AdsorptionIsotherm Orientation Alcohols SiO2 Kim

8/9/2019 AdsorptionIsotherm Orientation Alcohols SiO2 Kim

1/10

Adsorption Isotherm and Orientation of Alcohols on Hydrophilic SiO2 under Ambient

Conditions

Anna L. Barnette, David B. Asay, Michael J. Janik, and Seong H. Kim*

Department of Chemical Engineering, PennsylVania State UniVersity, UniVersity Park, PennsylVania 16802

ReceiVed: February 5, 2009; ReVised Manuscript ReceiVed: April 22, 2009

The adsorption isotherm and orientation of small alcohol molecules (n-propanol and n-pentanol) on clean,hydrophilic silicon oxide surfaces under ambient conditions were studied with attenuated total reflectanceinfrared (ATR-IR) spectroscopy and density functional theory (DFT). The ATR-IR study reveals that as thealcohol partial pressure relative to its saturation vapor pressure (P/Psat) increases from 0 to 10%, the isothermthickness ofn-propanol andn-pentanol increases rapidly to 0.3 nm and 0.6 nm, respectively. Upon furtherincrease in P/Psat, the isotherm thickness increases only slightly until the condensation occurs at the nearsaturation vapor condition. The alkyl chains of the alcohol molecules adsorbed at 5% P/Psat appear to betilted toward the surface, while the hydroxyl groups are oriented toward the surface normal direction. AsP/Psat increases further, the molecular orientation of adsorbed molecules drastically changes to a structurethat is either random or oriented 40-50 from the surface normal. The DFT calculations for n-pentanoladsorbed on SiO2 support the molecular structure determined from the ATR-IR experiment. The similartransition in molecular orientation with increase of the adsorbed thickness was observed for longer chainalcohols (n-decanol andn-octadecanol). The alkyl chain packing into a self-assembly like structure was observedonly when the chain length is long enough and the substrate is heated. These observations are discussed interms of the adsorbate-substrate interactions as well as intermolecular interactions.

Introduction

The thickness and molecular orientation of organic moleculesadsorbed or covalently bonded to solid surfaces play animportant role in many interfacial phenomena including catalyticbehavior,1 adhesion,2 wetting,3,4 protein adsorption,5,6 electro-chemistry,7,8 and tribology.9,10 In the cases of heterogeneouscatalysis and electrochemistry, metal surfaces are often partially

or fully covered with organic molecules and the interactionsbetween the organic molecules and the metal surface governschemical reactions occurring at the interface. Self-assembledmonolayers (SAMs) and Langmuir-Blodgett (LB) films areoften employed for the boundary lubrication of microscaledevices, and their molecular structures and orientations aregoverning factors for the performance and efficacy as alubricant.11,12 Such coatings can also be used to chemicallymodify surfaces of interest and thus change both the adhesiveforces and wetting behaviors of surfaces. Equilibrium adsorptionof short-chain linear alcohols in ambient conditions has beenproven to be efficient for lubrication of solid contacts from nano-to macroscales.9,10 Numerous spectroscopic, microscopic, and

theoretical studies have been utilized in the pursuit to understandthe structure of these organic films adsorbed to differentsubstrates of interest.13-21 Understanding the adsorption isothermand molecular orientation of organic molecules on solid surfacesmay give further insight into how such molecules affectinterfacial phenomena involved in catalyst reactivity, electro-chemisty, interfacial contacts, and so on.

Heterogeneous catalysis has spurred many experimental andtheoretical studies of simple, short-chain hydrocarbons adsorbedon metal and metal oxide surfaces under ultrahigh vacuum(UHV) conditions. Numerous studies on the structure and

orientation of the molecules adsorbed on these surfaces suggestthat the packing and orientation of adsorbed hydrocarbons arefunctions of surface coverage and adsorption temperature.19,20,22-25

For example,n-butane adsorbs with their molecular axis parallelwith the Pt(111) surface at low coverages. As the surfacecoverage increases, then-butane molecules reorient themselveswith their molecular axis tilted away from the Pt(111) surface.20,24

The structure and orientation of the adsorbed alkane moleculesis also greatly affected by the temperature of the substrate.23

For example, molecular dynamic simulations ofn-pentane andn-heptane adsorbed on graphite suggest the loss of the structuralorder observed at low temperatures as the substrate temperatureincreases.19

Extensive fundamental surface science studies have beencarried out for polar molecules such as short-chain alkylhalides and alcohols adsorbed on solid surfaces in UHVconditions.16,17,21,26-29 For instance, in the adsorption ofmethyl iodide on various metals and metal oxide surfaces,the methyl iodide molecules interact with the surfaceprimarily through the iodide atom rather than the methyl

group at coverages less than half a monolayer (0.5 ML).30,31

As the surface coverage of methyl iodide increases over 0.5ML, some molecules tend to absorb with the methyl groupfacing the surface instead of iodine.30,31 These changes inthe molecular orientation influence the desorption behaviorand subsequent reactions of photodissociation fragments.16,17,21

The alcohols can either chemisorb or physisorb onto solidsubstrates depending on the solid surface chemistry. Ingeneral, short-chain alcohols chemisorb at vacancy sites onmetal surfaces. Removal of the hydroxyl proton of the alcoholmolecule leads to the formation of an alkoxide group.32 Whenalcohols adsorb onto oxide surfaces, they can be adsorbedmolecularly via hydrogen bonding with the surface hydroxyl* Corresponding author. E-mail: [email protected].

