Van der Waals forces are the forces of attraction between molecules These are also known as intermolecular forces and are much weaker
However, these are weakest in nature as compared to other chemical interactions or forces This force is much weaker so it has no importance in our life The
Molecular solids become compre- hensible as soon as we recognize the contributions of the weak forces known as van der Waals attraction and hydrogen bonding
All weak intermolecular forces are called: van der Waals forces Page 8 van der Waals Forces Two major forms: • Dipole–
“Van der Waals forces - the relatively weak attractive forces that act on neutral atoms and molecules that arise because of the electric polarization induced in
(a) We define precisely what we mean by a van der Waals force and, to put our In the near- to mid-uv, water (14) shows a weak absorption at X = 1650 A or
the weak interaction: nuclear interaction (responsible for nuclear -decay / electron emission attractive intermolecular forces ( van der Waals forces) w r A
about different types of van der Waals forces 5 1 1 Dispersion weaker than ion -ion interaction because only weak, molecules do not cling together to make
PDF document for free
- PDF document for free
![[PDF] van der Waals Forces - CORE [PDF] van der Waals Forces - CORE](https://pdfprof.com/EN_PDFV2/Docs/PDF_8/19122_882486132.pdf.jpg)
19122_882486132.pdf
VANDERWAALSFORCES
SPECIALCHARACTERISTICSINLIPID-WATER
SYSTEMSANDAGENERALMETHODOF
CALCULATIONBASEDONTHELIFSHITZTHEORY
B.W.NINHAMandV.A.PARSEGIAN
FromthePhysicalSciencesLaboratory,DivisionofComputerResearchandTechnology, NationalInstitutesofHealth,Bethesda,Maryland20014.Dr.Ninham'spermanentaddress istheDepartmentofAppliedMathematics,UniversityofNewSouthWales,
Kensington,NewSouthWales,Australia2033.
ABsTRAcTApracticalmethodforexaminingandcalculatingvanderWaalsforcesisderivedfromLifshitz'theory.RatherthantreatthetotalvanderWaalsenergyasasumofpairwiseinteractionsbetweenatoms,theLifshitztheorytreatscomponentmaterialsascontinuainwhichthereareelectromagneticfluctuationsatallfrequenciesovertheentirebody.Itisnecessaryinprincipletousetotalmacroscopicdielectricdatafromcomponentsubstancestoanalyzethepermittedfluctuations;inpracticeitispossibletouseonlypartialinformationtoperformsatisfactorycalculations.Thebiologicallyinterestingcaseoflipid-watersystemsisconsideredindetailforillustration.ThemethodgivesgoodagreementwithmeasuredvanderWaalsenergyofinteractionacrossalipidfilm.Itappearsthatfluctuationsatinfraredfrequenciesandmicrowavefrequenciesareveryimportantalthoughtheseareusuallyignoredinpreferencetouvcontributions."Retarda-tioneffects"aresuchas todampouthighfrequencyfluctuationcontributions;ifinteractionspecificityisduetouvspectra,thiswillberevealedonlyatinteractions
across<200angstrom(A).DependenceofvanderWaalsforcesonmaterialelectricpropertiesisdiscussedintermsofillustrativenumericalcalculations.
INTRODUCTION
Thepurposeofthisandasucceedingpaper(forapreliminaryreport,seereference1) istodevelopinsomedetailageneralmethodforcalculatingvanderWaalsforces insituationsofbiologicalinterest.Fordefinitenessweshallreferinmostparttothe interactionofwateracrossathinhydrocarbonfilm.Theenergyofsuchasystem wasrecentlymeasuredbyHaydonandTaylor(2).Weshallshowthatvander Waals
forcesinthissystemcanbeanalyzedeasilyandaccuratelybythemacro-scopictheoryofLifshitz(3).Further,severalimportantqualitativefeaturesof
theseforces,unnoticedinearliertreatments,arerevealedbythisanalysis.The approachelaboratedhereallowspredictionofattractiveenergiesinothersystems,
646brought to you by COREView metadata, citation and similar papers at core.ac.ukprovided by Elsevier - Publisher Connector
aswellaspointedexaminationoftheprobableroleoftheseforcesinbiological processes.Inparticular,weshallfindthatacarefulanalysisprovidesastrong hintconcerningtheproblemofattractivespecificity.
Theoutlineofthepaperisasfollows:
(a)WedefinepreciselywhatwemeanbyavanderWaalsforceand,toputour investigationintoperspective,brieflysummarizeearlierworkontheproblem. (b)WegiveadescriptionoftheformulaeofLifshitzapplicabletothinfilms betweensemi-infinitemedia,anddiscussausefulrepresentationofdielectricsus- ceptibilitybymeansofwhichspectroscopicinformationmaybeadaptedforcal- culation. (c)Estimatesofthedispersionforceinlipid-watersystemsareobtainedand comparedwiththemeasuredvalues. (d)Adetailedexaminationismadeofthehighfrequencyspectrumofcontribu- tionstotheenergyinordertodevelopanintuitiveunderstandingofthebehavior ofvanderWaalsforcesacrosslipidlayers,andoftheexperimentalvariableswhich affectforces.Thisisimportantinasmuchasthisstudythrowssomelightonthe physicaloriginofspecificityinbiologicalsystems.Itwillbeshownthatthereisas greatacontributiontotheenergyfrominfraredfrequenciesasfromtheuvand thecontributionfromthemicrowaveregionisasgreatasinfraredanduvcombined.' Further,theso-calledretardationeffectsduetothefinitevelocityofsignalpropaga- tionemergeasaprogressivedampingofhigherfrequencycontributionsasfilm thicknessisincreased. (e)Finallywesummarizesomequalitativeimplicationsconcerningbiological systemswhichcanbedrawnfromadetailedstudyofvanderWaalsforcesby
Lifshitztheory.
