DTIC ADA533175: Effect of Humidity on the Interaction of Dimethyl Methylphosphonate (DMMP) Vapor with SiO2 and Al2O3 Surfaces, Studied Using Infrared Attenuated Total Reflection Spectroscopy

1 MB·English

by  Defense Technical Information Center

pubs.acs.org/Langmuir
ThisarticlenotsubjecttoU.S.Copyright.Published2010bytheAmericanChemicalSociety
Effect of Humidity on the Interaction of Dimethyl Methylphosphonate
(DMMP) Vapor with SiO and Al O Surfaces, Studied Using Infrared
2 2 3
Attenuated Total Reflection Spectroscopy
V.M.Bermudez*
NavalResearchLaboratory,Washington,D.C.20375-5347,UnitedStates
ReceivedAugust24,2010.RevisedManuscriptReceivedOctober13,2010
InfraredattenuatedtotalreflectionspectroscopyhasbeenusedtostudytheinteractionofDMMPvaporwithSiO ,
2
Al2O3, and AlO(OH) vs relative humidity (RH) and DMMP partial pressure (P/P0). For SiO2 the growth with
increasingRHofice-likeandliquid-likelayersisseeninagreementwithpreviouswork.HTDexchangeduringexposure
toH OandD Oindicatesthattheice-likelayerismoreresistanttoexchange,consistentwithstrongerH-bondingthan
2 2
intheliquid-likelayer.ExposureofnominallydrySiO toD OindicatestheexistenceofadsorbedH Othatdoesnot
2 2 2
exhibitanice-likespectrum.Theice-likelayerappearsonlyatafiniteRH.ExposureofSiO toDMMPintheabsenceof
2
intentionallyaddedH Oshowstheformationofastronglyboundmolecularspeciesfollowedbyaliquid-likelayer.The
2
stronginteractioninvolvesSiO;H333OdPbondstosurfacesilanolsand/orHO;H333OdPbondstopreadsorbed
molecularH2O.AtafiniteRHtheice-likelayerformsonSiO2eveninthepresenceofDMMPuptoP/P0=0.30.
DMMPdoesnotappeartopenetratetheice-likelayerundertheseconditions,andthetendencytoformasuchalayer
drivesthedisplacementofDMMP.AmorphousAl O andAlO(OH)donotexhibitanice-likeH Olayer.Bothhavea
2 3 2
highersurfaceOHcontentthandoesSiO ,whichleadstohighercoveragesofH OorDMMPatequivalentRHor
2 2
P/P0.AtlowP/P0,forwhichadsorptionisdominatedbyAl;OH333OdPbonding,a-Al2O3interactswithDMMP
morestronglythandoesAlO(OH)asaresultofthehigheracidityofOHsitesontheformer.UptoRH=0.30and
P/P0=0.30,DMMPappearstoremainbondedtothesurfaceratherthanbeingdisplacedbyH2O.H2Oappears
tohavelittleornoeffectonthetotalamountofDMMPadsorbedonanyofthesesurfaces,uptoanRHof0.30and
aP/P0of0.30.TheresultshaveimplicationsforthetransportofDMMPandrelatedmoleculesonoxidesurfacesin
theenvironment.
1. Introduction RH. The existence of this multilayer, particularly the ice-like
layer,hasimplicationsinsuchareasastribologyandbiology.In
Under ambient environmental conditions, oxide surfacesare
the former, a strongly bound surface layer of H O can affect
typically covered with a layer of adsorbed molecular H O in 2
2 slidingfriction.Inthelatter,suchalayerisimportantinenhan-
addition to surface hydroxyl (OH) groups. The transport of
cingbiocompatibilityofmaterialsbyinhibitingtheadhesionof
organicreagentsthroughthislayerandthecompetitionbetween
proteins,whichisadrivingmechanismforbiofouling.
suchspeciesandH Oforadsorptionsitesareimportantinseveral
areasoftechnology2.1,2Studyingsuchphenomena,whichdepend SimilarIRstudieshavebeenreportedforH2Ovaporinteract-
ingwiththe(0001)surfaceofbulkR-Al O usingATR(ref7)as
criticallyonthestructureoftheH O/substrateinterface,requires 2 3
2 wellastransmissionthrougha“pileofplates”(ref8).Thissystem
experimentaltechniquescapableofprovidingbond-specificche-
has also been investigated by X-ray diffraction,9 and X-ray
mical information under “practical” steady-state conditions.
photoemissionspectroscopy10(XPS)and desorption methods11
Infrared (IR) attenuated total reflection (ATR) spectroscopy
under steady-state conditions has been used previously3-6 to havebeenappliedafterhighexposurestoH2Ovapor.Evidenceis
againfound8,9forapartialorderingoftheH Olayerincontact
observetheinteractionofH Ovaporwiththinlayersofamor- 2
2 with the surface, which is fully O-terminated when exposed to
phous SiO2 (a-SiO2) on Si. These studies reveal a complex H O vapor9,10 and which exhibits a structure intermediate be-
structure thatdepends onrelative humidity (RH).The layer in 2
tweenthoseofR-Al O andγ-Al(OH) .However,inonestudy7a
closestproximitytothehydroxylatedSiO surfaceformshydro- 2 3 3
2 partialwettingofthesurface,indicatingdropletformation,was
genbonds(H-bonds)toSi-OHgroupsresultinginan“ice-like”
documented.WettingoftheAl O surfacemaybemorecomplex
orderedstructurewithtetrahedrallycoordinatedH O.Theout- 2 3
2 thanthatofSiO ,sincerecentwork12suggeststhatOHgroupsin
ermost layer is “liquid-like”, and there is a transition layer 2
thiscasecanactuallymakethesurfacehydrophobic.Itisnoted
between the two. The different layers are recognizable in the
thatseveralstudiesoftheH O/SiO andH O/R-Al O interfaces
structureoftheH Ovibrationalspectrumanditsvariationwith 2 2 2 2 3
2 havealsobeendoneusingsum-frequencygenerationtoprobethe
*Phone: þ1-202-767-6728. FAX: þ1-202-767-1165. E-mail: victor. (7)Thomas,A.C.;Richardson,H.H.J.Phys.Chem.C2008,112,20033.
[email protected]. (8)Al-Abadleh,H.A.;Grassian,V.H.Langmuir2003,19,341.
(1)Minakata,S.;Komatsu,M.Chem.Rev.2009,109,711. (9)Eng,P.J.;Trainor,T.P.;Brown,G.E.,Jr.;Waychunas,G.A.;Newville,M.;
(2)Ewing,G.E.J.Phys.Chem.B2004,108,15953.Chem.Rev.2006,106,1511. Sutton,S.R.;Rivers,M.L.Science2000,288,1029.
(3)Asay,D.B.;Kim,S.H.J.Phys.Chem.B2005,109,16760. (10)Liu,P.;Kendelewicz,T.;Brown,G.E.,Jr.;Nelson,E.J.;Chambers,S.A.
(4)Barnette,A.L.;Asay,D.B.;Kim,S.H.Phys.Chem.Chem.Phys.2008,10, Surf.Sci.1998,417,53.
4981. (11)Elam,J.W.;Nelson,C.E.;Cameron,M.A.;Tolbert,M.A.;George,S.M.
(5)Asay,D.B.;Barnette,A.L.;Kim,S.H.J.Phys.Chem.C2009,113,2128. J.Phys.Chem.B1998,102,7008.
(6)Anderson,A.;Ashurst,W.R.Langmuir2009,25,11549. (12)Gentleman,M.M.;Ruud,J.A.Langmuir2010,26,1408.