J. Phys. Chem. C2009, 113,106321064110632

10.1021/jp901064r CCC: $40.75 2009 American Chemical SocietyPublished on Web 05/21/2009


2/10

groups or chemisorbed as alkoxides after removal of hydroxylhydrogen.27 The chemisorbed alkoxide chains tend to orientthemselves parallel with the metal surfaces to which theyare adsorbed.26,28

While these studies carried out under UHV conditions providevaluable fundamental details about the molecule-surfaceinteractions as a function of surface coverage, the relevancy ofthese finding to ambient conditions remains to be tested. Thereare several examples where the structure of the adsorbed layers

determined in UHV is different from the one found in ambientconditions. For example, if CO and NO pressures are increasedabove the UHV conditions, they can be adsorbed onto Pt(111)and Rh(111) to a coverage and structure that cannot be obtainedin the UHV condition.33,34 CO has also been observed tochemisorb onto Au(110) in an ambient environment but not ina UHV environment.35 These discrepancies are often called apressure gap. In typical UHV experiments, the surface is dosedwith a controlled amount of molecules at cryogenic temperaturesand the adsorbed molecules are probed in the extremely lowpressure regime where the involvements of gas phase moleculescan be ignored. In this condition, the adsorbed molecules havea very long residence time on the surface at cryogenic

temperatures and can diffuse laterally to find the two-dimensional equilibrium configuration at a given cryogenictemperature. There are no collisions with the impingingmolecules from the gas phase or desorption from the substrateinto the gas phase. In ambient conditions, the substrate tem-perature is much higher than the cryogenic temperature regime;therefore, the adsorbed molecules will have much higher thermalenergy. Additionally, at equilibrium, the molecules at the surfaceare constantly desorbed and adsorbed at the same rate. In otherwords, the residence time of the adsorbed molecules is veryshort when they are in equilibrium with the gas phase molecules.

Regarding the organic molecules on solid surfaces in ambientconditions, a large number of studies found in the literature are

mostly focused on SAMs and LB films deposited on metal,metal oxide, and semiconductor surfaces.36-41 SAMs can bedeposited from the vapor phase or liquid phase, whileLangmuir-Blodgett films are deposited from the liquid phase.Once the films are removed from the deposition vessel, theyare no longer in equilibrium with the molecules in the bulkphase. The molecules deposited on the solid surface do notdesorb or have the ability to move laterally because theinteractions with the neighboring molecules as well as thesubstrate are so large. The molecule at the surface has an infiniteresidence time until they are physically (via heating, scratching,etc.) or chemically (through various chemical cleanings) re-moved from the surface. Thus, these systems are not ideal to

study the equilibrium behaviors of the adsorbed molecules underambient conditions.

In this study, the adsorption isotherm and orientation of smalllinear alcohol molecules (n-propanol andn-pentanol) and long-chain linear alcohols (n-decanol and n-octadecanol) on cleanhydrophilic SiO2surfaces in ambient conditions are discussedas a function of partial pressure of alcohol vapor relative totheir respective saturation vapor pressure (P/Psat). The adsorbedalcohols would interact with the silicon oxide substrate throughhydrogen bonding as well as van der Waals interactions andhave the capability of lateral diffusion on the surface anddesorption back to the gas phase. A silicon surface with anamorphous oxide layer was chosen as a substrate for this study

due to its great importance in the semiconductor and nanosciencetechnologies.10,42

Attenuated total reflection infrared (ATR-IR) spectroscopyand density functional theory (DFT) calculations were used asthe probing techniques. In ATR-IR, the evanescent fieldpenetration into the probing medium is very short (200-300nm in the alkyl and hydroxyl stretching vibration regions).Because the density of vapor-phase alcohol molecules is not

high enough, the alcohol vapors in the shallow probe volumeabove the substrate surface cannot generate measurable absor-bance signal in ATR-IR. Therefore, this technique allowsdetection of adsorbed alcohol molecules without interferencefrom gas phase molecules. In addition, the study with polarizedIR beams can provide the molecular orientation information ofthe adsorbed alcohol layer in equilibrium with the gas phase.Density functional theory (DFT) can predict adsorption energiesas a function of surface coverage. The molecular conformationof the adsorbed alcohols and surface coverage can be varied,therefore yielding the most probable orientation of adsorbedmolecules. These studies revealed interesting insights into theequilibrium structure of adsorbed molecules in equilibrium withthe gas phase.

