanol.Nodetectable
signalsstemmingfromotherconformationshavebeenobservedinourexperimentalresults.The
mostlikelyreasonforobservingpredominantlyonlyoneanddetectingothertwooutofnine
theoreticallypredictedconformationsisthecomplexshapeofthePESandtheconformational
relaxationchannelsleadingtointerconversionofthehigher-energyconformationstothelowest-
energyones.Theresultsshowthateveninthecaseoftheorthosubstitutionwithastronglyelec-
tronegativeatomleadingtoasignificantdistortionoftheelectrondensityinthearomaticring
andprovidingaconvenientconfigurationfortheformationofacompetingσbond,thenonclas-

6.5

Summary

and

conclusions

sicalπbondstillprevailsandremains

the

dominant

thisdemonstratestheimportanceandrobustness

conformational

structures

of

flexible

molecules.

intramolecular

of

the

OH∙

π∙

binding

bonds

motif.

for

the

In

111

conclusion,

stabilization

of

112

ormationalConf

obingPr

of

ortho2-

ophenylethanol-uor

7Chapter

ofClustersHydratedSinglyandophenylethanol-uorpara2-2-ortho-uorophenylethanolStudiedby
High-ResolutionUVSpectroscopyandab
Calculationsinitio

Thesinglyhydratedcomplexesoftheflexibleprototypemolecules,2-para-
fluorophenylethanol(2-pFPE)and2-ortho-fluorophenylethanol(2-oFPE)
havebeeninvestigatedinacoldsupersonicbeambycombinationofhigh-
resolutiontwo-colorR2PIspectroscopyandquantumchemistryabinitiocal-
culations.Wehaveidentifiedtheconformationalstructuresofthecomplexes
between2-pFPEandwater,and2-oFPEandwater,whichcorrespondtowa-
terbindingtothemoststablegaucheconformersinbothcases.Nodetectable
structuralchangesofthehostmoleculeshavebeenobserveduponattachment
ofasinglewatermolecule.Forthe2-oFPE∙H2Ocomplexwehaveobservedan
additionalstructurewithoneofthenext-in-energygaucheconformersofthe
monomerwhichpointedtothestabilizingeffectofthewaterformonomers
conformerwithlowerabundance.

113

114WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

oductionIntr7.1

Waterasabiologicaluniversalsolventhasattractedtheattentionofthescientificsociety.A
varietyofspectroscopicmethods[4,6–9,14,15,76,118]havebeenusedinthedeterminationof
themostabundantwaterclustersofdifferentbiologicallyrelevantmoleculesproducedbymi-
crosolvationinasupersonicjet[10–13,16,85,119].Astepforwardintheassignmentofthese
conformersisprovidedbygasphaseUVhigh-resolutionresonanceenhancedtwo-photonspec-
troscopyasasensitivetoolforrevealingthestructuralgeometriesofthehydratedclusters.In
Chapter3anassignmentofthedimericwaterclusterof2-phenylethanol(2-PE)wasmadeon
thebasisofcomparisonoftheparametersofthepartiallyrotationallyresolvedvibronicspectrum
withthepredictedonesfromhigh-levelabinitiocalculations.Here,thisapproachisusedfor
thedeterminationofthestructureofthesinglyhydratedclustersof2-para-fluorophenylethanol
(2-pFPE)and2-ortho-fluorophenylethanol(2-oFPE),moststablemonomericstructuresofwhich
werestudiedinChapters4and6.Asmentionedalready,themaingoalsofsuchinvestigation
aretheeffectofthewaterattachmentonthemonomersconformationalshape,thetypeofin-
termolecularinteractionandthepossiblestabilizingeffectofthewateronparticularconformers.
ThemassresolutionofourexperimentalsetupallowsmeasuringofUVhighlyresolvedrovibra-
tionalspectraatbothmasschannels:waterclusterandmonomer,respectively.Thus,issuesas
overlappingofcloselyingvibrationallevelsandfragmentationoflargerclusterscanbedirectly
ed.resolv

7.2ExperimentandDataProcessing

Themass-selectedresonanttwo-photonionization(R2PI)apparatususedinthepresentstudyhas
beenthoroughlydescribedinSection2.1.Liquid2-pFPE(97%)obtainedfromAldrichandused
withoutfurtherpurification,wasvaporizedat100◦Candmixedwith30mbarofwatervapors
inastreamofargonatastagnationpressureofabout3.2bars.Followingfree-jetexpansion
intovacuumthroughapulsed-operatedheatablenozzlewithorificediametersof500µm,the
rotationaltemperaturewasfoundtobeapproximately6-9Kundertheconditionsemployed.
Fortheone-colorlow-resolutionspectraphotonswereused,generatedbyfrequencydoubling
inaBBO−1I(BetaBariumBorate)crystaloftheoutputpulsesofacommercialbroadband(Δυ
∼0.4cm(FWHM))dyelaser(LambdaPhysikFL2002)operatedwithaCoumarin153laser
dye.Thefirstcolorexcitationphotonsinvolvedinthetwo-colorhigh-resolutionspectrawere
generatedbyfrequencydoublinginaKDPcrystalofthepulsedamplifiedoutputofacontinuous
wave(CW)single-moderingdyelaser(Coherent,CR699-21)operatedwithCoumarin334.
ThelimitedspectralresolutionofourscanningUVlaserpulsesisnotsufficientforresolution
ofsinglerotationallinesinthecongestedspectraoflargemolecularcomplexes.Forthisreason

DiscussionandResults7.3

115

weuseacomputer-basedmethodforfittingoftheexperimentalhighlyresolvedspectra.The
algorithmandthecomputerprogramelaboratedinourgrouphavealreadybeendescribedin
Section2.3.Crosscorrelationisemployedasanevaluationfunction.Fortheproductionofthe
simulatedspectraofthe2-pFPEwatercomplex,werantheprogramusing300generationswith
500individualsinageneration.

Results7.3Discussionand

7.3.12-pFPE-WaterCluster

Spectraw-ResolutionLo

Low-resolutionone-colorresonanttwo-photonionization(R2PI)spectrahavebeenrecordedat
threedifferentmasschannels(m/z=140,158and176)underwater-inconditionsneartheS1
←−S0electronictransitionof2-pFPE.ThespectraaredepictedinFig.7.1,whereinaddition
tothewater-inspectra,awater-freespectrumofthemonomer(m/z=140)isincludedforcom-
parison(SeeFig.7.1(a)).ThespectrashowninFigure7.1(b),7.1(c),7.1(d),weremeasured
underwater-inconditionsmonitoringthe2-pFPE∙H2Ocluster(m/z=158),2-pFPE∙(H2O)2clus-
ter(m/z=176)andmonomer(m/z=140)masschannels,respectively.Themostintensefeatures
ofthesinglyhydratedcomplexarelabeled-18cm−1,-2cm−1,53cm−1,and76cm−1,respectively.
Ithasbeenfoundthatmostofthesepeakshavecorrespondingonesatthedoublyhydratedcom-
plexmasschannel(Fig.7.1(b)).Forpeaks-2cm−1and53cm−1itisclearthattheyoriginatefrom
thefragmentationofthe2-pFPE∙(H2O)2complex.Peak-18cm−1coincideswiththeshoulderof
abroadpeakatthemasschannelofthedoublyhydratedcomplex.Thisleadustotheconclusion
that-18cm−1isamixtureoftwoclose-lyingbands:oneofthe2-pFPE∙(H2O)2andtheother
oneofthe2-pFPE∙H2Ocomplex.Onlyforpeak76cm−1thereisnodoubtthatitresultsfrom
thesinglyhydratedcluster.Theclearoriginofthisvibronicbandmadeitagoodcandidatefor
aninsighttotheconformationalshapeofthe2-pFPE∙H2Ospecies.Acomparisonbetweenthe
monomermasschannelsspectra(seeFig.7.1(a),7.1(d))recordedunderdifferentwatercondi-
tionscanelucidatethenatureofsomevibronicfeaturesobservedpreviouslyunderwaterfree
conditions[].Alladditionalpeaksappearingatthewater-inmonomerspectrum(Fig.7.1(d))
comparedtothewater-freespectrum(Fig.7.1(a))canbeeasilyattributedtothefragmentation
ofthewatercomplexesandthesmallbumpsinthewater-freespectrum(Fig.7.1(a))at-40cm−1
and-18cm−1areassignedasoriginatingfrom2-pFPE∙(H2O)nclustersfragmentation.Thenext
levelofanalysiswastorecordthehigh-resolutionspectraoftheabove-discussedbandsandto
comparethemwiththeresultsfromabinitiocalculationstodemonstratethepreferredbinding
moietyforthewaterinthesinglyhydratedstructure.

116

WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

Figure7.1:One-colorR2PIspectraoftheS1←−S0electronictransitionofthe2pFPEwatercluster
measuredatthemonomer(m/z=140)masschannelunderwater-free(a)andwater-present(d)conditions.
One-colorR2PIspectra:(c)recordedatm/z=158masschannelofthe2pFPE∙H2Ocomplex,and(b)
recordedatm/z=176masschannelofthe2pFPE∙(H2O)2complex.

SpectraResolutionHigh

Thebandsat-18cm−1,-2cm−1,53cm−1,and76cm−1werescannedwithhigherresolution
bothatthecluster(m/z=158)andmonomer(m/z=158)masschannels.Therotationallyresolved
spectraofthevibronicbands-18cm−1,-2cm−1,and53cm−1arecongestedandwithpoorquality,
whichprecludesthefittingprocedureforthesebandsandtheirdirectassignmenttopredicted
clusterstructures.Forcompleteness,theirexperimentalspectraarepresentedinFigs.7.2-7.3
andavisualassessmentismadeinthediscussionSection.Thenicelyresolvedspectrumofband
76cm−1allowsustoperformacomputerbasedfittingprocedureandtheresultingsynthetic
spectrumispresentedasininvertedtraceinFig.7.4.Theexperimentalspectrumcoversa
frequencyrangeofca.2.5cm−1andmanifestsawell-resolvedstructurebuiltupofseparatepeaks

7.3

Results

eFigur-2and

and-2

and

7.2:1−cm

1−cm

Discussion

High-resolution

o-colortw

UV

(lo7.1,Figureinpicture)wer

plausibleassignmentanddetails,seetext.

R2PI

spectrum

of

massatrecorded

vibronic

bands

withchannels

/-18

=mz

1

−cmand140

(upper
158.zmand

/=zmand140

picture)
F158.or

158.

For

117

118

aterW

Clusters

of

ophenylethanol-uorpara2-

and

2-

ortho

ophenylethanol-uor

1

Figure7.3:High-resolutiontwo-colorUVR2PIspectrumofvibronicbands53cm−1(upperpicture)
and76cm−1(lowerpicture)inFigure7.1,recordedatmasschannelswithm/z=140andm/z=158.For
plausibleassignmentanddetails,seetext.

DiscussionandResults7.3

Figure7.4:High-resolutiontwo-colorUVR2PIspectrumofvibronicband76cm−1inFigure7.1,
recordedatm/z=158.Thebandisassignedasthe000originbandofconformer8gofthe2-pFPE∙H2O
complexwiththerotationlesstransitioncenteredat37143.476(1)cm−1.Uppertrace:experimentalspec-
trum.Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable7.1(for
xt).teseedetails,

119

withalinewidthof250MHzresultingfromaggregationsofcloselysituatedsinglerotational
lines.ThespectrumhasapronouncedQbranchinthecenter,anintenseRbranchandlessintense
Pbranchbothwithirregularbutresolvedstructures.Itwasfittedbytheroutinebasedongenetic
algorithmsdescribedinSection2.3whichproducedthemolecularparametersofthesearched
conformericstructure.ThesimulatedspectrumwasconvolutedbyaGaussianprofilewithaline
widthof160MHz.Thesimulationisinagoodagreementwiththeexperimentreproducingthe
individualpeakpositions,intensitiesandoverallprofileoftheobservedspectrum.Theresulting
rotationalconstantsfortheground,S0andthefirstexcited,S1,electronicstates,thetransition
momentratio,thebandoriginposition,υ0,andtherotationaltemperature,T,aresummarizedin
7.1.ableT

CalculationsinitioAb

theThe2-pGaussianFPEsingly03hprogramydratedpackagecluster[]bywasabusedinitioforthequantumoptimizationchemistryofvariouscalculations.Theconformationsinitialge-of

120

WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

cBandarameterP1−cm760.0604(15)´´A0.01934(33)´´B0.01627(27)´´C0.0602(12)´A0.01891(24)´B0.01607(22)´Cµa2:µ2b:µc232:68:0
ν0,cm−137143.476(24)
6.1(6)(K)T97(%)correlationcrossBest-fitTable7.1:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,transitionmomentratio,µa2:µb2:µc2,bandoriginfrequency,ν0,rotational
temperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofthebandassignedto
the000originofconformer8gofthe2-pFPE∙H2Ocomplex,showninFigure7.5.Numbersinparentheses
arestandarddeviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevalues
ofµa2,µb2,andµc2inthetransitionmomentratiodoesnotexceed5(%).

ometriesofthewaterclusterswerecarefullyselectedbychemicalintuition,exploringpossible
bindingsitesofthewatermoleculetothemoststablegaucheandantimonomerstructures,ob-
servedexperimentallyanddescribedinChapter4.The2-pFPE∙H2Ocomplexeswerenumbered
bytheirprobabilityforappearance.Inthisway,fivewaterclusterswiththegauchemonomerand
twowaterclusterswiththeantimonomerwereoptimizedintheground,S0andthefirstexcited,
S1,electronicstatesattheMP2/6-311G∗∗andCIS/6-311∗∗levelsoftheory,respectively.Asa
result,fivestableclusterstructuresweredeterminedandsubjectedtofurtherfullgeometricalop-
timizationsforbothground,S0andandthefirstexcited,S1,electronicstatesatMP2/cc-pVDZand
CIS/cc-pVDZlevelsoftheory,respectively,withanimprovedelectroncorrelation(seeFig.7.5).
Thecalculatedvibrationalfrequenciesofthesefiveconformersarepositivewhichverifiesthat
theycorrespondtominimaonthepotentialenergysurface.Allrelevanttheoreticalparameters
ofthesestructuresaresummarizedinTable7.2.Themoststableconformerisconformer2g,
wherethewatermoleculeservesasabridgebetweentheterminalhydroxylgroupandthearo-
maticringformingaπ-hydrogenbondwiththebenzeneringandacceptingaprotonfromthe
OHgroupofthesidechain.Thesecond-in-energyconformerisconformer8g,inwhichthe
watermoleculebindssidewaystothe2-pFPEgauchestructurethroughtheformationofasingle
σbonddonatingaprotontothehydroxylgroupof2-pFPE.Thentheenergyorder,asfollows
fromtheinclusionofthebasissetsuperpositionerrors(BSSE)andzero-pointvibrationalener-
gies(ZPVE),is:conformer3a,conformer5a,conformer1g.Theinclusionofthesecorrections

DiscussionandResults7.3

121

isnotcandidatesaffectingfortheobservenerationgyinordertheofethexperiment.firstthreeItonlyclusterincreasesstructures,theenerwhichgyaredistancethemostbetweenprobablethe
firsttwoconformers,butstill61cm−1(seeTable7.2)isanegligibledifferenceandmayresultin
reorderinguponthechangeoftheoreticallevelandmodel.

Discussion

Thehigh-resolutionspectrumofband76cm−1wasrecordedsimultaneouslyatthemasschan-
nelsofthemonomerm/z=140andthemasschannelofthesinglyhydratedclusterm/z=158(see
Fig.7.4).Thetwotracesareidenticalwhichdemonstratesthenatureofthisbandasoriginating
formthe2-pFPE∙H2Ocluster.Forourfittingprocedure(seeSection2.3),thespectrummoni-
toredattheclustermasschannelwasusedforfurtherassignment.Thefitswerestartedprobing
therotationalconstantsfortheground,S0andthefirstexcited,S1,electronicstatesandthetran-
sitionmomentratiosobtainedfromthetheoreticaloptimizationsofconformers2g,8g,3a,5a
and1g(seeFig.7.5).Thecrosscorrelationratiobetweenexperimentalandsyntheticspectra
reached97%whenthefitwasinitiatedwiththeparametersforconformer8g.Thesoreceived
experimentalvaluesareinagoodagreementwiththecalculatedonesfortheabo−1ve-mentioned
conformer,whichmakestheassignmentstraightforward.Therefore,band76cmcorresponds
totailtheofwtheater2-pclusterFPEgauchestructurestructurewhereviatheawatersingleσbindshtoydrogentheterminalbondhdonatingydroxylagroupprotonoftothethefleoxygenxible
ofmotifthewOHasgroupobservedoffortheonesideofchain.theThishigh-enerpositiongywofaterthebandswaterofis2-PEnot(seeunusual,Chapterthe3).sameAnbindingaddi-
tionalargumentfortheassignmentistheproximityofthewatermoleculetohydrogenatomH10
ofthearomaticringthatfavorsaformationofanadditionalσhydrogenbondinwhichthewater
isaprotonacceptor.Thisassumptionissupportedfromthewell-resolvedrotationalstructureof
thisband(seeFig.7.4),typicalforrigidoralmostrigidcomplexesstabilizedbymultiplebonds
where−1tumblingmotionishindered.Asaconsequence,thepredictedsmallenergygapof61
cmfromthemoststableconformer(conformer2g)canbevalidated.Comparingthegeomet-
ricalparameters(interatomicdistances,planarangles,anddihedralangles)ofthe2-pFPEbare
molecule(conformer1,seeTable4.2inChapter4)andthecomplexwithwater(seeTable7.2),a
conclusionforthestabilityofthe2-pFPEspeciesuponcomplexationwithwatercanbemade.It
isclearthattheintramolecularπ-hydrogenbondofthehostgauchemonomerisunperturbedby
theattachmentofthewaterbyintermolecularσbonds.Aconsiderablechangeisobservedfor
Ttheablestransition4.2and7.2).momentTheratiochangeofthecan2-pbeFPE∙attribH2Outedtoclusteramassecomparedffecttoresultingtheonefromofmonomerreorientation1(seeof
theprincipleaxesofinertiaduetothebindingofawatermolecule.
Onolutiontheatredthesideofmassesthisofthevibronicsinglybandhseydratedveralclusterotherandintensethefeaturesmonomeratmeasuredthesameundertimehigheraresitu-res-

122

WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

:(O9-H21),−1µc2+ÅcmpconformersorFconformerorFb2−1−1−1
ZPVE),−1FPE∙cmH−Interatomic:1gr11.87aadistances,a:rand2g2gies,enerevRelatix.compleO1
12gS1)p2-theofconformersaaa
distancerepresents(F13-H21),antidplanar,angles,-23.3170.11.99hegauc(
0Conformerα5(µ1:elrEelrEelrEaµ2,C,B,A4τ5α1r3τ2τ1τ4α3α2α1α--2.762.56--2.822.543.182.92Å(H19-C1),rConformerarameterP
µb2)7.5).Fig.(see(C5F13H21O20)b7.5).Fig.(seeaxt.teseedetails,orFtable.the2-ofconformationsparentThe.theoryofelvleMP2fromobtainedas(ZPVE)gyenervibrationalzero-pointand(BSSE)errorsuperpositionsetbasistheofinclusionwithandwithout2-theofratio(TM)momenttransitiondipoleElectric.elyvrespectitheofstateelectronic7.2:ableT(BSSE(BSSE),cm,µ:cmcmcm-9.9gde(H19O9H21O20),170.7gde(O9H21O20),-179.6-179.6-67.1-63.8177.9173.6-65.1-58.8-87.4-89gde(H19O9C8C7),-179.8-179.966.560.7178.9163.664.158.464.758.5gde(C2C7C8O9),89.789.292.691.282.410286.787.58490.6gde(C3C2C7C8),109.4106.8109.3105.8109.5106.9109105.4109.3105.6gde(C8O9H19),107.3107.4113.1112.3107.5107.41113112.2112.8112gde(C7C8O9),113111.1113.7111112.7110.9113.4110.8114111.1gde(C2C7C8),120.0120.8120.1120.6119.9119.1119.8120.3120.1120.4gde(C1C2C7),
α:5elr-34.7144.5108.864.692.62.152.051.962.051.91S
S0Conformerand(O9H19O20)angleplanarrepresents8g
and(H19-O20),distancerepresentshegauc0.012060.012280.013830.014450.012400.013890.015820.016490.019650.020490.013550.013810.016230.016950.013760.016120.018690.019300.021820.022490.080120.080340.063230.063450.081830.074210.059990.061830.058840.06218
a2pMP2theatcalculatedmonomerFPEShegauc
5a-1836-1881-458-61-0-1330-1338-564-244-0-2362-1678-1126-10-04-0:93:7-3:91:6-8:86:6-0:80:20-88:8:0S
3a(S1anti)representsand(F13H21O20)angleplanarcµ2
CISfromobtainedasτground,theforconstantsrotationaland,
αfitheof2-ofxcompleydratedhsinglytheofstructuresconformationalstablemostevangles,dihedral,
ofheadertheinparenthesesinreportedareesxcompleydratedhevrespectithetorisevinggiFPEE146.3165157.9156.6161.1
SτConformer1grepresents4andcc-pVDZ/CISbbb0
S1-33.6138.12.33hegauc(
bbb)τdihedralS40geometryoptimizedtheofcalculationcc-pVDZ/
-38.3145.62.15Sangle0bConformerbb5aangledihedralrepresents,theoryofelvlecc-pVDZ/xcited,efirsttheand,
(-38137.72.33S1anti(O9H19O20H21)pcc-pVDZ/bbb)
FPES1,

DiscussionandResults7.3

Figure7.5:Electronicgroundstate,S0,structuresofthe2-pFPE∙H2OcomplexoptimizedattheMP2/cc-
pVDZleveloftheory.Typicalangles,bondlengths,inertialparameters,andbindingenergiesforthese
structuresarelistedinTable7.2.

123

nalsated.Thedoesnotspectraallowaredirectpresentedfitsforinthesefiguresbands7.2andwith7.3.ourThecomputerpoorsignal-assistedtonoiseprocedure.ratioofButtheselookingsig-
atthelowresolutiontracesofthedoublyhydratedclusters(seeFig.7.1)andtheirfragmentation
patternatlowermasseswecanreadilyassignbands-2cm−1and53cm−1asoriginatingfromthe
waterclusterwithtwowatermoleculesattachedtothe2-pFPEmonomer.Anadditionalargu-
mentisthesimilarhighlyresolvedrotationalprofileofthesebands,whichconsistsinbothcases
ofadominantQbranchandweakPandRbranches.Inthecaseofband-18(seeFig.7.2)the
situationiscompletelydifferent.Thesimultaneouslyrecordedsignalsatthemassesofthesingly
hydratedclusterandmonomerhavemismatchingrotationalcontours.Oneexplanationisthatthe
bandconsistsofamixtureofseveralrovibronicfeaturesoriginatingfromdifferentclustersizes.
ThestrongQbranchofthespectrummeasuredatmasschannelm/z=140leadstotwopossible
conclusions.Eitheritisproducedfromafragmentationofthedoublyhydratedwatercluster,as
isthecaseforbands-2cm−1and53cm−1oritisoriginatingfromafragmentationofacluster
withthebuffergasargon.InvestigatingtheprofileoftheRbranchesofthesignalsmeasured
atthemassesofthesinglyhydratedclusterandthemonomer,onecanseearesemblanceinthe

124WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

rotationalcontours.Thispointstoafragmentationofa2-pFPE∙H2Oclusteratthemonomer
masschannelandthesignalatm/z=158canbeassignedasoriginatingfromthemoststable
theoreticallypredicted2-pFPE∙H2Ostructure2g.

7.3.22-oFPE-WaterCluster

Spectraw-ResolutionLo

Theone-colorREMPIspectraof2-oFPEanditssinglyhydratedclusterrecordedatmasschan-
nelsm/z=140andm/z=158,respectivelyaredepictedinFig.7.6.Thereisonlyonestrong−1
vibronicbandintheS1←−S0electronictransitionofthewatercluster,blue-shiftedby62cm
relativetothemonomersoriginband.Therearesomeindicationsforothervibronicbandsbut
theirrelativelysmallintensityprecludesahighresolutionmeasurement.−1Asanextstepofeval-
uation,torevealtheconformationalgeometrygivingrisetothe62cmvibronicfeature,its
recorded.aswspectrumhigh-resolution

SpectraHigh-Resolution

Measuringband62cm−1underhigherresolution,wehaveseenthatitconsistsoftwoclosely
spaced(2cm-1)rovibronicfeatures,subband62Acm−1atlowerenergyandsubband62Bcm−1
athigherenergy.TheexperimentalspectraofthesetwosubbandsarepresentedinFigures7.7
and7.8(uppertraces).Theyhavecompletelydifferentrotationalprofiles,whichpointsoutthat
theyoriginatefromtwodissimilarsinglyhydratedclustersof2-oFPE.Band62Aischaracterized
byastrongQbranch,anintenseregularstructureofthePbranchwithwellresolvedindividual
peaksandaweakRbranchwithpoorresolution.Band62BconsistsofadominatingQbranch,
weakPbranchandagraduallyincreasingRbranchthatendswithacharacteristicsuddendrop
ofintensity,bothPandRbrancheshavingirregularrotationalstructure.Thespectraofthese
twobandsmanifestahybrida-,b-,andc-typecharacterwithdifferentcontributions.Theydo
notconsistofsinglerotationallinesbutweobserveacharacteristicpeakpatternproducedby
aggregationofseveralrotationallinesresultinginaminimumvalueforthepeakwidthof250
MHzfullwidthathalfmaximum(fwhm).Tofindtherotationalconstantsfortheground,S0and
thefirstexcited,S1,electronicstates,thetransitionmomentratio,therotationaltemperature,and
theoriginposition,ν0,weusedthecomputer-assistedrotationalfitprocedurebasedongenetic
algorithmdescribedinSection2.3.Asstartingvaluesforourfits,weusedthevaluesresulting
fromabinitocalculationsofthegroundstaterotationalconstantsforthefirstsixinenergy
singlyhydratedwaterclusters.Theproducedtheoreticalstickspectrumwasconvolutedusing
aGaussianlineshapewithafwhmof160MHz.Thebest-fitresultsforthebandsareshown
intheinvertedspectrainfigures7.7and7.8.Thesimulatedspectra(lowerinvertedpart)agree

DiscussionandResults7.3

Figure7.6:One-colorR2PIspectraoftheS1←−S0electronictransitionofthe2oFPEmonomerandits
singlyhydratedwatercluster,measuredatthemonomer(m/z=140)masschannelunderwater-present
conditions(b),andattheparent(m/z=158)masschannelofthe2oFPE∙H2Ocomplex(a).

125

wellachievinedbothcrosspeakcorrelationpositionsandbetweenpeaktheeintensitiesxperimentalwiththeandesimulatedxperimentalspectraonesis(upperashightrace).as95%.The
Theexperimentallyobtainedvaluesoftherotationalconstants,thetransitionmomentratios,and
therotationaltemperaturesaresummarizedinTable7.3.