18144 DOI:10.1021/la103381r PublishedonWeb11/11/2010 Langmuir2010,26(23),18144–18154
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
13 OCT 2010 2. REPORT TYPE 00-00-2010 to 00-00-2010
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Effect of Humidity on the Interaction of Dimethyl Methylphosphonate
5b. GRANT NUMBER
(DMMP) Vapor with SiO2 and Al2O3 Surfaces, Studied Using Infrared
Attenuated Total Reflection Spectroscopy 5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Naval Research Laboratory,Washington,DC,20375 REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT

15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 11
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
Bermudez Article
ν(O-H) stretchingmodes;however,thesehave dealtwithliquid surfaces.Theoxidesofinterest,SiO andAlO,arechosenbecause
2 2 3
HOandnotwiththevapor,whichisthesubjectofinteresthere. of their presence in many forms of soils, minerals, and building
2
One study13 applied another laser technique, cavity ring-down materials and because their surface chemistries have similarities
spectroscopy,toastudyofHOvibrationalmodesatthevapor/ withawiderangeofotheroxides.TheformofAlO usedhereis
2 2 3
bulk-a-SiO2interfaceandfoundevidenceforanorderedmonolayer amorphous (a-Al2O3) grown as described in section 2.3. The
similartothatdetectedinATRexperimentsonthinSiO2films. structure of a-Al2O3 has recently been studied27 using nuclear
Computational modeling of the interface between HO and magneticresonance(NMR)spectroscopy,whichshowsthatmost
2
eitherSiO orAlO hasbeencarriedoutusingmoleculardynamics Alatomsarefour-or5-foldcoordinated.Thus,thehydroxylated
2 2 3
(MD).ForSiO therehavebeennumerousstudiesusingeitherab surfaceisexpectedtoapproximatethatofhydroxylatedγ-AlO
2 2 3
initio density functional theory (DFT) or force-field methods, (ref28),whichalsocontainsfour-and5-fold-coordinatedsurface
crystalline or amorphous SiO models (with or without surface Alsitesbeforehydroxylation.
2
silanol),liquid-orvapor-phaseHO,andplanarorporousstruc- DMMPisimportantasasimulantfortoxicphosphorylcom-
2
tures. A wide range of temperatures is also encompassed in the poundsthatconstitutehazardstotheenvironmentandtoperson-
variousMDstudies.Here,wementiononlythoseworks14-21that nel. The interaction between DMMP and HO (either vapor or
2
aremostcloselyrelatedtothepresentexperiments.Thesereport liquid) has been studied previously using matrix-isolation IR
resultsatroomtemperatureforhydroxylatedsurfacesusingeither spectroscopycombinedwithabinitiocomputationalmodeling,29
an amorphous model or a series of different crystalline quartz bymolecularmechanicscalculations30(seealsoref31),byabinitio
surfaceplanestorepresentadisorderedsubstrate.Theresultsare theory,32 and by NMR and IR spectroscopies.33 In the vapor
generallyconsistentwithexperimentaldatashowingastructured phase29anH-bondformsbetweenasingleHOandtheOatomof
2
layercharacterizedbystrongH-bondingthattransitionstoaliquid thePdOgroupwithabondenergyof7.7kcal/molcomputedatthe
layer with increasing distance from the surface. However, the MP2/6-31þG(d,p)level.AthigherHOconcentrations,29,30,33i.e.,
2
detailed structure of this layer depends on the coverage and on inaqueoussolution,twoHOmoleculesbindtothePdOgroupofa
2
therelativepositionsofsurfaceSi-OHgroups.Thestructuredlayer singleDMMP.TheOatomsoftheCHOgroupsarefound29,30to
3
isconfinedtowithin∼10A˚ fromtheSiO surface,andthediffusion be largely ineffective in H-bonding to HO. Dissolving DMMP
2 2
of HO within this layer is slower than in bulk liquid HO. In inHOisfound,inabinitiomodels,32toaffecttherelativestability
2 2 2
contrast to SiO there has been relatively little computational ofdifferentconformersandtolowerthebarriertointerconversion
2
work22,23 done for high coverages ofHO on AlO.The results relativetothevaporphase.
2 2 3
forR-Al2O3(0001)concurwiththeexperimentallyobserved9high TheadsorptionofDMMPvaporona-SiO2(refs34-37)and
reactivity between HO and the OH-free surface and with the on OH-terminated organic self-assembled monolayers (SAMs,
2
resultingrestructuringofthesurface.Athighercoveragemolecular refs38-41)hasbeenstudiedusingprimarilyIRspectroscopy.A
HO forms H-bonded structures at OH sites resulting from the similarmolecule(DIMP),inwhichtheCH OgroupsofDMMP
2 3
initialdissociativeadsorptiononthebaresurface.Arecentreview24 arereplacedwith(CH ) CHO,hasbeenstudied42whileinteract-
32
discussesthepropertiesofvariousAlO surfacesunderambient ingwithOH-terminatedSAMs.Anabinitiocomputationalstudy
2 3
conditions. of DMMP/a-SiO2 has also been reported.43 The general con-
The interface between H O vapor and amorphous SiO or sensusisthatDMMPadsorbsnondissociativelybyformationof
2 2
Al2O3 has also been studied25,26 in situ using XPS, near-edge one or more O;H333OdP bonds, although one study35 has
X-ray absorption fine-structure (NEXAFS) and work function suggested H-bonding to the methoxy O atoms. Similar experi-
measurements.TheSiO resultscallintoquestiontheexistenceof mental studies have been performed for adsorption on various
2
the ice-like layer at room temperature. In particular, work- forms of Al O (refs 44-47), and an ab initio computational
2 3
functiondata25showlittleornoeffectduetoformationofthe
initialH Olayer.However,inthemodelproposedbyAsayetal.3
the ice-l2ike layer involves no net dipole moment, which is (27)Lee,S.K.;Lee,S.B.;Park,S.Y.;Yi,Y.Y.;Ahn,C.W.Phys.Rev.Lett.
2009,103,095501.
consistentwiththeabsenceofachangeinworkfunction. (28)Digne,M.;Sautet,P.;Raybaud,P.;Euzen,P.;Toulhoat,H.J.Catal.2002,
In the present work we are concerned with the interaction 211,1.J.Catal.2004,226,54.
(29)Ault,B.S.;Balboa,A.;Tevault,D.;Hurley,M.J.Phys.Chem.A2004,108,
betweenvaporsofdimethylmethylphosphonate((CHO) (CH)-
3 2 3 10094.
PdO,DMMP)andHOwhileincontactwithrepresentativeoxide (30)Vishnyakov,A.;Neimark,A.V.J.Phys.Chem.A2004,108,1435.
2
(31)Sokkalingam,N.;Kamath,G.;Coscione,M.;Potoff,J.J.J.Phys.Chem.B
2009,113,10292.
(13)Aarts,I.M.P.;Pipino,A.C.R.;Hoefnagels,J.P.M.;Kessels,W.M.M.; (32)Flori(cid:1)an,J.;S(cid:3)trajbl,M.;Warshel,A.J.Am.Chem.Soc.1998,120,7959.
vandeSanden,M.C.M.Phys.Rev.Lett.2005,95,166104. (33)Eaton,G.;Harris,L.;Patel,K.;Symons,M.C.R.J.Chem.Soc.Faraday
(14)Notman,R.;Walsh,T.R.Langmuir2009,25,1638. Trans.1992,88,3527.
(15)Argyris,D.;Cole,D.R.;Striolo,A.J.Phys.Chem.C2009,113,19591. (34)Siu,E.Y.;Andino,J.M.Nano2008,3,233.
Langmuir2009,25,8025. (35)Kanan,S.M.;Tripp,C.P.Langmuir2001,17,2213.Kanan,S.M.;Tripp,C.P.
(16)Lopes,P.E.M.;Murashov,V.;Tazi,M.;Demchuk,E.;MacKerell,A.D., Langmuir2002,18,722.
Jr.J.Phys.Chem.B2006,110,2782. (36)Henderson,M.A.;Jin,T.;White,J.M.J.Phys.Chem.1986,90,4607.
(17)Du,Z.;deLeeuw,N.H.DaltonTrans.2006,2623. (37)Ferguson-McPherson, M. K.; Low, E. R.; Esker, A. R.; Morris, J. R.
(18)Puibasset,J.;Pellenq,R.J.-M.J.Chem.Phys.2003,118,5613. J.Phys.Chem.B2005,109,18914.
(19)Warne,M.R.;Allan,N.L.;Cosgrove,T.Phys.Chem.Chem.Phys.2000,2, (38)Bertilsson, L.; Potje-Kamloth, K.; Liess, H.-D. Thin Solid Films 1996,
3663. 284-285,882.
(20)Chuang,I.-S.;Maciel,G.E.J.Phys.Chem.B1997,101,3052. (39)Bertilsson,L.;Engquist,I.;Liedberg,B.J.Phys.Chem.B1997,101,6021.
(21)Lee,S.H.;Rossky,P.J.J.Chem.Phys.1994,100,3334. (40)Bertilsson,L.;Potje-Kamloth,K.;Liess,H.-D.;Engquist,I.;Liedberg,B.
(22)Hass,K.C.;Schneider,W.F.;Curioni,A.;Andreoni,W.J.Phys.Chem.B J.Phys.Chem.B1998,102,1260.
2000,104,5527. (41)Bertilsson, L.; Potje-Kamloth, K.; Liess, H.-D.; Liedberg, B. Langmuir
(23)Thissen,P.;Grundmeier,G.;Wippermann,S.;Schmidt,W.G.Phys.Rev.B 1999,15,1128.