Experimental Details

Alcohol adsorption experiments were performed with aThermoNicolet Nexus 670 infrared spectrometer with an ATRsetup and a MCT-A detector. A silicon ATR crystal was usedin all experiments. The crystal had a 45bevel cut, providinga 45incidence angle and a total of 11 internal reflections atthe probing surface. The effective penetration depth of theevanescent IR field is 270 nm in the 2800-3000 cm-1 range,where the C-H stretching vibrational peaks of alkyl chains areobserved. A wire grid polarizer was used to generate s- andp-polarized IR beams. Here, s-polarization is perpendicular to

the plane of IR beam incidence (parallel to the substrate surface)and p-polarization is parallel to the plane of incidence. Figure1 illustrates the orientation of the s-and p-polarization electricfield vectors at the substrate/adsorbate/air interface with anincident angle of.

The amorphous oxide layer of the silicon ATR-IR crystalwas characterized with X-ray photoelectron spectroscopy (XPS).The XPS data showed only two Si 2p components: Si4+ at 103.1eV and Si0 at 99.3 eV, indicating that the surface of the ATRcrystal had a composition of SiO2. The relative intensity of Si4+/(Si4+ + Si0) is 18%. On the basis of the known inelastic meanfree path of photoelectrons, this intensity ratio corresponds toan oxide thickness of 2 nm. This was confirmed withellipsometry. This oxide layer was cleaned with a UV/ozone

treatment for 20 min and then placed in a 5:1:1 mixture ofwater:30% ammonium hydroxide:30% hydrogen peroxide at a

Figure 1. Illustration of s-polarized and p-polarized light for a three-layer system: (1) silicon, (2) adsorbed alcohol, (3) air or vacuum.

Isotherm and Orientation of Alcohols on Hydrophilic SiO2 J. Phys. Chem. C, Vol. 113, No. 24, 2009 10633


3/10

temperature of 75 ( 5 C (commonly known as RCA-1) for 15

min. Immediately following the cleaning step, the crystal wasrinsed with copious amounts of milli-Q water (resistance ) 18M/cm) and dried with high purity argon. This procedurecleaned the silicon oxide surface and produced the highlyhydrophilic oxide surface.43 The water contact angle on thissurface was measured to be


4/10

change. If the molecules are not in the perfect anti conformation,the actual geometry or average molecular tilt angle calculatedfrom the observed dichroic ratios could be different.

Computational Details

Density functional theory (DFT) calculations were performedto evaluate the energetics ofn-pentanol adsorption to the silicasurface. DFT can probe the structure and adsorption energies

as a function of coverage and molecular configuration, althoughthere are limitations in the use of an idealized crystal and surfacestructure and in the accuracy of representing nonbonding vander Waals attraction. Calculations were carried out using theVienna ab Initio Simulation Program (VASP), an ab initio total-energy and molecular dynamics program developed at theInstitute for Material Physics at the University of Vienna.51-53

The core electron wave functions were represented using theprojector augmented wave (PAW) method.54 The plane wavebasis set used to represent the valence electron density wascutoff at 450 eV. All calculations were performed spin-restricted.The Brillouin zone was sampled using a (2 2 1)Monkhorst-Pack grid,55 with the third latice vector perpen-dicular to the surface. The Perdew-Wang (PW91) version ofthe generalized gradient approximation (GGA) was used toincorporate exchange and correlation energies.56 This functionalhas been shown to provide accurate hydrogen bond formationenergies (within 20% over 12 test complexes),57 however, itdoes not accurately represent the energy associated with vander Waals interactions.58 The implication of this limitation isdiscussed in presenting the results. Structural optimizations wereperformed by minimizing the forces on all atoms to below 0.05eV -1.

The hydroxylated termination of the -cristobalite (111)surface was chosen to model the interaction ofn-pentanol withsilica. -Cristobalite represents a relatively stable polymorphof silica and exhibits a refractive index and density similar to

amorphous silicates.59 Because of the structure of the -cristo-balite (111) surface, only single hydroxyl groups that do notinteract with each other through hydrogen bonding are present.60

The surface was cleaved from an optimized cubic face-centered-cristobalite lattice with space group Fd3m and unit celldimensions a ) b ) c ) 0.745 nm (a 4% overpredictioncompared with experiment, in line with results using similarmethods in ref61). A slab representation was used to representthe surface, with eight silicon layers (16.1 thick) included inthe surface normal direction. A vacuum region of 1.8 nm(withoutn-pentanol adsorbed) was placed between slab layers.The bottom five silica layers were frozen at their bulk positionsduring structural optimization, whereas the top three layers were

relaxed. An asymmetric slab was used, with the top layerhydroxyl terminated and the bottom layer containing unsaturatedsilicon atoms. Evaluation of the local electrostatic potentialdistribution in the cell indicated that this configuration avoidedspurious dipole-dipole slab interactions between repeated slabs.Four hydroxyl groups are exposed per surface unit cell (2 2surface cell, 10.53 10.53 ), and the ratio ofn-pentanolmolecules adsorbed per hydroxyl group will be referred to as afractional coverage. The n-pentanol adsorption energy wascomputed for fractional coverages of 1/4, 1/2, 3/4, and 1, andwas calculated using the following equation:

Eads )Esurface,x-pentanol - Esurface - xEpentanol

x

(3)

wherexrepresents the number ofn-pentanol molecules adsorbed.