CalculationsinitioAb

Tosupportthestructuralanalysisofthe2-oFPE∙H2Ocomplexweperformedaseriesofabini-
tiostructuraloptimizationsusingGaussian03suitofprograms.Thehigh-resolutionspectrain
Chapter6clearlydemonstratethatthestrongestband00intheUVspectrumofthemonomer
originatesfromthemoststableandthemostabundant0conformationof2-oFPE,i.e.thegauche
structure1,inwhichthesidechainisbenttowardsthebenzenering(seeFig.6.9).Thisweused
asstartingstructureinthemodelingprocessofthewatercluster.Thewaterwasattachedto

126

WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

Figure7.7:High-resolutiontwo-colorR2PIspectrumofband62Acm−1inFigure7.6,recordedatm/z
0=the158.rotationlessThebandistransitionassignedcenteredastheat00origin37651.138(1)bandofcm−1.conformerUpperIIItrace:ofethe2-oxperimentalFPE∙H2Ospectrum.complexLowithwer
invertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable7.3(fordetails,seetext).

BandarameterP+62Acm−1+62Bcm−1
0.0568(14)0.0493(12)´´A0.02511(43)0.02486(43)´´B0.01968(35)0.02038(28)´´C0.0566(14)0.0486(11)´A0.02398(43)0.02415(31)´B0.01937(34)0.01996(17)´Cµa2:µb2:µc265:30:570:25:5
Originν0,cm−137651.14(5)37653.168(9)
TemperatureT(K)7.93(4)7.5(4)
Best-fitcrosscorrelation(%)9595
Table7.3:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,thebandoriginfrequency,ν0,the
rotationaltemperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofbands62A,
and+62Bcm−1ofthe2-oFPE∙H2Ocomplex,showninfigures7.7and7.8.Numbersinparenthesesare
standarddeviationinunitsoftheleastsignificantquoteddigit.Theuncertaintyfortherelativevaluesof
µa2,µb2,andµc2inthetransitionmomentratiodoesnotexceed5(%).

DiscussionandResults7.3

Figure7.8:High-resolutiontwo-colorR2PIspectrumofband62Bcm−1inFigure7.6,recordedatm/z
0=the158.rotationlessThebandistransitionassignedcenteredastheat00origin37653.168(2)bandcmof−1.conformerUpperXItrace:ofethe2-oxperimentalFPE∙H2Ospectrum.complexLowithwer
invertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable7.3(fordetails,seetext).

127

variouspossiblebindingsidesofmonomersgaucheconformer1.Theothermonomerstruc-
turesweprobedaspossiblebasesfortheformationofsinglyhydratedclustersweremonomers
gaucheconformers7and9,andanticonformer6,thenextinenergygeometries.Inthisway,
morethan15waterclusterconformersof2-oFPEwereproducedandweresubjecttofurther
optimization.WeperformedcalculationsattheMP2/cc-pVDZleveloftheoryfortheground,S0,
electronicstateandCIS/cc-pVDZleveloftheoryforthefirstexcited,S1electronicstate.After
thecalculationoftheBSSEcorrectionsfortheenergiesoftheabovestructuresandtheinclusion
oftheZPVE,onlythefirst6moststableclusterswereconsideredintheanalysisoftheexperi-
mentalhigh-resolutionspectra.TheoptimizedconformationalspeciesareshowninFigure7.9,
andtheresultsofthecalculationsaresummarizedinTable7.4.Theresultingenergyorderafter
theconsiderationofBSSEandZPVEcorrectionsis:conformerIIIg(0cm−1),conformerXVIa
(21cm−1),conformerIg(52cm−1),conformerXVIIa(128cm−1),conformerXIg(166cm−1)
andconformerVIg(288cm−1).Here,incontrasttothewaterclustersof2-PEand2-pFPEthe
moststablestructureistheoneinwhichthewaterbindssidewaystotheterminalOHgaucheof
thebaregaucheconformer1of2-oFPE.Asmentionedalready,thisordercanchangedepending
ontheleveloftheoryandfunctionalusedfortheoptimizationprocedure.Anotherdifference
fromthesinglyhydratedclustersof2-PEand2-pFPEisthatherewehaveaverystablewater
clusterwithanticonformerofthemonomer(XVIa).Thisisthestabilizingeffectofthewater

128

WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

Figure7.9:Electronicgroundstate,S0,structuresofthe2-oFPE∙H2OcomplexoptimizedattheMP2/cc-
pVDZleveloftheory.Typicalangles,bondlengths,inertialparameters,andbindingenergiesforthese
structuresarelistedinTable7.4.

thatselectsandlowerstheenergyofhigherenergyconformersofthemonomerafterwaterat-
tachment.Thesearemonomersstructureswhicharepresentwithnegligibleabundanceunder
theconditionsoffreejetexpansionofourexperiments.

Discussion

Theattemptstofitthewholerotationallyresolvedspectrumofband62cm−1,startingwiththe
predictedparametersforthefifteenthconformersof2-oFPE∙H2O,werenotsuccessful.Thiscon-
firmsourconclusionthatthisbandiscomposedoftwoseparatecloselyingrovibronicbands
(62Aand62B).Afteracomparisonoftherotationalconstantsfortheground,S0,andthefirst
excited,S1,electronicstatesandthetransitionmomentratioobtainedfromthefitofthehighly
resolvedspectrumwiththetheoreticallypredictedinternalparametersandtransitionmoment
ratiosoftheoptimizedstructuresofthecomplex,weassignedband62Atothe2-oFPE∙H2O
conformerIIIg.ThisclusterisanalogoustoconformerBinthecaseof2-PE∙H2Ocomplex(see
Chapter3)andconformer8ginthecaseof2-pFPE∙H2Ocomplex(SeeFig.7.5).Thebindingof

7.3DiscussionandResults

81.51-140.2-108.7-3.07111.0--4.10115.126.0-66.1123.8172.5--176.65.1470.2SXVIIaConformer78.900.01684104.5128103.60.021561.95112.1950.062733.18111.5-19.0646.9971.4123.071.299,9:0:0,1171.75.1062.1S

59.284.410.01639123.4-107.60.02109-0.06055XVIa2.11111.53.37113.52.3-73.9122.6-124.0--172.95.13SConformer063.278.90.01691115.521103.90.021531.95112.21340.062643.17111.51.8604.3972.4123.1-109.387:10:3-171.25.11S

171.597.10.01884136.5-109.30.02415-XIg2.01113.20.055303.13113.76.5--67.2121.5-107.2--61.04.52SConformer074.191.80.01952120.8166105.80.024641.87112.63470.057842.93111.17.0609.68-68.7121.4-85.673:18:9-58.04.48S

196.3--109.9-112.9--114.3-80.3121.3--76.73.95SVIgConformerS03.423.202.041.86170.8104.254.059.388.7-91.8-25.72.1
92.30.02740288106.40.02911112.250.04655110.6372.6785.9121.878:12:10-58.93.50

12.8886.40.01809-109.00.02459IIIg112.8-0.04803113.8--66.2122.1-61.6SConformerS02.051.913.233.15132.4112.4132.4161.520.398.360.162.6
2.6184.30.019210105.40.02601112.02610.04865110.549.71-59.1121.882:12:658.5

1IgS1.991.85123.9112.9-48.3-39.90.042880.044900.028320.030380.023630.02557-86:11:3
S0-0-0-52
1−3.112.99Å(H15-C1),122.0121.9gde(C1C2C7),1114.2110.6gde(C2C7C8),2112.8111.9gde(C7C8O9),3109.2106.1g(C8O9H19),de490.185.6gde(C3C2C7C8),62.760.2gde(C2C7C8O9),-88.2-88.6gde(C7C8O9H19),gde,(H19020H22)555.961.0gde,(O19H21O20)5gde(O9H19O20H22),21.715.4gde(H19O9H21O20),cm,cm,cm,cm,elrelrelr
ConformerarameterPdαααατ1τ2τ13.363.16Å(O9-H21),dÅ(H19-020),dαατ4τ4ABCratio,TMEEE
cmc2µZPVE),1:−b2µ:+a2µ−−−(BSSE),cm(BSSE
1−111

FPEoxcited,efirsttheand,S0
vlecc-pVDZ/.elyvrespecti,theoryofel
optimizedtheofcalculationcc-pVDZ/geometry
smallwithindicatedareesxcompleydratedhevrespectithetorisevinggiFPEo
CISandcc-pVDZ,ground,theforconstantsrotationaland,angles,dihedralα
/2-ofxcompleydratedhsinglytheofstructuresconformationalstablemostsixtheofgiesenerevrelatiThex.compleO2
MP2theatcalculatedFPEo2CISfromobtainedbeenhasratio,c
ableT7.4:2-ofesxcompleydratedhsinglytheofstateelectronic,S1(TM),momenttransitiondipoleelectricThe2-theofobtainedbeenevha(ZPVE)gyenervibrationalzero-pointandcorrection(BSSE)errorsuperpositionsetbasistheofinclusionwithandwithoutMP2frominnamessconformerthetoxtneletterxt.teseedetails,orFtable.theofheaderthe
µ:b2,angles,planard2µa
µ:distances,interatomicTheoreticalFPEo2-ofconformationsparentThecalculations.cc-pVDZ/
H∙

129

130WaterClustersof2-para-uorophenylethanoland2-ortho-uorophenylethanol

thewatermoleculeviaastrongσhydrogenbondtothemoststable(conformer1)of2-oFPEand
thepossibleexistenceofasecondσhydrogenbondwiththehydrogenatomH10ofthebenzene
ringinclustersconformerIIIg,havebeenalreadythoroughlydiscussedinthecasesfor2-pFPE∙
H2O(seeSection7.3.1)and2-PE∙H2Ocomplexes(seeChapter3).Here,againthestructureof
thehostmolecule(2-oFPE)isnotchangedupontheformationofthesinglyhydratedcluster(see
Fig.7.9andTables6.2and7.4).Themaindifferencefromtheabove-mentionedwaterclustersof
2-PEand2-pFPEisthattheexperimentaltransitionmomentratioofthemonomerisnotchanged
inthepresenceofwater.Previously,wehaveascribedsuchchangetoapuremasseffectdueto
theattachmentofawatermolecule.Thelackofchangeinthecaseof2-oFPE∙H2OcomplexIIIg
canbeattributedtothecompensatingpositionsofthefluorineandthewaterinthisconformer,
whichisnotleadingtoaremarkablereorientationoftheprincipalaxesofinertia.
Themoreintriguingsituationisobservedforband62A,wherewefoundaverygoodagreement
betweenexperimentalvaluesfortherotationalconstantsoftheground,S0,andthefirstexcited,
S1,electronicstates,thetransitionmomentratioandtherespectivetheoreticalvaluesforwater
conformerXIg.InconformerXIgwaterisinsertedbetweentheterminalhydroxylgroupandthe
fluorineatomofthebenzeneringofconformer7(seeChapter6)of2-oFPEthroughaformation
oftwostableσhydrogenbonds,whereinwaterdonatesaprotontothefluorineandacceptsone
fromthe−1OHgroupofthesidechain.Thisisnotoneofthemoststablewaterclusterstructures
(166cm)accordingtothetheoreticalresults.Theexistenceofthisconformerisreasonablebe-
causemonomer7wasidentifiedinthejet(seeChapter6).Thisagaindemonstratestheselective
effectoftheattachedwaterthatstabilizesconformersofthemonomer,leadingtoadetectable2-
oFPE∙H2Ocluster.Inthisindirectway,theabundanceofconformer7of2-oFPEcanbeproved.
Thepresenceoftheotherlow-energyconformers(XVIaandIg)cannotbefullyexcludedsince
therearesomesmallred-shiftedvibronicbandsinthelow-resolutionREMPIspectrummeasured
atthemasschannelofthesinglyhydratedclusterwhichcanoriginatefromtheseconformers.
Forouranalysistheyarenotsoimportantbecausetheycontributebylessthan5to10%tothe
concentration.clustertotal

andSummary7.4Conclusions

Withthecombinationofmassselectivelow-andhigh-resolutionone-andtwo-colorR2PIexper-
imentssupportedbyquantumchemistrycalculations,adetailedstudyofthe2-pFPE∙H2Oand
2-oFPE∙H2Ocomplexeshavebeenperformed.For2-pFPE∙H2Oclusteranassignmentofoneof
thevibronicbandshasbeenmadebasedonthecomparisonofthehigh-resolutionspectrumand
oneofthetheoreticallypredictedstructures.Mostoftheotherhighlyresolvedbandshavebeen
ascribedeithertofragmentsofthedoublyhydratedcomplexesof2-pFPEortooverlappingvi-
bronicbands.Similartothecaseof2-PEsinglyhydratedclusters(seeChapter3),hereagainthe

ConclusionsandSummary7.4

131

moststableconformersof2-pFPE∙H2Oconsistofawatermoleculeattachedtothemoststable
structureofthemonomer,withoutanydetectablechangeofthemonomerstructurecausedby
thewaterattachment.For2-oFPE∙H2Oweobserveasingledominantvibronicbandinthelow-
resolutionspectrumthatunderhigherresolutionshowstwocomponents.Eachofthesetwobands
hasbeenattributedtoadistinctsinglyhydratedcomplex.Duetotheproximityofthefluorine
atomtotheterminalOHgroupofthe2-oFPE,newmonomericstructuresariseprovidingdiffer-
entbindingsidesforthewatermolecule.Basedoncomparisonoftheexperimentalparameters
receivedfromthefitsofthered-shiftedbandofthe2-oFPE∙H2Ocomplexwiththeonesreceived
fromtheabinitiocalculationweassignedthisbandasoriginatingfromthemoststablecon-
formerofthesinglyhydratedcluster.Theotherbandwasascribedtoahigher-energyconformer
(accordingtotheory)ofthe2-oFPE∙H2Ocomplex,wherethewaterservesasabridgebetween
thehydroxylgroupandthefluorineatomofanothermonomersgauchestructure.Here,water
actsasastabilizingsolventforthemonomersconformerspresentwithlowerabundanceunder
theconditionsofourexperiment.Inallthecases,wefoundnosignificantstructuralchangesof
thehostmolecules,whichimpliesthatinthecaseofgaucheconformersof2-pFPEand2-oFPE,
theintramolecularhydrogenbondprevailsovertheintermolecularhydrogenbondresponsible
attachment.aterwthefor

132

aterW

Clusters

of

para2-

ophenylethanol-uor

and

2-

ortho

ophenylethanol-uor

8Chapter

IonizationResonance-EnhancedMassSpectrTwoscopyo-Photonof
Ephedrine:IndicationforaState-Selective
MoleculeFlexibleainFragmentation

R.Karaminkov,S.Chervenkov,P.H¨arter∗,H.J.Neusser,Chem.Phys.
Lett.442,238(2007).
ThevibronicstructureoftheS1←S0spectrumofephedrinewasmeasured
byresonance-enhancedtwo-photonionizationspectroscopywithmassreso-
lutionundercoldmolecularbeamconditions.Thespectrarecordedatfour
differentmasschannels,m/z=165(parent),58,andthehithertounknown
71,85fragmentionsshowdissimilarvibronicfinestructureandtheobserved
masspatternstronglydependsontheselectedintermediatevibrationalstate.
Thispointstoanintermediatestate-selectedprocessresultinginadifferent
fragmentationmasspattern.abinitiocalculationsattheMP2/6-311++G**
leveldemonstratethattheAG(a)conformerismorestableby238cm−1than
thenextstableGG(a)conformer.

∗AnorganischeChemie,TechnischeUniversit¨atM¨unchen,Lichtenbergstrasse4,D-85748Garching,Germany

133

134

oductionIntr8.1

R2PIMassSpectroscopyofEphedrine

Theclassoftheephedramoleculeshasattractedalotofattentionastheyplayanimportant
roleinthecentralandperipheralnervoussystemactivities.[120,121]Thepharmacologically
relevantephedrine(C10H15NO)(EPD)moleculeactsalsoasaneurotransmitter.[58,122]The
detectionandassignmentoftheconcentrationoftheneurotransmitterssuchascatecholamines
(epinephrine,norepinephrineanddopamine),serotonine,histamine,andephedrineinbodyflu-
idsisanimportantmarkerofabnormalitiesintheadrenergicbodyfunctions.EPDisusedas
aprecursorintheproductionofstimulantsinsport,andthedesiretounderstandingofthere-
sultingdopingpropertieshasinitiatedaseriesofmassspectrometricstudiesincludingavari-
etyofexperimentaltechniques[electronimpactionization(EI),chemicalionization(CI),fast
atombombardment(FAB),collisionallyinduceddecompositionmassanalyzedionkineticen-
ergy(CIDMIKE)].[123–125]Allthesemassspectrashowthattheephedrinecationundergoes
anefficientfragmentationwithadominatingfragmentionatmassm/z=58.UVspectraofthe
electronicoriginregionoftheS1←S0transitionofEPD,[66,67]norephedrine,[69]andpseu-
doephedrine[67]inthegasphasedetectingonlymass58uhavebeeninvestigatedbySimonsand
co-workersusingmass-selectedresonance-enhancedtwo-photonionizationspectroscopy(R2PI)
combinedwithUVholeburningandion-dipspectroscopy.Inarecentlypublishedworkfrom
ourgroup,[20]wehavepresentedthefirsthigh-resolutionspectrumofthestrongestoriginband
ofEPDmonomerandwewereabletodetermineaccuratevaluesfortherotationalconstants
oftheground,S0,state,theexcited,S1,state,andforthetransitionmomentratioofthisband.
Here,wepresentnewmass-selectedlow-resolutionR2PIUVspectraofEPDvibronicbands
measuredatfourmasschannels[parent(m/z=165)andfragments(m/z=58,71and85)]upto
10cm−1excessenergyabovetheelectronicorigin.Theydemonstrateaspecificfragmentation
behavioroftheEPDcationthatdependsontheselectedintermediatestateintheresonance-
enhancedtwo-photonionizationprocess.Ahighlystate-selectedfragmentationbehaviorhasnot
beenobservedinREMPIexperimentsoftherigidbenzenemolecule,whichisthearomaticba-
sisofEPD.[126,127]Thisfragmentationbehaviorseemstobeaconsequenceoftheflexibility
ofEPDbecauseofitssidechain.Experimentalresultsaresupportedbyhigh-levelabinitio
computationsofthestructuresandenergiesofthetwolowestenergyconformersofEPD.

8.2Experiment

Thedetailsoftheexperimentalapparatusaredescribedelsewhere.[10,79]Theexcitationscheme
fortronicstateproducingtothetheUVfirstespectraxcited,S1emplo,yselectronicpromotionstateofandthesubsequentmoleculesfromionizationtheirbyground,usingS0,photonselec-
withthesamefrequency(one-colorR2PI)generatedbyfrequencydoublinginaBBOIcrystal

Results8.3

135

oftheoutputpulsesofacommercialbroadband(∼0.4cm−1)dyelaser(LambdaPhysikFL2002)
withaCoumarin307dyeforrecordingspectraintheenergyregionfrom37450cm−1to38600
cm−1.Thelaserpulsesintersectperpendicularlythecoldmolecularbeamwhichwasproduced
byanadiabaticexpansionthroughapulsednozzleofephedrineseededinArbuffergasata
backingpressureof3.2bar.Wecarefullyevacuatedthevacuumchambersothatthepresenceof
waterwasminimized.Weusedasoftfocusedlaserbeamwithadiameterof2mminorderto
avoidsaturationofthetransitions.Todeterminethediameterofthelaserbeamintheinteraction
region,wedeflectedthebeamusingaprismplacedinfrontofthevacuum-chamberentrance
windowandmeasuredthebeamdiameteronamillimeter-grid-graph-paperscreenawayfrom
theprismatadistanceequaltothedistancebetweentheprismandtheinteractionpoint.Inthis
way,therelativediameterfordifferentfocusingconditionswasdetermined.Thenozzlecontain-
ingtheephedrinewasheatedupto110°C.Themolecularionswereacceleratedinadirection
perpendiculartoboththelaserandthemolecularbeambytheionopticsofahomebuiltWiley-
McLaren-typetime-of-flightmassspectrometeranddetectedbymicrochannelplates.Theion
signalinaselectedtime-of-flightrangewasintegratedbyagatedintegrator.Theabsolutefre-
quencywasmeasuredwithanaccuratewavemeter(Atos007,ν/Δν=108).Wealsorecorded
snapshotmassspectraofdifferentvibronicbandswithanoscilloscope(TektronixTDS360).
Thedataacquisitionandthecontroloftheexperimentwereperformedbyahomemadesoftware
operatinginaLABVIEWenvironment.Thepurityofthesample(1R,2S)-(–)-ephedrine99%
(EPD)wasadditionallycheckedbyHPLCtests.

Results8.3

calculationsinitioAb8.3.1

Tosupporttheexperimentalresultsandtofindtheenergeticallymostfavorablegeometriesof
theEPDmonomerintheground,S0,electronicstateabinitiocalculationsonthestructures
andenergeticshavebeenperformedattheMP2/6-311++G**leveloftheoryusingGAUSSIAN
03suiteofprograms.[82]Thenewcalculationswithextendedbasissetcorroboratestructures
AG(a)andGG(a)oftheEPDmonomerwithtwodistinctshapesofthesidechain.[67]The
A/Gnotationreferstothearrangement(antiorgauche)oftheCCCNandOCCNatomchains,
respectively.AGistheextendedandGGisthefoldedconformationalstructure.Theoptimized
geometriesareshowninFig.8.1.ThemoststableconformerhasbeenfoundtobetheAG
(a)one,ourcalculationspredictthattheenergygapbetweenthetwostructuresis0.68kcal/mol
(238cm−1),whichisbiggerthantheonecomputedattheMP2/6-31G*leveloftheory.[67]The
differencemostlikelystemsfromtheimprovedelectroncorrelationattainedwiththelargerbasis
set.TheenergydifferenceiscomparabletokTatroomtemperature,sothattheexistenceofboth

136

R2PIMassSpectroscopyofEphedrine

Figure8.1:Electronicgroundstate,S0,structuresofthetwolowestenergyAG(a)andGG(a)conformers
ofEPDoptimizedattheMP2/6-311++G**leveloftheory.Energieshavebeencalculatedatthesame
level.Relativeenergiesaregiveninkcal/mol.

conformersispossibleatroomtemperature.

8.3.2esultsrExperimental

Theone-colorR2PIspectrumoftheS1←S0electronictransitionspectrumofEPDatm/z=58
massearlierwchannelork[20]isbydepicted100cmin−1Fig.tothe8.2d.redItsideshoandwsanmoreethanxtended900scancm−re1togiontheinhigh-enercomparisongytosideourof
1−theenergiesoriginabobandvetheoftheSelectronic1←S0origintransitionenergy,locatedtwoatother37552.116(1)fragmentmasscm.[20]channelsAt(m/zhigher=71eandxcitationm/z
=vibronic85)arebandsobserved,appearwhichataboutarenot530knocm−wn1andfrom930precmvious−1emassxcessspectra.energyina[123–125]closeTwoagreementstrong
11thewiththefrequencypositionsdiffoferencetheisvibroniconly76cm0−1and.1All0bandsspectraintheSsimultaneously1←S0spectrumrecordedofatdiffbenzeneerent[128]:mass
channelsredshifted(mby/z25=cm165,−185,from71theand58)electronicareoriginpresentedobservinedFig.atmass8.2.Wechannelassignm/z=the58tosmallapeakband
1−theresultingoriginisfrommosttheprobablyfragmentationasequenceofanband.EPD-ArAclustermagnified,whileviewtheoftheband530cmredshifted−1byspectral8cmregionfromof
atthe519R2PIcm−1,spectrum526cmis−1sho,wnandin530Fig.cm−8.31rewithvealingstronglyadiffpronouncederingintensitystructurewhenwithsevmonitorederalatsubbandsthe
differentmasschannels.Thepeakat519cm−1isonlyobservedatm/z=85masschannel.Asa

Results8.3

Figure8.2:(a)One-color(R2PI)spectraofephedrine,measuredattheparent(m/z=165)(b-d)fragments
(m/z=85,m/z=71,m/z=58)masschannels,respectively.Allvibronicpeaksarelabeledwiththeirexcess
energyabovethe000bandrecordedatmasschannelm/z=58.

137

generaltrendthereisaredshiftoftheintensitymaximumfrom530to519cm−1fordetectionat
increasingfragmentmassesm/z=58,71and85.Asimilartrendisobservedfortheweakbands
atdifferentpositionsalloverthespectrum.Toexcludethatsomeoftheobservedpeaksoriginate
fromthefragmentationoftheEPD-H2Oclusterwerecordedsimultaneouslyasignalatm/z=
183masschannel(i.e.theEPD-H2Ocluster).Sincewewereperformingtheexperimentsunder
water-freeconditionstheintensityatthismasschannelwasapproximately200timessmaller
comparedtothesignalatthem/z=58masschannel.Onlyonebroadvibronicfeatureslightly
abovethenoiselevelisfoundinthe+530cm−1region.Itdoesnotcoincidewiththepeak
Fig.positions8.3.Toobserveddeterminewhenwhetherdetectingthethepeakmassesratioinofthethevotherariousspectrafragmentsof(mFig./z=8.385,71dependsandon58)thein
laserintensityweusedtwodifferentlaserbeamfocusingconditionsyieldingintensitiesdiffering
byafactorof16.Nopronouncedchangeintheintensityratioofthepeakheightsmeasuredatthe
threefragmentmasses(m/z=85,71and58)wasobserved.Wehavealsorecordedtheresulting
massspectrawhentuningthelaserfrequencytomostofthebandsshowninFig.18.2.InFig.
8.4,fourselectedsnapshotspectraareshownfortheintermediatevibronicbandsat00,173,205,

138

R2PIMassSpectroscopyofEphedrine

Figurmagnifiede8.3:parts(a)oftheOne-color(R2PI)(R2PI)spectraspectrashoofwninEPD-HFig.2O8.2atclusterthe,respectirecordedveatm/zfragment=183masses.masschannel;(b-d)

and225cm−1excessenergy.Theydemonstrateastronglychangingmasspatterndependingon
intermediateselectedthestate.

Discussion8.4

8.4.1Benzene-likeS1←S0spectrum
Weexcitethebenzeneringchromophorewhichappearstobenotstronglyperturbedbythepres-
enceofthesidechainbecausethemeasuredspectruminFig.8.2disclosetothatofbenzene,
featuringseveralcharacteristicvibronicbandsofthearomaticring.TheHerzberg-Tellerinduced
vibrationν6isthefalseelectronicorigininbenzene(shiftedby522.4cm−1fromtheforbidden
electronicorigin),whichisclosetothestrongbandat530cm−1excessenergyintheEPDspec-
trumofFig.8.2d.Inbenzene,thetotallysymmetricCCstretchν1vibrationleadstoastrong
progressionat923cm−1,whichagreeswiththeintensebandat930cm−1inEPD.Thereare,

Discussion8.4

Figure8.4:Massspectrarecordedatvibronicbands000,+173cm−1,+205cm−1and+225cm−1.