THEVANDERWAALSFORCE
ThevanderWaalsforceisfoundedontherecognitionthatspontaneous,transient
electricpolarizationcanariseatacenterduetothemotionofelectrons,moleculardistortion,ormolecularorientation.Thispolarizationwillactonthesurrounding
regiontoperturbspontaneousfluctuationselsewhere.Theinteractionresulting fromthisperturbationissuchastolowertheenergy.Theclassesofsuchinteraction havebeenextensivelyreviewedbyKauzmann(4)andJehle(5). Theresultingattractive"dispersion"forceforelectronicfluctuationswasfirst calculatedbyF.London(6)inhisfamous1930paper.Heshowedthattheforce
betweentwoisolatedatomsisproportionaltotheinverseseventhpowerofdistance,andtheproductoftheirpolarizabilities.London'stheoryprovidedthe
basisforsubsequenttheoreticalestimates(7)ofdispersionforcesbetweencon-
1Thecharacteristicsofthemicrowave-frequencyfluctuationsaresufficientlydistinctfromthoseofhigherfrequenciessothattheyareconsideredseparatelyinthesucceedingpaperreferredtoas"two";Parsegian,V.A.,andB.W.Ninham.1970.Biophys.J.10:664.
NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems647 densedmedia.Theseestimateshavebeenseverelylimitedbyseveraladhocassump- tionswhichare: (a)theassumptionofpairwiseadditivityofindividualinteratomicinteractions inacondensedmedium; (b)theapproximationthatcontributionscenteredaroundasingledominant frequencyoftheelectromagneticfieldintheultravioletareimportant;and (c)thatthedifficultyofdealingwithanintermediatesubstance(e.g.hydrocarbon betweenaqueousregions)canbehandledbytheinsertionofanarbitrary"dielec- tricconstant"correctionatasinglefrequency.Inaddition,detailedinformation aboutatomicpolarizabilitiesandrelaxationfrequencieswasrequiredforcom- putations.Asemphasizedbylaterworkers(3)acalculationofvanderWaals forcesbasedontheseassumptions(validfordilutegases)isintrinsicallyunsound. Thismethodobscuresalmosttotallythemostinterestingqualitativefeaturesof vanderWaalsforces. In1955Lifshitz(8)andlaterheandhiscoworkers(3)developedatheorybased
onideasofCasimirandPolder(9)whichovercomesthesedifficulties.Histheoryincludesallmany-bodyforcesthroughacontinuumpicture,retainscontributions
fromallinteractionfrequencies,anddealscorrectlywiththeeffectsofintermediate substances.Moreimportant,theinformationrequiredforcalculationsiscontained inthedielectricpropertiesofcomponentmaterials-informationavailableinprin- ciplefromindependentspectroscopicmeasurements. Foracondensedmedium,wheretherangeofstronginteractionexceedsthe distancebetweenatomiccenters,Lifshitzregardstheentiresetoflocalspontaneous electricfieldfluctuationsasanelectromagneticfieldwhichextendsoverthewhole system.Thistime-varyingfieldcanbefrequencyanalyzed;thestrengthofafield ofagivenfrequencyisdirectlydependentontheresponseofthematerialtoan appliedfieldofthatfrequency,i.e.,itsdielectricsusceptibilitye(W).Boundary surfacesbetweenunlikematerialswillaffecttheseelectricfieldsandconsequently theelectromagneticenergyofthesystem.Thisapproachtodispersionforces examinesthechangeinenergyofasystemwithmovementoftheboundarysur- facesbetweenunlikeregions. Itwasthought(10)thatthetheoryofLifshitzcouldnotbeappliedtothecalcu- lationofdispersionforcesintheabsenceofcompletespectralinformation.Wefind thatsuchcompletedataareunnecessary,particularlywhencomponentsubstances
areofsimilarweightdensity.Infact,itispossibletomakeseveralsimplifyingassumptionsregardingspectra.Allthatappearsnecessaryforthecalculationofvan
derWaalsforcesinthepresentthinhydrocarbonfilmsystemaresingleaverage absorptionfrequenciesintheinfraredanduvforwater,oneaveragefrequencyin theuvforhydrocarbon,indicesofrefractionatvisiblefrequencies,alimiting
valueforthedielectricsusceptibilityofwaterbetweenmicrowaveandinfraredfrequencies,andthesimplestformforthedielectricdispersionofliquidwaterinthemicrowaveregion.