2009,80,245403.Wippermann,S.;Schmidt,W.G.;Thissen,P.;Grundmeier,G.Phys. (42)Crooks,R.M.;Yang,H.C.;McEllistrem,L.J.;Thomas,R.C.;Ricco,A.J.
StatusSolidiC2010,7,137. FaradayDiscuss1997,107,285.
(24)Kelber,J.A.Surf.Sci.Rep.2007,62,271. (43)Bermudez,V.M.J.Phys.Chem.C2007,111,9314.
(25)Verdaguer,A.;Weis,C.;Oneins,G.;Ketteler,G.;Bluhm,H.;Salmeron,M. (44)(a) Mitchell, M. B.; Sheinker, V. N.; Mintz, E. A. J. Phys. Chem. B
Langmuir2007,23,9699. 1997, 101, 11192. (b) Sheinker, V. N.; Mitchell, M. B. Chem. Mater. 2002, 14,
(26)Deng,X.;Herranz,T.;Weis,C.;Bluhm,H.;Salmeron,M.J.Phys.Chem.C 1257.
2008,112,9668. (45)Aurian-Blajeni,B.;Boucher,M.M.Langmuir1989,5,170.
Langmuir2010,26(23),18144–18154 DOI:10.1021/la103381r 18145
Article Bermudez
studyofadsorptiononthehydroxylatedγ-Al O surfacehasbeen Theinternal-reflectionangleofincidenceofθ=60(cid:1),whichis
2 3
done.48 Adsorption in this case is nondissociative at room wellabovethecriticalangle(θ ≈17(cid:1)forSiinthemid-IR),wasa
c
temperatureandoccursbyH-bondingbetweenanacidicOHsite compromise.55 This choice of θ gave a total of eleven internal
andtheOatomofthePdO;however,decompositionbeginsata reflectionssamplingthevapor-prisminterface.Asmallerθwould
give more reflections per unit length of prism and thus higher
littleaboveroomtemperature.AlmostalltheSiO andAlO studies
2 2 3 sensitivity.However,thelongeropticalpathlengththroughtheSi
wereperformedinvacuoonhigh-surface-area(HSA)powders.
would then move the transmission cutoff (due to multiphonon
Theobjectiveofthepresentworkistogaininsightintohow
absorption)tohigherenergy,whichwouldobscuremoreofthe
RH affects the adsorption and transport of DMMP (and, by
DMMPspectrum.Furthermore,theamplitudeoftheevanescent
itmanpcleicoatfioRnH,ohfassimbeileanrnreoatgedenptsr)eviniotuhselye4n9viinroansmtuednyt.oTfhaecitmivpatoerd- ealmecbtireicntfmieelddiudmec,a5y7sanadsthexepp(e-nze/tdrpa)tiownitdhepdtihst(adnpc)ein(czr)eaisnetsowtihthe
carbon, where adsorption of H O was seen to inhibit that of decreasingθaboveθ,thusincreasingtheunwantedcontribution
2 c
DMMP.Asimilarinhibitioneffecthasbeenreportedrecently50 from vapor-phase absorption. With the present configuration
forDMMPadsorptiononSiO2withpreadsorbedH2O.Onthe dp/λIR ≈ 0.056 and data could be obtained with essentially no
other hand, the interaction of DMMP with OH-terminated contributionfromthevapor.
SAMshasbeenseen41toincreasewithRH.Basedontheabove Therangecoveredinthiswork,∼4000-1000cm-1,islimitedat
thelowendbythetransmissionoftheSiATRprism.Thisrange
discussion one expects a complex three-way interaction among
includesasignificantpartofthemid-IRspectrumofDMMP,and
DMMP,H OandsurfaceOHgroups,whichwillalsodependon
2 thestructureandmodeassignmentsfortheliquid58,59andvapor
theacid/basecharacter48,51oftheOH.Itisnoted,forexample, phases60,61arediscussedelsewhere.Featuresoftenseenat1152
thatthecomputedadsorptionenergy(ΔEads)forDMMPonthe and 1206 cm-1 are the ν(C-F) stretching modes in Teflon
hydroxylatedSiO surfaceis20.0kcal/mol(ref43),whichisonlya contacting the optically accessible edges of the Si prism. These
2
littlegreaterthanthatcalculated52forthemoststableconfigura- are strong and, due to small instabilities in the optical system,
tionofH Odimersonsuchasurface(748mev/H O=17.4kcal/ oftendonotcompletelycancelintheratio.StrongSimultiphonon
2 2
mol).ThissuggeststhatcompetitionbetweenDMMPandH O absorptions alsodonot completelycancelinall spectra, which
foradsorptionsitesmaybeanimportantfactor. 2 leadstobaselineartifacts,particularlynear1450cm-1.
2.2. Reagent Preparation and Handling. Reagent-grade
DMMP(PfaltzandBauer,99.5%),D O(Aldrich,99%D),and
2. ExperimentalDetails 2
locally prepared deionized (DI) H O were used as received.
2
2.1. ATRSpectroscopy.Thebasicexperimentalapproachis TypicallythetotalN flowthroughtheIRcellwas100mL/min
2
reviewed elsewhere.53 Silicon parallelepipeds (termed “ATR andwaskeptconstantduringaseriesofexperiments.Partofthis
prisms”)measuring25(cid:2)15(cid:2)1mm3wereobtainedfromHarrick flow waspure, dry N obtainedfromliquid-N boil-off, which
2 2
Scientific(Pleasantville,NY)orfromSpectralSystems(Hopewell passed through a coil of Cu tubing immersed in a constant-
Junction,NY).Thesewerepreparedfromfloat-zonematerialin temperaturebathat25.0(cid:1)C.AnotherpartwasN fromthesame
2
ordertoavoidthestrongabsorptionbandat1106cm-1duetothe sourceafterpassagethroughaporousglass“bubbler”(orfritted
bulk oxygen impurity54 found in Si grown by the Czochralski gas-washingbottle)filledwithDMMP.Athirdpartoftheflow
(silicacrucible)method.Theprismswerecoveredonbothsides consistedofN throughasecondbubblerwithH OorD Oand
2 2 2
with an oxide layer, sandwiched between two hollow Teflon wasusedtoadjusttheRH,whichwascheckedwithatempera-
blocks55 and the whole assembly squeezed tightly together to ture/dew-point/humidity meter (Model 4080; Control Co.;
formavapor-tightsealaroundtheedgesoftheprism.Thecellwas Friendswood,TX).Inthefollowing,RH=1.00meansarelative
fitted with Teflon tubes through which vapor flowed over the humidityof100%.Allreagentswerepurgedofdissolvedgases
exposedprismfaces. priortousebyflushingN throughthebubblersatahighflow
2
SpectrawererecordedusingaFouriertransforminfrared(FTIR) rate.Thepartial-pressures(P/P0,whereP0isthevaporpressureat
spectrometerwitha“narrow-band”Hg Cd Te(MCT-A)detec- 25(cid:1)C)andRHvaluesarebasedontherelativeflowrates.For
x 1-x
tor. No polarizer was used, and the intrinsic polarization of the DMMP,P0=0.84TorrintheabsenceofH2Ovapor;62forH2O
beamproducedbyanFTIRspectrometeristypicallyfairlyweak.56 (D2O),P0=23.8(20.5)Torr.63
Usually2000scanswereaveragedat4cm-1resolution,and2-or Reagentpuritycouldbecheckedbypassingtheeffluentfrom
4-fold zero filling and triangle apodization were applied to the the ATR cell through a 10 cm gas cell (with KBr windows)
interferogrambeforetransformation.Datawereobtainedbyratio- mountedinthespectrometersamplecompartment.Noevidence
ingsingle-beamspectrarecordedwithandwithoutreagentintheN oftraceimpuritieswasseenintheIRspectrumofthevaporstream
2
stream(seefollowingsection),fromwhichwasobtainedδR/R,the exitingtheATRcell.Inperformingthisanalysisthespectrumofthe
fractionalchangeinreflectancecausedbyadsorption.Nosmooth- DMMP effluent wascompared with vapor-phase data for dried
ing and, except where noted, no background subtraction were andvacuum-distilledDMMPgivenelsewhere.60Recentlyananal-
appliedtothedata.Allspectraaredividedbythenumberofinternal ysisoftraceimpuritiesincommerciallyavailableDMMPhasbeen
reflectionstogiveδR/Rperreflection. reported;64 however, a systematic search for evidence of these
speciesinthepresentmaterialwasnotperformedhere.