Results and Discussion

A. ATR-IR Study of Short-Chain Linear Alcohol Ad-

sorption on Amorphous SiO2. The nonpolarized ATR-IRspectra were used to calculate the adsorption isotherm thick-nesses. The O-H stretching vibration intensities relative to theC-H vibration peaks in the alkyl chain did not changesignificantly compared to those in the bulk alcohols.32 Thisimplies that the alcohols are molecularly intact and did not formalkoxides. Isotherm thicknesses were estimated in two methods.One is to compare the experimentally measured intensity ofthe adsorbed layer with the theoretical intensity simulated usingthe optical constants of bulk liquids.62 The other is to dividethe intensity of the adsorbed species with the absorbance perunit path length determined from the penetration depth of theIR beam into the alcohol liquid and the ATR-IR intensitymeasured for the bulk liquid.49 These two methods produced

essentially the same adsorption isotherm curves within ourexperimental error range.

Figure 4 displays the adsorption isotherms, averaged overmultiple peaks and spectra, for n-propanol and n-pentanol as afunction ofP/Psat. In both cases, the isotherm thickness increasesfast as P/Psatincreases from 0 to 10% and then grows slowlyor almost levels off whenP/Psat>10%. The adsorption isothermatP/Psat > 10% appears to level off somewhat faster in the caseofn-pentanol thann-propanol. When the alcohol partial pressureapproaches the saturation vapor pressure (P/Psat > 80%), theisotherm thickness increases fast again, indicating the occurrenceof multilayer condensation.63

The observed adsorption isotherm behavior can be rational-

ized with differences in the substrate-

molecule and intermo-lecular interactions. The molecules in the first layer will interactwith the oxide surface primarily through hydrogen-bondinginteractions between the hydroxyl group in the alcohol moleculeand the surface silanol groups. The strong hydrogen bondinginteraction with the surface silanol group leads to the adsorptionof the alcohol molecules on the surface even at low P/Psat. Theamount of these strongly bound alcohol molecules on the surfaceis governed by (or equilibrated with) the alcohol partial pressurein the gas phase. Once the substrate surface is coveredsubstantially with the hydrogen-bonded molecules, the incomingmolecules from the gas phase will interact with the adsorbedmolecules mostly through van der Waals interactions. Becausethe van der Waals interaction is weak, the impinged molecules

will not be held by the adsorbed molecules and desorb back tothe gas phase. As the alcohol partial pressure approaches the

Figure 4. Adsorption isotherm thicknesses for n-propanol and n-pentanol layers on clean hydrophilic SiO2as a function of alcohol partialpressure relative to the saturation pressure.



5/10


6/10

dichoric ratios of the three peaks that have dipole momentsperpendicular to the molecular axis using eqs 1 and 2b. At 5%P/Psat, the propyl and pentyl chains are tilted by 80 (12and70 ( 14from the surface normal, respectively. When P/Psat >10%, the average tilt angle of the alkyl groups decreases andapproaches a value of approximately 40-50from the surfacenormal.

The observedP/Psatdependence of the alkyl chain tilt angle() is believed to be related to the change in the adsorbedmolecule amount from incomplete coverage to complete cover-age. In the case of SAMs where the adsorbate molecules arecovalently attached to the surface, it is known that the alkylchains are tilted toward the surface when the surface coverageof the organic molecules is low.71 This geometry provides themaximum interaction between the alkyl chain and the surface.As the surface coverage increases, the alkyl chains have moreintermolecular interactions and become erected toward the

surface normal direction. If the alkyl chains are fully self-assembled, the alkyl chain tilt angle decreases to


7/10

alkyl-surface interaction. The DFT calculations showed thatthe gauche confirmation remains preferred at fractional coverageup to 1/2.

At fractional coverages higher than 1/2, steric repulsionbetween the alkyl chains prevents rotation about the C1-C2bondto the gauche conformation and the alkyl chains orient closerto the surface normal. Parts c and d of Figure 7 illustrate theoptimal conformation that maximizes the number of surfacehydroxyl groups being involved in the adsorption ofn-pentanolvia hydrogen-bonding interactions. At this coverage (1 ML),the average adsorption energy remains close to the low coverage

value. Although there are still lone electron pairs on the surfacenot involved in hydrogen bonding, the surface is too crowdedto admit further adsorption ofn-pentanol. The alkyl chain tiltangle in the anti conformation changes from 30at low coverageto 14 at high coverage due to perturbation of the hydrogenbond geometry at higher coverage. At low coverage, thereceiving and donating bonds are distinct, whereas at highcoverage, with each surface hydroxyl group forming twohydrogen bonds, each of the hydrogen bonds become close toequivalent in length and angle. Though C1-C2rotation is notpossible without considerable steric interactions with otherpentanol molecules at this coverage, rotation about the C3-C4dihedral remains possible. This can contribute to explaining the

greater alkyl chain tilt angle observed experimentally incomparison with the calculated value at 3/4 and 1 ML coverages.With consideration of the idealized-cristobalite (111) surface

as a reasonable representation of the amorphous surface usedin the experimental work, the DFT calculation predicts that thehydrogen bonding interaction fixes the orientation of the C1inthe pentanol molecule relative to the surface and that anti/gaucherotations account for the relative orientation to the surfacenormal. It also suggests that the monolayer layer coveragerepresents approximately one pentanol molecule adsorbed persurface hydroxyl. Overall, the DFT results are in agreement withthe ATR-IR experimental observations regarding the decreaseof the alkyl chain tilt angle as the coverage approaches thesaturation of the first layer anchored through hydrogen bondingwith the surface hydroxyl group. However, there are some minordifferences between the DFT calculation and the experimentaldata.