139

however,twodifferences:(i)theelectronicoriginwhichissymmetryforbiddeninbenzeneis
appearingintheEPDspectrumbecausetheC6symmetryisbrokenbythesidechain;and(ii)the
spectrumisshiftedby534cm−1totheredfromthatinbenzene(38086cm−1).[129]Ashiftof
similarsizeisobservedforseveraln-alkylbenzenes:ethyl-(501cm−1),propyl-(504cm−1),and
butylbenzene(511cm−1),[130]andfortoluene(609cm−1).[131]Thisisatypicalbehaviorfor
asidechainwithoutadoublebond(nomesomericeffect)whennoinductiveeffectfromanelec-
tronegativesubstituentispresent.Whilethedeviationoftheelectronicoriginis534cm−1,the
twoprominentvibronicbandsat530and930cm−1excessenergydifferonlyby7cm−1fromthe
excessenergiesofthecorrespondingonesinbenzenethoughthesidechainhasalargemassand
atom.nitrogenacontains

pathwaysFragmentation8.4.2

58InourwhichR2PIdoesenotxperimentsoriginatethefromstrongestaionfragmentationsignalisoftheobservedbenzeneattheringasfragmentisknomasswnfromchannelprem/viousz=

140

R2PIMassSpectroscopyofEphedrine

R2PIexperimentsonbenzene.[126,127]Asdiscussedinourrecentwork,[20]thefragmentation
ofephedrinecouldnotbeavoidedinaR2PIexperiment.Inone-colorexperimentstheenergy
oftheionizingphotonisfixed,thisleadingtoatwo-photonexcitationenergyofmorethan9.3
eV,andmostlikelytohighexcessenergyabovetheadiabaticionizationenergy.Fromthat
weconcludethatthefragmentionspreferentiallyresultfromafragmentationoftheionicEPD
ratherthanfromdissociationintheneutralintermediatestate.Thefragmentatmassm/z=58is
alsothedominatingfragmentinEI,CI,FABandCIDMIKEmassspectraofEPD,[123–125]
supportingtheaboveargumentofanionicfragmentation.However,wecannotexcludefast
dissociationofEPDafterexcitation(e.g.theejectionofanH2)ofthevibronicbandsinS1during
thelaserpulseleadingtoalargeneutralfragmentthatcanbeionizedbyasecondphoton.The
observedlinewidthinourrecenthigh-resolutionexperiments[20]wouldallowfordissociation
timesslowerthan3nsandthustheabsorptionofafurtherphotonduringthelaserpulsewouldbe
possible.Thiscouldfinallyleadtoafragmentationoftheresultingionintotheobservedfragment
ions.ThedifferingionizationpathwaysinR2PIandEIrequireacloserdiscussionofthecommon
fragmentionatmassm/z=58(seeFig.8.5a).TheresonantintermediateelectronicstateofR2PI
isaππ∗stateofthebenzenechromophoreandthustheabsorptionofthesecondphotonleads
totheionizationofaπ-electronofthebenzenering.ForthisreasonimmediatelyafterR2PI,
thepositivechargeislocatedatthebenzeneringandchargetransferhastotakeplacebeforethe
fragmentationintofragmentionwithmassm/z=58.Thisleadsustotheconclusionthatan
electrontransferoccursfromthenitrogenatomtothebenzenering.Thenthedissociationintoa
neutralfragmentofmassm/z=107andachargedsidechainfragmentofmassm/z=58occurs
afterabreakingofthesinglecarbonbond.Asimilarchargetransferprocesshasbeendescribed
fortheclustersofphenethylamineandamphetaminewithpolar,nonpolar,andhydrogen-bonding
solventsstudiedbymass-resolvedexcitationspectroscopy.[100]Twopossiblefragmentation
pathwaysoftheEPDparentionyieldingthefragmentionwithmassm/z=71originatingfrom
thetwomoststableconformericstructurespredictedfromtheabinitiocalculationsaredrawnin
Fig.8.5bandc.InthecaseofthemorestableconformerAG(a),thereisanelectronmigration
fromthesubstituentside(partc),whereasinthecaseofthesecondlowestenergyconformer
GG(a)theelectrontransferfromthenitrogenismorelikelydirectbecauseofthefavorable
positionoftheelectronegativeelementclosetothebenzenering.InFig.8.5daplausibleionic
fragmentationpathwayyieldingthefragmentatmassm/z=85isshown.Herewehaveatypical
forthebenzylalcoholsfragmentationwiththedetachmentofH2andformationofatriplebond
betweenthecarbonandtheoxygen.ThisfragmentationpathwayisonlypossiblefortheGG(a)
.conformer

Discussion8.4

Figure8.5:FragmentationpathwaysoftheEPDcationafterresonance-enhancedtwo-photonexcitation:
(a)m/zresultingfromAG(a)conformer;(b)m/z=71resultingfromAG(a)conformer;(c)m/z=71
resultingfromGG(a)conformerand(d)m/z=85resultingfromGG(a)conformer.

fragmentationevState-selecti8.4.3

141

Asdiscussedabove,wehavetoconsidermorethanoneconformericstructureofEPDtoexplain
thefragmentationleadingtothem/z=85fragmention.Mostlikelymorethanoneconformeric
structureispopulatedintheheatednozzlepriortothefreeexpansionofthemoleculesintothe
vacuumchamber.However,duringtheexpansionunderourjetconditions,arelaxationfromthe
GG(a)tothemoststableAG(a)structuretakesplace,owingtothetheoreticallypredicteden-
ergygapof238cm−1betweenthemassumingthatthepotentialenergybarrierseparatingthetwo
conformersisnottoohigh.Thisisinlinewithourrecenthigh-resolutionexperimentsbyanaly-
sisofthepartlyresolvedrotationalbandstructuresofthestrongestband.[20]Wecannotexclude
theexistenceofseveralconformersinthemolecularbeambuttheanalysisofthevibronicbands

142

R2PIMassSpectroscopyofEphedrine

observedatthefragmentmasschannelsdoesnotsupporttheirpresence.Thesmalldistancesof
thevibronicfeaturesat-4cm−1and-11cm−1observedatfragmentmassesm/z=71andm/z=
85(butnotatm/z=58)leadustotheconclusionthatthesearehotbands,ratherthanelectronic
originsofotherconformericstructures.Inaddition,theirintensityismuchsmallerthanthein-
tensityofthe000bandatfragmentmasswithm/z=58whichistheelectronicoriginoftheAG(a)
conformer.[20]Indeed,thebandat+46cm−1whichhasbeenassignedtoadifferentconformer
origininRef.[67]doesnotshowanyfragmentationtothem/z=71,85masschannels(seeFig.
8.2).Anotherargumentthatthevibronicstructureatm/z=71andm/z=85masschannelsis
notproducedfromthefragmentationofdifferentconformersinthebeamisthatthefragment
intensityratioofthepeaksinthe000cm−1regioncomparedtotheratioofthecorresponding
peaksinthe+530cm−1rangeisnotpreserved.Thiswouldbeexpectedforthecloselyingorigin
andthe101progressionbandsofdifferentconformers.Thustwodifferentprocessesleadingto
theobservedvaryingfragmentationbehaviorofEPDareplausible:(i)anexcessenergyspecific
fragmentationintheionaftertheabsorptionoftwoorthreephotons.Scanningtohigherenergies
couldleadtotheopeningofnewfragmentationchannels.Intheionizationstep,accordingtothe
FranckCondonfactors,thephotonenergyisdistributedbetweenkineticenergyoftheelectron
andinternalenergyoftheion,producingionsindifferentenergystates,ratherthanleadingtoa
sharpinternalenergy.Forabroadenergydistribution,asuddenchangeofthefragmentationbe-
haviorwithinafewcm−1isnotexpectedforastatisticalunimoleculardissociationprocess.[127]
(ii)Themostlikelyexplanationisastate-selectiveisomerizationprocessintheintermediatestate
oftheneutralvibronicallyexcitedEPD.Thiswewouldliketodiscussforthebandsat530,526,
and519cm−1(seeFig.8.3b-d),eachleadingtoadifferentfragmentationpattern.Weassume
thattheredshiftedbandsat+519cm−1(m/z=85)and526cm−1arehotsequencebandsofthe
AG(a)conformerleadingtodifferentvibronicstatesofincreasingenergyintheS1electronic
state.Thiscouldleadtoastate-selectedisomerizationordissociationprocessinS1andtothe
ionizationofanotherconformer[e.g.GG(a)]oradissociationproductwithnewfragmentation
pathways.Anotherevenmorepronouncedstate-selectedfragmentationisobservedinthelow
excessenergyregionbetween100and400cm−1.Herespecialvibronicbandscanbeobserved
onlywhenmonitoringdifferentmassesandthespectrameasuredatspecificmassesarecom-
pletelydissimilarinthisfrequencyrange.Thisisalsoseenfromthemassspectrawhichare
obtainedafterexcitationofthevibronicstatesat000,173,205,and225cm−1excessenergyas
showninFig.8.4.Whilethemassspectrameasuredviathe000andthe225cm−1bandsarethe
oneswithadominatingm/z=58masspeak,themassspectrameasuredvia173cm−1and205
cm−1bandsentirelydifferfromthemandfromeachother.Theidentityofthevibronicstatesis
notyetclearbutfutureabinitiocalculationsmayrevealthenatureofthesevibronicbands.It
isparticularlyinterestingifvibrationswithamotionparalleltothebenzeneplanearetheones
leadingtothefragmentmassesm/z=71andm/z=85.Theycanactasinducingmodesofan
isomerizationprocessinS1electronicstate.

conclusionsandSummary8.5

conclusionsandSummary8.5

143

WehaveshownthatthemainfeaturesintheR2PIspectrumofephedrinearesimilartothe
onesinthespectrumofbenzene.Theytestifythattheelectronicexcitationtakesplaceinthe
benzenechromophore.However,finestructureofthemainbandsandweakeradditionalbands
originatesfromtheflexiblesidechainofEPD.Whiletheexcitationofthestrongestbenzene-
likebandsintheR2PIspectrumyieldsthefragmentwithmassm/z=58knownfromEIand
CImassspectrometry,excitationoftheadditionalweakerEPDspecificbandsresultsinnew
fragmentmasschannelsintheR2PIspectrum.Fromthisresult,wemayconcludethatastate-
selectiveprimaryprocesse.g.isomerizationordissociationtakesplaceafterexcitationofthe
intermediatestateandpriortothesuccessiveionizationwithspecificfragmentationchannelsfor
thedifferentionizedconformers.Toelucidatethenatureoftheintermediatestateandtheprimary
processhigh-levelcalculationsoftheexcited,S1,statepotentialenergysurfaceandthevibronic
frequenciesofthelowestenergystructuresareinprogress.Inconclusion,wehavedemonstrated
thatflexiblemoleculesshowanintermediatestate-selectivefragmentationpatternafterR2PIvia
vibronicstatestypicalforamotionofthesidechain.

144

R2PI

Mass

oscopySpectr

of

Ephedrine

9Chapter

MassSelectiveHigh-ResolutionUV

Pseudoephedrine:ofoscopySpectr

EvidenceforConformerSpecific

Fragmentation

Usingresonance-enhancedtwo-photonionizationspectroscopywithmass
resolution,alow-resolutionS1←S0vibronicspectrumofpseudoephedrine
wasrecordedatdifferentm/zchannels.Thevibronicspectrumhasaspecific
fragmentationalpatternscatteredoverthreefragmentmasschannels.Two
ofthefragmentswithm/z=71andm/z=85areobservedforthefirsttime
forthismolecule.Highlyresolvedspectraofallprominentvibronicfeatures
weremeasuredandafteracomparisonwiththetheoreticalresultsfromhigh
levelabinitiocalculations,assignedtotwodistinctmolecularconformations
AG(a)andGG(b).Thedifferentconformersshowdistinctandspecificfrag-
mentationalpathways,whichstronglydependonthestartinggeometry.

145

146

Pseudoephedrine:EvidenceforConformerSpecificFragmentation

oductionIntr9.1

(1S,2S)-(+)-Pseudoephedrine(pseudoEPD)isthediastereoisomerofthe(1R,2S)-(-)-Ephedrine.
Thetwomoleculesdifferbythechiralityofonecenter.Thisnonhumanneurotransmittersshare
anumberofstructuralcharacteristicswiththenaturalneurotransmitter[58,122]adrenalineand
havesimilarpharmacologicaleffectsonthecardiovascularsystem.Forstudyingbiologicalre-
actionmechanisms,itisimportanttoknowtheorientationofdifferentreactivechemicalgroups,
sincethetopologyofthemoleculeplayskeyroleforthemetabolism.Theinteractionbetween
suchmoleculesandanenzymebindingsitedependontheshapeofthemoleculewhichisdeter-
minedbyasubtleinterplaybetweenbondedandnon-bondedforces.
Inthelastyears,anumberofstudieshavebeenperformedtodeducetheconformationalprefer-
encesofsmallbiologicalflexiblemolecules[3,4,66,67,69,85,108,132](seeChapters3,4and
6).Inalltheseworks,ithasbeenshownthatthecombinationofdifferentUVlaserspectroscopy
techniquesinthegasphasewithtime-of-flightmassspectrometryisapowerfulmethodforprob-
ingtheconformationallandscape.Theadditionofhigh-levelquantumchemicalcalculationsin
theconformationalinvestigationsallowstheassignmentofdiscreteconformationalstructuresto
experimentalfeatureswithahighdegreeofconfidence.

9.2ExperimentandDataProcessing

AdetaileddescriptionoftheexperimentaltechniqueisgiveninChapter2.Theexcitation
schemeforproducingthehigh-resolutionUVspectraemployspromotionofthemoleculesfrom
theirground,S0,electronicstatetothefirstexcited,S0,electronicstatebyscanninganarrowband
pulsedlaser(Δυ∼0.003cm−1)andsubsequentionizationbyabsorptionofphotonswithfixed
frequencyfromabroadband(Δυ∼0.4cm−1)dyelaser.Theexcitationphotonsinvolvedinthe
two-colorhigh-resolutionspectraweregeneratedbyfrequencydoublinginaKDPcrystalofthe
pulsedamplifiedoutputofacontinuouswave(CW)single-moderingdyelaser(Coherent,CR
699-21)operatedwithCoumarin334dye.Thesubstance((1S,2S)-(+)-Pseudoephedrine,98%)
obtainedfromAldrichwasplacedinahome-buildheatablenozzleandheatedupto1100C.The
so-producedvaporsofthemoleculeweremixedwithbuffergasargonatastagnationpressureof
2bars.Thegasmixturewasexpandedintovacuumthroughapulsed-operatedheatablenozzle
withorificediameterof300µm,leadingtorotationaltemperaturesforPseudoEPDofaround
7-9K.Fortheone-colorlow-resolutionspectraphotonswiththesamefrequencywereused,
generatedbyfrequencydoublinginaBBOI(BetaBariumBorate)crystaloftheoutputpulses
ofacommercialbroadbanddyelaser(LambdaPhysik,FL2002)withaCoumarin153dye.The
congestedrotationalstructureofvibronicbandsoflargefloppymolecules,precludesadirect
singlerotationallineassignmentforthespectralresolutionofthescanningUVlaserpulses(70

Results9.3

147

MHz(FWHM)).Forthisreasonweuseacomputer-basedmethodforfittingoftheexperimental
highlyresolvedspectra.Thealgorithmandthecomputerprogramelaboratedinourgrouphave
alreadybeendescribedinSection2.3.Crosscorrelationisemployedasapenaltyfunction.
FortheproductionofthesimulatedspectraofpseudoEPD,werantheprogramusinga300
generationswith500individualsinageneration.

Results9.3

CalculationsinitioAb9.3.1

TheGaussian03programpackage[105]wasusedfortheoptimizationoffourpreviouslyintro-
statesducedat[67]theMP2/conformationsaug-cc-pVDZofandpseudoEPDCIS/incc-pVDZtheleground,velsSof0,andtheory,therespectifirstevelyxcited,.TheS1,notationelectronicof
thereferstomonomerthesarrangementconformers(isantitakorengaucfromhe)theofthepublicationCCCNofandButzOCCNetal.atom[67].chains,TheA/Grespectivnotationely.
AGistheextendedandGGisthefoldedconformationalstructure.Theoptimizedgeometriesare
inshoawnindecreaseFig.of9.1.energyThegapsimprovbetweenedtheelectronconformerscorrelationinattainedcomparisonwithtothethelarabgerinitiobasissetcalculationsresults
∗∗atthevibrationalMP2/ener6-311Ggieslev(ZPVE)elofistheoryunchanged:[67].TheAG(a)relati0vecmener−1,gyGG(b)order62withcman−1,includedGG(a)161zero-pointcm−1,
1−AaddsG(b)about30560cmcm.−1ThetotheinclusionofuncorrectedtheZPVEenergyvcorrectionalues.isHonotwevaerff,itectingmaytheresultenergyinaorder,itreorderingonly
whenthetheoreticallevelofcalculationorthemodelischanged.Thecalculatedvibrational
minimafrequenciesontheofthepotentialfourenerconformersgysurfareace.allAllpositirelevveantwhichvtheoreticalerifiesthatparameterstheyofcorrespondthesetostructuresstable
aresummarizedinTable9.2.

Spectraw-ResolutionLo9.3.2

Low-resolutionone-colorresonanttwo-photonionization(R2PI)spectrainthespectralregion
between37350cm−1and37700cm−1havebeenrecordedatthreedifferentmasschannels
(m/z=58,m/z=71andm/z=85)underwater-freeconditionsneartheS1←S0electronictransition
ofpseudoEPD.ThespectraaredepictedinFigure9.2.Thetwoadditionalmassfragments
m/z=71andm/z=85wereobservedforthefirsttimeinthecaseofephedrine,asdiscussed
alreadyinthepreviouschapter.Itisinterestingtoseethatthevibrationalsignalsoffragments
withm/z=71andm/z=85aremuchmoreintenseandpronouncedinthelow-frequencyregion

148

Pseudoephedrine:EvidenceforConformerSpecificFragmentation

Figure9.1:Electronicgroundstate,S0,structuresofthefourlowestenergyconformersofpseudoEPD
optimizedattheMP2/aug-cc-pVDZleveloftheory.Energieshavebeencalculatedatthesamelevel.
Relativeenergiesaregivenincm−1.

neartotheoriginbandofpseudoEPDatm/z=58incomparisontothoseofephedrine(seeChapter
8).ThemostintensevibronicfeaturesofpseudoEPDarelabeled000cm−1,126cm−1,135cm−1,
181cm−1and212cm−1,respectively.Thepositionsofthebandsaregivenrelativetothelowest
energybandat37417.939(1)cm−1measuredatthemainfragmentmasschannelm/z=58.To
elucidatethenatureoftheabovevibronicbandsandtoinvestigatethespecificfragmentational
patternofpseudoEPDweperformedhigh-resolutionscans.

SpectraHigh-Resolution9.3.3

Thehigh-resolutionUVspectrumofthelowestenergybandmeasuredatthemostabundant
fragmentmasschannelm/z=58ispresentedinFigure9.3(uppertrace).Wehaveappliedour
computer-assistedfitprocedure(describedindetailsinSection2.3)todeterminetheexperimen-

Results9.3

149

BandarameterP000cm−1+126cm−1+135cm−1+181cm−1+212cm−1
A´´0.0665(05)0.0541(27)0.0655(07)0.0675(61)0.0548(01)
CB´´´´0.0158(61)0.0173(98)0.0192(94)0.0199(79)0.0156(34)0.0173(43)0.0157(31)0.0170(37)0.0189(11)0.0195(55)
A´0.0655(39)0.0532(54)0.0637(11)0.0656(72)0.0532(34)
CB´´0.0160(53)0.0173(43)0.0190(39)0.0201(42)0.0161(98)0.0167(28)0.0155(88)0.0169(72)0.0188(57)0.0192(96)
µa2:µb2:µc268:14:1823:60:1726:63:1117:67:1642:47:11
ν0,cm−137417.939(1)37545.410(5)37553.639(2)37597.612(3)37631.392(1)
T(K)8.5(2)8.1(4)9.2(3)7.2(3)11.9(4)
Best-fitcrosscorrelation(%)9495949794

Table9.1:Experimentalrotationalconstantsfortheground,S0(A´´,B´´,C´´),andforthefirstexcited,S1
(A´,B´,C´),electronicstates,thetransitionmomentratio,µa2:µb2:µc2,thebandoriginfrequency,ν0,
therotationaltemperature,T,andthebest-fitcrosscorrelationobtainedfromtherotationalfitofbands
000cm−1,126cm−1,135cm−1,181cm−1and212cm−1,showninFigures9.3-9.4-9.5(fordetails,seetext).
Thenumbersinparenthesesrepresentonestandarddeviationinunitsoftheleastsignificantquoteddigit.

talrotationalconstants,transition-momentratio,bandoriginandrotationaltemperature.The
programproducesasyntheticspectrum,whichiscomparedwiththeexperimentaloneonthe
basisofcrosscorrelationandthevaluesoftheaboveparametersaredetermined.Theseartificial
spectraareshownasinvertedtracesinFigures9.3-9.5.Inordertoreducethesearchingspace
fortheexperimentalrotationalconstantswehavestartedthefittingprocedurewiththeoretical
valuesreceivedfromMP2/cc-pVDZlevelofabinitiocalculations.Thebestagreementforthe
experimentalhighly-resolvedspectrumoftheband000isobtainedwhenthetheoreticalrotational
constantsfortheground,S0,electronicstateoftheAG(a)conformerwastakenasstartingpa-
rameter.Thesimulatedspectrumagreewellbothinpeakpositionsandinintensitywiththe
experimentalone,whenconvolutingthetheoreticalstickspectrumwithaGaussianlinewidthof
200MHzFWHM,withanachievedcrosscorrelationof94%.Thefitshowsthattherotational
spectrum(seeFigure9.3)consistsofweak,butwell-resolvedP-andR-branchesinthewings
andaprominentQ-branchinthecenter.Itwasassignedasa-,b-,andc-typehybridwithasig-
nificanta-typecontribution(seeTable9.1).Thestructureofthehighlyresolvedspectraofbands
135cm−1(scannedatfragmentmasschannelm/z=71)and181cm−1(scannedatfragmentmass
channelm/z=58)differfromthatoftheabovedescribedband(seeFigure9.4).Thelattertwo
spectrafeaturenicelystructuredandpronouncedP-andR-brancheswithaminorcontribution
fromtheQ-branch.Fromtheirappearance,thespectracanbeassignedaspredominantlyb-type
withsmallcontributionsalsofroma-andc-types(seeTable9.1).Toreceivetherotationalcon-
stantsfortheground,S0,andthefirstexcited,S1,electronicstates,thetransition-momentratios,
thebandoriginsandtherotationaltemperatures,theabove-mentionedcomputer-aidedfitrou-

150

Pseudoephedrine:EvidenceforConformerSpecificFragmentation

Figure9.2:(a)One-color(R2PI)spectraofpseudoephedrine,measuredatthefragmentmasschannels:
m/z=85,m/z=71,andm/z=58,respectively.Allvibronicpeaksarelabeledwiththeirexcessenergy
abovethe000bandrecordedatmasschannelm/z=58.

tinewasemployed.Inseparatefitprocedurethetheoreticalrotationalconstantsfortheground,
S0,electronicstateofallfourtheoreticallypredictedconformersofpseudoEPDwereusedas
startingvaluesforthefits.Theresultingbest-fitsimulationsareshowninFigure9.4(lower
invertedtraces)withanachievedcrosscorrelationof94%and97%,respectively.Themolecular
parametersbringingforththesefitsaresummarizedinTable9.1.Thefitswereabletoreproduce
verywellboththepeakpositionsandthepeakintensities,inparticulartheagreementforband
181cm−1isexcellentevenforthesmallpeaksinthewingsoftheP-andR-branches.Here,
againasinthecaseoftheband000,thebestfitsforbothbandswereproducedwhenstarting
theprocedurewiththetheoreticalrotationalconstantsfortheground,S0,electronicstateofthe
AG(a)conformer.Thesameapproachwasemployedforbands126cm−1(scannedatfragment
masschannelm/z=85)and212cm−1(scannedatfragmentmasschannelm/z=58),respectively.
Startingthefittingprocedureforbothexperimentalspectrawiththetheoreticalrotationalcon-

Results9.3

∗∗631-theand,0631-G/
/CISbbbbbbbaug-cc-pVDZ/
SandG(b)11:94:5Sground,theforCCISandcc-pVDZ/
cc-pVDZAConformer2.212.05110.2122.258.7748.440.068220.067680.017510.017690.015950.01624and
/2:bofstructuresconformationalstablemostfourtheof
0-SBbbbbbbbCISaug-cc-pVDZ,
,AGG(a)S2CISfromobtainedc
/1µS054:11:35-2a
Conformer2.212.05115.5121.7-55.91-54.980.055250.053910.018290.018990.017730.01824constants,rotationaland,
elrµE:µτS070:10:20--241-96-9-0305-161-62-0-
angles,aaaaaa
a1GG(b)SdihedralS0-48.96-5348:8:44-
100.2120.3(OHN)50.451.57(OCCN)cmcmcm,ZPVE),(incl.9.2:ableTxcited,efirst∗∗Ggies,enerevRelati.monomerpseudoEPDtheofgeometriesoptimizedtheofcalculationMP2theatcalculatedbeenevha(ZPVE)gyenervibrationalzero-pointtheofinclusionwithandwithoutmonomerpseudoEPDthe.theoryofelvleaCISatcalculatedaluesVbCISatcalculatedaluesV
Conformer2.142.12118.4119.90.054160.054370.019560.020630.019100.01990
,α2:bpseudoEPDtheofconformersfourofstateelectronic,MP2theatcalculatedmonomer1
aaaaaaaangles,planar,d
G(a)1∙∙∙1−1−1−−ratio(TM)momenttransitiondipoleelectricThe.elyvrespecti,theoryofelsvle
SA0.067280.065850.016960.016990.015720.01596theoryofelvlecc-pVDZ/theoryofelvle631-G**/
2:adistances,interatomicTheoretical
2.382.09ÅN)µ1
1−c2cmµConformerarameterPdα1τ1,A,B,Cratio,TMEelrEelr
Sµcm(OH

151

152

Pseudoephedrine:EvidenceforConformerSpecificFragmentation

Figure9.3:High-resolutiontwo-colorUVR2PIspectrumofthevibronicband000inFigure9.2,recorded
atmasschannelwithm/z=58.ThebandisassignedasanoriginoftheAG(a)conformerofthepseu-
doEDPmonomerwithitsrotationlesstransitionscenteredat37417.939(1)cm−1.Uppertrace:experimen-
talspectrum.Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable9.1
xt).teseedetails,(for

stantsfortheground,S0,electronicstateofthefourpredictedstructuresattheMP2/cc-pVDZ
levelofquantumchemicalcalculations,thebestcrosscorrelation(95%and94%,respectively)
betweenexperimentalandsyntheticspectrawasachievedforconformationGG(b).Thespectra
aredifferent(seeFigure9.5):forband126cm−1oneobservesaresolvedP-branchwithsev-
eralwellpronouncedpeaks,smallQ-branchandanintenseR-branchwithnicelydefinedpeaks
whileforband212cm−1therotationalstructureconsistsofdenselyspacedrotationalpeaksinthe
P-branch,astrongQ-branchandapoorlyresolvedR-branchwithafloatingintensityoftherota-
tionalpeaks.Despitethedifferenceinrotationalstructure,thefittedrotationalconstantsforboth
peaksareverysimilar;asmalldifferenceoriginatesfromthemismatchingtransition-moment
ratios(seeTable9.1).Inthecaseofband126cm−1theb-typedominatesoverthea-andc-type
parts,whileinband212cm−1thecontributionsfroma-andb-typearealmostequalwithsmall
c-typeparticipation.Asaresult,wemayconcludethatthetwobandsrepresentdifferentvibronic
bandsofthesameconformer.