BIOPHYSICALJOURNALVOLUME101970648
FORMULARY
LifshitzExpression
Thegeneralformulafortheattractiveforceperunitareabetweentwosemi-infinite mediaofsubstance1acrossaplanarslabofsubstance2ofthicknessI(Fig.1)(3)is
F(l)=0CE823n$f2{[(S+p)exp(2Pne2)]-1
F(1)~[(e-pl(=C~ex]+[(SE2+;::)2ex2P~nIE2)f}d/21
wherekisBoltzmann'sconstant,Tabsolutetemperature,cthevelocityoflightin vacuum,qandE2thedielectricsusceptibilitiesevaluatedontheimaginaryfre- quencyaxisatco=itn, s=vi2_17/2,(2) and 2irkT with2wrhPlanck'sconstant.Thesumistakenoverintegralnandtheprime(')on thesumsignindicatesthatthen=0termbemultipliedbyY2. Byintegrationoftheforcewithrespectto1,thecorrespondingfreeenergyof interactionperunitarea(takingG=0atI=X)is
G(l,T)kTEI((n1)(4)870ln=O
waterhydrocarbonwater
FIGURE1Twosemi-infinitemediaofsubstance1separatedbyaplanarslab,thickness1,ofsubstance2.Inthetextweconsiderthecasewhere1iswaterand2ishydrocarbon.
NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems649 where
1(n1)p(2nle2)f{[1-i2exp(2ptnIE2
+In[-A2expe2)]}dp(5) with
Se2-=l-P(s6)
SE2+pC-1s+p
Toconformwiththenotationofcolloidchemistry,itwillbeconvenientinthe followingtodefinea"Hamakerfunction,"A=A(l,T),suchthat
G(l,T)=A(1,)7
Comparingequation7withequation4,then
00
A(1,T)=1.5kTElI(n21).(8)n==O
LimitingCases.Lowtemperature:whenkT<
ForI< IU(n0)=fxln[1-(::)2e~ld(10)
(Inthiscasespsincevaluesofp>>1contributetotheintegralinequation5.)SometimesA2<<1and /2o0)'(2-61 InthedoublelimitI->0,T->0
A(O)_3h4JolxIun[1_(2'Ele-dxdt
,3hlo(f2-fld(112 BIOPHYSICALJOURNALVOLUME101970650
DielectricDispersionA
Inordertousetheaboveformulae,weneedaconvenientrepresentationforthe dielectricsusceptibility. Thefunctione(co)isacomplexfunctionE!+iE"ofacomplexfrequencyco= (wR+it.Thedispersionenergydependsonlyonthevaluesofeontheimaginary frequencyaxise=e(iQ)and,fromequations4and5,aknowledgeofE(i#)isim- portantonlyforthosevaluesoftforwhichthereisadifferencebetweenthedi- electricsusceptibilitiesofthetwomaterialsatagivenfrequency.Weseekasuitable representationfore(it)whichwillincorporateexperimentaldataaswellassatisfy generalconstraintswhichmustbesatisfiedbyanydielectricsusceptibility.These constraintsarethatE(it)beapurerealquantitymonotonicdecreasingwithi, andthatthefunctione(w)(C=WR+it)havenozerosorpolesintheupperhalfW planeinordertosatisfytheKramers-Kronigrelations(11).Asuitablerepresenta- tionis e(X)1+1-ixjx+E1-(co/c)2+i(*(12') ThisrepresentationincludessimpleDebyerelaxationformicrowavefrequencies plustheclassicalformofLorentzelectrondispersionforinfraredthroughmid-uv frequencies.WewilldescribebelowhowthenecessaryconstantsCmwandCy(proportionalto"oscillatorstrengths")andresonancefrequenciescomw,WXmay
bedeterminedfrommeasurementsmadeontherealfrequencyaxisco=WR.On theimaginaryaxis(co=it),wehavefromequation11 +1mwCE1j(13)1+~/cmwjtc)Y Nowthedampingterminyjcoinequation12issignificantonlywhencoCojsince bandwidthsarealwaysmuchlessthanabsorptionfrequencies.Thecontributionofthisterminequation13willbenegligible(1+(/Coj)2>>yjy)sothatwemay
take CmwC,e(it)=1+1+/jm+E1+(/coj)(14)
Clearlythen,onlytheoscillatorstrengthsandabsorptionfrequenciesofcom- ponentmaterialswilldeterminedispersionenergies. Atveryhighfrequenciesthedielectricdispersionhasthelimitingform(11) (O)1-_4xA2(15) Hereeandmareelectronchargeandmass,andNisthenumberofelectrons/cc. NINHAMANDPARSEGIANVanderWaalsForces:CakulationforLipidWaterSystems651 Ontheimaginaryaxisequation15becomes
=1+42Ne(16)mi Thisformholdsformaterialsmadeoflightelementsatfrequenciesinandabove thefar-uv.Itisimportanttoemphasizethatmaterialsoflightelementsandsimilar weightdensitywillhavesimilarsusceptibilitiesatthesehighfrequencies,andcon- sequentlygiveaverysmallcontributiontotheforceandenergyintegrals.Thisis theusualsituationinsystemsofbiologicalinterest. Betweenthelow-tomid-uvregiondescribedbyequation14andthefar-uvto X-rayregiondescribedbyequation16,thesimplestprocedureistoconstructaninterpolationformulaforE(it).GiventheconditionthatE(it)bemonotonicde-
creasing,thereislittleambiguityinsodoing. However,becauseofthesimilarityofthesusceptibilitiesofcomponentsubstances inthemid-uv,theintegrandsorsumtermsoftheforceandenergyexpressions tendtozeroveryrapidlyinthisregion,andgivelittlecontribution.Inpractice therefore,itisusuallysufficienttousetheformofequation14forsusceptibilities, andnointerpolationisrequired.Indeed,inspectionofthefullexpressionsequations 1,4,and5fortheforceandenergyshowsthattheeffectoftheexponen-tialsexp(-2ptle2112/c)istodiminishcontributionstotheintegrandseverelyforfrequenciest.>c/(2lV\e2).Forexample,andtypicallywhenI150A,t,>(3X