Although the OH orientations of the adsorbed alcoholmolecules for high coverages are in good agreement betweenthe experimental and theoretical analyses, the results for thelow coverages are somewhat different. In the ATR-IR experi-ment, the OH orientation for P/Psat >10% is determined to be30from the surface normal, while it is predicted to be 60in the DFT calculation for the 1/4 ML coverage. This discrep-ancy might originate from the fact that amorphous surface isnot fully terminated with the -cristobalite (111) surface. Atvery low coverages, water molecules may preferentially ad-

sorbed on defect sites and these species may dominate theobserved vibrational spectrum. As the water coverage increases,

the molecules occupy the more stable and prevalent sites thatare similar to the sites on the -cristobalite (111) surface. Inaddition, there might be some changes in the surface silanolgroups or the strongly bound water molecules present prior tothe initial water adsorption.72 These slight changes could alterthe background of the spectrum, but they cannot be determinedindependently.

The tilt angles of the pentyl chains at high P/Psat(40-50)are much larger than those predicted by the DFT theory (14)for the n-pentanol molecule with the anti conformation at highfractional coverages. On the basis of the adsorption isotherm

curve shape, it is expected that the first full layer is formed atP/Psat 10%. At this pressure, the effective thickness of theadsorbed layer is determined to be 0.60 nm, which is shorterthan the value of 0.79 nm predicted for n-pentanol with theanti conformation at 3/4 and 1 ML coverages in the DFTcalculation (Table 1). The DFT results indicate that C3-C4rotations are possible even at the surface-saturation coverageand this will reduce the thickness from 0.79 to 0.64 nm. Inaddition, some corrugation at the atomic level and the lack ofregularly spaced surface hydroxyls of the amorphous silicasurface may lead to less dense packing of the adsorbed pentanol.These could lead to higher tilt angles of the alkyl chains.

C. Comparison with Long-Chain Linear Alcohol Adsorp-

tion on Amorphous SiO2.The orientation of long-chain alcoholmolecules (n-decanol and n-octadecanol) on SiO2was investi-gated with the polarized ATR-IR to compare with the trendobserved for short-chain alcohol adsorption. In this experiment,the amount of the adsorbed alcohols was controlled by exposingthe ATR crystal surface to the n-decanol and n-octadecanolvapor for varying periods (as in the dose control in UHVexperiments). Figure 8 displays the average tilt angle of the

TABLE 1: DFT Calculated Adsorption Energy, Hydroxyl and Alkyl Chain Orientations, and Thickness for n-PentanolAdsorbed on the -Cristobalite (111) Surface as a Function of Fractional Coverage and Conformation of the(HO)-C1-C2-Propyl Dihedral Angle

alkyl chain conformation fractional coverage Eads (kJ mol-1) OH orientation (deg) (Figure 7b) (deg) d(Figure 7a) (nm)

gauche 1/4 -58 61 70 0.51anti 1/4 -57 60 30 0.73gauche 1/2 -57 53 71 0.52anti 1/2 -58 52 14 0.79anti 3/4 -59 52 14 0.79

anti 1 -

51 52 14 0.80

Figure 8. Average tilt angle of alkyl chains () with respect to thesurface normal for adsorbed n-decanol and n-octadecanol layers as afunction of adsorption isotherm thickness. Note that the tilt angle isthe value estimated by assuming the alkyl chains are in the perfectanti configuration.

10638 J. Phys. Chem. C, Vol. 113, No. 24, 2009 Barnette et al.


8/10

decyl and octadecyl chains as a function of the adsorbed layerthickness. The effective thickness of the adsorbed n-decanoland n-octadecanol layers varied from 0.25 to 1.6 nm. Notethat 1.6 nm is the maximum thickness obtained after the alcohols

were dosed on the SiO2surface at room temperature and thenthe substrate surface was purged with Ar until there was nomore decrease in the alcohol peak intensity in ATR-IR. Thisthickness is comparable to the poorly packed n-decyl silaneSAMs (1.5 nm) and much lower than the well-packed n-octadecyl silane SAMs (2.5 nm).37,73,39 The general trendobserved for long-chain alcohols with varying thicknesses issimilar to the trend observed for the short-chain alcoholadsorption under equilibrium conditions. The alkyl tilt anglesof n-decanol and n-octadecanol are initially high at lowcoverages and then approach 40(which corresponds to DR 0.8) as the adsorbed layer thickness increases to its maximumvalue.