Indesigningthecoadsorptionexperiments,attentionmustbe
(46)(a)Templeton,M.K.;Weinberg,W.H.J.Am.Chem.Soc.1985,107,97.(b) giventothesuppressionofDMMPvolatility49,65byH Ovapor.
J.Am.Chem.Soc.1985,107,774. 2
(47)Davies,P.R.;Newton,N.G.Appl.Surf.Sci.2001,181,296.
(48)Bermudez,V.M.J.Phys.Chem.C2009,113,1917.
(49)Kaplan,D.;Nir,I.;Shmueli,L.Carbon2006,44,3247. (57)Harrick,N.J.InternalReflectionSpectroscopy;HarrickScientific:Ossining,
(50)Hearn, J.; Weber, R.; Henley, M.; Cory, M.; Runge, K.; Hurley, M.; NY,1987.
Taylor, DeC.; Burns, D. 2009 Chemical and Biological Defense Science and (58)vanderVeken,B.J.;Herman,M.A.PhosphorousSulfur1981,10,357.
TechnologyConference(Dallas,TX;Nov.16-20,2009),PosterT201. (59)Moravie,R.M.;Froment,F.;Corset,J.Spectrochim.Acta1989,45A,1015.
(51)Leung,K.;Nielsen,I.M.B.;Criscenti,L.J.J.Am.Chem.Soc.2009,131, (60)vanderVeken,B.J.;Herman,M.A.J.Mol.Struct.1983,96,233.
18358. (61)Bermudez,V.M.J.Phys.Chem.C2007,111,3719.
(52)Yang,J.;Meng,S.;Xu,L.;Wang,E.G.Phys.Rev.B2005,71,035413. (62)Butrow,A.B.;Buchanan,J.H.;Tevault,D.E.J.Chem.Eng.Data2009,54,
(53)Hind,A.R.;Bhargava,S.K.;McKinnon,A.Adv.ColloidInterfaceSci. 1876.
2001,93,91. (63)Besley,L.;Bottomley,G.A.J.Chem.Thermodyn.1973,5,397.
(54)Kaiser,W.;Keck,P.H.;Lange,C.F.Phys.Rev.1956,101,1264. (64)Hoggard,J.C.;Wahl,J.H.;Synovec,R.E.;Mong,G.M.;Fraga,C.G.
(55)Queeney,K.T.;Fukidome,H.;Chaban,E.E.;Chabal,Y.J.J.Phys.Chem. Anal.Chem.2010,82,689.
B2001,105,3903. (65)Tevault,D.E.;Buchanan,J.H.;Buettner,L.C.Int.J.Thermophys.2006,
(56)Sperline,R.P.Appl.Spectrosc.1991,45,677. 27,486.
18146 DOI:10.1021/la103381r Langmuir2010,26(23),18144–18154
Bermudez Article
Data reported by Tevault et al.65 show the maximum DMMP
partialpressurethatcanbesustainedforagiventemperatureand
RH. Attempting to increase P/P0 beyond this point leads to
precipitationintheformofanaerosol.49Thus,forexample,at
25(cid:1)CandRH=0.33themaximumP/P0forDMMPisabout
0.41. The dew point for such a mixture65 is ∼25.0 (cid:1)C. In the
present work, the semitransparent Teflon tubing downstream
from the region of mixing was checked for any indication of
aerosolformation.Thiswasdonebyilluminatingwithabright
lightandlookingforevidenceoflightscattering.49
2.3. Sample Preparation and Characterization. Follow-
ing previous3-5 work, the a-SiO2/Si samples were grown using
ultraviolet/ozone(UV/O )oxidation.Thiswasdoneinorderto
3
facilitate comparison with these studies. Some of the data re-
ported in ref 6, with which the present results will also be
compared, were obtained for a-SiO2 films grown by plasma
oxidation.Datafora thinthermal oxide were alsoobtained in
the course of the present work in order to check for any
dependence of the results on oxide growth method. The ATR
prismwasfirstcleanedbyimmersioninwarm(50(cid:1)C)acetonefor
∼3010h(cid:1)Can(dMthoednelsuUbVje-c1t;edSatmoUcoVI/nOte3ronxaitdioantiaoln;Kfoyro1toh,aJtanpoamn)i.nTalhlye aFdigsuorrebe1d.oInRUAVT/RO3dSaitOa2fvosrRthHeaνt(Oro-oHm)tesmtreptecrhaitnugreb(a2n4d(cid:1)Cof).HT2hOe
verticalscalegivesδR/Rperinternalreflection.Thenumbersgive
feed-gas was O enriched in O by passage through a “silent
2 3 theRHforeachspectrumstartingwiththeupper-most,whereRH
discharge” prior to entering the UV-irradiated area. The UV
=1.00meansarelativehumidityof100%.Thespectrahavebeen
radiation, which produces additional O , was provided by a
3 displacedverticallyforclarity,butthesmallslopingbackground
low-pressureHg-vapor/inert-gaslampemittingmainlyatλ =
UV presentinsomespectrahasnotbeenremoved.Twoofthespectra
254nm.TheoxidewasthenstrippedinaqueousHFsolutionand
areshownincolortohelpguidetheeye.Theweakfeaturelabeled
thesamplerinsedinDIH OanddriedinastreamofN .TheUV/
2 2 “freeOH”isdiscussedinsection3.1.1.
O treatmentwasrepeatedfor1h,followedbya15-minimmersion
3
in“RCA-1”solution(H O:NH OH(30%):H O (30%)ina3:1:1
2 4 2 2
ratiobyvolume)at75-80(cid:1)C,followedbyrinsinginDIH Oand measured since contact with liquid HO is known to alter the
2 2
dryinginN . surfacepropertiesofAlO (discussedinsection3.1.4below).After
2 2 3
Thecontactangle(CA)forDIH Owasmeasuredatseveral mountingintheATRcell,thesamplewasexposedovernighttoa
2
spotsonthecleanoxidizedsurface.TheCAwasalways<10(cid:1)and 100mL/minflowofN saturatedwithHO(RH=1.0)inorderto
2 2
often<5(cid:1),whichindicatesahighlyhydrophilicsurface,i.e.,one hydroxylatethesurfaceandtoprecludeanysignificantchangeinthe
whichisbothfreeofsignificanthydrocarboncontaminationand surface condition during subsequent experiments involving HO
2
highinsilanolcoverage.66Previouswork67,68forUV/O3oxida- vapor.Thecharacterizationofthea-Al2O3samplesusingXPSand
tionofSiindicatesalimitingSiO thicknessofabout2.5nmfor IRtransmissionisdescribedintheSupportingInformation.
2
irradiation in pure O (some of which is photochemically con-
2
vertedtoO )atasubstratetemperatureof300(cid:1)C.Thepresent
3
sampleswereanalyzedusingXPS,whichshowedanSiO thick- 3. ResultsandDiscussion
2
nessof∼2.1nm.SurveyXPSscanswerealsodonetocheckfor 3.1. Amorphous SiO . The discussion will begin with an
contamination,andthesampleswerefurthercharacterizedbyIR examinationofthepure-H2Oandpure-DMMPsteady-stateIR
transmission(seetheSupportingInformation).Filmsofa-SiO2 ATRdatawhichisnecessar2yforanappreciationofthecoadsorp-
werealsopreparedbythermaloxidationindryO at800(cid:1)C.Prior
2 tionresults.TheH Odatawillalsoservetovalidatethepresent
to growth the substrate was cleaned in acetone and in RCA-1 2
samples and methods by comparison with previous results.
solution as described above. The oxide thickness, determined
from XPS measurements, was ∼8.1 nm. Thermal oxides of InfrareddataforDMMPathighP/P0havenot,toourknowl-
thickness g3 nm are essentially equivalent to bulk a-SiO2, as edge,beenreportedpreviouslyforSiO2orAl2O3.
shownbyIRspectroscopy.69TheCAresultsafterRCA-1clean- 3.1.1. H2OinPureN2.Figure1showsdataforSiO2vsRH
ingwerethesameasthosedescribedabovefortheUV/O oxide. (in the absence of DMMP). The results essentially reproduce
3
TheoxidizedsamplewascleanedagaininRCA-1solutionbefore those given previously,3-6 which reveal a band at about 3240
mountingintheATRcell. cm-1duetotheice-likelayer andanother atabout 3400cm-1
AmorphousAl O filmswerepreparedcommerciallybyradio
2 3 assignedtotheliquid-likelayer.Incontrast,theδ(HOH)bending
frequency magnetron sputter deposition (PVD Products;
modeat1640cm-1(notshown)doesnotindicateanydistinction
Wilmington,MA)ontoSiATRprisms.Thefilmthicknesswas
∼8.5nm,whichhasbeenshown70tobesufficienttoachieveIR between the ice- and liquid-like phases.3-6 The weak, sharp
optical properties equivalent to those of bulk material. The structureinthe3600-3900cm-1rangeisduetoH2Ovapor.In
samples were cleaned by immersion in warm (∼50 (cid:1)C) hexane the lower-RH spectra this arises from the incomplete dry-N
2
andwarmacetonefor10mineachfollowedbya30-minUV/O purging of the optical path. At higher RH it represents the
3
exposureat nominalroomtemperature. The H O CA was not contribution from H O vapor in the cell under steady-state
2 2
conditions.Thelowintensityrelativetothatofthesurfacespecies
indicates the excellent rejection of vapor contributions in the
(66)Cras, J. J.; Rowe-Taitt, C. A.; Nivens, D. A.; Ligler, F. S. Biosens.