1016)~10166,frequenciesintheuvbandsatisfy10el.8.,1016.8sothatfrequencies
atandabovethefar-uvwillbeinanyeventunimportant.Thispointisillustrated indetailinthesectionAnalysisofIntegralswherewecarryoutananalysisofthe integralofequation9. DATAANDCALCULATIONS
SpectroscopicExperimentalData
Forwaterandhydrocarbonsthefollowingdielectricdataareknown. Water.Fromaudiothroughmicrowavefrequencieswaterexhibitssimple Debyerelaxationfromitsstaticvalueof80.4downto5.2withacharacteristicrelaxationwavelength1.78cm.Thus,Cmw=(80.4-5.2)=75.2,andwmw=27rC/XmW=(1.06X1011)=1011.026radians/sec.(Datahere(12)areforT=
20°C.)Theinfraredspectrumisdominatedbythreecloselyspacedabsorption
peaks(13)atcir=3.0X1014,6.89X1014,and7.08X1014,afterwhichere- laxestovaluesobservedasthesquareoftheindexofrefractionntwIwhere 2nw=1.78.WeshalluseCir=(5.2-1.78)=3.42andapproximatethevibrationfrequencybyanaveragevaluewir=[U3(3+6.89+7.08)X1014]=1014.75radians/
sec.Insensitivityofthecalculatedenergytothisapproximationischeckedbelow. Inthenear-tomid-uv,water(14)showsaweakabsorptionatX=1650Aor BIOPHYSICALJOURNALVOLUME101970652
wuv=(1.14Xl01)=1016.058andanapparentlymuchstrongerabsorptionat aboutX=1250Aorwuv=(1.507X101w)=1016.178.However,therelativestrength ofthesepeaksandthepresenceofotherabsorptioninthefurtheruvarenotknown. Becauseoftheslowvariationofe(it)withtonemayuseanaveragefrequencyw,uvforthenear-tomid-ultravioletregion.Acommonapproximationistousea
frequencyequivalenttothefirstionizationpotential(4).Inthiscase12.62ev-hcouv,orcouv=(1.906X101")=1016.28rad/sec.(Throughoutpaperradusedfor
radian.)Itwillbeshownthatuseoftheionizationpotentialvalueorthestronger absorptionbandvaluehaslittleeffectontheestimatedenergy.ThevalueofCu, forthisrelaxationis(n2-1)=(1.78-1)=0.78.Wehavethen Ewater=ew(i)=1+1+7mW+1+C+1+(17)
wheretheconstantsaregivenabove. Hydrocarbons.Thereappearstobelittledielectricrelaxationshownby hydrocarbonliquidsbetweenaudioandvisiblefrequencies(12)where(ha=nhe(n=indexofrefraction).Byexaminingthereflectionoflaserbeamsfromathin
lipidfilm,CherryandChapman(15)havemadeaprecisemeasurementofan anisotropicindexofrefraction:nhe=1.486forpolarizationperpendiculartotheplaneofthefilmandnho- 1.464paralleltothefilm.ThesegiveEhe=2.208and2.143,respectively.Other
2possibleestimatesatopticalfrequenciesareEhc=nhc=1.89forn-hexane(12)
andehe=nhc=2.0,anaveragevalueoftenusedastypicalforhydrocarbons. Eachofthesenumberswillbeconsideredbelow.Againwewillsummarizerelaxationasoccurringatanaveragefrequencycor-
respondingtotheionizationpotential.Anappropriatevalueofthisforathin hydrocarbonfilmisnoteasilydeterminedfromavailabledata.Valuesfordecane(16),10.19evorCuv=1.54X1016,andforethane(16,14),11.65evorcouv=
1.76X1016,willsufficeinthepresentinstance.Usingthesevalueswehave
2-1Ehe(lt)=1+1he+v)2
TheoreticalEstimates
WefirstcalculatethevanderWaalsenergyofwateractingacrossalipidfilm (Fig.1)50Athickat20°C.Usingequations17and18fore.(it)andfhc(it)with anaverageinfraredabsorptionfrequencyforwaterandaverageuvabsorption derivedfromionizationpotentialsofwateranddecaneorethane,wefindA(50A)byequation8forfourvaluesofn20(TableI).Forthesedata,valuesofA(=50A,
T=20°C)rangefrom5.5to7.1X1014erg.WenotethatAisnotsimplymono- tonicinnhhc NINHAMANDPARSEIGANVanderWaalsForces:CalculationforLipidWaterSystems653 Theseestimatesareslightlyhigherthanthoseinferredfromexperimentsonthinlipidfilms(2)(4.7X10-14erg)butmuchlowerthanthosederivedfromex-
perimentsonsuspensionsofparaffinsinaqueoussuspension(17)(r1.6X10-'1 erg).Inprinciplebothexperimentalestimatesshouldbeapproximatelythesame sothattheoreticalestimatesherearelessambiguousthanexperimentsbutwithin thecorrectranges. TableIIAsuggeststhatAisrelativelyinsensitivetotheassumedcUs,aslong astheyarechoseninaconsistentwayforthetwomaterials.Usingstrongestnear- uvabsorptionpeaksorionizationpotentialsforbothmaterialsgiveAestimates 6.1-7.1X10-14forn2=2.208thataredistinctfromthevalueobtainedbyusing
thesecond-strongestwaterpeak(9X10-14).Also,(TableIIB)itmakeslittle differenceifoneaverage(Dirisusedforwaterorifthethreemaininfraredpeaksare equallyweighted,andseparatelycontributetoew(it). Acarefulattempttousee(it)thatsatisfybothlowandhighfrequencybehavior(equations14and16)makesnegligibledifferencetotheestimatedA.Bystraight-
lineinterpolationbetweenformsfornear-uvandX-rayregionsandusingthe TABLEI
ESTIMATESOFTHEHAMAKERFUNCTIONAUSINGTHEGENERALEQUATION7 Takelr=5.66X1014rad/sec,C1°r=3.4cow=1.06X1011rad/sec,C;mw=75.2.2.Wuv=1.906X10"rad/sec,n2=1.78.I=50A,T=20°C.