Further insight into the reason that the DR values for adsorbedalcohol molecules at high P/Psat or high coverages are 0.8can be obtained from theoretical simulations of the ATR-IRspectra of the adsorbed layers. In these simulations, the complexrefractive indices of the bulk alcohols determined from theirrespective transmission spectra were used. Figure 9 displaysthe s- and p-spectra simulated for a 1 nm thick film ofn-pentanoland n-octadecanol on a silicon surface. When the ATR-IRspectra are simulated from the refractive index of the liquidalcohol, the dichroic ratio is 0.8 for both asymmetric andsymmetric alkyl stretching vibration peaks. This dichroic ratioof0.8 is neither a function of the adsorbed layer thickness(as long as it is much smaller than the effective penetrationdepth of the evanescent field) nor the type of alcohol simulated.

Note that as long as the molecular orientation in the adsorbedlayer remains identical to that in the bulk liquid (which is

random), the dichroic ratio of the adsorbed layer should bethe same as the one simulated with the bulk refractive index.The dichroic ratio can be sensitive to changes or deviation ofthe adsorbate optical properties from the bulk properties.47 Theexact value of the refractive index of the adsorbed molecule isnot readily available in the literature. However, it should benoted that the adsorption isotherm thicknesses predicted fromthe bulk refractive indices were essentially the same as thosecalculated from the comparison of the bulk liquid and adsorbedlayer spectra. This suggests that there is little to no deviationbetween the bulk and adsorbed phase optical properties of thealkyl chains in the alcohols.40,74 This is reasonable because thealkyl chains have only weak van der Waals interactions withthe surface. This simulation study reveals that the polarizationATR-IR experiment cannot distinguish between the alkyl chains

tilted 40 from the surface normal and the randomly orientedalkyl chains.

Last, the degree of interalkyl chain packing via van der Waalsinteractions was estimated from the peak positioned at 2940-2920cm-1, which has been assigned as the CH2asymmetric stretchingvibration for SAMs in the literature.38 This peak position is thehighest for the gas phase molecules and decreases from thisvalue when either liquid or solid phases are formed. The lowestpeak position is observed for a crystalline phase where the CH2chains are in the transconformation and packed in an orderedfashion.14,38 Figure 10 displays the peak position of the CH2vibration of the adsorbed phase alcohols as a function of chainlength. The figure also shows the vibration peak positions for

the gas-phase n-propanol and n-pentanol as well as the highlyordered n-octadecanol film obtained via heating to 70 Cduring the vapor-phase deposition. The peak positions of theCH2vibrations of the adsorbed n-propanol andn-pentanol layersare much closer to the gas-phase values than the liquid phasevalues. These short-chain alcohols do not seem to haveconsiderable interchain interactions. In the case of n-decanolandn-octadecanol, the CH2vibration peak position is 5 cm-1

red-shifted from their bulk values, indicating some degree ofpacking. However, the average of the observed alkyl chainorientation is still close to the random orientation (Figure 8).These results may imply that the long-chain alkyl groups areweakly packed in small domains and these domains arerandomly oriented in the laboratory coordinate.

When the n-octadecanol film is deposited and annealed at70 C, the alkyl chain tilt angle decreases to 20, the OH

Figure 9. s- and p-Polarization ATR-IR spectra for a 1 nm thick filmof (a) n-pentanol and (b) n-octadecanol simulated with bulk liquidoptical constants.

Figure 10. Peak position of CH2asymmetric stretching vibration (orFermi resonance in the case of short chain alcohols) as a function ofchain length for gas phase, bulk liquid, and adsorbed alcohol species.



9/10

dipole is oriented 52 from the surface normal, and the CH2peak position decreases to 2919 cm-1 (data shown in theSupporting Information). These values are congruent with thosepredicted from DFT calculations for the well-packed alcoholmolecules (e.g., the complete coverage of n-pentanol in theanticonformation). This implies that thermal activation is neededto rearrange the local packing into a long-range self-assembly.However, the thermal annealing cannot be used to preferentiallyalign the short-chain alcohols because they desorb to the gas

phase rather than being packed.

Conclusions

The structure and orientation of alcohol molecules in theadsorbed layer on silicon oxide in equilibrium with the vaporphase under ambient conditions was studied with attenuated totalreflectance infrared spectroscopy (ATR-IR). This study revealedthat asP/Psatincreases from 0 to 10%, the isotherm thicknessofn-propanol andn-pentanol increases rapidly to 0.3 and 0.6nm, respectively, which roughly corresponds to the saturationof the substrate surface. Upon further increase in P/Psat, theisotherm thickness increases only slightly until the condensationoccurs near the saturation vapor condition. The molecular

orientation of the adsorbed alcohols was also found to be afunction of surface coverage. The OH dipoles of the adsorbedalcohol molecules at 5% P/Psatexhibited a low dichroic ratiocorresponding to near perpendicular orientation from the surface.As the partial pressure was increased, the average OH dipolebecomes oriented 50 from the surface normal. The alkylchains of the alcohol molecules adsorbed at 5% P/Psat(lowestmolecular coverage) were found to be tilted toward the surface;however, as theP/Psator coverage increases, the alkyl orientationgradually changes to a structure that is either random or oriented40from the surface normal. DFT calculations corroboratedthat the alkyl chains are oriented toward the surface plane atlow fractional coverages (less than 1/2 ML). As the fractional

surface coverage increases to complete coverage, the tilt angleof alkyl chains changes toward the surface normal direction.Similar molecular orientation results were observed for adsorbedn-decanol and n-octadecanol molecules from incomplete tocomplete coverage of the alkyl groups adsorbed onto the SiO2surface.