9.3

Results

Figure9.4:High-resolutiontwo-colorUVR2PIspectrumofvibronicbands+135cm−1and+181cm−1
inFigure9.2,recordedatmasschannelswithm/z=71andm/z=58,respectively.Thebandsareassigned
asaprogressionoftheAG(a)conformerofthepseudoEDPmonomerwiththeirrotationlesstransitions
centeredat37553.639(2)cm−1and37597.612(3)cm−1,respectively.Uppertrace:experimentalspectrum.
Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable9.1(fordetails,see
xt).te

153

154

Pseudoephedrine:

Evidence

orf

ormerConf

Specific

Fragmentation

Figure9.5:High-resolutiontwo-colorUVR2PIspectrumofvibronicbands+126cm−1and+212in
Figure9.2,recordedatmasschannelswithm/z=85andm/z=58,respectively.Thebandsareassigned
asoriginatingformtheGG(b)conformerofthepseudoEDPmonomerwiththeirrotationlesstransitions
centeredat37545.410(5)cm−1and37588.040(1)cm−1,respectively.Uppertrace:experimentalspectrum.
Lowerinvertedtrace:thebest-fitsimulatedspectrumyieldingtheparametersinTable9.1(fordetails,see
xt).te

Discussion9.4

Discussion9.4

155

Themoststablecalculatedconformer,withoutandwiththeinclusionoftheZPVE,corresponds
totheextendedAG(a)geometrywithitsHatomfromtheOHgrouppointingtotheneighboring
(methyl)aminogroup.Thestructureisassumedtobestabilizedbyanintramolecularhydrogen
bond,asisthecaseintheEPDmonomer.[67]Thenext-in-energyconformerlies62cm−1above
themoststablestructureandbelongstothefoldedGG(b)conformer,whichdiffersfromcon-
formerAG(a)bytheformationofanadditionalπhydrogenbondbetweentheN-methylgroup
andthearomaticring.Thethird-in-stabilityconformationcorrespondtotheGG(a)geometry
withitsNHgrouppointingtotheπelectronsofthebenzeneringwhichishigherby161cm−1
fromthelowest-energyconformation.Thestabilityofthisstructureisdistortedbythecom-
petitionbetweentheformationofanintramolecularNH∙∙∙πhydrogenbondandthepossible
existenceofahydrogenbondbetweentheOHgroupandtheN-methylgroup.Thehighest-
energyconformationat+305cm−1istheAG(b)conformerwitharotatedterminalmethylgroup
at900incomparisontoAG(a).
Weassignthelowestenergypeakatfragmentmassm/z=58tothe000originbandofthemost
stableconformerAG(a).Thesameconformergivesrisealsotobands135atfragmentmass
m/z=71and181atthemainfragmentmassm/z=58.Theassignmenthasbeenmadeonthebasis
oftheverygoodagreementofthetheoreticallypredictedrotationalconstantsforthisconformer
(seeTable9.2)withtheonesobtainedfromthefitofthehighlyresolvedspectrabothforthe
ground,S0,andthefirstexcited,S1,electronicstates(seeTable9.1).Itisinterestingthatde-
spitethesimilarexperimentalrotationalconstantsforthesethreebands,thetransition-moment
ratioforthe000band(68:14:18)(whichisinaverygoodagreementwiththetheoreticalone
(70:10:20)),differssignificantlyfromtheexperimentaltransition-momentratiosoftheothertwo
bands135cm−1(26:63:11)and181cm−1(17:67:17),respectively.Atrustworthyexplanationfor
thechangeofthetransition-momentratiocanbededucedinacloserinspectionofthetheoretical
resultsforconformerAG(a).Itisvisiblefromtheabinitiooptimizationsinthefirstexcitedstate
(seeTable9.2)thatasignificantstructuralchangeoccursuponelectronicexcitation.Thedis-
tanced(OH∙∙∙N)increasesfrom2.09to2.38Åandtheplanarangleα1(OHN)decreasesfrom
120.3oto100.2o,thistrendcanbeeasilyinducedbylow-frequencyvibrationalmodes,which
includetorsionofthesidechain,particularlyadisplacementoftheN-methylgroup.These
changesresultinaredistributionofthemolecularmassandinterchangeoftheinertialmoments,
henceandtoachangeofthetransition-momentratiofortheabovelow-frequencymodes.Two
suchvibrationalmodesfortheAG(a)conformerarepredictedfromthetheoreticalcalculations
inthefirstexcited,S1,electronicstate:theβ1-bendingoftheN-methylgroupvs.ringat71
cm−1andtheβ2-bendingoftheN-methylgroupvs.ringat112cm−1.Bothmodesincludea
displacementoftheN-methylgroupandcaneasilyleadtointerchangeoftheinertialmoments
(seeTable9.2).Theyareinagoodagreementwiththeexperimentalvibrationalfrequenciesat

156

Pseudoephedrine:EvidenceforConformerSpecificFragmentation

135cm−1(correspondingtoβ2)and181cm−1(correspondingtothecombinationmodeβ1+β2).
Asimilarsituationisobservedandforthe−1experimentalhighlyresolvedspectraoftheothertwo
bandsunderconsideration:band126cmrecordedatfragmentmasschannelm/z=85andband
212cm−1measuredatthemainfragmentchannelwithm/z=58.Theassignmentthatthesebands
originatefromconformerGG(b)isbasedagaininthegoodagreementbetweentheexperimen-
talandtheoreticallypredictedvaluesoftherotationalconstantsfortheground,S0,andthefirst
excited,S1,electronicstates(seeTables9.1and9.2).Thedeviationsintheexperimentalval-
uesforthetransition-momentratiofromthetheoreticalonesandbetweeneachofthebandscan
beattributedagaintostructuralchangesuponexcitationduetolow-frequencymodesinvolving
movementoftheN-methylgroup.Thetheoreticalcalculationsforthefirst−1excited,S1,electronic
stateofconformerGG(b)predictabendingvibrationalmodeat89cm(β3)whichincludes
movementoftheN-methylgroupvs.benzenering.−1Thismodeis−1inavery−1goodagreementwith
thedistancebetweenthevibronicbandsat126cmand212cm(86cm).Unlikeephedrine
(seeChapter8),wherethepronouncedfragmentationbehaviorwasobservedatexcessenergies
largerthan500cm−1abovethe000vibrationandwasexplainedasvibrationallyspecific,herein
thecaseofthediastereoisomerpseudoephedrine,forthefirsttime,evidencesforaconformer
specificfragmentationarepresented.Asalreadydiscussedintheprevioussection,therelatively
highintensityofthevibronicsignalsatfragmentmasseswithm/z=71andm/z=85,allowedthe
measurementofthesebandsunderhigherresolution.Inacomparisonwiththetheoretically
predictedmolecularparametersoffourenergeticallylowlyingconformationalstructures,anas-
signmentoftheexperimentalspectrawasmadeasoriginatingfromtwodistinctconformers:
GG(b)andAG(a).Asitwasshowninthepreviouschapterforephedrine(seeFigure8.5)that
afterelectronicexcitation,theAGconformationscouldbeobservedbothatmasschannelswith
m/z=58(themainone)andm/z=71whileinadditiontothemostabundantfragmentchannel
withm/z=58forconformationsGGthereexistsafragmentmasschannelwithm/z=85.This
clearlydemonstratesaconformerspecificfragmentationbecausemasschannelwithm/z=85is
onlyobservedfortheGG(b)conformer.

ConclusionsandSummary9.5

Theflexiblebiologicalmoleculepseudoephedrine,adiastereoisomeroftheneurotransmitter
ephedrine,hasbeeninvestigatedbycombinationoflow-andhigh-resolutionR2PItwo-color
mass-selectivelaserspectroscopyandhigh-levelabinitiocalculationsofitsstructureandener-
getics.Inthiswork,allprominentvibronicbandsinthelow-resolutionspectrumofthepseu-
doEPDmonomerupto300cm−1havebeenmeasuredunderhighresolutionof70MHzatthe
observedfragmentmasschannelswithm/z=58,m/z=71andm/z=85,respectively.Comparing
thevaluesoftherotationalconstantsobtainedfromthefitsofthehighlyresolvedspectrawith

ConclusionsandSummary9.5

157

thetheoreticallypredictedonesforthefourlowest-energyconformations,wewereabletoassign
allofthesevibronicbandstotwoconformationalgeometries:thelowest-energystructureAG(a)
andthenext-in-energystructureGG(b).Differentfromthesituationinephedrinethemasschan-
nelwithm/z=85ismoreefficientwithstrongsignals.Therotationalanalysisofthevibronic
bandsmeasuredatdifferentmassesclearlydemonstratesthatm/z=85channelisonlypresentfor
theconformerGG(b).Weobserveonlytwooutoffourpredictedstructuralgeometries,which
Tisoinestimateindicationthethatpotentialarelaxationbarriersofandthethehigherrelaxationenergypathwconformersayshigh-letovtheelstablecalculationsonesofmaytheoccurpo-.
tentialenergysurfaceareinprogress.Theobserveddeviationsofthetransition-momentratios
intherotationallyresolvedspectraofthevibronicbandsoriginatingfromthetwoconformers
areattributedtogeometricalchangesduringtheelectronicexcitationacceleratedbystructure
distortinglow-frequencyvibrationalmodes.Fromtheseresults,weconcludethatconformer
specificprocessestakeplaceaftertheexcitationoftheneutralconformersofflexiblemolecule.
Thisrepresentsaspectroscopicproofofconformerselectiveprocessesinbiologicallyrelevant
molecules.

158

Pseudoephedrine:

Evidence

orf

ormerConf

Specific

Fragmentation

artP

Concluding

159

IV

Remarks

ConclusionsandSummary

Thepresentworkcontainsnewresultsonmodelflexiblemolecules,theirbinarywatercom-
plexesandtheirfragmentationbehavior,acquiredthroughthepowerfulcombinationofhigh-
resolutionmultiphotonUVandthresholdionizationspectroscopybothwithmassselectivity,
genetic-algorithm-basedrotationalfittingprocedure,andhigh-levelabinitioquantumchemistry
computations.Theseresultselucidateseveralimportantphenomena:oneistheeffectoffluorina-
tionontheconformationalpreferencesofbiologicallyrelevantprototypemolecules;anotherone
istheformationofintramolecularhydrogenbonds,whichstabilizeconformationalstructures
andofintermolecularhydrogenbonds,whichparticipateintheformationofwatercomplexes.
Anotherinterestingeffectthatwasrevealedinthisworkistheoriginofthenewlyobservedfrag-
mentationalpatternintwoimportantneurotransmittermolecules.
Inthebeginningofthethesis,ageneraldescriptionofthespectroscopictechniquesalongwith
thedetailsoftheexperimentalsetupandconditionsusedfortheinvestigationoftheabovemen-
tionedmolecularsystemshasbeenpresented.Particularemphasishasbeenputontheexcitations
schemesofmassanalyzedthresholdionizationandresonanceenhancedtwo-photonionization,
theattemptofpulsestretchingofthelaserpulses,ontheproductionofsupersonicmolecular
beams,thelasersystemsandthegenetic-algorithm-basedfittingroutine.
Theexperimentalsectionisfollowedbyadetailedsurveyoftheobtainedresults.Thatcanbe
dividedintothreemaingroups.Thefirstoneconsiderstheconformationallandscapeofflexible
biologicallyprototypearomaticmoleculesandtheeffectoffluorinationontheconformational
behavior.Thesecondgroupofresultsconsistsoftheinvestigationoftheformationofwaterbi-
narycomplexesofthesespecies.Finally,thethirdgroupencompassesevidencesthatshedlight
onthespecificfragmentationalpatternoftwoneurotransmittermolecules.
ThenonconventionalO-H∙∙∙πtypehydrogenbondanditsstabilizingeffectonthemostabundant
conformericstructureshasbeenaddressedthroughtheinvestigationoftheflexible2-phenylethanol,
2-para-fluorophenylethanoland2-ortho-fluorophenylethanol.Theexperimentalhigh-resolution
resonanceenhancedtwo-photonionizationspectroscopicresultstogetherwithfitsoftherota-
tionalstructureofvibronicbandsandabinitocalculationshavedemonstratedthatinallcases
thepreferredgeometryisthegaucheoneinwhichtheOHgroupofthebentsidechainiscon-
nectedandstabilizedthroughanintramolecularhydrogenbondwiththeπelectronsoftheben-

161

162

zenering.For2-phenylethanoland2-para-fluorophenylethanol,presenceofantistructuresis
identified.Despiteoftheirhigherenergy,theseconformersareavailablesincerelaxationtothe
lowergaucheonesisnotpossiblebecauseofhighpotentialbarriers.Theotherimportantresults
isthatthepresenceofastronglyelectronegativeatominthemoietyofthearomaticringdoesnot
affectthestabilityoftheO-H∙∙∙πinteractionandthegaucheconformersoftheabovementioned
ed.preservaremoleculesThecompetitionbetweenσandπhydrogenbonds,conditionedbytheproximityoftheFatom
tothesidechain,andhencetotheterminalhydroxylgroupin2-ortho-fluorophenylethanolhas
beensubjectedtoadetailedspectroscopicinvestigation.Itshowedadisappearanceoftheanti
conformers,observedinthecaseof2-para-fluorophenylethanol,andapossiblepresenceofa
newgaucheconformerwheretheOHgroupin2-ortho-fluorophenylethanolislocatedcloseto
thefluorineatom,facilitatingtheformationofaσhydrogenbondwithnearbyelectronegative
atom.ThefateoftheO-H∙∙∙πhydrogenbondduringionizationhasbeenexploredbymassanalyzed
thresholdionizationspectroscopyoftheprototypesystem2-para-fluorophenylethanol.Ioniza-
tionleadstotheejectionofaπelectronandthustoadecreaseoftheπ-electrondensityinthe
ring.FromabinitocalculationswiththerecentlydevelopedMO5DFTfunctionalandthreshold
ionvibronicspectra,ithasbeenfoundthattheresultingionicstructureoftheneutralgauche
conformerissignificantlyinfluencedwhichleadstoitstransformationtoastructurewherethe
stabilizingO-H∙∙∙πhydrogenbonddoesnolongerexists.
Thebindingpositionsofthewaterandtheresultingsinglyhydratedcomplexesof2-phenylethanol,
2-para-fluorophenylethanoland2-ortho-fluorophenylethanolhavebeenelucidatedthroughthe
analysisofthehighlyresolvedrovibrationalspectraofwaterbinaryclustersinthesupersonic
jetandtheoreticallybyquantumchemicaloptimizationsofthemoststableconformericstruc-
tures.For2-phenylethanoland2-para-fluorophenylethanol,thetheoreticallymoststablebinary
watercomplexisone,wherethewateractsasaprotonacceptorfromtheOHgroupofthehost
andaprotondonortotheπelectronsofthebenzenering.Thesewerenotdetecteddirectly
throughtheirhighlyresolvedUVspectra,butevidencefortheirpresencewasdeducedfrom
massselectivevibrationallow-resolutionspectra.Ontheotherside,thenext-in-energycomplex
structureswherethewateractsasaprotondonortotheterminalOHgroupofthemonomer
weredirectlyidentifiedbyrotationalanalysisoftheirhigh-resolutionspectra.Inthecaseof
2-ortho-fluorophenylethanolthetheoreticallypredictedmoststablestructurewasobservedex-
perimentally,togetherwithahigherenergyonethatisformedthroughtheinsertionofthewater
moleculebetweenthefluorineatomandtheOHgroupofthesidechainofthehostmolecule.
Hithertounknownfragmentsofthebiologicallyactivemoleculesephedrineanditsdiastereoiso-
merpseudoephedrinehavebeenobservedforthefirsttimeinourtwophotonionizationmass
spectra.Theypointedtotwointerestingeffects:Inephedrine,thefragmentationalbehavioris
foundtodependonthenatureofthevibrationalintermediatestateofasingleconformerinthe
two-photonionizationprocess.Ontheotherhand,inthecaseofpseudoephedrinethefragmen-

163

tationalpatterndependsonthedifferentstructuralgeometriesoftwoconformers.
Inconclusionwehaveshownthatthecorrelatedanalysisofhigh-resolutionR2PIandMATI
spectroscopydata,supportedbyhigh-levelabinitiocalculations,providesveryaccurateinfor-
mationonthestructureofmediumsizedmonomers,andtheirwaterbinaryclustersanddetailsof
theintramolecularandintermolecularinteractionprocesses,bothintheexcitedandintheionic
state.Oncethespectroscopyandconformationalpreferencesofthemonomersareavailable,the
moleculespresentedhereandtheirlargerderivativeswillbeinterestingcandidatesforfurther
investigationoftheeffectsofwaterattachmentontheirconformationalpreferences.Asthesize
ofthesolutemoleculegrows,thepotentialforintramolecularhydrogenbondsalsowillincrease.
Complexationwithwatercouldleadtoacompetitionbetweenintramolecularandintermolecular
hydrogenbonds,ortoastrengtheningoftheintramolecularhydrogenbondsbywaterbridges.
Theinfluenceofwatersolvation,andmoreparticularlywaterbridgeformation,ontheconfor-
mationalpreferencesofflexiblemonomersofevengreatercomplexitywillbeafascinatingand
challengingarenainwhichtoapplytheprecisemethodshighlightedinthiswork.Theyalsowill
presentaparticularchallengetocomputationalchemists,andthoseseekingtoimprovesemiem-
fields.forcepiricalInaddition,reliableinformationforthecationicstructurescanserveasabasefortheunder-
standingofprocesseslikechargedelocalizationinbiologicallyactivemoleculeslikepeptides,
aminoacidsandneurotransmitterswhereultrafastchargemobilitycanbehighlycorrelatedwith
reactivityasshownbytheoreticalmodels.Asafirstpromisingsteptowardsthisgoal,wehave
presentedthresholdionizationspectrawithvibrationalresolutionofaprototypeflexiblemolec-
cation.ularBothfromtheexperimentalandtheoreticalpointsofview,thewayisopentoreachsystemsof
higherdimension,startingfromchromophorecontainingbiologicallyrelevantflexiblesystems
withshortsidechainsliketheneurotransmitteramphetamineandtheneuralhormonemelatonin,
andcontinuingtolargersystemslikecarbohydrates,includingmono-andoligosaccharides,pep-
glycopeptides.andtides

164

ppendicesA

165

AAppendix

InteractionsMolecular

Thisappendixdiscussesthemolecularinteractionsandtheirmostimportant
characteristics.Itpresentsaclassificationoftheweakintermolecularbonds
alongwithexamplesoftheirtypicaloccurrences.Stabilizationofmolecular
structuresandbindingpatternsofmolecularclustersresultingfromthesubtle
interplaybetweenvariousweakinteractionsarehighlighted.

A.1BondsChemical

Theinteractionformationoftwofoatomschemicalandbondsresultingisainquantumthebuildingmechanicalofmorecomplephenomenonxobjectsmanifestingsuchasitselfmoleculesinthe
andsolids.Chemicalbondscanbesubsumedintothreemaincategories:covalent(homopolar),
[133].and,metallicionic•Covalentbondsareformedbysharinganelectronpairbetweentwoatoms,thusleading
tocovaalentminimizationbond,forofinstancethetotalH2,Oener2,gyetc.,ofthethiselectronsystem.densityWhenistwodistribidenticalutedatomssymmetricallyforma
twobetweenatomsthemandparticipatingthebondintheinthiscovcasealentisbondreferredformationtoashavnonpolaredi.ffInerentgeneral,electronhoaweffiver,nities,the
andhencetheelectrondensityisshiftedtowardstheonewithhigherelectronegativity,
thusresultingtoapolarbond:HCl,HF,etc.
•Inlarge,theethextremeelectoncase,pairiswhenshiftedtheditofftheerencestronglyofelectronelectroneaffigatinitiesveofatom,thetwthusogiatomsvingisriseveryto
theattraction.formationInthisofwtwayotheelectricstronglychargespolarcowhichvalentarebondheldtogethertransformsbyintoCoulombanionicbondelectrostaticsuch

167

168

InteractionsMolecular

asinthecaseofNaCl,forinstance.Theborderbetweenstronglypolarcovalentandionic
bonds,however,isnotrigorouslyset.
•whichMetallicareinbondstheareconductionstronglyband,delocalized.betweenTheallyatoms,resultthusfromformingsharingthethevso-calledalentelectrons,electron
as.gChemicalbondenergiesrangebetween1and10eV,andtheirtypicaldistancesarebetween1
and3Å.Theydeterminetheskeletalstructureofmolecules.

InteractionsMoleculareakWA.2

Inisolatedmoleculesthenegativelychargedelectronshellshieldsoffthepositivelycharged
molecularcore,thusyieldinganeutralsystem.Whenmoleculesareclosetooneanother,how-
ever,theirelectronshellsexperiencenotonlytheinfluenceoftheirownpositivecoresbutalso
thepresenceoftheelectronshellsofthesurroundingmolecules.Thisbringsaboutadeformation
andredistributionoftheelectronchargewithinthemolecules,leadingtoanincompleteshielding
ofthepositivecoresand,respectively,totheappearanceoflocallychargedmolecularspecies.
Thus,moleculesinteractwithoneanotherbyelectrostaticforcesactingbetweenthepositively
chargedmolecularcoresandthenegativelychargedshells.Weakmolecularinteractionsarealso
referredtoasnonbondingornoncovalenttodistinguishthemfromchemical(alsocalledbond-
ing)interactions.Nonbondinginteractionsaretwotypes:attractiveandrepulsive.Attractive
weakmolecularinteractionsareclassifiedaselectrostatic,inductive,anddispersion.Repulsive
interactionsstemfromthePauliexclusionprinciple.

A.2.1AttractiveMolecularInteractions

Interactionsutionge-Distribermanent-CharP

Staticchargedistributionscanbeexpandedinaseriesofmultipolemoments.Thus,twomolecules,
AandB,interactwitheachotherthroughtheinteractionofthestaticmultipolemomentsoftheir
chargedistribution.Thatiswhy,thistypeofintermolecularinteractionisreferredtoaselec-
trostatic,andtheresultinginteractionenergyisdesignatedasEel.Thistypeofintermolecular
interactionsissubjectofthefirst-orderperturbationtheory.Theelectrostaticenergycanbe
presentedinthefollowingway:

E=qAqB+qA|µB|+qB|µA|+µA∙µB+qAQB+qBQA+µA∙QB+µB∙QA+QA∙QB+∙∙∙(A.1)
el|r|r2r2|r|3|r|3|r|3r4r4|r|5

InteractionsMoleculareakWA.2

169

whereq,µ,Q,andrstandfortheelectriccharge,electricdipolemoment,electricquadrupole
moment,andthepositionvectorofthesecondmoleculerelativetothefirstone,respectively.Itis
importanttoemphasizethattheenergy,Eeldependsnotonlyonthemagnitudeofthemultipoles
butalsoontheirmutualorientation(thisisencodedinthedotproductsinEq.A.1).Thatis
why,innaturalsystemsthemoleculesareorientedsuchastominimizethetotalelectrostatic
energy.ItisobviousfromEq.A.1thatthemagnitudeofthehigher-ordertermsinthemultipole
expansiondwindlesdownveryrapidlywiththeintermoleculardistance(inversepowerlaw)and
forthisreason,theinteractionenergycanbedescribedfairlyaccuratelyonlybyretainingthe
firstnonvanishingtermsinthemultipoleseries.Somemolecules,duetosymmetries,donothave
low-rankpermanentmultipoles,andhencenecessitatetheinclusionofhigher-orderterms.There
existalsomolecularsystems,e.g.,complexeswithraregasatoms,thatdonothavepermanent
multipolesatall,andforthemtheelectrostaticinteractionsdonotcometothescene.

InductionInteractions

Theinductioninteractionbetweentwomolecules,AandB,originatesfromtheinteractionbe-
tweenthepermanentdipolemomentofoneofthemoleculeswiththeinduceddipolemoment
(itinduces)intheotherone.Themagnitudeoftheinduceddipolemomentdependsonthe
magnitudeofthepermanentdipolemoment,thespacingbetweenthetwomolecules,andthe
polarizabilityαofthemoleculeinwhichthedipolemomentisinduced.Thismodelconstitutes
theso-calledsecond-orderperturbationtheory.Thepotentialenergyofinductioninteractions,
Eind,canbepresentedintheform:

1q2µ2Q2
Eind=−2αBr4A+f1(µA,µB)rA6+f2(µA,µB)r8A,(A.2)
whereα,q,µ,andQdesignatethepolarizability,theelectriccharge,theelectricdipolemoment
,andtheelectricquadrupolemoment,respectively.Themoleculewithpermanentmultipoles
isdenotedbyA,andtheonewithinducedmultipolesbyB.Functionsf1andf2dependonthe
mutualorientationofthetwomolecules,AandB.

InteractionsDispersion

Dispersioninteractionsconstitutethethirdtypeofattractiveintermolecularinteractions.They
derivefromthemutualpolarizationoftheinstantaneouselectrondensitydistributionsofthetwo
monomers,AandB.TheywereforthefirsttimerationalizedanddescribedbyLondon[134],
andhencecalledalsoLondoninetractions.Thepotentialenergyofsuchaninteraction,Edisp,can
formtheincastbe

170

InteractionsMolecular

Edisp=−C66−C88−C1010−∙∙∙(A.3)
rrrC6,C8,C10,etc.areempiricalconstants,andristhepositionvectorofthemoleculeBrelative
tomoleculeA.London[134]hasdiscoveredarelationbetweentheabove-mentionedconstants
andthepolarizabilitiesαAandαBofthetwointeractingmolecules,AandB,andtheirionization
energies.Notwithstandingitistheweakestamongtheattractiveinteractions,thedispersion
interactiongainsimportanceinnonpolarmolecularcomplexeswhereitgivesrisetotheonly
bindingforce.Hence,itisofparamountimportanceinlargemoleculeswithlargepolarizability,
andincomplexescontaininganoble-gasatom,asisshowninthiswork.