Bioelectron.1999,14,683. presentATRexperiment.Aweakfeatureat∼3740cm-1,labeled
(67)Ishikawa,Y.;Shibamoto,T.;Nakamichi,I.Jpn.J.Appl.Phys.1992,31, “free OH”, is seen in most IR data for thin H O films and is
2
1148. assigned to dangling (i.e., not H-bonded) O-H bonds on the
(68)Saitoh,T.;Kobayashi,D.;Kimura,D.;Asai,K.Mater.Res.Soc.Symp.
Proc.1999,569,101. surfaceoftheH2Olayer.Figure2showstheisotherm,derived
(69)Queeney,K.T.;Weldon,M.K.;Chang,J.P.;Chabal,Y.J.;Gurevich, fromthedatainFigure1,whichissimilartothatofAsayetal.3
A.B.;Sapjeta,J.;Opila,R.L.J.Appl.Phys.2000,87,1322.
(70)Bru€esch,P.;Ko€tz,R.;Neff,H.;Pietronero,L.Phys.Rev.B1984,29,4691. Thisisotherm,andthosereportedlaterforothermaterials,areof
Langmuir2010,26(23),18144–18154 DOI:10.1021/la103381r 18147
Article Bermudez
Figure2. Adsorption isothermconstructedbyplotting the inte- Figure3. DataforexposureofSiO2tomixedH2OandD2O.The
gratedbandareainFigure1vsRH.Thesmoothcurvethroughthe H O was maintained at RH = 0.50 while the RH of D O was
2 2
pointsissimplyavisualaid.Theverticalscaleisinunitsoftotal increased in stages. The sloping background present in some
absorbance (not divided by the number of internal reflections) spectrahasnotbeenremoved.Assumingcompleteisotopicmixing,
times wavenumbers and is given explicitly in order to facilitate the H O:HDO:D O relative concentrations are (a) 1:0:0, (b)
2 2
quantitativecomparisonwithisothermsforothermaterials. 0.83:0.16:0.01,(c)0.69:0.28:0.03,and(d)0.59:0.36:0.05.
thetypeIIform71characteristicofahydrophilicsurface,onefor Figure3showsdataforSiO inH OvaporatanRHof0.50
2 2
which the heat of H O adsorption is greater than that of and anincreasingRH ofD O.An RH of0.50issufficientfor
2 2
condensation. completion of the H O ice-like layer (cf. Figure 1). The IR
2
There are, however, some differences from previous results. spectrumofisotopicallymixedwateriscomplexandisdiscussed
Asay et al.,3,5 working with a sample at slightly below room indetailelsewhere,mostrecentlyinref76.InFigure3,additionof
temperature,observedthesequentialformationofice-like,mixed D OleadstothegradualattenuationoftheH Oice-likebandat
2 2
(ice-like and liquid-like), and liquid-like phases with increasing ∼3240 cm-1. The loss of ice-like intensity is interpreted as the
RH.At20.8(cid:1)CandRH<0.30,thespectrumwasdominatedby resultofdisorderingcausedbytheadmixingofHDOandD O.
2
the ice-like phase. On the other hand, Anderson and Ashurst6 ForhigherRHsofD O,interpretationisdifficultduetothefact
2
observedthesimultaneouspresenceofbothformsatallvaluesof thatthespectrainvolverelativelylargecontributionsfromHDO,
RH, with the ice-likebandalways being themoreintense.The forwhichν(O-H)andν(O-D)differonlyslightlyfromthoseof
resultsinFigure1exhibitcharacteristicsofbothsetsofdata.Even H OandD O(ref77).TheformationofHDOisrevealedbythe
2 2
atthelowestRHthespectrumappearstobeasumofapproxi- appearance of the δ(HOD) bending mode77 at 1450 cm-1 (not
matelyequalcontributionsfrombothforms;however,theliquid- shown). It is noteworthy, however, that the HO ice-like band
2
likelayerbecomesdominantathigherRHvalues. remainsdetectableevenforaDO/HOratioof∼1/3andbeyond,
2 2
It is possible to propose a tentative interpretation for these disappearingcompletelyonlyforaratioof∼1/1(notshown),for
differences.TheresultsofAsayetal.3,5suggestauniformgrowth whichtherelativeHO:HDO:DOconcentrationis0.25:0.50:0.25.
2 2
ofanice-likelayeratlowRH,followedbygrowthofaliquid-like Thefactthattheice-likebanddoeseventuallydisappearindicates
multilayerathigherRH.ThoseofAndersonandAshurst,6onthe thattheshoulderat∼3240cm-1inFigure3isnotduetoice-like
other hand, are consistent with the formation of droplets that HDO.
increaseinsizetowardhigherRHbutwhichnevercoalesceintoa Theseresultssuggestthatisotopeexchangeinvolvingtheice-like
continuouslayer.Thepresentresultsseemtoindicatetheinitial layerisslowerthanfortheliquid-likelayer,whichisconsistentwith
formation of droplets that, with increasing RH, merge into a astrongerintermolecularinteractionwithintheice-likelayer.Some
continuous multilayer. These differences probably arise from MDworkhasalsofoundthatexchange17ofHObetweenthetwo
2
variationsintheSiO preparationand/orsurfacecondition.Here layersisslowandthatlateraldiffusion14withintheice-likelayeris
2
“surface condition” refers to the coverage and identity (e.g., slowerthaninbulkHObyafactorof∼2.Afurtherdiscussionof
2
isolatedvsgeminal)ofSi-OHgroupsaswellastothecoverage HO/DOisotopeexperimentsisgiveninthefollowingsection.
2 2
ofadsorbedH2OremainingafterpurgingindryN2atroomtem- 3.1.2. DMMPinDryN2.Figure4showsresultsforDMMP
perature.Itiswellestablished72-75thatevacuationatanelevated over a range of partial pressures in the absence of intentionally
temperatureisneededtoremoveallH OwhichisH-bondedto addedHO.Themostsignificanteffectisablue-shiftinthePdO
2 2
the SiO2 surface. Hence, any layer growth under the present stretch(ν(PdO))withincreasingP/P0.Thelimitingpeakpositions
conditionsoccursinadditiontoapreadsorbedH Olayer,further vary somewhat from run to run, falling in the 1222-1230 and
2
evidenceforwhichwillbegiveninthefollowingsection. 1246-1250cm-1rangesforlowandhighP/P0respectively.This
variabilitycouldberelatedtotheamountofpreadsorbedmolecular
(71)Morrison,S.R.TheChemicalPhysicsofSurfaces;Plenum:NewYork,1977; H2O (discussed later in this section). The ν(PdO) mode is
Chapter7. known33,41tobesensitivetothesurroundingmedium.Inthevapor,
(72)McDonald,R.S.J.Phys.Chem.1958,62,1168. it occurs at 1276 cm-1; whereas, in the liquid, intermolecular
(73)Zhdanov,S.P.;Kosheleva,L.S.;Titova,T.I.Langmuir1987,3,960.
(74)Burneau,A.;Barr(cid:4)es,O.;Gallas,J.P.;Lavalley,J.C.Langmuir1990,6,
1364.Gallas,J.P.;Lavalley,J.C.;Burneau,A.;Barr(cid:4)es,O.Langmuir1991,7,1235.
(75)Feng,A.;McCoy,B.J.;Munir,Z.A.;Cagliostro,D.E.J.ColloidInterface (76)Riemenschneider,J.;Wulf,A.;Ludwig,R.Z.Phys.Chem.2009,223,1011.
Sci.1996,180,276. (77)Mar(cid:1)echal,Y.J.Chem.Phys.1991,95,5565.