Acknowledgment.The work was financially supported bythe National Science Foundation (grant no. CMMI-0625493).

Supporting Information Available: Calculation of effectiveisotherm thickness from ATR-IR spectra; determination ofmolecular orientation from s- and p-polarized ATR-IR spectra;determination of refractive indices from bulk transmission

spectra; ATR-IR spectra ofn-octadecanol/SiO2after annealingat 70 C. This material is available free of charge via the Internetat http://pubs.acs.org.

References and Notes

(1) Kissin, Y. V. J. Mol. Catal., A 1998, 132, 101.(2) Mowery, M. D.; Kopta, S.; Ogletree, D. F.; Salmeron, M.; Evans,

C. E. Langmuir1999, 15, 5118.(3) Bains, C. D.; Whitesides, G. M. Langmuir1989, 5, 1370.(4) Ulman, A.; Evans, S. D.; Shnidman, Y.; Sharma, R.; Eilers, J. E.;

Chang, J. C. J. Am. Chem. Soc. 1991, 113, 1499.(5) Pale-Grosdemange, C.; Simon, E. S.; Prime, K. L.; Whitesides,

G. M. J. Am. Chem. Soc. 1991, 113, 12.(6) Prime, K. L.; Whitesides, G. M. Science 1991, 252, 1164.(7) Bradbury, C. R.; Zhao, J.; Fermin, D. J. J. Phys. Chem. 2008,112,

10153.(8) Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M.J. Am. Chem. Soc. 1990, 112, 4301.

(9) Asay, D. B.; Dugger, M. T.; Ohlhausen, J. A. Langmuir2008, 24,155.

(10) Asay, D. B.; Dugger, M. T.; Kim, S. H. Tribol. Lett.2008, 29, 67.(11) Ashurst, W. R. Y., C.; Carraro, C.; Maboudian, R.; Dugger, M. T.

J. Micro/Nanolithogr., MEMS2001, 10, 41.(12) Jun, Y. Z., X. J. Adhes. Sci. Technol. 2003, 17, 593.(13) Ahn, D. J.; Franses, E. I. J. Phys. Chem. 1992, 96, 9952.(14) Allara, D. L.; Swalen, J. D. J. Phys. Chem. 1982, 86, 2700.(15) Bent, S. F.; Schilling, M. L.; Wilson, W. L.; Katz, H. E.; Harris,

A. L. Chem. Mater. 1994, 6, 122.(16) Kim, S. H.; Stair, P. C.; Weitz, E. Langmuir1998, 14, 4156.(17) Kim, S. H.; Stair, P. S.; Weitz, E. J. Chem. Phys. 1998,108, 5080.(18) Mugele, F.; Persson, B.; Zilberman, S.; Nitzan, A.; Salmeron, M.

Tribol. Lett.2002, 12, 123.(19) Pint, C. L. Phys. ReV. B 2006, 73, 045415.(20) Raut, J. S.; Sholl, D. S.; Fichthorn, K. A. Surf. Sci. 1997,389, 88.(21) White, J. M.; Henderson, M. A.J. Phys. Chem. B 2005,109, 14990.(22) Ihm, H.; White, J. M. J. Phys. Chem. B 2000, 104, 6202.(23) Syomin, D.; Wang, J.; Koel, B. E. Surf. Sci. 2001, 495, L827.(24) Weaver, J. F.; Ikai, M.; Carlsson, A.; Madix, R. J. Surf. Sci. 2001,

470, 226.(25) Zhao, X.; Kwon, S.; Vidic, R. D.; Borguet, E.; Johnson, J. K.

J. Chem. Phys. 2002, 117, 7719.(26) Dai, Q.; Gellman, A. J. J. Phys. Chem. 1991, 95, 9443.(27) Kittaka, S.; Umeza, T.; Ogawa, H.; Maegawa, H.; Takenaka, T.

Langmuir 1998, 14, 832.(28) Street, S. C.; Gellman, A. J. Surf. Sci. 1997, 372, 223.(29) Zhang, R.; Gellman, A. J. J. Phys. Chem. 1991, 95, 7433.(30) Fairbrother, D. H.; Briggman, K. A.; Stair, P. C.; Weitz, E.J. Chem.

Phys. 1995, 102, 18.(31) McCarthy, M. I.; Gerber, R. B.; Trentelman, K. A.; Strupp, P.;

Fairbrother, D. H.; Stair, P. C.; Weitz, E. J. Chem. Phys. 1992, 97, 5168.(32) Cai, S.; Sohlberg, K. J. Molec. Catal., A 2003, 193, 157.(33) Rider, K. B.; Hwang, K. S.; Salmeron, M.; Somorjai, G. A. Phys.