A.2.2RepulsiveMolecularInteractions

Theonlyrepulsiveintermolecularinteractionarisingwhentwomolecules,AandB,arecloseto
eachotherstemsfromtheexchangeinteraction.Thelatterisamanifestationofthefundamental
Pauliexclusionprinciple,whichprecludesthepenetrationofelectronsfromoneofthemolecular
moietiesintotheoccupiedorbitalsoftheotherone.Thedescriptionoftheexchangeinteraction
isbasedcompletelyonquantummechanics.Theresultingenergy,Eexchhasbeenmathematically
formulatedbyHeitlerandLondon[133–135]throughanexponentialorinversepowerlaw:

r2−Eexch=A∙ea0

(A.4)

BEexch=rn(A.5)
InEq.A.4a0istheBohrradiusofthehydrogenatom,andAisanempiricalconstant.InEq.A.5
Bisanempiricalconstant,andthepowernrangesbetween10and20.Theinverse-powerlow
describestheexchangerepulsionincasesofverysmallseparationdistancesbetweenthetwo
B.andAmolecules,

A.3TheConceptofWeakMolecularBonds

Weakmolecularinteractions(descibedabove)arenotisolatedbutusuallyco-existandactco-
operatibetweenvely.them,Thenetultimatelyeffectofleadingthetotheconcurringformationweakofaweakinteractionsbondisawhoseresultofnature,ahodelicatewever,isbalancedif-
ferentfromthenatureofchemicalbonds(seeSec.A.1above)altogether.Itsstrengthistypically

A.3TheConceptofWeakMolecularBonds

171

fromonetotwoordersofmagnitudeweakercomparedtotheoneofchemicalbonds,depending
onthenatureoftheparticularmolecularsystemandtheweakinteractionsinvolved.Usually,the
energyofweakmolecularbonds,Ebond,isintherangeof0.01-1eV,andtheinteratomicdistances
arelonger(2-5Å)thantheonesinthecaseofchemicalbonds.
Thetotalenergyofaweakmolecularbond,Ebondcanbepresentedasasumoftheenergiesof
theinteractionsinvolved

Ebond=Eattr+Erep=Eel+Eind+Edisp+Eexch(A.6)

Theinteractionpotentialanditstwoconstituents,therepulsiveandtheattractivepotentials,are
A.1Fig.inschematicallypresented

FigureA.1:Schematicrepresentationofaweaktwo-atommolecularinteractionpotentialcomposed
ofarepulsiveandattractivepotentials.r0designatestheinteratomicdistancecorrespondingtothethe
minimumofthepotentialwell,andD0isthedepthofthepotentialwell.

Itisinterestingtoobservethatduetothebalancebetweentheattractiveandrepulsivepotentials,
inmostcases,thereexistsanequilibriuminteratomicdistance,r0,atwhichthenetinteraction
energyhasminimum.Itmayhappen,however,thattheattractivepotentialcannotmakeupfor
theactionoftherepulsivepotentialatanypoint,thusleadingtoanonbondingpotential,i.e.,
apotentialthatdonotpertainaminimum.Inthiscasenoweakbondcanbeformed.Itisim-

172

InteractionsMolecular

portant,however,topointoutthatforabondtobeformed,notonlymustthepotentialhavea
minimum,butalsothezero-pointvibrationallevelmustbebelowthedissociationenergy.
Veryoften,forpracticalneeds,theshapeoftheinteractionpotentialmustbeknown.Obtaining
theshapeofthepotentialstartingfromfirstprinciplesandconsideringalltypesofinteractions
involvedisquitedemanding,andinmostcases,untractableproblem.Anotherapproachtosolv-
ingtheproblemistodeviseamodelpotential,i.e.,aspecialfunctioncontainingfreeparameters,
astheonesshowninFig.A.1,andtofitthoseparameterstoexperimentaldata.Thesimplest,yet
sustainablypopularone,wassuggestedbyMorse[136]in1929.Itisazeroth-orderapproxima-
tiontoexperimentalpotentials.Tomoreaccuratelydescriberealpotentials,onehastoresortto
morecomplexpotentials,themostcommonlyusedonebeingthefamousLennard-Jonespoten-
tial[133,137,138].Itincludestwotermsaccountingforrepulsiveandtheattractiveinteractions.
Inrealsystems,theinteractionsarenotconstrainedtotwo-bodyones.Foranadequatedescrip-
tionofweakmolecularinteractionsinvolvingmorethantwoparticles,themany-bodyinteraction
isreducedtoatwo-bodyinteraction,whereintheabove-discussedtreatmentapplies.

A.4ClassificationofWeakMolecularBonds

Thecombinationofthevarioustypesofweakmolecularinteractionsgivesrisetoabigvarietyof
bondingpatterns.Thoughthereisnorigorousborderlinebetweenthem,ageneralclassification
accordingtothenatureoftheinteractionsinvolved,bondenergies,andbondlengthscanbe
made.

A.4.1BondsogenHydr

Hydrogenbonds(H-bonds)areoneofthemostabundantandimportanttypesofweakmolecu-
larinteractionsinnature.Theydeterminethepropertiesandbehaviorofliquidsandbiological
systems.HydrogenbondshavethegeneralmotifX-H∙∙∙Y,wheretheX-Hgroupisaproton
donor,andYisaprotonacceptor.Usually,Xisanelectronegativeatom(O,N,Cl,F,etc.),
andYismosttypicallyeitheranatompossessingaloneelectronpairoritisaπ-electronsystem
(benzenering,ordoubleortriplechemicalbond),whichhasanexcessofelectrondensity.Weak
hydrogenbondsinvolvingaπ-electronsystemhaveattractedscientistsattentionformorethan
60yearssinceDewarswork[139]appeared,andtheirsignificancehasbeenrecognizedinmany
biologicallyrelevantsystems.Itwasnotuntilrecently,whentheimportanceofanotherspe-
cifictypeofhydrogenbondswasrealized;thesearetheC-H∙∙∙Y(nonconventional)hydrogen
bonds[81,140–143],whichplayanimportantroleforthestabilizationofmolecularstructures.
Usually,Yisanelectronegativeatomoraπsystem.Theyaremuchweakercomparedtothe
conventionalhydrogenbond,andwithrespecttothebindingenergytheyareattheborderline

A.4ClassificationofWeakMolecularBonds

173

betweenhydrogenbondsanddispersionbonds(seebelow).
Ithaslongbeenknownthattheπelectronsofalkenes,alkynes,andaromaticcompoundsmay
actashydrogenacceptors[144,145].Alcoholsandphenols,aswellasprimaryandsecondary
amines,areknownasgoodhydrogendonors.Forexample,hydrogenbondsoftenprovidethe
strongestintermolecularforcesbetweenmoleculesinorganicmolecularcrystalsandhenceof-
tendictatethepreferredpackingarrangement.Thegeneralprinciplesunderlyinghydrogen-bond
formationarereasonablywellunderstoodandthestructuresofhydrogen-bondedcrystalscanof-
tenberationalizedinpreferredcombinationsofhydrogen-bonddonorsandacceptors[146–148].
Ingeneral,thestrongesthydrogen-bonddonorspairoffwiththestrongesthydrogen-bondaccep-
tors.Similarpairingprocessesarerepeateduntilallthehydrogen-bonddonorsandacceptors
havebeenutilized.However,whenasystemcontainsexcessdonorsoracceptors,atleasttwo
hydrogen-bondingstrategiesareavailabletoaccommodatethemismatch[149]:(i)changeinhy-
bridizationor(ii)theformationofhydrogenbondsinvolvingtheπsystemofanaromaticgroup
astheacceptor.Interactionbetweenπbasesandhydrogendonorshavebeenexaminedmainlyby
IRspectroscopy[150].AsaconsequenceofweakeningoftheOHorNHbond,theOHorNH
stretchingabsorptionofhydrogendonorsshowlow-frequencyshifts.Theshiftsare,however,
considerablysmallerthaninOH∙∙∙Oandotherhydrogenbonds,indicatingthatthestrengthof
anOH∙∙∙πinteractionisweakerthanthatofanOH∙∙∙Ohydrogenbond.
RelativelyfewstructurestudiesofcompoundswithintramolecularOH∙∙∙πhydrogenbondsare
known.Inthisworkwehavestudiedtheformationandtheroleofsuchπhydrogenbondsas
astabilizingfactorintheconformationalpreferencesofmodelorganicmolecularsystems.(See
6.)4,3,ChaptersThemostsignificantfeaturesandpropertiesofhydrogenbondsarehighlightedinthefollowing.

•Hydrogenbondsarestabilizedbyasubtleinterplaybetweenelectrostatic,induction(charge-
transfer),anddispersioninteractions[89].Themostimportantcontribution,however,is
theelectrostaticone,realizedmainlythroughdipole-chargeanddipole-dipolecoupling.
Thisbringsaboutthenexttraitsofhydrogenbonds.
•Hydrogenbondsprovideatypicalexampleofatwo-bodyinteraction.Duetotheelectro-
staticattractiontheH-bondlengthisshorterthanthesumofthevanderWaalsradiiofthe
twoatoms,X,andY.
•Directionality.Hydrogenbondsarestronglydirectional,i.e.,thethreeatoms,X,H,and
Y,liedownastraightline.
•TypicalH-bondlengthsrangebetween2and3Å.
•TheenergyofH-bondsvariesbetween0.1and1eV,dependingonthenatureofXandY
atom(groups),andconsequently,ontherelativecontributionsoftheconstitutinginterac-

174

MolecularInteractions

tions.Thus,someH-bondsrenderthemselvesatthebrinkofchemicalbonds
Altogetherhydrogenbondscanbesubsumedintothreemajorgroups:proper(red-shifting),
improper(blue-shifting),anddihydrogenbonds.

operPrBondsogenHydr

Proper(red-shifting)hydrogenbondsweakenthecovalentX-Hbondthusleadingtoitselonga-
tion,andrespectively,toaredshiftoftheX-Hstretchingvibrations.Thenatureofthiseffect
hasbeenexplainedonthebasisofbondorbitalanalysis[89,151].Thelattershowsthatacharge
transfertakesplacefromtheloneelectronpairortheπ-electronsystemoftheprotonacceptor
totheantibondingorbitalsoftheprotondonor.Thisleadstoanincreaseoftheelectrondensity
intheantibondingorbitals,whichcausestheweakeningoftheX-Hchemicalbond,andtoits
ation.elong

ImprBondsogenHydroper

hImproperydrogenbonds(blue-shifting)(seeabohve),ydrogeni.e.,bondsuponthe[89,152]formationexhibitofbehasuchviorweakoppositebondstothetheonechemicalofproperX-H
bondwereshortenstheoreticallyandthepredictedfrequencforyofthecarbon-protonX-Hdonorvibration-benzeneincreases.complexesBlue-shifting[153]andhlaterydrogenon,bondstheir
e[154].xistenceAswinasethecasexperimentallyofvred-shiftingerifiedforbonds,theamodelchargesystemtransferprocesschloroform-fluorobenzeneunderliestheobservcompleedx
phenomena.Inthiscase,however,thechargetransferisdirectedtotheremotepartoftheX-H
bond,followedbyastructuralchangeintheproton-donormolecule.

BondsogenydrDih

Dihydrogenbonds[89,155]havethepatternX-H∙∙∙H-Y,whereXisanatomofametalelement,
andYisanelectronegativeatom.Theywerediscoveredonlyrecentlyinahydrogen-bonded
complexcontainingIr[156].Themechanismofthesebondswasrationalizedabitlater[157],
anditisquitestraightforward.Themetalatomdonateselectrondensitytothecovalentlyattached
toitHatom,thuscreatingapartialnegativechargeuponit.Ontheotherhandtheelectronegative
atomYwithdrawstheelectrondensityfromtheadjacentHatomsothatthelatterbecomes
positivelycharged.Inthiswaythedihydrogenbondisstabilizedbyamultipoleinteraction.

A.4ClassificationofWeakMolecularBonds

BondsQuadrupole-QuadrupoleA.4.2

175

Theoccurenceofquadrupole-quadrupolebondsislimitedtosystemscontainingbenzenerings,
zenewhichdimerdo,notandhaveplayanpermanentimportantdipoleroleformoments.theTheystabilizationdetermineofthesecondaryT-shapeandstructuretertiaryofthestructuresben-
macromolecules.of

Bondsransferge-TCharA.4.3

Charge-transfer(CT)bondsemergewhenoneofthebondedmoietiesisagoodelectrondonor
(ithasalowionizationpotential),andtheotheronehashighelectronaffinity.Donorspertain
antibondingorbitalsdesignatedasn,σ,andπ,respectively,andacceptorshavevacantorbitals
labelledasn*,σ*,andπ*,respectively.Thestrongestoftheso-formedCTcomplexesarethe
onesofn-vtype.

BondsIon-MediatedA.4.4

Ion-mediatedbondsstemfromthepresenceofmetalliccations,whichhavehighelectronaffinity,
andhenceformpolarbondsonthebasisofmultipoleinteractions.

InteractionsophobicHydrA.4.5

Hydrophobicinteractionsareaspecialclassofinteractionsrepresentingassociationsofnon-
polargroupsinpolarsolvents.Ithasbeenfound[89,158,159]thatthechangeoftheenthalpy
isalmostvanishing,andthedrivingforceforthereorganizationofthemoleculesinasolutionis
theentropy.Thisistheirmostdistinctivefeaturecomparedtotheabove-describedbonds,which
areformedasaresultofanenergyminimization.

BondsDispersionA.4.6

bDisperution,sionandhencebondsareonlytheformeddispersionwhenthetermininteractingtheattractimoietiesvedopotentialnothaisveaccountedpermanentfor.chargeDispersiondistri-
Theirbondslackmanifestationdirectionalityismostandhelppronouncedfortheinmolecularstabilizationofcomplexesmoleculescontainingandamolecularnoble-gascompleatom.xes.

176

A.5

BindingMotifs

InteractionsMolecular

Weakinteractionsandweakbondscanbeintramolecularorintermolecular.Intramolecularweak
bondsplayanimportantroleforthestabilizationofmolecularstructures,inparticular,those
thathavemanyinternaldegreesoffreedomandcanassumedifferentconformationalshapes.
Themostabundantintramolecularbondsarethehydrogenones,thoughquadrupole-quadrupole
bondsarealsoobserved.Themosttypicaloccurrenceofsuchbondsisinbiologicallyrelevant
molecules,asisdiscussedinChapters3,4and6.Intermolecularbondsareresponsibleforthe
formationofsupramolecularstructures.Whenbenzene-ring-containingmoleculesareinvolved
intheformationofacomplexwithanothermolecule,twobindingpatternsarepossible.Abond
thatisformedbetweenanatom/moleculeandtheπelectronsofthehostmoleculeisreferredto
asaπbond.Abondthatisrealizedbetweenanatom/moleculeandabenzene-ringsubstituentis
termedaσbond.AnexampleforanintermolecularσbondispresentedinChap.7.

BppendixA

InitioAbCalculations:ChemistryQuantumStructure,Energetics,andFrequency
SpeciesMolecularIsolatedofAnalysis

Theinterpretationoftheexperimentalresultsissupportedbytheoreticalstud-
ies.Thetheoreticalpredictionofmolecularpropertieshaseverbeenachal-
lengingtask.Manymodelsdescribingoneorafewaspectsofthemolecular
behaviorhavebeendevelopedovertheyears.Untilrecently,manytheoret-
icalstudiesevenonsmallmoleculeswereuntractableduetotheir,mainly,
mathematicalcomplexity.Therapidenhancementofcomputationalpower,
however,hasstimulatedthedevelopmentofsophisticatedalgorithms,and
enabledthetreatmentofmedium-sizedandlargemoleculesandmolecular
complexes[160–162].Theexistingmethodsforthecalculationofmolecu-
larstructuresandpropertiescanbegroupedinfourcategories,whicharede-
scribedinthefollowing.Themainemphasis,however,isputonthequantum-
mechanicalmethods,whichwillbepresentedinmoredetails[163].

MethodsSemiempiricalB.1

SemiempiricalmodelsuseforthedescriptionofmoleculesasimplerHamiltonianthanthereal
one,andemployasetofparameterswhosevaluesareadjustedtomatchtheexperimentaldata,
andhencethenamesemiempirical.AtypicalexampleofsuchamodelistheH¨uckelmolecular
orbitalmodel,whichemploysaone-electronHamiltonian,andthebondintegralsplaytherole
parameters.adjustabletheof

177

178

CalculationsChemistryQuantumInitioAb

B.2AbInitioQuantumChemistryCalculations

Abinitioquantumchemistrycalculationsarebasedontheuseofthefirstprinciplesandthe
fundamentalphysicalconstants.TheyemploythetruemolecularHamiltonian,whichaccounts
foralltheinteractionswithinthemolecularsystem.Itshouldbepointedout,however,thatdue
totheapproximationsinevitablyintroducedinthesolvingoftheSchr¨odingerequation,abinitio
calculationsdonotprovidetheexactsolutiontothetreatedproblem.Themainideaonhowthe
Schr¨odingerequationcanbesolvedalongwithsomeimportantconceptsaredescribedbelow.

B.2.1TheMøller-PlessetPerturbationTheory

TheMøller-Plesset(MP)perturbationtheory[164]isaparticularcaseofthemany-bodyper-
turbationtheorywhentheinterelectronicrepulsionistreatedasaperturbationtotheinteraction
energybetweenthenucleiandtheelectronsofamolecularsystem.Thefirstpracticalapplica-
tionsoftheMPtheorycametothesceneonlyin1975withtheworkPopleandco-workers[165].
Nowadays,itiswidelyusedforaccuratemodelingofmoleculesandmolecularcomplexes,and
itisofparticularsignificancewhenelectroncorrelationsmustbeaccountedfor.Theoverview
presentedbelowisrestrictedonlytoclosed-shellground-statemolecules.Forpracticalreasons,
thespin-orbitals,ratherthanthespatialorbitals,areused.Forspin-orbitals,theHFequationsfor
electronminann-electronmoleculehavetheform

fˆ(m)ui(m)=iui(m)
Thespin-orbitalHamiltoniancanbecastinthefollowingform

(B.1)

nfˆ(m)≡−21m2−rZi+[jˆj(m)−kˆj(m)].(B.2)
mi1=jijˆandkˆaretheCoulombandtheexchangeoperators,respectively.Theyaredefinedas

jˆj(α)f(α)=f(α)|φj(β)|2r1αβdvβ(B.3)
∗kˆj(α)f(α)=φj(α)φj(α)f(β)dvβ(B.4)
rαβInfunction,bothanddefinitionsαandβabolabelve,thetheinteelectrons.grationTheisCoulombperformedovoperatorertheshowholewsthespace,interactionfisanofarbitraryelectron

B.2AbInitioQuantumChemistryCalculations

179

αphwithysicalthemeaning.spread-outItcharoriginatesgeoffromelectrontheβ.conditionTheeofxchangethewaveoperatorfunctiondoesuponnothatheveeanxchangeintuitivofe
electrons.otwisItwhcanybetheprofirstvedthatsignificantthesumimproofvtheementzeroth-intheandMPenerfirst-ordergyisenergiesintroducedequalsthroughtheHFtheenerinclusiongy.Thatof
thereferredtosecond-orderLevinesenertegyxtbookcorrection.onFquantumorachemistrycomprehensiv[163].eThediscussiononsecond-orderthematter,correctionthereadertotheis
isgyener

(B.5)=E(2)|ψs(0)|Hˆ|Φ0|2
0s0E0(0)−Es(0)
wherethesummationisperformedovertheexcitationofthezeroth-orderwavefunction,i.e.,
xcitations.eetc.double,single,

MethodCoupled-ClusterTheB.2.2

Thecoupled-cluster(CC)methodwasinventedin1958byCoesterandK¨ummel.Anicediscus-
siononthismethodcanbefoundin[165].Thefundamentalequationthecoupled-clustermethod
is,onfoundedis

ψ=eTˆΦ0(B.6)
Inthisformula,ψistheexactnonrelativisticground-statemolecularelectronicwavefunction,
andΦ0isthenormalizedground-stateHFfunction.Theexponentialoperatorispresentedbyits
taylorization,

∞eTˆ≡1+Tˆ+Tˆ2+Tˆ3+∙∙∙=Tˆk(B.7)
2!3!k=0k!
Tˆisreferredtoasaclusteroperator,andTˆ=Tˆ1+Tˆ2+∙∙∙+Tˆn,wherenstandsforthenumber
ofelectronsinthemolecule.Onlythelow-orderclusteroperatorsareofpracticalimportance.
Thetwomostcommonlyusedclusteroperatorsaretheone-particleexcitationoperator,Tˆ1,and
thetwo-particleexcitationoperator,Tˆ2,whicharedefinedas

n∞aaTˆ1Φ0=tiΦi
a=n+1i=1

(B.8)

180

InitioAbCalculationsChemistryQuantum

ˆ∞∞nn−1abab
T2Φ0=b=a+1a=n+1j=i+1i=1tijΦij(B.9)
Intheabovedefinitions,Φia,andΦijabdesignatesingly-,anddoubly-excitedSlaterdeterminants,
wheretheoccupiedspin-orbitaluiisreplacedbythevacantorbitalua,andtheoccupiedspin-
orbitalsuijarereplacedbythevacantorbitalsuab,respectively.Thecoefficientstia,andtijabare
namedamplitudes.ThegoalofaCCcalculationistofindtheamplitudes.Itisworthpointing
outthattwoimportantapproximationsarepresentintheCCcalculations:i)thebasissetisnot
infinitebuthassomefinitenumberofmembersincluded;ii)onlythefirstcoupleoftermsin
thecluster-operatorexpansionareconsidered.TheoryshowsthatthemaincontributiontoTˆis
providedbyTˆ2,andhencetheapproximation

Tˆ≈Tˆ2(B.10)
isjustified.Withtheapproximationsmade,theground-statewavefunctionψbecomes

ψCCD=eTˆ2Φ0(B.11)
Coupled-clustermethodsprovideveryaccuratetreatmentofelectroncorrelations,buttheyare
alsoquitecomputationallyexpensive,andhenceareappliedpredominantlytosmallandmedium-
sizedmolecules.Asitwasshownfor2-para-fluoropheylethanolinChapter6.

TheoryDensity-FunctionalB.3

Density-functionaltheory(DFT)emergedin1964withtheHohenberg-Kohntheorem[166].The
theoremstatesthattheground-stateenergy,E0,thewavefunction,andallmolecularproperties
areuniquelydeterminedbythetheelectronprobabilitydensity.Mathematically,thismeansthat
theground-stateenergyisafunctionaloftheelectrondensitydistribution,ρ0.Thelatterdepends
onthethreespatialcoordiantesx,y,andz.

E0=E0[ρ0(x,y,z)](B.12)
Thisrelationbetweenthemolecularpropertiesinthegroundstateandtheelectronprobability
densitylaysthegroundforthedevelopmentoftheDFT.TheapproachoftheDFTistoderive
ground-statemolecularpropertiesfromtheelectronprobabilitydensity.
IntheDFTthepotentialnucleicreateonelectronsisconsideredasanexternalpotentialsinceit

Density-FunctionalB.3Theory

181

comesfromoutsidetheelectronsystem.Itisclearthattheground-stateenergydependsonthe
typesofthenucleiandontheirconfiguration,hencetheelectronicground-stateenergycanbe
presentedinthefollowingmanner

E0=Ev[ρ0]=ρ0(r)v(r)dr+T[ρ0]+Vee[ρ0]=ρ0(r)v(r)dr+F[ρ0](B.13)
Theindexvshowsthattheenergydependsonthenuclearpotential,andFisafunctionalthat
dependsontheaveragekineticenergyoftheelectronsandontheinterelectronicinteractions,
butitdoesnotdependontheexternalpotential.Thisresultasastand-alonedoesnothaveany
practicalapplicationsincethefunctionalF[ρ]andtheprobabilitydistributionareunknown.To
Letbegivenatrialelectronprobabilitydensity,ρtr(r),thatobeystherelationsρtr(r)dr=n(nis
harnessEq.B.13forpracticalneeds,HohenbergandKohnhaveproventhefollowingtheorem.
thetotalnumberofelectronsinthemolecularsystem),andρtr(r)≥0.Foreverytrialprobability
holdsinequalitywingfollothe,densityρ

≥densityρtr,thefollowinginequalityholds

E0=E[ρ0(r)]≤Ev[ρtr(r)],

(B.14)

whereρ0isthetruemolecularelectronprobabilitydensity.Theinequalitycanbealsoreenun-
ciatedthatthetruemolecularelectronprobabilitydistributionminimizestheenergyfunctional
Ev[ρtr(r)].ThistheoremisastepaheadtowardsthepracticalimplementationoftheDFTsinceit
claimsthatinprincipleallgroundstatemolecularpropertiesdescendfromtheelectrondensity
distributionbut,itstilldoesnotprovidearecipeonhowtodothat.Thepracticalaspectofthe
problemwasresolvedbyKohnandSham[167].ManyfunctionalshavebeenproposedforF,
andtheyhaveprovedthemselvestobeefficientindescribingagreatvarietyofmolecularsys-
tems.ThemosttypicalandwidelyusedfunctionalistheB3LYPone[168].
TherecentlydevelopedDFTfunctionalM05canbecalledahybridmetageneralizedgradient
approximation,becauseitincorporateselectronspindensity,densitygradient,kineticenergy
density,andHartree-Fock(HF)exchange.[106,169,170]TheM05functionalperformswellfor
kinetics,main-groupthermochemistry,andnoncovalentinteractions,includingthoseinnonpo-
larweaklyinteractingsystemsandcharge-transfercomplexesandalsofortransition-metalbond
energies,ionizationpotentials(IPs),andelectronaffinities(EAs).Thisfunctionalincorporates
kineticenergydensityinabalancedwayintheexchangeandcorrelationfunctionals;itsatisfy
theuniformelectrongaslimit,anditisaselfcorrelation-free.Thismakesitapreferredchoice
fortheoreticaloptimizationsofconformericstructuresstabilizedbynonconventionalweakinter-
actions.Inthepresentwork,wehaveusedthefunctionalfortheoptimizationoftheconformers
of2-para-fluorophenylethanol(seeChapter5).

182

MolecularB.4Mechanics

CalculationsChemistryQuantumInitioAb

Themolecularmechanics(MM)methodisnotaquantummechanicalmethod.Ittreatsmolecules
asasystemofatomsheldtogetherbyelasticbondscharacterizedbyforceconstants,andobeying
theequationsofclassicalmechanics.Onthebasisoftheforceconstants,themolecularenergy
andthevibrationalfrequenciesarecalculated.Thismethodiscomputationallycheap,andhence
sometimesprovidesagoodstartingpointtoquantum-mechanicalcalculations.

GeometryMolecularB.5

EquilibriumB.5.1Geometry

Findingoftheequilibriumgeometryofamoleculeisoneofthemajorgoalsofthetheoretical
calculations.Theequilibriumgeometryofamoleculerepresentsthearrangementoftheatomic
tasknucleiofthatfindingminimizesthetheequilibriummoleculargeometryenergy,Ebecomes,includingmorealsochallengingtheinternuclearwiththeincreaserepulsions.ofThethe
manymoleculardegreessize.ofThefreedom,problemandisoftenaggrasevveralateddibyfferentthefactatomicthatarrangementspolyatomicmaymoleculesresulthainvaevsim-ery
ilarmolecularenergy,whichposesverystringentrequirementsontheaccuracyoftheenergy
calculations.

B.5.2PotentialEnergySurface

Thegeometricalstructureofamoleculeisuniquelydeterminedbydefining3N-6independent
coordinates,whereNstandsforthenumberofatomsinthemolecule.Asmentionedinthepre-
cedingsection,theenergyofamoleculedependsontheatomicconfiguration.Thatiswhy,it
isconvenienttopresentthemolecularenergy,E,asafunctionofthe3N-6independentcoordi-
nates.Thiswillresulttoahypersurface(potentialenergysurface(PES))inthe3N-6-dimensional
space.ItcanbeformallywrittenasE=−E(r).Theminimaonthissurfacecorrespondtostable-
equilibriumstructuresofthemolecule.Thelowest-energyminimumisreferredtoastheglobal
minimum.Allotherminimaarecalledlocalminima.Thepotential-surfacemaximacorrespond
tounstable-equilibriumstructuresthatrelaxtothenearestminimum.Therearealsopointsonthe
PEScorrespondingtominimafor3N-7coordinates,andtoamaximumforonecoordinate.Such
pointsarereferredtoassaddlepoints.Thecalculationoftheenergyatafixedconformationis
calledasinglepointenergycalculation.