18148 DOI:10.1021/la103381r Langmuir2010,26(23),18144–18154
Bermudez Article
Figure5. Data showing the effects of adsorption on ν(O-H)
Figure4. SimilartoFigure1butforexposuretoDMMPvaporat
roomtemperatureintheabsenceofintentionallyaddedH O.The modes.(a)DMMPindryN2(P/P0=0.40)aftersubtractionof
2 apolynomialbackground(seesection3.1.3).(b)pureD O(RH=
nlaubmelbeder,sa0n.d05t,heetcsp.egcivtreaPh/aPv0efobreeenacdhisspplaeccterdumve.rPticeaalklyenfoerrgcileasriatrye. 1.0).(c)pureH2O(RH=1.0).TheδR/Rscaleappliesto2traces(b)
Thenoisy,irregularregionfromabout1100to1150cm-1corre- and(c).δR/Rfortrace(a)hasbeenexpandedbyafactorof10.The
δ(HOH)bendingmodeappearsatslightlydifferentenergiesin(b)
spondstoaregionofstrongmultiphononabsorptionintheSiATR
and(c),asindicated.Theregionnear1500cm-1isaffectedbya
prism in which the sample is almost completely opaque. The
slight miscancellation of the strong Si multiphonon absorption
verticaldashedlinesindicatetheshiftinν(PdO)withcoverage.
at∼1450cm-1.
TheinsetshowsamodelforDMMP.
interactionsshiftitto1245cm-1(refs58-61).InHOsolution,33 derivative-like structure,the exact positions ofthe bands being
2
ν(PdO)isfoundat1206cm-1,andthecorrespondingmodeofa addedandremovedaredifficulttodetermine.Mostofthechange
secondconformerappearsat1185cm-1.AtlowP/P0thelargered- in the ν(O-H) spectrum is complete at low DMMP partial
shiftrelativetothevapor(Δν(PdO)g-46cm-1)suggestsastrong pressure(P/P0e0.20),forwhichtheintensityofthe1230cm-1
bonding interaction involving the phosphoryl O atom. At high ν(PdO)peak(Figure4)reachessaturation.Thissuggeststhatthe
P/P0,ν(PdO)isslightlyhigherthanforliquidDMMPbutmuch changes arise from a replacement of Si-OH333OH2 bonds in
lowerthanforthevapor.ThisindicatesthattheDMMPisinan favor of Si;OH333OdP bonds. Previous studies of DMMP
environmentessentiallylikethatofthepureliquid.Similarshifts adsorptiononotherformsofsilicafoundabroadbandat3223
withincreasingcoverage39atasampletemperatureof100Korwith cm-1 (ref 35) or3380 cm-1 (ref 37) for ν(O-H) in an Si-OH
increasingP/P0atroomtemperature40,42wereseenforDMMPand bondedtoDMMP.
DIMPonOH-terminatedSAMs. Figure5bshowsthespectrumofOHgroups(intheabsenceof
Previous studies ofDMMP adsorption onother forms ofa- DMMP) that can exchange with D2O and that are, therefore,
SiO ,undervacuumconditions,foundΔν(PdO)=-19(ref35) chemicallyaccessible.Thesewillnowbediscussedinthecontext
2
and-41cm-1(ref37)relativetothevaporphase.Inviewofthe ofpreadsorbedH O,whichispresent72-75atroomtemperature
2
present results, this difference could be related to the DMMP andRH=0.Thebroadbandcenteredat∼3350cm-1represents
coverage,theextentofsurfacehydoxylationorthenatureofthe stronglyinteractingSi-OHgroupsandadsorbedmolecularH O.
2
Si-OHsites(e.g.,isolatedvsgeminal).Computationalresults43 A shoulder is seen at ∼3620 cm-1, which can be assigned to
indicatethatH-bondingbetweenthephosphorylOatomandtwo Si-OHgroupsweaklyinteractingwitheachotherorwithH O.A
2
Si-OHgroupsisthemoststablemodeofadsorptionona-SiO2. second D2O spectrum, completed about 20 min after the one
Forthisstructure,Δν(PdO) =-29cm-1 iscalculated.Other, shown, exhibited only a small (∼10%) further increase in the
less environmentally sensitive modes are the δ(P-CH ) sym- intensity(butnochangeintheshape)oftheOHandD Obands.
s 3 2
metricdeformationat1312cm-1,theF(O-CH )rockingmodeat ThisshowsthatmostoftheHTDexchangewascompleteduring
3
1185 cm-1 and the out-of-phase and in-phase ν(P-O-CH ) thefirstscan.Followingthisexperimentthecellwasthoroughly
3
stretchesat1034and1058cm-1respectively.Noneoftheseshow purgedwithH OvaportoremoveresidualD.Forcomparison,
2
any pronounced shift with changing P/P0. In particular, the Figure5cshowsdataforpureH2OatRH=1.0.
ν(P-O-CH )modesappearattheliquid-phasefrequenciesover Note that Figure 5a is shown with the δR/R scale expanded
3
thefullrangeofP/P0(vs1050and1075cm-1inthevaporphase). 10-foldrelativetospectrabandcinFigure5.Theresultsshowthat
Sincethemethoxygroupsarenotdirectlyinvolvedinadsorption only a small fraction (at most ∼10%) of the total surface OH
thissuggeststhatthereisasignificantintermolecularinteraction content(Si-OHandHO)interactsstronglywithDMMP.Results
2
involvingeitherDMMPorpreadsorbedH O. for other forms of silica,35,37 obtained under conditions of low
2
Figure 5 shows the effects of adsorption on the ν(O-H) adsorbed-H Ocoverage,indicatethatamuchlargerfractionofthe
2
spectrum.72-75 Upward- (downward-) pointing features corre- totalOHcontentisaffectedbyDMMP.Thepresenceofahigh
spond to species removed (added) by adsorption. The DMMP coverageofH OrelativetoSi-OHcanbeseenbyexaminingthe
2
data (Figure 5a) show the replacement of “moderately-strong” δ(HOH)bendingmodeinthe1600-1700cm-1range.Itisdifficult
H-bonds72 (the upward-pointing band at ∼3500 cm-1) with todistinguishSi-OHfromH Ointheν(O-H)spectrum,whereas
2
strongerSi;OH333OdP bonds givingthe downward-pointing theδ(HOH)modeisuniquetomolecularH2O.InFigure5bthe
band at ∼3250 cm-1. Since the two bands overlap, giving a δ(HOH)/ν(O-H) ratio of integrated areas (∼0.06) is essentially
Langmuir2010,26(23),18144–18154 DOI:10.1021/la103381r 18149
Article Bermudez
thesameasthatofpureHO(Figure5c),indicatingthatmostof
2
the exchangeable OH is in the form of molecular HO. These
2
resultsconfirmtheexistenceofahigh(relativetoSi-OH)coverage
of molecular H O after prolonged purging in dry N at room
2 2
temperature.
It is noteworthy that this preadsorbed H O layer does not
2
exhibit an ice-like ν(O-H) spectrum since there is no clear
indicationofabandat∼3240cm-1(cf.Figure1).Theice-like
structureappearstoformonlyinanambientwithafiniteRH.
This can be understood if the dominant interaction for pread-
sorbedH OiswithSi-OHgroups.Theice-likelayer,ontheother
2
hand, involves a strong intermolecular interaction leading to
tetrahedrally coordinated H O. Thus, at finite RH, the pread-
2
sorbedH Omayactasaconnective-oratransitionlayerbetween
2
thehydroxylatedSiO surfaceandtheice-likelayer.Itispossible
2
that,atfiniteRH,adsorbedH Omodifiesthepreadsorbedlayer,
2
ineffectincorporatingitintotheice-likelayer.However,dataat
lowRH(Figure1)shownoevidenceoftheremovalofaband
near 3350 cm-1, which suggests that preadsorbed H2O is not Figure6. Dataintheν(O-H)andν(C-H)rangeforthecoad-
stronglyaffectedbythegrowthoftheice-likelayer. sorptionofH2OandDMMP.RHandP/P0givethevapor-phase
164F0orcmH-21O, wahticRhHis=cha1r.a0ct(eFriisgtuicreof5cli)q,uδid(HHOHO); waphpeeraearss,aitt bcoeenncednitsrpaltaiocendsovferHti2cOallayndfoDrMclaMritPy.reTshpeecdtiavsehlye.dTlhineespinecttrraaceha(vbe)
2 illustratesthetypeofbackgroundthatwassubtractedtoproduce
occursat1660cm-1fortheH Oremainingafterpurgingthecell
2 spectralikethatinFigure5a.Thiswasobtainedbyfittingasingle
with dry N2 (Figure 5b). This blue-shift indicates a relatively polynomialtothe2000-2700and3800-4000cm-1regions.