2A(erg),wuv=1.54X1016A(erg),Wuv=1.76X1016nho(decaneI.P.*)(ethaneI.P.)
ergerg 1.9(n-hexane)5.8X10-145.8X10-14
2.0("typical"value)5.55.82.143(laserexpt.)5.76.52.208(laserexpt.)6.17.1
*I.P.=ionizationpotentials. TABLEIIA
DEPENDENCEOFAONWufv
Takenhw=2.208,otherwiseasinTableI.
uhoX10-16-vX10-16AX1014Sourceofw01 rad/secrad/secerg 1.41.5076.6Strongestabsorption
1.41.149.2ndstrongestH20absorption
1.761.9067.1Ionizationpotentials(ethaneforhc)
1.541.9066.1Ionizationpotentials(decaneforhc)
BIOPHYSICALJOURNALVOLUME101970654
TABLEIIB
EFFECTOFAVERAGINGVS.EQUALWEIGHTINGOFTHREEWATERINFRAREDFREQUENCIES Dataasabove.
WjyA,(couh=1.54X1016)A,(coh"=1.76X10")
ergerg 3.0,6.89,7.88X1014rad/secequally5.9X10-147.0X1-14weighted=5.66X1014md/sec6.17.1
TABLEIII
A(l)VS.I
Usen2hO=2.208,V=1.54X1016andasinTableI.
I(A)A(l)X1014A(l)/A(O)A(l)/A(O)oldtheory(18)
erg 06.3656.351.1.106.331.0.95506.090.960.841005.780.910.725004.740.750.2810004.430.70.1550003.650.570.03100003.330.520.015500003.160.50.003
electrondensityofmaterialhaving1g/ccweightdensity,wefindthattheeffectis alwayslessthan6%.(Themethodofinterpolationisdescribedindetailelsewhere forthecaseofsoapfilmsinair[18].) AasaFunctionofThickness1.Byvirtueofequations8and10thevalue ofAas1->0convergesto A(I=0,T)=l.5kT>2'fxIn[1-(2)e-x]dx.
nO0O2+C-1 DeviationsofAfromthislimitingformwhenI>0arecalled"retardationeffects" sincetheyarecausedbythefinitetraveltimeofelectromagneticradiationacross thegap1. InTableIIIwehavelistedcalculationofA(L)forI=0to50,000A.There- tardationfactorofA(1)/A(0)isalsolistedforcomparisonwiththefunctionusedincolloidchemistry(19).Clearlytheneglectofnon-uvfrequencyfluctuations,typical
oftheoldertheory,givesvastlylowerestimatesatlargeseparationdistances. NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems655 Theproperanalysisofretardationeffectsisdiscussedinthefollowingsectionand intheAppendix.AtverylargedistancesthecalculationofAisdominatedbythe n=0terminequation8;thistermisthesubjectofthesucceedingpaper(two). ANALYSISOFINTEGRALS
SpectralContributions
Inordertoclarifythedependenceofthedispersionenergyonfluctuationfrequencies suchthathwkT,weexaminetheintegrandsoftheintegralswhichoccurin equations7and9.Theseareoftheform JI(t,1)dt,(19)
whereI(,1)definedbyequations5and10measuresthecontributiontotheenergy 1-infrared-l1-.UV--.1
6-n2n=2.208
5I(C;0)x)
Q:0 1j21031415161718
x=logioCP FIGURE2SpectrumofrelativeinfraredanduvcontributionstovanderWaalsenergyinthelimitI=0.Useequations17,18,and10withdataforwaterandethane.Datafordecanegivesharplyreduceduvpeak.Notelargeinfraredcontribution,sharpdecreasein
uvcontributionwhenn2h,isreducedfrom2.208to2.0tobringitclosertontO2=1.78for water.ThefactortmultiplyingI(Q,I=0)intheordinateistocorrectforthelogarithmicscaleintheabscissa. BIOPHYSICALJOURNALVOLUME101970656
offrequenciesintherange(Q,t+dc).Changingvariablestox=logio(Q),the relativecontributionfromthespectralrange(x,x+dx)isthen2.303(IQ,1)dx. Fig.2showsplotsofI(Q,I=0)fortwovaluesofn2,2.208and2.0withI= 0.(OtherdataareasinTableI.)Thefinitefrequencycontributiontothetotal