MolecularB.5Geometry

183

Thesetofthedihedralanglesaboutallbondsinamoleculedefinesitsconformation.Thecon-
largerformationthemolecule,correspondingthetobiggerantheenergynumberminimumofconformers(localoritglobal)has.isThatistermedwhya,theconformerprediction.Theof
molecularconformersisoneofthemajorgoalsofthetheoreticalcalculations.

OptimizationsGeometryB.5.3

Ageometryoptimizationorenergyminimizationistheprocedureoffindingthemolecularcon-
formationcorrespondingtoanenergyminimuminvicinityofsomeinitiallydefinedmolecular
geometry(conformation).Fromtheviewpointofmathematics,thisistheproblemofthemultidi-
mensionaloptimization.Thereexistvariousalgorithmstohandletheproblem.Themostefficient
onesarethegradientmethod,theNewton-Raphsonmethod,thesteepest-descendmethod,etc.
Tofinddifferentmolecularconformers,onehastoprobedifferentstartinggeometries,andlocally
searchfortheenergyminimum.Theproblemisaggravatedinmoleculeswithmanyinternalde-
greesoffreedom,whichmayhavemanydifferentconformers.Thefindingoftheglobalenergy
minimumisnotatrivialtask,andingeneral,thisproblemhasnotbeensolvedyet.Theguessof
thestartingconformationinthevicinityofwhichtheglobalenergyminimumisexpectedmight
bemisleadingsinceoftenitmayhappenthatthemoststableconformer(thisistheconformer
correspondingtotheglobalenergyminimum)isratherunconventional.Thisnecessitatesthe
developmentofalgorithmsforglobaloptimization.

B.5.4ConformationalSearch

Manyalgorithmshavebeeninventedtotreattheproblemoftheglobaloptimizationthough,none
ofthemcanprovideacompletereliability.Forthisreason,sometimes,acombinationofdifferent
algorithmsisusedtosolveaparticularproblem.Thegeneralnameofthesealgorithmsisglobal
optimizers.Inthefollowing,theyarelistedwithoutpresentingthedetailsonhowtheywork.A
comprehensivesurveyofthesetechniquescanbefoundin[163].
•Systematic(grid)searchmethod..Thisisoneofthecommonlyusedtechniquesfor
global-energy-minimumsearch.ThismethodsamplesthePESwithacertainstepand
ateachsamplepointitcalculatesthesinglepointenergy.Forpracticalreasonsusually
onlyafewmolecularcoordinatesarescanned(mosttypicallydihedralangles),whilethe
othercoordinatesarekeptfixedattheirinitialvalues.Thenalocalsearchisappliedtothe
lowest-energysamplepoint.Inthiswaytheglobalenergy-minimumcanbefound.This
methodwasusedintheconformationalinvestigationofthesystemsinthepresentwork.
•Random(StochasticorMonteCarlo)searchmethod.

184

InitioAbCalculationsChemistryQuantum

•Distance-geometrymethod.Inthismethodthemoleculeisdescribedasadistancematrix
whoseelements,dij,arethedistancesbetweenatomsiandj.
•Geneticalgorithm.ThismethodisdescribedindetailinSec.2.3inconjunctionwiththe
analysisofhighlyresolvedmolecularspectratofindtheglobalmaximumforthecross
correlation.ch.seardynamicsMolecular••Metropolis(MonteCarlo)method.
•Thediffusion-equationmethod.
•Thelow-modeconformationalsearch.

BSSEB.6

Supposewewishtocalculatetheenergyofformationofabimolecularcomplex,suchasthe
energyofformationofahydrogen-bondedwaterdimer.Suchcomplexesaresometimesreferred
toassupermolecules.Onemightexpectthatthisenergyvaluecouldbeobtainedbyfirstcal-
culatingtheenergyofasinglewatermolecule,thencalculatingtheenergyofthedimer,and
finallysubtractingtheenergyofthetwoisolatedwatermolecules(thereactants)fromthatof
thedimer(theproducts).However,theenergydifferenceobtainedbysuchapproachwillin-
variablybeanoverestimateofthetruevalue.Thediscrepancyarisesfromaphenomenonknown
asbasissetsuperpositonerror(BSSE).Asthetwowatermoleculesapproacheachother,the
energyofthesystemfallsnotonlybecauseofthefavorableintermolecularinteractionsbutalso
becausethebasisfunctionsoneachmoleculeprovideabetterdescriptionoftheelectronicstruc-
turearoundtheothermolecule.Onewaytoestimatethebasissetsuperpositionerrorisvia
thecounterpoisecorrectionmethodofBoysandBernardi[171],inwhichtheentirebasissetis
includedinallcalculations.Thus,inthegeneralcase:

A+B→AB

(B.15)

ΔE=E(AB)−[E(A)+E(B)](B.16)
ThecalculationoftheenergyoftheindividualspeciesAisperformedinthepresenceofghost
orbitalsofB;thatis,withoutthenucleiorelectronsofB.Asimilarcalculationisperformedfor
BusingghostorbitalsonA.Thecalculationcanbepresentedandinamoredetailway.Letus

BSSEB.6

185

considerasupermoleculeABmadeupoftwointeractingsubsystemsAandB.Thestabilization
energycanbewrittenas:

ΔE=EAABB(AB)−[EAA(A)+EBB(B)](B.17)
WhereEYZ(X)istheenergyofasubsystemXatgeometryYwithbasissetZ.Thestabilization
energycanbespiltinthefollowingway

ΔE(AB)=ΔEint(AB)+ΔErel(A,B)(B.18)

Thefirsttermrepresentstheinteractionenergycontribution,whichdependsonlyonthesuper-
ABparameters,geometricalmolecule

ΔEint(AB)=EAABB(AB)−[EAAB(A)+EABB(B)](B.19)
whereasthesecondtermrepresentstherelaxationcontribution,whichcompensatesforthege-
ometrydistortionofthesubsystemsinthesupermolecule,EAAB(A)andEABB(B),withregardtothe
isolatedoptimumgeometry,EAA(A)andEBB(B).

ΔErel(A,B)=EAAB(A)−EAA(A)+EABB(B)−EBB(B)(B.20)
NotethatErel(A,B)dependsonboththesupermoleculeandsubsystemparametersAB,A,B.Ac-
cordingtothecounterpoiseidea,sincethesamebasissetisusedintherelaxationtermfor
eachsubsystem,onlytheinteractionenergycontributiontermbringsaboutBSSE.Thus,the
counterpoise-correctedinteractionenergyshouldbepresentedas:

ΔECP(AB)=[EAABB(AB)−EAABB(A)−EAABB(B)]+[EAAB(A)+EABB(B)−EAA(A)−EBB(B)](B.21)
Rearrangingtermsoftheexpressionaboveoneobtains

ΔECP(AB)=[EAABB(AB)−EAA(A)−EBB(B)]+[EAAB(A)+EABB(B)−EAABB(A)−EAABB(B)]=ΔE(AB)+δABBSSE
(B.22)wheretheCP-correctionexpressedasδABBSSEpresentsoneimportantpropertythattheBSSEis
notanadditivetermtothestabilizationenergy.Actuallycanbestronglygeometry-dependentso
thatitshouldbetakenintoaccountateverypointofthepotentialenergysurfacewhenlooking
forstationarypoints.Therelevanceofthebasissetsuperpositionerroranditsdependenceupon

186

CalculationsChemistryQuantumInitioAb

thebasissetandtheleveloftheoryemployedremainsasubjectofmuchinterest(forreference
seePhDthesisofSedano[172]).

B.7MolecularVibrationalFrequencies

Theconformationalsearchandthegeometryoptimizationofamoleculeprovidetheelectronic
energyofthismolecule.Since,however,atomicnucleiareneveratrestbutperformsmallos-
cillationsabouttheirequilibriumpositions,itisofimportancetocalculatealsothemolecular
frequencies.vibronicTheenergyofaharmonicoscillatorEvibis

1Evib=v+2hν(B.23)
wherevstandsforthevibrationalquantumnumberthatcantakeonvaluesfrom0toinfinity,
andνisthevibrationalfrequency.Thevibrationalenergycorrespondingtov=0isknownas
thezerothvibrationalenergy.Fora3N-atomicmoleculethereare3N-6vibrationalmodesif
themoleculeisnotplanar,and3N-5modesifthemoleculeislinear.Eachofthemodeshasa
frequencyvn,whichiscalledafundamentalfrequency.Intheharmonicapproximation,thetotal
vibrationalenergyofamoleculeisthesumofthevibrationalenergiesofallvibrationalmodes
presentedasindividualharmonicoscillators

MEvib=vn+21hνn,(B.24)
1=nwhereM=3N−6inthegeneralcase,andM=3N−5forlinearmolecules.Thesumofall
thezerothvibrationalenergiesyieldsthetheso-calledzero-pointvibrationalenergy(ZPVE).
Thisisthelowestenergylevelamoleculecanoccupyforacertainconformer.Inpractice,the
weightedfundamentalforce-constantvibrationalmatrixfrequencieselementsofa.3NForce-atomicconstantsmoleculearearealsocalleddeterminedHessiansthrough,themass-

1∂2U
Fij=√mimj∂xi∂xj(B.25)
Therealtreatmentofmoleculesrequiresalsotheinclusionofanharmoniccorrections.The
methodologyisdesribedinverydetailbyWilson,Decius,andCross[173].

B.8ProgrammePackagesforTheoreticalMolecularInvestigations

187

B.8ProgrammePackagesforTheoreticalMolecularInvesti-

gations

Variouscommercialprogramsformodelingofmolecularstructuresandproperties,andtheoreti-
calinvestigationofmolecularphenomenaareavailablenowadays.Oneofthemostconventional
programpackagesisGaussian[82].Ithasbeenusedalsotoprovidethetheoreticalresultsinthis
work.Other,alsopowerful,programsareMolpro[174],Turbomol[175],GAMESS(General
AtomicandMolecularElectronicStructureSystem),Q-Chem,whichisoptimizedforcalculation
oflargemoleculescontainingseveralhundredatoms,Jaguar,ACESII,CADPAC,SPARTAN.

188

Ab

Initio

Quantum

Chemistry

Calculations

CppendixA

MatterandLightbetweenInteraction

Spectroscopydealswithtransitionsbetweenstatesinatomsandmoleculesactivatebyanexternal
electromagneticfield(light,inthecaseoflaserspectroscopy).Thestudiedquantumsystems
(atomsormolecules)startfromsomestationarystate,andasaresultoftheinteractionwith
thelight,theyendupinsomeotherstationarystate.Quantumtransitionsaredescribedbythe
time-dependentSchr¨odingerequation

i∂ψ∂(tq,t)=[Hˆ]ψ(q,t),(C.1)
whereqdesignatesthe3NspatialandtheNspincoordiantesofasystemconsistingofNparti-
cles,andtistime.TheHamiltonianofthejointsystem(0)molecule-electromagneticwavecanbe
presentedasthesumoftheisolated-moleculeHamiltonianHandtheHamiltoniandescribing
theinteractionbetweenthemoleculeandtheexternalfieldHint(thelower-caseindexintstands
e.i.interaction),for

Hˆ=Hˆ(0)+Hˆint(t)

(C.2)

AsseenfromEq.C.2,theinteractionHamiltoniandependsontime.Itcanbeshown[176,177]
thatthetime-dependentwavefunctionψ(q,t)canbeexpandedintermsofthetime-independent
(stationary-state)wavefunctioncorrespondingtoHamiltonianHˆ(0),

ψ(q,t)=ai(t)ψi(q)(C.3)
iAftersomestandardmathematicaltrasformations,andbearinginmindthatthestationary-state

189

190

InteractionMatterandLightbetween

wavefunctionsψi(q)areorthonormal,constituteacompletebasis,andEq.C.2canbetrans-
formedintoitsequivalentform,

where

dcmiiωt(0)(0)
=−ckemnψm|Hˆint|ψn,m=1,2,3,...
tdn

(C.4)

(0)(0)ωmn≡Em−En(C.5)
Eq.C.4issolvedonthebasisofperturbationtheory.TheinteractionHamiltonianistreatedasa
perturbationtothemolecularHamiltonian.Iftheperturbationisappliedtothemoleculeatthe
timet=0instationarystatenthenatt=0,ψ(q,0)=ψn(0),andfromEq.C.3,itcanbeinferred
thatcn(0)=1,andci(0)=0,forin.Assumingalsothattheperturbationissmallandacts
forshorttimefromt=0tot=t1,itisreasonabletoinferthatddctmisalsosmall,andhencethe
coefficientcmisobtainedas

t1icm(t1)=−eiωmntψm(0)|Hˆint|ψn(0)dt(C.6)
0wherecm(0)=δmn.Theprobabilityforthesystemtocommutefromtheinitialquantumstaten
intoanewstatemisgivenby|cm(t)|2.Thisformulaistheonsetforthederivationofthefamous
Fermisgoldenrule[178],whichcalculatesthetransitionrate(probabilityoftransitionperunit
time)fromacertainenergyeigenstateofaquantumsystemintoacontinuumofeigenstates,
causedundertheactionofperturbation.Thisruleapplieswhenthereisnodepletionofthe
initial-statepopulation.Fermisgoldenrulehastheform

Tn→f=2πδ(Ef−En)|f|Hˆpert|n|2ρ(C.7)
Intheformulaabovenistheinitialstate,fdesignatesthecontinuumofstates,andρshowsthe
states.finaltheofdensityFormulaC.6isirrelevantofthenatureoftheperturbationapplied.Theinteractionoflightwith
quantumsystemsisrationalizedintheframeofthesemiclassicaltheory,whereinlightistreated
asanelectromagneticwave.Inmostcases,theinteractionbetweenlightandatoms/molecules
canbevieweduponasaninteractionbetweenanelectromagneticwaveandthedipoleitinduces
inmolecules.Thisistheso-calledelectricdipoleapproximation.TheinteractionHˆinthasthe
form

MatterandLightbetweenInteraction

191

Hˆint=−µ∙E(t)(C.8)
E(t)representstheelectric-fieldvectorasafunctionoftime.Itisconvenienttorepresentatime-
varyingelectricfieldasasuperpositionofplane-polarizedwaveswithdifferentfrequencies,ωi,
time-dependentamplitudes,fi(t),andplanesofpolarizatione.Theexpansionintermsofplane
wavesisdescribedbythenextformula

∞E(t)=eifi(t)e−iωit(C.9)
1=iOnthebasisofformulaeC.6andC.7,thefollowingimportantforspectroscopyconclusionscan
wn.drabe•Anopticalone-photontransitionispossibleonlywhentheenergyofthephotonequals
theenergydifferencebetweenthetwolevels,nandm.Thisistheso-calledresonance
.condition•Iftheresonenceconditionismet,theintensityofthetransitiondependsonthematrix
element|ψm(0)|Hˆint|ψn(0)|2.Therearetwocaseswhenatransitionmayberesonance-
allowedbuttheintensityofthetransitionmaybezero,orverysmall.i)Whentheinterac-
tionHamiltonianHintisverysmallduetothesmallinduceddipolemomentµtheabove
matrixelementisalsosmall.ii)Becauseofmolecular-symmetryconsiderations,thema-
trixelementmayvanishevenincaseswhentheinduceddipolemomentisnotzero.This
laysthegroundfortheestablishmentofselectionrules,i.e.,thereareallowedandfor-
biddentransitions[176].Theparticularselectionrulesthatapplytorotationallyresolved
spectroscopyarediscussedinAppendixD.

192

Interaction

between

Light

and

Matter

ADppendix

MolecularCharacteristicsQuantumandMolecularMechanicalStructure

Moleculesarecomplexquantum-mechanicalobjects.Fortheunderstandingoftheirstructures
andproperties,theentangledinteractionsbetweentheirconstituentshavetobeexplained.The
Hamiltonianofamoleculewithmnucleiandnelectronsis

Hˆ=HˆN+HˆE+HˆNN+HˆEE+HˆNE(D.1)
InthisformulaHˆNisthekineticenergyofthenuclei,HˆEisthekineticenergyoftheelectrons,
HˆNNistheinteractionenergyofthenuclei,HˆEEstandsfortheinetarctionbetweentheelectrons,
andthelasttermdescribestheinteractionbetweenthenucleiandtheelectrons.
ThesubstitutionoftheaboveHamiltonianintotheSchr¨odingerequation(seeAppendixC),and
thesolutionofthelatter,inprinciple,yieldsalltheinformationontheenergylevelsandwave
functionsofthismolecule.DuetothecomplicatedformoftheHamiltonian,however,theso-
producedSchr¨odingerequationdoesnotrenderitselftoadirectsolution.Thatiswhy,some
reasonableassumptionshavetobemadeinordertomakethesolutiontractable.

D.1Born-OppenheimerApproximation

Themoststraightforwardandwidely-usedapproximation,whentreatingmoleculesquantum-
mechanically,istheBorn-Oppenheimerapproximation[179].Electronsare1837timeslighter
thanprotons/neutrons.Thismeansthatforasmalltomedium-sizedmoleculecontainingseveral
tensofatoms,theratiobetweenthemassoftheelectronsandthatofthenucleiisintherange
10−3−10−5.Duetothehigher,massnucleonsareintrinsicallymoreinertthanelectrons,and
193

194MolecularQuantumMechanicalCharacteristicsandMolecularStructure

hencetheirresponsetoforcesisslower.Thisimpliesthatateveryinstantoftimetheelectronsin
amoleculeexperienceastaticpotentialcreatedbythenuclei.Thus,theelectronicwavefunction
reactsadiabaticallytoanychangeinthenuclearconfiguration,i.e.,thenuclearconfigurationis
encodedintheelectronicwavefunction.Actually,theequilibriumconfigurationoftheatomic
nucleiinamoleculeistheonethatyieldsminimumofthesumoftheenergiesresultingfrom
thetheelectrons,interactionENE.betweenThethebindingnuclei,ofENNatoms,tobetweenformathemoleculeelectrons,EbringsEE,andforththebetweenelectronicthenucleienerandgy
ofamolecule,Eel.Theatomicnucleiinamoleculeoscillateabouttheirequilibriumpositions,
sotheseoscillationshavevibrationalenergy,Evib.Moleculesalsorotatewithrotationalenergy,
Erot.Whenconsideringmolecularproperties,theoveralltranslationofthemoleculeisirrelevant,
andhencetranslationenergyisomitted.
Thus,theBorn-Oppenheimerapproximationstatesthattheelectronicmotions,thevibrations,
andtherotationsofamoleculecanbetreatedseparately,andthetotalenergyofamolecule
(excludingitstranslationalcomponent)isthesumofitselectronic(thepotentialthatkeepsthe
atomsboundtogether),vibrational,androtationalenergies:

E=Eel+Evib+Erot

(D.2)

Thereisarelationbetweenthemagnitudesoftheelectronic,vibrational,androtationalenergies:

mmErot≈MEvib≈MEel(D.3)
Theenergiesdifferbytwo-threeordersofmagnitudefromeachotherandthattheelectronic
energyconstitutesthelargestcontributiontothetotalenergy.
ItisimportanttopointoutthattheBorn-Oppenheimerapproximationisnotalwaysvalid.There
arecaseswhichcannotbedescribedcorrectlywithinthisapproximation.Suchcasesarere-
ferredtoastheBorn-Oppenheimerapproximationbreak-down.TypicalexamplesoftheBorn-
Oppenheimerapproximationbreak-downarethevibration-rotationinteractionthroughtheCori-
oliscoupling[180],ortheinteractionbetweentheelectronicwavefunctionsandthevibrations,
whichisknownastheHerzberg-Tellereffect[181].

MoleculesofRotation.2D

Theanalysisandinterpretationofrotationallyresolvedmolecularspectranecessitatescognizance
ofthetheoryofthemolecularrotation.Thistheoryrelatesthestructureofmoleculeswiththeir
spectra.rotational

ofRotation.2DMolecules

D.2.1MolecularRotationalEnergy

195

Ifoneassumesthattheinteratomicdistancesinamoleculearefixedtosomevibrationallyav-
eragedvalues,thenthemoleculecanbetreatedasarigidbodyandhencetheenergylevels
correspondingtoitsoverallrotationscanbecalculated.Theassumptionofrigidityiswelljusti-
fiedinmanycases.When,however,thisassumptionfailsduetomoleculardistortions,thelatter
areaccountedforbyanexplicitinclusionofdistortioncoefficients.
EveryrigidmoleculeischaracterizedbyitstensorofinertiaIˆ[182,183]

IxxIxyIxz

IzxIzyIzz
Iˆ=IyxIyyIyz
Thematrixelementsaredefinedinthefollowingway

(D.4)

Ixx≡mi(yi2+zi2)etc.,Ixy≡mixiyietc.(D.5)
iiwherexi,yi,andziarethecoordinatesofatomiwithrespecttoanarbitrarycoordinatesystem,
andmiistheatomicmass.Whenthistensorisdiagonalized(alloff-diagonalelementsIxybecome
zeros),principalitaxyieldsesoftheinertia,threea,b,principalandc,momentsrespectivofely.inertiaThe,principaldesignatedaxasesIofa,Ib,inertiaandIcare,aboutlabeledthesothreethat
inequalitytheproduceto

Ia≤Ib≤Ic(D.6)
Withrespecttotheirprincipalmomentsofinertiamoleculescanbedividedintofourgroups.
•Sphericaltop.Ia=Ib=Ic,forinstanceCH4
•Prolatesymmetrictop.Ia<Ib=Ic,forinstanceCH3Br.Linearmoleculesconstitutea
particularcaseofprolatesymmetrictops,whenIa=0andIb=Ic.Alldiatomicmolecules
belongtothisclassofmolecules.
•Oblatesymmetrictop.Ia=Ib>Ic
•Asymmetrictop.IaIbIc,mostoflargermolecules
TherotationalkineticenergyHamiltonianofarigidmoleculecanbeexpressedthroughitstensor
ofinertiaandangularvelocity

196

MolecularQuantumMechanicalCharacteristicsandMolecularStructure

1Hˆrot=2Iijωiωj

(D.7)

Intheprincipalcoordinatesystem,therotationsaboutthethreeprincipalaxesofinertiaarede-
coupledandthetotalrotationalkineticenergybecomesthesumoftherotationalkineticenergies
esaxthreetheabout

Hˆrot=1(Iaωa2+Ibωb2+Icωc2)=APa2+BPb2+CPc2(D.8)
2

ThecoefficientsA,B,andCarecalledrotationalconstants,andobeytheinequalityA≥B≥C.
Theyareinverselyproportionaltothemomentsofinertia,andaredefinedbytheexpressions

222
A=2IaB=2IbC=2Ic

(D.9)

Asanalysiswillbeofshownrotationallyintheresolvedsubsequentspectra,chapters,andtherotationalassignementconstantsofmolecularplayanstructuresimportantonroletheinbasisthe
ofspectroscopicdata.ThequantitiesPα,α=a,b,c,aretheangularmomentaofthemolecule
relativetoitsprincipalaxesofinertia.

D.2.2MolecularRotationalEnergyLevels

Toobtaintheenergylevelsofarigidmolecule,itisconvenienttorepresentthequantum-
mechanicalHamiltonianthroughtheangularmomentaofthemoleculeasshowninthelast
equalityofEq.D.8.Severalimportantpropertiesandrelationsoftherigid-moleculeangular
momentaarehighlightedinthefollowing.Thedetailedderivationofthesepropertiesandrela-
tionscanbefound,forinstance,in[176,184,185].LetXYZbeaspace-fixedcoordinatesystem
withitsorigininthecenterofmassofthemolecule,andabcbethecoordinatesystemofthe
principalaxesofinertiaofthemolecule.Thenthefollowingrelationsbetweentheprojectionsof
thesquaredangularmomentum,Pˆ2,ontotheaxesX,Y,andZ,anda,b,andc,andtherotational
holdHamiltonianHrot

MoleculesofRotation.2D

Pˆ2=Pˆa2+Pˆb2+Pˆc2=PˆX2+PˆY2+PˆZ2
[PˆX,PˆY]=iPˆZetc.
[Pˆa,Pˆb]=−iPˆcetc.
[Pˆ2,Pˆc]=0etc.
[Pˆ2,PˆZ]=0etc.
[PˆZ,Pˆc]=0etc.
[Hˆrot,Pˆ2]=0etc.
[Hˆrot,PˆZ]=0etc.
Itcanbeshownthatforeveryrotor,thereexistsomefundamentalrelations

197

(D.10)(D.11)(D.12)(D.13)(D.14)(D.15)(D.16)(D.17)

Pˆ2ψ=J(J+1)2ψ,J=0,1,2,...(D.18)
PˆZψ=Kψ,K=0,±1,...,±J(D.19)
where√J(J+1)isthemagnitudeofthetotalrotationalangularmomentumandKisits
componentalongaspace-fixedaxis.
Theenergylevelsforthedifferenttypesofrotors,sphericaltop,symmetrictop,andasymmetric
top,arelistedbelow.

Spherical-TopEnergyLevels
E=AJ(J+1)
whereA=B=Cistherotationalconstantofthemolecule.

Symmetric-TopEnergyLevels

E(J,K)=BJ(J+1)+(C−B)K2oblatetop
J=0,1,2,...K=0,±1,±2,...,±J
E(J,K)=BJ(J+1)+(A−B)K2prolatetop
J=0,1,2,...K=0,±1,±2,...,±J
E=BJ(J+1)linear

(D.20)

(D.21)(D.22)(D.23)

198MolecularQuantumMechanicalCharacteristicsandMolecularStructure

Asseenfromtheaboveformulae,therotationalenergyofsymmetrictopsdependsnotonlyonJ
butalsoonasecondquantumnumber,K,whichdeterminestheprojectionsofthetotalangular
momentumalongamolecule-fixedaxisofthesymmetrictop.

Asymmetric-TopEnergyLevels

Theformulaefortheenergylevelsofasymmetric-topmoleculesarequitecomplicated,andthe
energylevelscanbeobtainedanalyticallyonlyforsmallvaluesofJ.ForlargevaluesofJ,the
energylevelsarecalculatednumerically.Animportantparameterintheanalysisofasymmetric
rotorsistheRaysasymmetryparameter,definedas

κ=2B−A−C(D.24)
CA−Thisparametershowsthedeviationoftheparticularasymmetrictopfromeithertheprolateorthe
oblatesymmetrictop.Theenergylevelsofanasymmetrictopare,usually,designatedasJKprKob,
whereKprandKobdesignatethethevaluesofKfortheprolateandoblatesymmetrictops,
respectively,thatcorrelatewiththeasymmetry-toplevelunderconsideration.Itisimportantto
emphasize,however,thatKprandKobarenottruequantumnumbersfortheasymmetrictop.An
alternativenotationfortheasymmetricenergylevelsisJτ,whereτ=Kpr−Kob.
Thefollowingtable(takenfromRef.[183])showstheanalyticalsolutionsforthefirstfewenergy
levelsofanasymmetrictop.