strongH-bondinginteraction78,79consistentwiththeresistanceof
this H O to easy desorption. H-bonding shifts ν(O-H) and
2
δ(HOH)inoppositedirections,andtheenergiesareknown79to
becorrelatedindifferentformsofice.Aν(O-H)of∼3350cm-1,
asinFigure5b,correlateswithaδ(HOH)of∼1680cm-1,whichis
fairlyclosetothevalueof1660cm-1seenhere.Inthepresenceof
DMMP, the δ(HOH) region becomes complex. The sharp,
derivative-likestructurenear1700cm-1isbelievedtoresultfrom
theinteractionofDMMPwithasmallconcentrationofimpurity
CdOgroups,whichevidentlydoesnotaffecttheformationofthe
ice-like layer. C 1s XPS data (see the Supporting Information)
indicateacoverageof∼0.08monolayersofsuchspecies,anda
slight blue-shift in the intense ν(CdO) absorption band, as a
resultofDMMPadsorption,wouldgivetheobservedderivative-
likeline shape.Thisfeature appears to besuperimposedonan
upward-pointingpeakat∼1660cm-1indicatingthedisplacement
ofadsorbedH O.
2
Theconclusionatthispointisthat,atRH≈0andP/P0<
∼0.50,acomplexsurfacephaseisformedconsistingofDMMP
and preadsorbed H Ostronglyinteracting with eachother and Figure7. Datainthelow-frequencypartofthespectrumcorre-
2 spondingtothoseinFigure6.Thelabelscorrespondtothosein
withSi-OH.Asimilarmodelhasbeenproposed41forDMMPat
Figure6.Theupward-pointingfeaturelabeledTat∼1145cm-1
P/P0=0.02interactingwithOH-terminatedSAMsinhumidN2. is due to miscancellation of a strong Teflon absorption (see
In the present case, DMMP and H2O may both H-bond to section2.1).
Si-OH, forming a mixed monolayer, or displaced H O may
2
adsorb on top of the DMMP. It is also possible that both was subtracted to obtain the spectrum shown in Figure 5a.
processes occur simultaneously. At higher P/P0 and RH ≈ 0 SpectralikethoseinFigures6dand7dwerealsorecordedusing
(Figure4),aliquid-likeDMMPlayerformsontopofthisinitial D O in place of H O. There was no sign of H T D exchange
2 2
layer. involvingtheCH groupsofDMMP,indicatingthatexchangeof
3
3.1.3. DMMP in Wet N2. Figures 6 and 7 show data for HbetweenDMMPandH2Oisnotinvolvedunderthepresent
differentRHsandDMMPP/P0values.Thesewereobtainedby conditions.
mixingseparateflowsofpureN andofN saturatedwitheither Twopointsarenoteworthy.First,thepresenceofH Odoesnot
2 2 2
H OorDMMP.ThemaximumRHof0.30wasselectedsoasto appeartohaveastrongeffectonthetotalamountofadsorbed
2
allowasufficientlyhighDMMPP/P0forclearobservationofthe DMMP,asshownbythecomparableDMMPbandintensitiesat
DMMP spectrum without exceeding the dew-point restriction P/P0 = 0.30withand withoutadded H2O vapor. Conclusions
discussedinsection2.2.Forreference,traces(a)and(b)ineither aboutadsorbatecoveragebasedonIRdatamustbequalifiedby
figure show data for pure H O and pure DMMP respectively. noting that interaction between coadsorbed species can affect
2
Figure6balsoshows,forillustration,thetypeofbackgroundthat intensities.Forexample,relativeintensitiesdifferintheIRspectra
ofDMMPvaporvsliquid.Furthermore,theFTIRbeamisnot
completelydepolarized.56Hence,ifanadsorbateisnotrandomly
(78)Paul,S.O.;Ford,T.A.J.Cryst.Spectrosc.Res.1986,16,811.
(79)Hernandez,J.;Uras,N.;Devlin,J.P.J.Chem.Phys.1998,108,4525. oriented,thenreorientationwithrespecttothesurfaceinresponse
18150 DOI:10.1021/la103381r Langmuir2010,26(23),18144–18154
Bermudez Article
tocoadsorbedspeciescanalsochangebandintensities.Incon-
trast,H OenhancestheadsorptionofDMMPonOH-terminated
2
SAMsasshown41usingbothIRdataandmeasurementswitha
surfaceacousticwavedevice,whichissensitivetochangesinthe
totaladsorbatemass.
Second,DMMPdoesnotinhibittheformationofanice-like
layer,asshownbythecontinuedclearpresenceofthe∼3240cm-1
bandatP/P0=0.30.Theredoesappeartobeaslightred-shiftor
broadeningoftheliquid-likeband,probablyduetoH-bondingto
DMMP.Matrix-isolationstudies29showthatina1:1H-bonded
complex of H O and DMMP, ν(O-H) red-shifts by 203 cm-1
2
relativetothevalueforH OisolatedinanArmatrix(3639cm-1).
2
However,thereisapparentlynosignificantdisruptionoftheice-
like layer. This suggests that, under these conditions, DMMP
does not penetrate into or through this layer and that ice-like
bonding is sufficiently strong to overcome SiO;H333OdP
bonding. For OH-terminated SAMs, in which the OH groups
formanH-bondednetwork,ν(O-H)isfound40at∼3350cm-1.
DisruptingthisnetworkbyH-bondingtoadsorbedDMMPshifts Figure8. SimilartoFigure1butshowingdatafora-Al2O3.Onlya
few selected RH values are shown. For comparison, (a) shows
νth(Oe-prHes)etnot3i4c5e-0lickme-la1yseurggthesetiνn(gOa-wHea)koefr∼H3-b2o40ndctmo-D1MinMdicPa.tIens smimulitliaprlideadtabyfoarfHac2tOoraodfs5orrpeltaiotinveotnotthheormseafloar-ASilOO2a.tTRhHed=iffe0r.e3n0t,
2 3
strongerH-bondingthanforanOH-terminatedSAM,onewhich traceshavebeenshiftedverticallyforclarity.Notethedifferencein
persistsatleastuptoaDMMPP/P0of0.30.Thepossibilityhas theR/RscalevsFigure1.
been considered that the apparent ice-like band in Figure 6c,d
mightinsteadactuallyresultfromaninteractionbetweenDMMP
3.1.4. ThermalvsUV/O Oxide.Thedatapresentedthusfar
and H O. However, corresponding results (section 3.2) for 3
2 were all obtained for a UV/O oxide. All experiments were
3
a-Al2O3andAlO(OH),whichshownoice-likelayer,alsoshowno repeated for a thermal oxide, and the results were virtually
bandnear3240cm-1forcoadsorbedH OandDMMP.
2 identical to those given above. One small exception was in the
ThedatainFigures6and7wererecordedinthesequence(b),
growthoftheH Olayer(Figure1).Forthethermaloxideatlow
(c),and(d),i.e.,byintroducingH Ovaporintoaninitiallydry 2
2 RH(<0.30)theν(O-H)spectrumwasdominatedbytheice-like
flowofDMMP.Data(notshown)recordedinthesequence(a),
bandandcloselyresembleddataforaUV/O oxidegiveninref3.