energyiscleanlydividedintoaninfraredandanuvpeak.Therelativemagnitude ofthesetwopeaksobviouslydependsonthevalueofnh.Forexample,withnh= 2.208,35%oftheintegralcomesfromtheinfrared;butwithnhe=2.0,80%of
theenergyintegralisduetoelectromagneticfluctuationsatinfraredfrequencies! Itisremarkablethattheroleoftheinfraredspectrumofwaterhasbeenignoredin previousdiscussionsofvanderWaalsforces.Similarly,asalreadyemphasized, thereisanimportantcontributionfromthemicrowaveregionwhosequalitative featuresaresufficientlyuniquetobeconsideredseparately(two). ForfixedIandfixedwatercomposition,theonlytwoparameterswhichcan2-hoaltersignificantlywithhydrocarboncompositionarenhoand&Dhc.TheplotsofI(Q,1)inFig.2alsoshowthattheeffectofchangingn20istoaltertheentirespectrum
ofcontributingfrequencies.Similarly,variationinabsorptionfrequenciescO, willshiftthepositionofmaximaofthespectrum,andaltertheheightofpeaks. Forthepresentexamplec,>Elhiintheinfrared,butc,RetardationEffects ThecaseI=0consideredaboveshowsthenatureofspectralcontributionstothe energywithvariationofmaterialpropertiesofcomponentsubstances.Thereis anotherconditionimposedontheallowedcorrelationsinelectromagneticfluc- tuationsacrossagapduetothefinitevelocityoflight.Whenthetraveltimeofa fluctuationsignalacrossthegapiscomparabletoafluctuationfrequency(21E2'12)/c1/i,thecorrelationinfluctuationsbetweeneithersideisdiminished.Thein-
tegrandI(Q,I)satisfiestheunequalityI(Q,I)1=0,50Aand500A,andfornh0=2.208and2.0;theyillustratetheprogressive removalofhighfrequencymodeswithincreasing1.Onthesameabscissawehave alsoplotted(Fig.3c)theratioI(Q,I)/I(I,0)<1.Thisrathercomplicatedfunction isweaklydependentonmaterialpropertiesandisthemultiplicativefactorfordampingthesehighfrequencymodes. Thefrequencyregimeoverwhichtheintegrandintisdampedoutisaboutonedec- NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems657 -infrared-- II(C;i)xC
x10-13 nh2a2.208n.2=1.78A Xc-I0910-
FIGuRE3aTheinfluenceofthefinitevelocityofpropagationonthespectrumofinfraredanduvcontributionstovanderWaalsenergies.n2hc=2.208,dataforethaneandwater.LargeuvcontributionatI=0andsignificantdecreaseofAwith1.
IF-*---infrared-1-UV---
tI(C;l)4X Xto-s3
n2L2.0 n2=1.78 FIGURE3bTheinfluenceofthefinitevelocityofpropagationonthespectrumofinfraredanduvcontributionstovanderWaalsenergies.nfhc=2.0,dataforethaneandwater.SmalluvcontributionatI=0andweakdependenceofAon1.
BIOPHYSICALJOURNALVOLUME101970658
1. 0. 0. 0 0..0CA=__QOA
I(C;O).4%Q=50A
.2Q-500A\\ I~II}_I_ FIGURE3cTheinfluenceofthefinitevelocityofpropagationonthespectrumofinfraredanduvcontributionstovanderWaalsenergies.Retardationfactorfordampingcontribu-tionsatdifferentfrequencies.ArrowsIindicatefrequencyatwhichE=c/21.NotethatretardationdampingiseffectivelyashiftofthecurveI(t;I)/I(I;0)tothelefttocutouthigherfrequencies.Therangeofdampingisapproximatelyonedecadewideandcenteredaboutt=c/21.