JKprKobτ=Kpr−KobE(J,Kpr,Kob)
000001101A+B
1110A+C
101-1B+C
22022A+2B+2C+2(B−C)2+(A−C)(A−B)
22114A+B+C
2110A+4B+C
212-1A+B+4C
202-22A+2B+2C−2(B−C)2+(A−C)(A−B)
TableD.1:Rigidasymmetric-topenergylevels

.2DMoleculesofRotation

D.2.3ElectricDipoleTransitionMoment

199

Itisproved[176]intheelectricdipoleapproximation(seeChap.C)thatthetransitionprobability
ofamoleculefromstatentostatemundertheactionofplane-polarized(inthexdirection)
monochromaticlightwithelectromagneticenergydensityuxatfrequencyνmnforatimetis

|cmn(T)|2=2π−2T|m|µˆx|n|2ux(νmn)(D.25)
Intheaboveequationµxisthexcomponentofthemoleculardipolemoment.Thematrixelement
|m|µˆx|n|canbewritteninthemoregeneralform

|m|µˆ|n|2=|m|µˆx|n|2+|m|µˆy|n|2+|m|µˆz|n|2(D.26)
whereµˆ=iµˆx+jµˆy+kµˆz.|µˆiscalledtransitiondipolemomentortransitionmoment.The
transitionmomentisavectorquantity,anditisusuallydefinedwithrespecttotheprincipalaxes
ofinertiaofthemolecule.Dependingonwhetherthetransitiondipolemomentisorientedalong
thea,b,orcprincipalaxisofinertia,onedistinguishesbetweenthreetypesoftransitions,a−,
b−,andc−type.

RulesSelection.2.4D

Notalltransitionsbetweentheenergylevelsofanasymmetrictoparepossible.One-photon
transitionsareallowedonlyfortransitionsforwhichΔJ=0,±1.Inasetoftransitionsthosefor
whichΔJ=−1formthePbranch,thoseforwhichΔJ=0giverisetotheQbranch,andΔJ=+1
formtheRbranch.Fortwo-photontransitions,theselectionrulesaredifferentfromthosefor
one-photontransitions.InthiscaseΔJ=−2,+2,correspondingtoOandSbranches[186].

IntensityransitionT.2.5D

Rotationaltransitionintensitiesaredeterminedbystatisticalweightsdependingonthedegener-
acyinthequantumnumberK,nuclearspin,thepolarizationofthelightinducingthetransition,
andonthethermaldistributionofthespeciesgivenbytheBoltzmannformula

expEΔmn−Tkrot

(D.27)

200MolecularQuantumMechanicalCharacteristicsandMolecularStructure

whereΔEmnistheenergydifferencebetweenlevelsmandn,kistheBoltzmannconstant,andT
.temperaturerotationaltheis

D.2.6DeviationsfromtheBorn-OppenheimerApproximation

ThediscussionhithertowasonrigidmoleculesobeyingtheBorn-Oppenheimerapproximation
(seeSec.D.1).Sometimes,however,theseapproximationsbreakdown,andonehastoeleborate
differentmodelstotreatthemolecularbehaviorproperly.Themosttypicaldeviationsfromthe
aboveapproximationsarelistedhere.
•Centrifugaldistortion.WhenconsideringhighvaluesofJ,molecularbondsstretchand
hencemoleculescanbenolongertreatedasrigidobjects.Inthiscase,theirrotational
energylevelsarecalculatedbyincludingintherespectiverigid-molecule-approximation
formulaeacorrectionforthecentrifugaldistortionexpressedbycentrifugaldistortioncon-
184][176,stants•Corioliscoupling.TheCorioliscouplinginmoleculesresultsfromthecouplingbetween
themolecularvibrationsandtheoverallrotationofthemolecule.TheCoriolis-coupling
correctionintheBorn-Oppenheimerapproximationisintroducedthroughthetheoryof
perturbations.•Herzberg-Tellereffect.TheHerzberg-Tellereffect[181]isthegainofintensitybyan
electronicallyforbiddenbutvibronicallyallowedtransitionfromanotherbothelectroni-
callyandvibronicallyallowedtransitionthroughvibrationalcoupling.
•Jahn-Tellereffect.Ifatacertainsymmetricalnonlinearnuclearconfigurationofapoly-
atomicmolecule,twoelectronicstatesaredegenerate,thelattercanbeliftedbysome
nucleardistortion.Thesplittingoftheelectronicdegeneracybyanucleardistrosionac-
companiedbytheinteractionbetweentherotation-vibrationlevelsofthetwoelectronic
statesisreferredtoastheJann-Tellereffect[187].

D.3RotationalConstantsandStructureofMolecularSpecies

Fromtherotationallyresolvedspectroscopyonecanderivethevaluesofmolecularrotational
constants,whicharerelatedtotheprincipalmomentsofinertiaofthestudiedspecies.Unfortu-
nately,rotationalconstantsdonotprovideuniqueinformationonthemolecularstructuresince
itmayhappenthatdifferentstructuresmayhaveverysimilarrotationalconstants.Thatiswhy,
itisnecessarytofindamethodtoassignstructuresonthebasisoftheexperimentallyobserved
constants.rotational

D.3RotationalConstantsandStructureofMolecularSpecies

201

Molecularstructuresaredeterminedbydefiningtheinteratomicdistances,planar,anddihe-
dralangles.Regardingbondlengths,thefollowingbond-lengthdefinitionshavetobedistin-
[183].guished•reisthebondlegththatcorrespondstoahypotheticvibrationlessstateofthemolecule.
•r0istheeffectivebondlengthcorrespondingtoaground-statevibration.Itisthesedis-
tancesthatareinvolvedinthemeasuredrotationalconstants.
•rsisthebondlengthcorrespondingtoanisotopicsubstitution.Theso-definedbondlength
isdiscussedinconjuctionwiththeKraitchmanequations[188]
•risanaveragedbondlengthcorrespondingtosomeatompositionsresultingfrompartial
correctiontothevibrationaleffects.
•rmisthemass-weightedbondlengthderivedfromtheaveragingofthebondlengthsofa
largenumberofisotopicallysubstitutedspecies.
Kraitchmanequations[183,188,189]areaconvenientanalyticaltoolfordeterminingthecoor-
dinatesofanisotopicallysubstitutedatominarigidmoleculewithrespecttothecenter-of-mass
(COM)principalaxissystemoftheparentmolecule(themoleculebeforetheisotopicsubstitu-
tion).Kraitchmanequationsallowthecoordinatesofthesubstitutedatomtobefoundonthe
basisofthemeasuredrotationalconstantsfortheparticularmolecularspeciesbeforeandafter
theisotopicsubstitution.Thisapproachhasbeensuccessfullyappliedtotheanalysisofmany
noble-gasclustersofbenzeneandbenzenederivatives[79,190,191].Thecoordinatesofaniso-
topicallysubstitutedatomwithrespecttotheprincipal-axissystemoftheparentmoleculeare
formulaethebycalculated

|x|=ΔPx1+ΔPy1+ΔPz
µPy−PxPz−Px
|y|=ΔPy1+ΔPz1+ΔPx
µPz−PyPx−Py
|z|=ΔPz1+ΔPx1+ΔPy
µPx−PzPy−Pz

whereµisthereducedmassofthemolecularspecies

mΔMµ=M+Δm

(D.28)

(D.29)

202MolecularQuantumMechanicalCharacteristicsandMolecularStructure

Misthemassofthemoleculebeforetheisotopicsubstitution,andΔmisdifferenceofthemasses
substitution.thebeforePx,Py,andPzinEq.D.3aredefinedas

(D.30)

1Px=2(−Ix+Iy+Iz)
Py=1(−Iy+Iz+Ix)(D.30)
21Pz=(−Iz+Ix+Iy)
2Ix,IyandIzdesignatetheprincipalmomentofinertiaoftheunsubstitutedmolecule.Differences
ΔofparameterPinEq.D.3aredefinedas

(D.31)

ΔPx=Px−Px
ΔPy=Py−Py(D.31)
ΔPz=Pz−Pz
Theprimedparametersintheabovedefinitionscorrespondtothemolecularspeciesafterthe
substitution.isotopic

.3D

Rotational

Constants

and

eStructur

of

Molecular

Species

203

204

Molecular

Quantum

Mechanical

Characteristics

and

Molecular

eStructur

yBibliograph

[1]F.A.CareyandR.J.Sundberg,AdvancedOrganicChemistry.PartA:Structureand
Mechanisms,5ed.(Springer,US,2007),p.169ff.
[2]A.Scha¨fflerandN.Mencke,BiologieAnatomiePhysiologie,DasNervengewebe,4ed.
(UrbanandFischer,ADDRESS,2000),Chap.10,p.145ff.
[3]E.G.RobertsonandJ.P.Simons,Phys.Chem.Chem.Phys.3,1(2001).
[4]T.S.Zwier,J.Phys.Chem.A105,8827(2001).
[5]G.Scoles,Atomicandmolecularbeammethods(OxfordUniversityPress,NewYork,
1988).[6]R.N.PribbleandT.S.Zwier,Science265,75(1994).
[7]T.Ebata,A.Fujii,andN.Mikami,Int.Rev.Phys.Chem.17,331(1998).
[8]R.M.HelmandH.J.Neusser,Chem.Phys.239,33(1998).
[9]M.Gerhards,B.Kimpfel,M.Pohl,M.Schmitt,andK.Kleinermanns,J.Mol.Struct.270,
(1992).301[10]H.J.NeusserandK.Siglow,Chem.Rev.100,3921(2000).
[11]R.M.Helm,M.Clara,T.L.Grebner,andH.J.Neusser,J.Phys.Chem.A102,3268
(1998).[12]P.F.Barbara,T.J.Meyer,andM.A.Ratner,J.Phys.Chem.100,13148(1996).
[13]J.R.Carney,F.C.Hagemeister,andT.S.Zwier,J.Chem.Phys.108,3379(1998).
[14]J.R.CarneyandT.S.Zwier,J.Phys.Chem.A103,9943(1999).
[15]S.GeorgievandH.J.Neusser,Chem.Phys.Lett.389,24(2004).
[16]M.Mons,E.G.Robertson,L.C.Snoek,andJ.P.Simons,Chem.Phys.Lett.310,423
(1999).

205

206

BIBLIOGRAPHY

[17]J.P.Reilly,MassSpectrometryReviews28,425(2009).
[18]B.Lucas,M.Barat,J.A.Fayeton,M.Perot,C.Jouvet,G.Gregoire,andS.B.Nielsen,
TheJournalofChemicalPhysics128,164302(2008).
[19]J.A.Stearns,S.Mercier,C.Seaiby,M.Guidi,O.V.Boyarkin,andT.R.Rizzo,Journalof
theAmericanChemicalSociety129,11814(2007).
[20]S.Chervenkov,P.Q.Wang,J.E.Braun,andH.J.Neusser,J.Chem.Phys.121,7169
(2004).[21]D.H.Levy,Ann.Rev.Phys.Chem.31,197(1980).
[22]N.F.Ramsey,MolecularBeams,InternationalSeriesofMonographsonPhysics(Oxford
UniversityPress,NewYork,1985).
[23]I.Holvan,Diplomarbeit,AufbaueinerMolekularstrahlapparaturzurErzeugungultra-
kalterMolek¨ulclusterf¨urdieSub-Doppler-Laserspektroskopie(1992).
[24]R.M.Helm,Diplomarbeit,Hochau߬osendeSpektroskopiedurchVier-Wellen-Mischen
(1995).-Photonen-IonisationMehrresonanteund[25]K.Siglow,Diplomarbeit,Fakult¨atf¨urChemie,BiologieundGeowissenschaftenderTech-
nischenUniversit¨atM¨unchen(1996).
[26]R.Gerthner,Diplomarbeit,Hochau߬osendeSub-Dopper-UVSpektroskopiezurUnter-
suchungderSchwingungsdynamikinmolekularenClustern(1997).
[27]R.Sussmann,Dissertation,Koh¨arenteIonendipSpektroskopieundrotationsaufgel¨oste
UV-LaserspektroskopiezurUntersuchungderintermolekularenWechselwirkungin
Aromat-EdelgasVanderWaals-Komplexen(1994).
[28]S.Chervenkov,Dissertation,InvestigationofWeakIntra-andIntermolecularInteractions
andConformationalStructuresofFlexibleMoleculesandComplexesbyMassSelective
HighResolutionResonanceEnhancedTwoPhotonIonizationLaserSpectroscopy(2007).
[29]D.R.Miller,FreeJetSourcesinAtomicandMolecularBeamMethods(OxfordUniversity
Press,NewYork,1988),Vol.1,Chap.2.
[30]M.Knerich,Diplomarbeit,Fachbereichf¨urChemie,BiologieundGeowissenschaftender
TechnischenUniversit¨atM¨unchen(1990).
[31]T.Weber,Dissertation,Fachbereichf¨urChemie,BiologieundGeowissenschaftender
TechnischenUniversit¨atM¨unchen(1991).
[32]W.C.WileyandI.H.McLaren,Rev.Sci.Instr.26,1150(1955).

BIBLIOGRAPHY

207

[33]R.WallensteinandT.W.H¨ansch,Opt.Comm.14,353(1975).
[34]R.WallensteinandT.W.H¨ansch,Phys.Rev.A22,192(1980).
[35]M.M.Salour,Opt.Commun.22,202(1977).
[36]E.Riedle,R.Moder,andH.J.Neusser,Opt.Commun.43,388(1982).
[37]A.E.Siegman,Lasers,1ed.(UniversityScienceBooks,MillValley,CA,1986).
[38]R.Neuhauser,Diplomarbeit.Fachbereichf¨urChemie,BiologieundGeowissenschaften
derTechnischenUniversit¨atM¨unchen(1993).
[39]U.Schubert,Dissertation,Institutf¨urChemie,BiologieundGeowissenschaftenderTech-
nischenUniversit¨atM¨unchen(1988).
[40]F.Tr¨ager,HandbookofLasersandOptics(Springer,NewYork,2007),Chap.12.1,pp.
938–945.[41]R.WehrensandL.M.C.Buydens,TrendsAnalyt.Chem.17,193(1998).
[42]J.A.Hageman,R.Wehrens,R.deGelder,W.L.Meerts,andL.M.C.Buydens,J.Chem.
Phys.113,7955(2000).
[43]A.E.Nikolaev,G.Miszkiewicz,G.Berden,W.LeoMeetrs,J.F.Pfanstiel,andD.W.
Pratt,J.Chem.Phys.122,084309(2005).
[44]S.Chervenkov,P.Q.Wang,J.E.Braun,S.Georgiev,C.K.Nandi,T.Chakraborty,and
H.J.Neusser,J.Chem.Phys.122,244312(2005).
[45]M.Schmitt,D.Kr¨uger,M.B¨ohm,C.Ratzer,V.Bednarska,I.Kalkman,andW.L.Meerts,
Phys.Chem.Chem.Phys.8,228(2006).
[46]S.Chervenkov,R.Karaminkov,J.E.Braun,H.J.Neusser,S.S.Panja,andT.Chakraborty,
J.Chem.Phys.124,234302(2006).
[47]R.G.Neuhauser,K.Siglow,andH.J.Neusser,J.Chem.Phys.106,896(1997).
[48]T.L.GrebnerandH.J.Neusser,Chem.Phys.Lett.245,578(1995).
[49]W.A.Chupka,J.Chem.Phys.98,4520(1993).
[50]W.A.Chupka,J.Chem.Phys.99,5800(1993).
[51]L.ZhuandP.M.Johnson,J.Chem.Phys.94,5769(1991).
[52]H.KrauseandH.J.Neusser,J.Chem.Phys.97,5923(1992).
[53]J.E.BraunandH.J.Neusser,MassSpectrom.Rev.21,16(2002).

208

BIBLIOGRAPHY

[54]K.M¨uller-Dethlefs,M.Sander,andE.W.Schlag,Chem.Phys.Lett.112,291(1984).
[55]B.Ernstberger,H.Krause,A.Kiermeier,andH.J.Neusser,J.Chem.Phys.92,5285
(1990).[56]J.E.Braun,Dissertation,SelektiveIonisationvonWasserstoffbr¨ucken-gebundenenClus-
ternundMolek¨ulen:BindungsenergienundLadungstransfer(2007).
[57]M.F.Zigmond,FundamentalNeuroscience(AcademicPress,London,1998).
[58]S.H.Snyder,DrugsandtheBrain(ScientificAmericanLibrary,NewYork,1999).
[59]P.D.Godfrey,R.D.Brown,andL.D.Hatherley,J.Am.Chem.Soc.117,8204(1995).
[60]S.SunandE.R.Bernstein,J.Am.Chem.Soc.118,5086(1996).
[61]J.A.Dickinson,M.R.Hockridge,R.T.Kroemer,E.G.Robertson,J.P.Simons,J.Mc-
Combie,andM.Walker,J.Am.Chem.Soc.120,2622(1998).
[62]S.J.MartinezIII,J.C.Alfano,andD.H.Levy,J.Mol.Spectrosc.158,82(1998).
[63]N.Guchhait,T.Ebata,andN.Mikami,J.Am.Chem.Soc.121,5705(1999).
[64]M.HockridgeandE.G.Robertson,J.Phys.Chem.A103,3618(1999).
[65]Y.Lee,J.Jung,B.Kim,P.Butz,L.C.Snoek,R.T.Kroemer,andJ.P.Simons,J.
Phys.Chem.A108,69(2004).
[66]P.Butz,N.A.Macleod,L.C.Snoek,F.O.Talbot,andJ.P.Simons,CentralLaserFacility
AnnualReport2000/2001,97(2002).
[67]P.Butz,R.T.Kroemer,N.A.Macleod,andJ.P.Simons,J.Phys.Chem.A105,544(2001).
[68]R.Karaminkov,S.Chervenkov,P.H¨arter,andH.J.Neusser,Chem.Phys.Lett.442,238
(2007).[69]P.Butz,R.T.Kroemer,N.A.Macleod,E.G.Robertson,andJ.P.Simons,J.Phys.Chem.
(2001).1050,A105[70]P.D.Godfrey,R.N.Jorissen,andR.D.Brown,J.Phys.Chem.A103,7621(1999).
[71]R.Weinkauf,F.Lehrer,E.W.Schlag,andA.Metsala,FaradayDiscuss.115,363(2000).
[72]R.D.BrownandP.D.Godfrey,J.Phys.Chem.A104,5742(2000).
[73]S.S.PanjaandT.Chakraborty,J.Phys.Chem.A107,10984(2003).
[74]D.R.Borst,P.W.Joireman,D.W.Pratt,E.G.Robertson,andJ.P.Simons,J.Chem.
Phys.116,7057(2002).

BIBLIOGRAPHY

209

[75]G.M.FlorioandT.S.Zwier,J.Phys.Chem.A107,974(2003).
[76]M.Schmitt,M.B¨ohm,C.Ratzer,C.Vu,I.Kalkman,andW.L.Meerts,J.Am.Chem.
(2005).10356,Soc.127[77]T.V.Nguyen,J.T.Yi,andD.W.Pratt,Phys.Chem.Chem.Phys.8,1049(2006).
[78]L.Alvarez-ValtierraandD.W.Pratt,J.Chem.Phys.126,224308(2007).
[79]R.Sussmann,R.Neuhauser,andH.J.Neusser,Can.J.Phys.72,1179(1994).
[80]S.Chervenkov,P.Q.Wang,R.Karaminkov,T.Chakraborty,J.E.Braun,andH.J.Neusser,
Proc.SPIE5830,246(2005).
[81]S.Chervenkov,P.Wang,J.E.Braun,T.Chakraborty,andH.J.Neusser,Phys.Chem.
Chem.Phys.9,837(2007).
[82]M.J.Frisch,G.W.Trucks,H.B.Schlegel,G.E.Scuseria,M.A.Robb,J.R.Cheeseman,
J.Montgomery,J.A.,T.Vreven,K.N.Kudin,J.C.Burant,J.M.Millam,S.S.Iyengar,
J.Tomasi,V.Barone,B.Mennucci,M.Cossi,G.Scalmani,N.Rega,G.A.Petersson,
H.Nakatsuji,M.Haga,M.Ehara,K.Toyota,R.Fukuda,J.Hasegawa,M.Ishida,T.
Nakajima,Y.Honda,O.Kitao,H.Nakai,M.Klene,X.Li,J.E.Knox,H.P.Hratchian,
J.B.Cross,C.Adamo,J.Jaramillo,R.Comperts,R.E.Stratmann,O.Yazyev,A.J.
Austin,R.Cammi,C.Pomelli,J.W.Ochterski,P.Y.Ayala,K.Morokuma,J.A.Voth,P.
Salvador,J.J.Dannenberg,V.G.Zakrzewski,S.Dapprich,A.D.Daniels,M.C.Strain,
O.Farkas,D.K.Malick,A.D.Rabuck,K.Raghavachari,J.B.Foresman,J.V.Ortiz,Q.
Cui,A.G.Baboul,S.Clifford,J.Cioslowski,B.B.Stefanov,G.Liu,A.Liashenko,P.
Piskorz,I.Kamaromi,R.L.Martin,D.J.Fox,T.Keith,M.A.Al-Laham,C.Y.Peng,
A.Nanayakkara,M.Challacombe,P.M.W.Gill,B.Johnson,W.Chen,M.W.Wong,C.
Gonzales,andJ.A.Pople,Gaussian03,RevisionB.04,2003.
[83]J.C.Pearson,K.V.L.N.Sastry,E.Herbst,andF.C.DeLucia,Astrophys.J.480,420
(1997).[84]M.R.Hockridge,S.M.Knight,E.G.Robertson,J.P.Simons,J.McCombie,andM.
Walker,Phys.Chem.Chem.Phys.1,407(1999).
[85]T.S.Zwier,J.Phys.Chem.A110,4133(2006).
[86]H.S.Im,E.R.Bernstein,H.V.Secor,andJ.I.Seeman,J.Am.Chem.Soc.113,4422
(1991).[87]D.W.Pratt,Annu.Rev.Phys.Chem.49,481(1998).
[88]D.W.Pratt,Science296,2347(2002).

210

BIBLIOGRAPHY

[89]K.M¨uller-DethlefsandP.Hobza,Chem.Rev.100,143(2000).
[90]R.Karaminkov,S.Chervenkov,andH.J.Neusser,J.Phys.Chem.A112,839(2008).
[91]S.S.Panja,P.Biswas,andT.Chakraborty,Chem.Phys.Lett.411,128(2005).
[92]W.Dietz,H.J.Neusser,U.Boesl,E.W.Schlag,andS.H.Lin,Chem.Phys.66,105
(1982).[93]P.Sinha,S.E.Boesch,C.Gu,R.A.Wheeler,andA.K.Wilson,J.Phys.Chem.A108,
(2004).9213[94]J.P.Merrick,D.Moran,andL.Radom,J.Phys.Chem.A111,11683(2007).
[95]R.Karaminkov,S.Chervenkov,andH.J.Neusser,workinpreparation.
[96]C.M.Western,PGOPHER,AProgramforSimulatingRotationalStructure,2007.
[97]G.Herzberg,Molecularspectraandmolecularstructure,III.Electronicspectraandelec-
tronicstructureofpolyatomicmolecules(KriegerPublishingCompany,Malabar,1991),
III.ol.V[98]L.C.Snoek,E.G.Robertson,R.T.Kroemer,andJ.P.Simons,Chem.Phys.Lett.321,49
(2000).[99]X.Tong,M.S.Ford,C.E.H.Dessent,andK.M¨uller-Dethlefs,J.Chem.Phys.119,12908
(2003).[100]J.Yao,H.S.Im,M.Foltin,andE.R.Bernstein,J.Phys.Chem.A104,6197(2000).
[101]C.E.H.Dessent,W.D.Geppert,S.Ullrich,andK.M¨uller-Dethlefs,Chem.Phys.Lett.
(2000).375,319[102]J.E.Braun,T.Mehnert,andH.J.Neusser,Int.J.Mass.Spectrom.203,1(2000).
[103]S.Georgiev,T.Chakraborty,andH.J.Neusser,J.Phys.Chem.A108,3304(2004).
[104]R.Karaminkov,S.Chervenkov,andH.J.Neusser,Phys.Chem.Chem.Phys.10,2852
(2008).[105]M.J.Frisch,G.W.Trucks,H.B.Schlegel,G.E.Scuseria,M.A.Robb,J.R.Cheeseman,
J.Montgomery,J.A.,T.Vreven,K.N.Kudin,J.C.Burant,J.M.Millam,S.S.Iyengar,
J.Tomasi,V.Barone,B.Mennucci,M.Cossi,G.Scalmani,N.Rega,G.A.Petersson,
H.Nakatsuji,M.Haga,M.Ehara,K.Toyota,R.Fukuda,J.Hasegawa,M.Ishida,T.
Nakajima,Y.Honda,O.Kitao,H.Nakai,M.Klene,X.Li,J.E.Knox,H.P.Hratchian,
J.B.Cross,C.Adamo,J.Jaramillo,R.Comperts,R.E.Stratmann,O.Yazyev,A.J.
Austin,R.Cammi,C.Pomelli,J.W.Ochterski,P.Y.Ayala,K.Morokuma,J.A.Voth,P.