(c),and(d),byintroducingDMMPintoaflowofhumidN ,gave 3
2 AthigherRH,ontheotherhand,theν(O-H)spectraforboth
essentially the same results. However, unlike in Figure 6, the
typesofoxidewereessentiallyidentical.
liquid-likeν(O-H)bandremainedwell-definedasDMMPwas
3.2. Amorphous Al O . Figures 8 and 9 show data for
added.WhenDMMPisaddedtohumidN theliquid-on-iceH O 2 3
2 2 amorphous Al O vs RH (in the absence of DMMP). The
layerstructureisalreadywell-formedbeforeintroductionofthe 2 3
ν(O-H)spectrumshowsonlyliquid-likeH Owithnoevidence
DMMP; whereas, in the reverse sequence, H O must displace 2
2 ofanice-likebandatanyRHintherangeof0.02to1.0.Aweak
adsorbedDMMPinordertoformtheice-likelayer.
shouldernear∼3250cm-1,seenatveryhighRH,canbeassigned
Inthelow-frequencyrange(Figure7)thereisnostrongeffect
to the overtone of the 1640 cm-1 δ(HOH) mode (not shown).
ontheDMMPspectrumduetoadditionofH O.Under“dry”
2 Figure8ashows,forcomparison,thespectrumforSiO atRH=
conditions (Figure 7b) ν(PdO) appears at 1226 cm-1 with an 2
asymmetrytohigherenergysuggestingtheonsetoftheliquid-like 0.30.Inadditiontotheabsenceofanice-likelayer,a-Al2O3shows
a higher H O coverage than does SiO at the same RH, as is
DMMPlayer.Thereislittleornochangeinν(PdO)asH Ois 2 2
2 evidentincomparingtheisothermsinFigures2and9.Following
added.Bertilssonetal.41foundν(PdO)at1235cm-1forDMMP
theseexperiments,XPSshowednosignificantincreaseintheOH
bondedtoanOH-terminatedSAMintheabsenceofH O.With
2 content(seetheSupportingInformation).
theadditionof4H OperDMMPthismodeshiftstoabout1226
2 Figure 10 shows results for pure DMMP (in the absence of
cm-1,andafurtherincreaseinH Ocoverageshiftsittoabout
2 H O)andforthecoadsorptionofH OandDMMP.Asinthe
1222 cm-1. The present results suggest that DMMP already 2 2
caseforH O,theDMMPcoverageindicatedbytheIRband
interactswithpreadsorbedH O(discussedinsection3.1.2)before 2
theRHisincreasedbythein2tentionaladditionofH O. intensities is higher than on SiO2 at the same P/P0. Both
2 observations are consistent with the higher coverage of OH
Takentogether,thesedatasuggestmixingofDMMPwiththe
liquid-like H O layer, forming in effect an ultrathin aqueous groupsona-Al2O3asseenintheO1sXPS(seetheSupporting
2 Information)andwiththeimportanceofH-bondingtoOHin
solutionontopofanintactice-likelayer.ThestrongH-bonding
theadsorptionofbothH OandDMMP.Thetrendinν(PdO)
withintheice-likelayerapparentlymakesitresistanttodisruption 2
byDMMPand,furthermore,promotesdisplacementofadsorbed withincreasingP/P0issimilartothatseenforSiO2inFigure4.
DMMPbyH O,atsufficientlyhighRH,infavoroftheforma- However,atverylowP/P0,ν(PdO)appearsabout1226cm-1,
2 whichissomewhatlowerthanforSiO andsuggestsastronger
tionoftheice-likelayer.ThesolubilitylimitofDMMPinH Ois 2
2
estimated80tobe4.95Mat25(cid:1)CandpH7.Whenthisisexceeded, O;H333OdP bond. Results obtained under vacuum condi-
a layer of liquid-like DMMP (with ν(PdO) ≈ 1245 cm-1) is tions for DMMP adsorption on HSA Al2O3 are useful for
expected to form. Some evidence for this might be seen in the comparison.HerebondingoccursviaAl333OdPinteractionat
asymmetryoftheν(PdO)peakinFigure7d. coordinativelyunsaturatedAlsitesgivingν(PdO)at1216cm-1
(refs 44a and 46a), which is significantly red-shifted from the
(80)Lu,X.;Nguyen,V.;Zeng,X.;Elliott,B.J.;Gin,D.L.J.MembraneSci. presentresultsforAl;OH333OdPbonding.Afurtherdiscus-
2008,318,397. sionofthea-Al2O3resultsisgiveninthefollowingsection.
Langmuir2010,26(23),18144–18154 DOI:10.1021/la103381r 18151
Article Bermudez
Figure11. Similar to Figure1 but showing ν(O-H) vs RH for
Figure9. SimilartoFigure2butshowingdataforAl2O3.Notethe H O/AlO(OH). Only a few selected RH values are shown. The
differenceinthebandareascalevsFigure2. 2
spectrum for SiO at RH = 0.30 is the same as that shown in
2
Figure8aandisincludedforcomparison,multipliedbyafactorof
10relativetotheAlO(OH)data.
Thisisknown83,84toproduceahigh-qualitya-Al2O3filmuptoa
limitingthicknessofabout2nm.Thesamplewasthenimmersed
inboilingDIH Ofor15min.Characterizationoftheresulting
2
film using XPS and transmission IR data (see the Supporting
Information) indicated the presence of Al oxyhydroxide, AlO-
(OH),intheso-called“pseudoboehmite”or“poorly-crystalline
boehmite(PCB)”formwithnometallicAlremaining.
Therearemanysimilaritiesbetweenthe resultsforPCBand
thosefora-Al2O3discussedintheprecedingsection.Theν(OH)
spectrum vs RH (Figure 11) shows only liquid H O with no
2
obviousice-likelayeratanyRH.ThisconcurswithpreviousIR
transmissionresults85forPCBincontactwithH Ovapor.The
2
weak shoulder seen at ∼3250 cm-1 at high RH has been
assigned85totheovertoneoftheδ(H-O-H)bendingmodeat
1640 cm-1 (not shown). The integrated band area at any RH
(Figure12)issignificantlygreaterthanforeitherSiO (Figure2)
2
Figure10. Similar to Figures 4 and 6 but showing results for ora-Al2O3(Figure9)atthesameRH,andtheisothermissimilar
in shape to that reported earlier85 for H O on PCB. These
DMMP on a-Al2O3. The numbers 0.10, etc. give the DMMP 2
P/P intheabsenceofintentionallyaddedH O.The“T”marksa observations are consistent with a high density of OH groups
0 2
miscanceled Teflon absorption from the ATR cell. The bottom (documentedintheSupportingInformation)thatactasadsorp-
traceshowsdataforcoadsorptionofDMMP(P/P0=0.30)and tionsites.TheinitialexposureofthissurfacetoH2O(Figure11)
H2O(RH=0.30).Thespectrahavebeendisplacedverticallyfor results in an upward-pointing peak at 3660 cm-1 due to the
clarity,andthedashedlinesshowthepositionoftheν(PdO)mode,
removal, through adsorption, of free OH (i.e., groups not
whichchangeswithP/P0.NotethedifferencebetweentheδR/R involved in H-bonding to each other or to H O). Very weak
scaleshereandinFigures4and6. 2
structure in the 2800-3000 cm-1 range arises from either mis-
3.3. Highly-Hydroxylated a-Al O . It is known81,82 that cancellationoffeaturesinthesingle-beamspectraduetoorganic
2 3 contaminationintheFTIRbeamsplitterortoasmallamountof
prolongedexposureofAl O toliquidH Oconvertsthesurface
2 3 2 impurityactuallypresentonthesample.
to a mixture of AlO(OH) and Al(OH) , with the detailed
3 Adsorbing DMMP on this surface (Figure 13) gives results
chemistry depending on the method of treatment (e.g., on the
H Otemperature).Apreliminarystudywasdoneoftheeffectsof similartothosediscussedabove.AstrongO;H333OdPinter-
su2ch“weathering”ontheinteractionwithDMMP.Tosimulate actionoccursatlowP/P0,leadingtoalargered-shiftinν(PdO)
relative to the vapor (1231 vs 1276 cm-1), followed by the
oneformofahighlyhydroxylatedAl O surface,4nmofAlmetal
wasvapor-depositedoneachsideofa2Si3ATRprismandeachside formationofliquid-likeDMMPathigherP/P0.AthighP/P0a
shoulderisseenonthehigh-energysideofthe1246cm-1liquid-
subsequentlyexposedtoUV/O foronehouratnominalroom
3 likeν(PdO)peak,whichisnotevidentinthecaseofadsorption
temperature. (Radiative heating by the UV lamp raised the
sampletemperaturetoabout40(cid:1)Cbytheendoftheexposure.) onSiO2(Figure4).Thisfeature,whichisalsoseenfora-Al2O3
(Figure 10), might indicate the formation of DMMP clusters
(81)Lef(cid:4)evre,G.;Duc,M.;Lepeut,P.;Caplain,R.;F(cid:1)ederoff,M.Langmuir2002, (84)Chang,C.-L.;Engelhard,M.H.;Ramanathan,S.Appl.Phys.Lett.2008,
18,7530. 92,263103.
(82)Desset,S.;Spalla,O.;Lixon,P.;Cabane,B.ColloidSurf.A2002,196,1. (85)Wang,S.-L.;Johnston,C.T.;Bish,D.L.;White,J.L.;Hem,S.L.J.Colloid
(83)Popova,I.;Zhukov,V.;Yates,J.T.,Jr.Surf.Sci.2002,518,39. InterfaceSci.2003,260,26.
18152 DOI:10.1021/la103381r Langmuir2010,26(23),18144–18154