adewideandroughlycenteredabout=c/(21E'12)-'(c/21)whereeho--s1.Byvirtue ofthelargeinfraredcontributiontotheenergy,evenat500A(where,10155< uvfrequencies),almosthalftheoriginalintegralisintactwhennh.,=2.208(Fig. 3a).Fornh.2=2,(Fig.3b)some78%oftheoriginalcontributiontotheintegral
remains.ItwouldappearthenthatvanderWaalsinteractionsinhydrocarbon- watersystemsarepeculiarlylongrangebecauseoftheinfraredandmicrowave spectrumofwater.Theeffectofdampingouthighfrequencymodesbyretardation isintimatelyconnectedwiththemagnitudesofoscillatorstrengthsaswellasab-sorptionfrequencies;bothmaterialpropertiesandgeometricfactorssetsimul-
taneousconditionsontheenergyspectrumandmustbeconsideredtogetherin addingupthemodalenergies. CONCLUSIONS
Thetimeislongoverdueformakingsystematicinquiryintothenatureofvander Waalsforcesinbiologicalsystems.ThepresentmethodderivedfromtheLifshitz theoryallowsonetolearnthequalitativefeaturesoftheseinteractionsaswellas makesatisfactoryestimatesoftheirmagnitude.Inadditionitmakesclearthede- pendenceofvanderWaalsforcesonbothdielectricpropertiesanddimensionsofa system. Inprincipleitisnecessarytoknowthedielectricdispersionofallcomponent substancesforallfrequencies;inpracticeitisimportantonlytoknowoscillator strengthsandmeanabsorptionfrequenciestogetagoodideaoftheelectromagnetic NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems659 forces.WefindthatfittingthesimpleformE(ic)=1+(Cmw/[l+(Q/cmmw)])+2(Cj/[I+(/lwj)2])givesanadequaterepresentationforthedielectricdispersion
asitisneededhere. Thereisaverystrongcontributiontothetotaldispersionenergyfromthein- fraredspectrumofwater.Thisfeature,certaintoholdinbiologicalsituations,is usuallyignoredinpreferencetotheuvcontributionsandisimportantincalculating thelong-distancebehaviorofthevanderWaalsforce. Ontheotherhand,twophenomenamilitateagainstimportantcontributions fromthefar-uvfluctuationfrequenciesinbiologicalsystems.First,materialsof similarweightdensitywillhavesimilardielectricsusceptibilitiesathighfrequencies (i.e.far-uvtoX-rayregion).Sincetheinteractingsurfacesareelectromagnetically definedonlywhentherearedifferencesindielectricsusceptibilityofinteracting speciesandinterveningmedium,therewillnotbeimportantbehaviorfromthe far-uv.Thisisqualitativelydifferentfromthecasewherebodiesinteractatshort distanceacrossavacuum. Second,thefinitevelocityoflight,c,causesretardationdampingacrossgaps1 atandabovefrequenciestsuchthat1/t<21/corwavelengthsX<4irl.This saysthatatlongdistance,e.g.I=50m,u=500A,thematerialpropertiesonlyfor X<500m,u=5000Aareneeded.Atshortdistancesshorterwavelengthsbegin tocontribute. Itisclearthatsubstancesthatchangetheindexofrefractionofwatercanchange thedispersionforces.Wehavecalculatedelsewhere2bythepresentmethod thatproteinorsaccharidematerialsonthebiologicalcellsurfacecangreatlyin- creaseintercellattractiveforces. Ifthecoatingmaterialsondifferentkindsofcells havedifferentuvspectra,weexpectspecificityinattractiveforcestoappearfor intercellulardistanceslessthan200A.(Itwouldbemostfortunateifthedistinctive spectrawereintheeasilyaccessible150m,u oo)limit.Inotherpapers(18)wehavederivedformulaefortriplefilms(e.g.soap bubblesinair),multilayers3(e.g.myelinfiguresandnervemyelin),andalarge numberofconfigurations2appropriatetotheinteractionofbiologicalcells witheachotherorwithasemi-infinitesubstratum. Two,therepresentationforeandnumericalintegrationasusedhereareapractical andsimpleprocedureforexaminingvanderWaalsforcesquantitatively.Webelieveitwillbeusefulinmanyconnections.
ItisourhopethattheLifshitztheorycanbeusedforausefultheoreticalandexperimentalframeworkinwhichtoviewelectromagneticforcesinawidevariety
2Ninham,B.W.,andV.A.Parsegian.Manuscriptsubmittedforpublication.Ninham,B.W.,andV.A.Parsegian.Manuscriptsubmittedforpublication.
BIOPHYSICALJOURNALVOLUME101970660
ofbiologicalproblems.Betterspectraldatamayalterthenumericalestimatesmade inthispaperbutshouldnotinvalidatethemethodofcalculationintroducedhere, ortheprincipalqualitativefeatureswhicharealreadyapparent. APPENDIX
RetardationEffects:ExplicitAnalysis
TheprecisenatureofthedampingduetoretardationdiscussedinAnalysisofIntegralscanbemadeexplicitbyacarefulanalysisoftheintegralI(t,1)whichgivesthespectrumoffrequencycontributions.Werecalleq.(8)intheform
I(t;l)=-(t/28)2fpdp{-(5P)e-(In'I)P]
+InLl(SE2+P)e-')PJ).X(A) where 2=e212]=V(1/e2)-1±p2.(A2)
Intermsofthisintegral,theHamakerfunctionandenergyaregivenas A(l)=4I(;1)d{;E(l)=-A/(12X12).(A3)
ToagoodapproximationthelogarithmswhichoccurinequationA1canbereplacedbytheirleadingterms,andweconsiderfirst
Iapprox=(t/t*)2Jpdp(SE2e-E8)P(A4)
Forafixedfrequencyi,thefunction([SE2-PEL]/[SE2+pei])isslowlyvaryingasafunctionofp.Sincetheexponentialisadecreasingfunctionofpwhilepisincreasing,theintegrandofequationA4hasamaximumatp=pogivenby
dcdp(Inp-[p/l.81)=O;POrl%W/=2tl-,*/e2(A5) Thismeansthatforafixedvalueof1,andsmalli,themajorcontributiontothepintegralcomesfromlargevaluesofp.Ontheotherhandforthesamegivenvalueof1,forlarget,Po<1andthemaincontributiontothepintegralcomesfromtheregionp-1.Two
casesthenarise (a)t2v1. ThiscorrespondstothecaseofsmalldistancesdiscussedbyLifshitz(3).Forthesevaluesoftwemayputptsintheintegrandofequation19.Thefirsttermdisappears.Inthe
NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems661 remainingintegrandthelowerlimitcanbereplacedbyzeroandwehave J(t;1)(<)2dpapiE2-Ele-tPlt8'V~~~~~~~-E2+'El/
=-fxdxln-1_)2e-] 5m=::).(A6)E2+'El
Thecorrectiontermsareoftheorderof(Q,/t)2.