BIBLIOGRAPHY

211

Salvador,J.J.Dannenberg,V.G.Zakrzewski,S.Dapprich,A.D.Daniels,M.C.Strain,
O.Farkas,D.K.Malick,A.D.Rabuck,K.Raghavachari,J.B.Foresman,J.V.Ortiz,Q.
Cui,A.G.Baboul,S.Clifford,J.Cioslowski,B.B.Stefanov,G.Liu,A.Liashenko,P.
Piskorz,I.Kamaromi,R.L.Martin,D.J.Fox,T.Keith,M.A.Al-Laham,C.Y.Peng,
A.Nanayakkara,M.Challacombe,P.M.W.Gill,B.Johnson,W.Chen,M.W.Wong,C.
Gonzales,andJ.A.Pople,Gaussian03,RevisionE.01,2003.
[106]Y.Zhao,N.E.Schultz,andD.G.Truhlar,J.Chem.Phys.123,161103(2005).
[107]M.D.PateyandC.E.H.Dessent,J.Phys.Chem.A106,4623(2002).
[108]G.M.Florio,R.A.Christie,K.D.Jordan,andT.S.Zwier,J.Am.Chem.Soc.124,10236
(2002).[109]E.Nir,C.Janzen,P.Imhof,K.Kleinermanns,andM.S.deVries,J.Chem.Phys.115,
(2001).4604[110]E.Nir,M.M¨uller,L.I.Grace,andM.S.deVries,Chem.Phys.Lett.355,59(2002).
[111]J.C.Lopez,V.Cortijo,S.Blanco,andJ.L.Alonso,Phys.Chem.Chem.Phys.9,4521
(2007).[112]W.L.Meerts,M.Schmitt,andG.C.Groenenboom,Can.J.Chem.82,804(2004).
[113]W.L.MeertsandM.Schmitt,Int.Rev.Phys.Chem.25,353(2006).
[114]R.Karaminkov,S.Chervenkov,H.J.Neusser,V.Ramanathan,andT.Chakraborty,The
JournalofChemicalPhysics130,034301(2009).
5-9-1)..(VTURBOMOLE,[115][116]O.Christiansen,H.Koch,andP.Jørgensen,Chem.Phys.Lett.243,409(1995).
[117]C.H¨attigandF.Weigend,J.Chem.Phys.113,5154(2000).
[118]M.Becucci,G.Pietraperzi,M.Pasquini,G.Piani,A.Zoppi,R.Chelli,andE.Castellucci,
J.Chem.Phys.120,5601(2004).
[119]P.Butz,R.T.Kroemer,N.A.Macleod,andJ.P.Simons,Phys.Chem.Chem.Phys.4,
(2002).3566[120]G.H.Ragab,H.Nohta,andK.Zaitsu,AnalyticaChimicaActa403,155(2000).
[121]M.-H.E.Spyridaki,C.J.Tsitsimpikou,P.A.Siskos,andC.G.Georgakopoulos,Journal
ofChromothographyB758,311(2001).
[122]G.A.JeffreyandW.Saenger,HydrogenBondinginBiologicalStructures(Springer,
1991).Berlin,

212

BIBLIOGRAPHY

[123]T.L.BarryandG.Petzinger,BiologicalMassSpectrometry4,129(1977).
[124]Y.MardeandR.Ryhage,ClinicalChemistry24,1720(1978).
[125]F.Brazzarola,M.Tubaro,P.Bravo,M.Zanda,A.Volonterio,andP.Traldi,IIFarmaco
(2003).69,58[126]U.Boesl,H.J.Neusser,andE.W.Schlag,J.Chem.Phys.72,4327(1980).
[127]H.K¨uhlewind,A.Kiermeier,andH.J.Neusser,J.Chem.Phys.85,4427(1986).
[128]C.S.Parmenter,inAdvancesinChemicalPhysiscs,1sted.,editedbyI.Prigogineand
S.A.Rice(Wiley-Interscience,NewYork,1978),Vol.22.
[129]R.H.Page,Y.R.Shen,andY.T.Lee,J.Chem.Phys.88,4621(1988).
[130]D.R.Borst,P.W.Joireman,D.W.Pratt,E.G.Robertson,andJ.P.Simons,J.Chem.Phys.
(2002).7057,116[131]A.Boatwright,N.A.Besley,S.Curtis,R.R.Wright,andA.J.Stace,J.Chem.Phys.123,
(2005).021102[132]B.C.Dian,A.Longarte,S.Mercier,D.A.Evans,D.J.Wales,andT.S.Zwier,
J.Chem.Phys.117,10688(2002).
[133]G.C.Maitland,M.Rigby,E.B.Smith,andW.A.Wakeham,Intermolecularforces,The
InternationalSeriesofMonographsonChemistry,2nded.(ClarendonPress,Oxford,
1987).[134]F.London,Z.Phys.63,245(1930).
[135]W.HeitlerandF.London,Z.Physik44,455(1927).
[136]P.M.Morse,Phys.Rev.34,57(1929).
[137]J.E.Lennard-Jones,Proc.Phys.Soc.43,461(1931).
[138]R.M.Helm,Hochaufgel¨osendeUV-Laserspektroskopie:StrukturundDynamiknachelek-
tronischerAnregungvonwasserstoffbr¨uckengebundenenAromat-WasserKomplexen(Her-
bertUtzVerlag,M¨unchen,1998).
[139]M.J.S.Dewar,Nature(London)156,784(1945).
[140]E.Carrasquillo,T.S.Zwier,andD.H.Levy,J.Chem.Phys.83,4990(1985).
[141]M.Nishio,M.Hirota,andY.Umezawa,TheCH/I¨Interaction:Evidence,Natureand
Consequences(Wiley-VCH,NewYork,1998).
[142]S.Djafari,H.-D.Barth,K.Buchhold,andB.Brutschy,J.Chem.Phys.107,10573(1997).

BIBLIOGRAPHY

213

[143]A.Fujii,S.Morita,M.Miyazaki,T.Ebata,andN.Mikami,J.Phys.Chem.A108,2652
(2004).[144]L.P.Hammett,PhysicalOrganicChemistry(McGrawHill,NewYork,1940),p.294.
[145]L.J.Andrews,Chem.Rev.54,713(1954).
[146]M.C.Etter,Acc.Chem.Res.23,120(1990).
[147]M.C.Etter,J.C.McDonald,andJ.Bernstein,ActaCryst.B46,256(1990).
[148]M.C.EtterandS.M.Reutzel,J.Am.Chem.Soc.113,2586(1991).
[149]L.R.Hanton,C.A.Hunter,andD.H.Purvis,J.Chem.Soc.Chem.Commun.1134(1992).
[150]N.Sheppard.,I.D.Hadzi,andE.H.W.Thompson,HydrogenBonding(PergamonPress,
85105.p.1959),London,[151]E.Reed,Alan,A.Curtiss,Larry,andF.Weinhold,Chem.Rev.88,899(1988).
[152]P.HobzaandZ.Havlas,Chem.Rev.100,4253(2000).
[153]P.Hobza,V.Spirko,H.L.Selzle,andE.W.Schlag,J.Phys.Chem.A102,2501(1998).
[154]P.Hobza,V.Spirko,Z.Havlas,K.Buchhold,B.Reimann,H.-D.Barth,andB.Brutschy,
Chem.Phys.Lett.299,180(1999).
[155]G.N.Patwari,T.Ebata,andN.Mikami,J.Phys.Chem.105,8642(2001).
[156]C.LeeJr.,Jesse,P.Eduardo,L.Rheingold,Arnold,andH.Crabtree,Robert,J.Am.Chem.
(1994).11014,Soc.116[157]Q.LiuandR.Hoffmann,J.Am.Chem.Soc.117,10108(1995).
[158]W.Kauzmann,Adv.Prot.Chem.14,1(1959).
[159]D.Chandler,Nature437,640(2005).
[160]P.HobzaandR.Zahradnik,Chem.Rev.88,871(1988).
[161]J.L.CalaisandE.S.Kryachko,Conceptualperspectivesinquantumchemistry,1sted.
1997).Amsterdam,Netherland,(Springer[162]A.R.Leach,Molecularmodellingprinciplesandapplications(AddisonWesley,Sussex,
1996).[163]I.N.Levine,QuantumChemistry,2ndindianreprinted.(PearsonEducation,Singapore,
2004).[164]C.MøllerandM.S.Plesset,Phys.Rev.40,618(1934).

214

BIBLIOGRAPHY

[165]R.J.Bartlett,Ann.Rev.Phys.Chem.32,359(1981).
[166]P.HohenbergandW.Kohn,Phys.Rev.136,864(1964).
[167]W.KohnandL.J.Sham,Chem.Rev.140,1133(1965).
[168]C.Lee,W.Yang,andR.G.Parr,Phys.Rev.B37,785(1988).
[169]Y.Zhao,N.E.Schultz,andD.G.Truhlar,J.Chem.TheoryComput.2,364(2006).
[170]Y.ZhaoandD.G.Truhlar,J.Phys.Chem.A109,5656(2005).
[171]S.F.BoysandF.Bernardi,Mol.Phys.19,553(1970).
[172]P.S.Sedano,Dissertation,ImplementationandApplicationofBasisSetSuperposition
Error-correctionSchemestotheTheoreticalModelingWeakIntermolecularInteractions
(2001).[173]E.B.WilsonJr,J.C.Decius,andP.C.Cross,MolecularVibrations,1955ed.(Dover
Publications,NewYork,1955).
[174]H.J.Werner,P.J.Knowles,R.Lindh,F.R.Manby,andM.Sch¨utz,MOLPRO,version
2006.1,apackageofabinitioprograms.
[175]R.Ahlrichs,M.B¨ar,M.H¨aser,H.Horn,andC.K¨olmel,Chem.Phys.Lett.162,165
(1989).[176]I.N.Levine,MolecularSpectroscopy,1sted.(Wiley-Interscience,NewYork,1975).
[177]S.H.Lin,Y.Fujimura,H.J.Neusser,andE.W.Schlag,(1984).
[178]E.Fermi,NuclearPhysics(ChicagoUniversityPress,Chicago,1950).
[179]M.BornandR.Oppenheimer,Ann.Physik84,457(1927).
[180]E.Teller,Hand-u.Jahrb.Chem.Phys.9,43(1934).
[181]G.HerzbergandE.Teller,Z.Phys.Chem.B21,410(1933).
[182]L.D.LandauandE.M.Lifschitz,LehrbuchdertheoretischenPhysikI,Mechanik
1982).Leipzig,erlag,(Akademie-V[183]W.GordyandR.L.Cook,Microwavemolecularspectra,Vol.XVIIIofTechniquesof
Chemistry,1sted.(JohnWileySons,NewYork,1984).
[184]J.E.Wollrab,RotationalSpectraandMolecularStructure,Vol.13ofAseriesofMono-
graphsonPhysicalChemistry,1sted.(AcademicPress,NewYork,1967).
[185]N.Zare,Richard,AngularMomentum.UnderstandingSpatialAspectsinChemistryand
Physics,1sted.(WileyInterscience,NewYork,1988).

BIBLIOGRAPHY

[186]

[187]

[188]

[189]

[190]

[190]

[191]

H.J.Neusser,Fachbereichf¨urChemie,BiologieundGeowissenschaftender
Universit¨atM¨unchen(1977).

H.A.JahnandE.Teller,Proc.R.Soc.LondonAA161,220(1937).

J.Kraitchman,Am.J.Phys.21,17(1953).

C.C.Costain,J.Chem.Phys.29,864(1958).

T.M.Korter,J.K¨upper,andD.W.Pratt,J.Chem.Phys.111,3946(1999).

K.SiglowandH.J.Neusser,Chem.Phys.Lett.343,475(2001).

215

echnischenT

216

BIBLIOGRAPHY

esFigurofList

1.1

2.12.22.32.42.52.62.72.8

3.13.23.33.43.53.63.73.84.1

EpoxidationofE-2-butene.............................

Resonanttwo-photonionization(R2PI)scheme...................
Schematicdrawingoftheheatablenozzleforproductionofmolecularbeams...
SchematicdrawingoftheIonoptic.........................
Schematicrepresentationofthelasersystemforlow-andhigh-resolutionres-
onanceenhancedtwo-photonionizationexperiments.Enlargement:Schematic
viewofthemolecular-beamsetup..........................
Schematicviewofthelaserpulsedelayline.....................
Graphicalrepresentationoftheelongatedexcimerlaserpulse.Thetimewidthis
measuredatFWHM.................................
Mass-analyzedthresholdionization(MATI)scheme................
Schematicrepresentationofthelasersystemformass-analyzedthresholdioniza-
tion(MATI)experiments...............................

2Atomlabelsofthe2-PE∙HOcomplex.......................
01One-colorR2PIspectraoftheS←Selectronictransitionofthe2-PEmonomer.
1High-resolutiontwo-colorUVR2PIspectrumofvibronicband+42cminFig-
−ure3.2,recordedatm/z=122............................
1High-resolutiontwo-colorUVR2PIspectraofvibronicbands+48cmand+58
−1cminFigure3.2,recordedatm/z=122.....................
−1High-resolutiontwo-colorUVR2PIspectraofvibronicbands+136cmand
−1+213cminFigure3.2,recordedatm/z=122..................
−High-resolutiontwo-colorUVR2PIspectrumofvibronicband1winFigure
13.2,recordedatm/z=140..............................
Potentialenergysurface(invertedforthesakeofabettervisualization)asafunc-
tionoftheC2C7C8O9andC7C8O9H19dihedralanglesofthe2-PEmonomer..
Electronicgroundstate,S0,structuresofthe2-PE∙H2Ocomplex..........
Atomlabelsof2-pFPE...............................

217

16

2426272931313436

414244454749515262

218

4.24.34.44.54.65.15.25.35.45.55.66.16.26.36.46.56.66.76.86.97.17.27.37.47.57.67.7

OFLISTFIGURES

One-colorR2PIspectraof2-pFPE.........................64
0High-resolutiontwo-colorUVR2PIspectrumofthe0bandofthe2-pFPE
0monomerrecordedatm/z=140...........................66
1−High-resolutiontwo-colorUVR2PIspectrumofvibronicbandat+42cmof
the2-pFPEmonomerrecordedatm/z=140.....................67
1−High-resolutiontwo-colorUVR2PIspectrumofthe+80cmbandofthe2-
pFPEmonomerrecordedatm/z=140........................68
Potential-energysurface(invertedforthesakeofabettervisualization)asafunc-
tionoftheC2C7C8O9andC7C8O9H19dihedralanglesofthe2-pFPEmonomer.69
ExtendedOne-colourS1←S0REMPIspectrumof2-pFPE............81
TICspectraof2-pFPE...............................82
MATIspectraof2-pFPEviagauche(a)andanti(b)origins............83
11−−MATIspectraof2-pFPEviathe+632cm(a),the+147cm(b),andtheanti
00(c)bands.....................................84
0Theoreticallypredictedcationconformersof2-pFPE...............85
HOMOelectrondensitydistributioncomparisonof2-pFPEconformersinthe
neutralandthecation................................86
Atomlabelsof2-oFPE...............................94
One-colorR2PIspectraof2-oFPE.........................97
0High-resolutiontwo-colorUVR2PIspectrumofthe0bandof2-oFPE.....98
01−High-resolutiontwo-colorUVR2PIspectrumofthe+3cmbandof2-oFPE..98
1−High-resolutiontwo-colorUVR2PIspectrumofthe+43cmbandof2-oFPE.100
1−High-resolutiontwo-colorUVR2PIspectrumofthe+86cmbandof2-oFPE.100
1−High-resolutiontwo-colorUVR2PIspectrumofthe+109cmbandof2-oFPE.101
1−High-resolutiontwo-colorUVR2PIspectrumofthe+125cmbandof2-oFPE.101
PESofthe2-oFPEmonomer............................103
One-colorR2PIspectraofthe2pFPEwaterclustermeasuredatdifferentmasses.116
11−−2High-resolutionR2PIspectraofvibronicbands-18cmand-2cmof2pFPE∙HO
cluster........................................117
11−−High-resolutionR2PIspectraofvibronicbands53cmand76cmof2pFPE∙H2O
cluster........................................118
1−High-resolutiontwo-colorUVR2PIspectrumofvibronicband76cmofthe
22-pFPE∙HOcomplex................................119
Theoreticalelectronicgroundstate,S0,structuresofthe2-pFPE∙H2Ocomplex..123
One-colorR2PIspectraofthe2oFPE∙H2Ocomplexrecordedattwomasschannels.125
1−High-resolutiontwo-colorR2PIspectrumofband62Acmofthe2-oFPE∙H2O
complex.......................................126

OFLISTFIGURES

7.87.98.18.28.38.48.5

9.19.29.39.49.5

A.1

219

1−2High-resolutiontwo-colorR2PIspectrumofband62Bcmofthe2-oFPE∙HO
complex.......................................127
Theoreticalelectronicgroundstate,S0,structuresofthe2-oFPE∙H2Ocomplex..128
Electronicgroundstate,S,structuresofthetwolowestenergyAG(a)andGG
0(a)conformersofEPD...............................136
(a)One-color(R2PI)spectraofephedrine,measuredattheparent(m/z=165)
(b-d)fragments(m/z=85,m/z=71,m/z=58)masschannels,respectively.....137
(a)One-color(R2PI)spectraofEPD-H2Ocluster,recordedatm/z=183mass
channel.......................................138
110Massspectrarecordedatvibronicbands0,+173cm,+205cmand+225
−−01cm.........................................139
−FragmentationpathwaysoftheEPDcationafterresonance-enhancedtwo-photon
excitation......................................141

Electronicgroundstate,S,structuresofthefourlowestenergyconformersof
0pseudoEPD.....................................148
(a)One-color(R2PI)spectraofpseudoephedrine,measuredatfragmentmass
channels:m/z=85,m/z=71,andm/z=58,respectively...............150
0Highlyresolvedspectraofband0ofpseudoEPD..................152
011Highlyresolvedspectraofbands+135cmand+181cmofpseudoEPD....153
−−11Highlyresolvedspectraofbands+126cmand+212cmofpseudoEPD....154
−−

Bondingpotentialofweakmolecularinteractions..................171

220

LIST

OF

FIGURES

ablesTofList

3.1Experimentalparametersofbands+42,+48,+58,+136,and+213cm−1,shown
inFigures3.3-3.5..................................48
3.2Experimentalparametersofthe00originbandofconformerBofthe2-PE∙H2O
complexshowninfigure3.6...0..........................50
3.3Theoreticalparametersfortheground,S,andthefirstexcited,S,electronic
stateofthegaucheandanticonformersof0the2-PEmonomer...1.......54
3.4Theoreticalvibrationalfrequenciesforthefirstexcited,S1,electronicstateof
conformationsofthe2-PEmonomer........................55
3.5stateTheoreticaloftheconformersparametersofforthethe2-PEground,singlyS0h,andydratedthefirstcompleexes.xcited,..S.1,...electronic....57
4.1Experimentalparametersforbands000,+42and+80cm−1of2-pFPE.......67
4.2Theoreticalparametersofthegaucheandanticonformersofthe2-pFPEmonomer.71
5.1Theoreticallypredictednormal-modevibrationalfrequenciesofthe2-pFPEcation
conformers.....................................87
5.2AssignmentoftheobservedvibronicbandsintheMATIspectraof2-pFPE
showninFig.5.3b.................................89
5.3ExperimentalandtheoreticalvaluesoftheAIEsofthe2-pFPEobservedconformers89
6.1Experimentalparametersfortheground,S0andforthefirstexcited,S1electronic
statesofbands000,+43,+86,+109,and+125cm−1of2-oFPE...........99
6.2Theoreticalparametersforthegaucheandanticonformersofthe2-oFPEmonomer
1046.3Theoreticalvibrationalfrequenciesfortheground,S0,andthefirstexcited,S1,
electronicstateofthegaucheandanticonformersofthe2-oFPEmonomer...108
7.1Experimentalparametersforband76cm−1ofthe2-pFPE∙H2Ocomplex......120
7.2Theoreticalparametersofthepredictedstructuresofthe2-pFPE∙H2Ocomplex..122
7.3Experimentalparametersofbands62Aand62Bcm−1of2-oFPE∙H2Ocomplex..126
221

7.4

9.1

9.2

D.1

Theoreticalparametersofthepredictedstructuresofthe2-oFPE∙H2Ocomplex..129
0−1−1−1−1−1
Experimentalparametersforbands0cm,126cm,135cm,181cmand212cm
0ofpseudoEPD....................................149
TheoreticalparametersfortheoptimizedconformersofpseudoEPD........151

Rigidasymmetric-topenergylevels

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

198

PublicationsofList

S.Chervenkov,P.Q.Wang,R.Karaminkov,T.Chakraborty,J.E.Braun,andH.J.Neusser,
HighResolutionUVResonanceTwoPhotonIonizationSpectroscopywithMassSelectionof
BiologicallyRelevantMoleculesintheGasPhasein13thInternationalSchoolonQuantum
Electronics:LaserPhysicsandApplications,P.A.Atanasov,S.V.Gateva,L.A.Avramov,A.A.
,eds.afetinides,SerProc.SPIE5830,246(2005).
S.Chervenkov,R.Karaminkov,J.E.Braun,H.J.Neusser,S.S.Panja,andT.Chakraborty,
SpecificandNonspecificInteractionsinaMoleculewithFlexibleSideChain:2-phenylethanol
andits1:1ComplexwithArgonstudiedbyHighResolutionUVSpectroscopy,
J.Chem.Phys.124,234302(2006).
R.Karaminkov,S.Chervenkov,P.H¨arter,andH.J.Neusser,Resonance-EnhancedTwo-Photon
IonizationMassSpectrometryofEphedrine:IndicationofState-SelectiveFragmentationina
,MoleculexibleFleChem.Phys.Lett.442,238(2007).
R.Karaminkov,S.Chervenkov,andH.J.Neusser,IdentificationofConformationalStructures
of2-PhenylethanolandItsSinglyHydratedComplexbyMassSelectiveHigh-ResolutionSpec-
troscopyandabInitioCalculations,
J.Phys.Chem.A112,839(2008).
R.Karaminkov,S.Chervenkov,andH.J.Neusser,FluorineSubstitutionandNonconventional
OH∙∙∙πintramolecularbond:High-ResolutionUVSpectroscopyandabInitioCalculationsof
,ophenyl)ethanol-uorρ2-(Phys.Chem.Chem.Phys.10,2852(2008).
R.Karaminkov,S.Chervenkov,H.J.Neusser,V.Ramanathan,andT.Chakraborty,Competition
betweenπandσHydrogenBondsandConformationalProbingof2-orthouorophenylethanol
byLow-andHigh-ResolutionelectronicSpectroscopy,
J.Chem.Phys.130,034301(2009).
R.Karaminkov,S.Chervenkov,andH.J.Neusser,MassAnalysedThresholdIonisationSpec-
troscopyofFlexible2-para-uorophenylethanolConformerswithandwithoutanIntramolecular
OH∙∙∙πBond,
Phys.Chem.Chem.Phys.11,2249(2009).
S.Georgiev,R.Karaminkov,S.Chervenkov,V.Delchev,andH.J.Neusser,Mass-Analyzed
ThresholdIonizationSpectroscopyof2-Phenylethanol:ProbingofConformationalChanges

CausedJ.

Phys.

byIonizationChem.

A

,

113,

12328

(2009).

PosterandOralPresentations

19thColloquiumonHighResolutionMolecularSpectroscopy
Salamanca,Spain,11.09.-16.09.2005
PosterCONFORMATIONANALYSISOFBIOLOGICALMOLECULESINTHEGASPHASEBYMASS
SELECTIVEREMPIHIGH-RESOLUTIONUVSPECTROSCOPY
19thInternationalConferenceonHighResolutionMolecularSpectroscopy
Prague,TheCzechRepublic,29.08.-02.09.2006
PosterFRAGMENTATIONBEHAVIORANDCONFORMATIONSOFSMALLBIOLOGICALMOLECULES
STUDIEDBYMASSSELECTIVEHIGH-RESOLUTIONTWO-PHOTONUVLASERSPEC-
OSCOPYTRDiscussionMeeting,SpectroscopyandDynamicsofMoleculesandClusters
CorbettRiverViewRetreat,Uttarkhand,India,23.02.-25.02.2007
PosterSTATE-SELECTIVEFRAGMENTATIONOFEPHEDRINEANDCONFORMATIONALSTUD-
IESOFSMALLBIOLOGICALMOLECULESANDTHEIRHYDRATEDCLUSTERSBYMASS
OSCOPYSPECTRUVHIGH-RESOLUTIONSELECTIVETechnischeUniversit¨atM¨unchen,PhysikalischeChemieI,Kaffeeseminar
M¨unchen,Germany,10.05.2007
PresentationOralSTATESELECTEDFRAGMENTATIONAFTERR2PI:HOWDOESITOCCUR?
20thColloquiumonHighResolutionMolecularSpectroscopy
Dijon,France,03.09.-07.09.2007
PosterINFLUENCEOFTHEπ-ELECTRONDENSITYIN2-(ρ-FLUOROPHENYL)ETHANOLAND
2-PHENYLETHANOLONTHECONFORMATIONALPREFERENCESSTUDIEDBYMASS
OSCOPYSPECTRUVHIGH-RESOLUTIONSELECTIVEGordonResearchConference,MolecularandIonicClusters
CentrePaulLangevin,Aussois,France,7.09.-12.09.2008
PosterMATISPECTROSCOPYOFAFLEXIBLESYSTEM:GAUCHEvs.ANTICONFORMATION
OF2-(ρ-FLUOROPHENYL)ETHANOLANDHIGHRESOLUTIONUVSTUDYOFITSWATER
DIMER

DiscussionMeeting,SpectroscopyandDynamicsofMoleculesandClusters
Mandarmoni,WestBengal,India,20.02.-22.02.2009
PresentationOralandPoster

MASSANALYSEDTHRESHOLDIONISATIONSPECTROSCOPYOFFLEXIBLE2-PHENYLETHANOL

ANDITSortho-ANDpara-FLUORO-DERIVATIVES:THEEFFECTOFCONFORMATION

HIGH-RESOLUTIONUVANDMASSANALYZEDTHRESHOLDIONIZATIONSPECTROSCOPY

MOLECULESFLEXIBLEOF

TechnischeUniversit¨atM¨unchen,LehrstuhlfrExperimentalphysikE11,Institutsseminar
M¨unchen,Germany,12.11.2009
PresentationOral

CONFORMATIONSANDFRAGMENTATIONOFBIOLOGICALLYRELEVANTMOLECULES

ANDTHEIRBINARYCOMPLEXESWITHWATERPROBEDBYHIGH

ANDMASSANALYZEDTHRESHOLDIONIZATIONSPECTROSCOPY

RESOLUTION

UV

wledgementsAckno

Iwouldliketothankallthosewhohavecontributedtothesuccesfullcompletionofthiswork.
IamgratefultoProf.Dr.HansJ¨urgenNeusserfortheopportunitytoworkinhisgroup,forhis
involvementandkeeninterestinthescientificresearch,forhistoleranceandattention,forhis
constantsupportandunderstanding.Iappreciatethestimulatingdiscussionsandthevaluable
adviceshehasgiven.
Iwarmlythankmycolleaguesandgoodfriends,M.Sc.StoyanGeorgiev,Dr.SotirChervenkov,
andDr.AntonTrifonovfortheirhelpinthebeginningofmyresearch,fortheirtrust,understand-
ing,responsiveness,attention,andhelp.IamobligedtoDr.SotirChervenkovforhisintroduction
inthespecificaspectsofourwork,forhiswillingnesstohelp,forhisvaluablediscussionsand
comprehensiveanswers.Heisoneofthemainreasonstostayandfulfillmyworkinthegroup.
IwouldliketocordiallythankProf.TapasChakrabortyforourfruitfulcooperationandforthe
valuablesuggestions,discussions,andideashehasgiven,andforhishospitalityduringmyvisits
toIndia;Dr.HeinrichSelzleforhisaffability,attention,andtheexperthelpinvariousareas,
especiallyincomputersandquantumchemistrycalculations;S.Kn¨orinthegroupofProf.H.
Kessler,whokindlycarriedouttheHPLCpuritytestofephedrine;Dr.VasilDelchevfromthe
groupofProf.Dr.WolfgangDomckeforhishelpinthetheoreticalcalculationsandpriceless
advices.Isincerelythankoursecretary,Mrs.Thiem,forherresponsivenessandassistance.
IthankalsoMr.MaxWiedemanandMr.Hans-ArnulfM¨ullerandthestaffoftheElectroncs
workshopfortheirindispensablehelpinthelab,Mr.WernerTauchmanforhisassistancefor
solvingcomputerproblems,Mr.OttoStrasserandhiscolleaguesfromtheMechanicsWorkshop.
Ithankallmyfriendsfortheirsupport.IappreciatetheresponsivenessandhelpofDr.Hristo
.vIgleIthankwholeheartedlyallofmyfamilyfortheirlove,encouragement,underastanding,and
support.

227