AtI=50A,t8t3X1016whichhesinthemid-uv.Fort>3X1016thefunction(1C2-1lJ/[E2+El])2isalreadyverysmallcomparedwithitsmaximumvalue,sothateffectsofretardationarealsosmall.OntheotherhandforI=500A,t8liesinthenear-uvandweexpectthesehighfrequencycontributionstobediminished.
(b)t/A="<<1 ConsiderfirstthesecondintegralofequationA1
Jqt;1)(>)(Z/Z8)2pdp(5E2ElPC-pQlt.).A7)
Writingx=p-1,wehave~(~,~j2f[si(+X)(El/E2)I(t;l)(>)-"t/JoL5+(1++X))(E12)Jexp[-(t/t,)(I+x)]dx.(A8) Afterafurtherchangeofvariabletoxt/l8=y,notingthatthemajorcontributioncomesfromtheregiony=0,wecanexpandtheintegrandinascendingpowersofytoget
00(;)>)t0)exp(/8)adyo~~~~~~~~~
Xexp(-Y)(_23_/)(1+O(YAk/ID))
(tt)exp(-2/_VAsimilarcontributioncomesfromtheremainingintegrandofequationA1.Thusfort>t,(I;1)hasaformessentialyequivalenttoequationA6butwithexponentialdamp-ing.Completeasymptoticexpansionswhichconvergerapidlyforcomputationcaneasilybeconstructed.However,thereappearstobelittlepointinsodoing-theexpansionsarerathercomplicated,anditissimplertocalculatethespectrumnumericallyforeachcase. RequestsforreprintsshouldbeaddressedtoDr.Parsegian. Receivedforpublication14November1969andinrevisedform3April1970. BIOPHYSICALJOURNALVOLUME101970662
REFERENCES
1.PARSEGIAN,V.A.,andB.W.NINHAM.1969.Nature(London).224:1197.
2.HAYDON,D.A.,andJ.L.TAYLOR.1969.Nature(London).217:739.3.DZYALOSHINSKII,I.E.,E.M.LIFSHITZ,andL.P.PITAEVSKII.1959.Advan.Phys.10:165.4.KAUZMANN,W.1957.QuantumChemistry.AcademicPress,Inc.,NewYork.514.
5.JEHLE,H.1969.Ann.N.Y.Acad.Sci.158:240;JEHLE,H.,W.C.PARKE,andA.SALYERS.1964.InElectronicAspectsofBiochemistry.AcademicPress,Inc.NewYork.313.6.LONDON,F.1930.Z.Phys.63:245.7.VERWEY,E.J.W.,andJ.TH.G.OVERBEEK.1948.TheoryoftheStabilityofLyophobicColloids.N.V.UitgeversMij.Elsevier,Amsterdam.8.LIFSHrrz,E.M.1955.Zh.Eksp.i.Teor.Fiz.29:95;1956.Sov.Phys.J.E.T.P.2:73.9.CASIMIR,H.G.B.,andD.POLDER.1948.Phys.Rev.73:360.10.OVERBEEK,J.TH.G.1966.Discuss.FaradaySoc.42:7.11.LANDAU,L.D.,andE.M.LIFSHITZ.1960.ElectrodynamicsofContinuousMedia.Addison-WesleyPublishingCo.,Inc.,Mass.12.TablesofDielectricDataforPureLiquidsandDiluteSolutions.1958.NationalBureauofStandards,Washington.Circular589.13.BRAND,J.C.,andJ.C.SPEAKMAN.1960.MolecularStructure.ArnoldLtd.,London.14.HERZBERG,G.1966.ElectronicSpectraofPolyatomicMolecules.D.VanNostrandCo.,Inc.Princeton.15.CHERRY,R.J.,andD.CHAPMAN.1969.J.Mol.Biol.40:19.16.REED,R.I.1962.IonProductionbyElectronImpact.Table1.AcademicPress,Inc.,NewYork.Table1.17.SRIVASTAVA,S.N.,andD.A.HAYDON.1964.Trans.FaradaySoc.60:971.18.NINHAM,B.W.,andV.A.PARSEGIAN.1970.J.Chem.Phys.52:4578.19.OVERBEEK,J.TH.G.1952.InColloidScience.H.Kruyt,editor.N.V.UitgeversMij.Elsevier,Amsterdam.1:271.
NINHAMANDPARSEGIANVanderWaalsForces:CalculationforLipidWaterSystems663
Van Der Waals Forces Documents PDF, PPT , Doc