ReV. Lett. 2001, 86, 4330.(34) Jensen, J. A.; Rider, K. B.; Salmeron, M.; Somorjai, G. A. Phys.

ReV. Lett. 1997, 80, 1228.(35) Jugnet, Y.; Aires, F. J. C. S.; Deranlot, C.; Piccolo, L.; Bertolini,

J. C. Surf. Sci. 2002, 521, L639.(36) Bent, S. F.; Schilling, M. L.; Wilson, W. L.; Katz, H. E.; Harris,

A. L. Chem. Mater. 1994, 6, 122.(37) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D.

J. Am. Chem. Soc. 1995, 117, 3145.(38) McGuiness, C. B., D.; Masejerwski, J.; Sundararajan, U.; Cabarcos,

O.; Smilgies, D.; Allara, D. ACS NANO 2007, 1, 30.

(39) Sung, M. M.; Sung, K.; Kim, C. G.; Lee, S. S.; Kim, Y. J. Phys.Chem. B 2000, 104, 2273.(40) Tillman, N.; Ulman, A.; Penner, T. L. Langmuir1989, 5, 101.(41) Ulman, A. An Introduction to Ultrathin Organic Films from

Langmuir-Blodgett to Self-Assembly; Academic Press: San Diego, 1991.(42) Stumm, W.; Sigg, L.; Sulzberg, B.Chemistry of the Solid-Water

Interface: Processes at the Mineral-Water and Particle-Water Interfacein Natural Systems; Wiley: New York, 1992.

(43) Graubner, V.-M.; Jordan, R.; Nuyken, O.; Schnyder, B.; Lippert,T.; Kotz, R.; Wokaun, A. Macromolecules 2004, 37, 5936.

(44) Cassie, A. B. D. Discuss. Faraday Soc. 1948, 3 , 11.(45) Lamb, R. N.; Furlon, D. N. J. Chem. Soc., Faraday Trans. 1982,

78, 61.(46) Asay, D. B.; Kim, S. H. J. Phys. Chem. B 2005, 109, 16760.(47) Barnette, A. L.; Asay, D. B.; Kim, S. H. Phys. Chem. Chem. Phys.

2008, 10, 4981.(48) Harrick, N. Internal Reflection Spectroscopy; Interscience: New

York, 1967.(49) Urban, M. W. Attenuated Total Reflectance Spectroscopy ofPolymers: Theory and Practice; American Chemical Society: Washington,D.C., 1996.

(50) Hasegawa, T.; Nakano, Y.; Ishii, Y. Anal. Chem. 2006, 78, 1739.(51) Kresse, G.; Furthmuller, J. Comput. Mater. Sci. 1996, 6, 15.(52) Kresse, G.; Furthmuller, J. Phys. ReV. B 1996, 54, 11169.(53) Kresse, G.; Hafner, J. Phys. ReV. B 1993, 47, 558.(54) Kresse, G.; Joubert, D. Phys. ReV. B 1999, 59, 1758 LP.(55) Monkhorst, H. J.; Pack, J. D. Phys. ReV. B 1976, 13, 5188 LP.(56) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;

Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B 1992, 46, 6671.(57) Tsuzuki, S.; Luthi, H. P. J. Chem. Phys. 2001, 114, 3949.(58) Gerber, I. C.; Angyan, J. G. J. Chem. Phys. 2007, 126, 044103.(59) The Surface Properties of Silica; Legrand, A. P., Ed.; Wiley: New

York, 1998.(60) Chuang, I.-S.; Macial, G. E. J. Phys. Chem. B 1997, 101, 3052.

(61) Demuth, T.; Jeanvoine, Y.; Hafner, J.; Angyan, J. G. J. Phys.:Condens. Matter1999, 11, 3833.(62) Milosevic, M.; Berets, S. L. Appl. Spectrosc. 1993, 47, 566.

10640 J. Phys. Chem. C, Vol. 113, No. 24, 2009 Barnette et al.


10/10

(63) Adamsom, A.Physical Chemistry of Surfaces, 5th ed.; John Wileyand Sons: New York, 1990.

(64) Bondi, A. J. Phys. Chem. 1964, 68, 441.(65) Tong, Q. Y.; Gao, G. H.; Han, M.-H.; Yu, Y.-X.Int. J. Thermophys.

2001, 23, 635.(66) Pimentel, G. C.; McClellan, A. L. Annu. ReV. 1971, 22, 347.(67) Ketteler, G.; Yamamoto, S.; Bluhm, H.; Andersson, K.; Starr, D. E.;

Olgetree, D. F.; Ogasaware, H.; Nilsson, A.; Salmeron, M. J. Phys. Chem.C2007, 111, 8278.

(68) Sobocinski, R. L.; Pemberton, J. E.Langmuir1992, 8 , 2049.(69) Lu, R.; Gan, W.; Wu, B. H.; Zhang, Z.; Guo, Y.; Wang, H. F. J.

Phys. Chem. B 2005, 109, 14118.

(70) Michniewiez, N.; Muszynski, A. S.; Wrzeszcz, W.; Czarnecki,M. A.; Golec, B.; Hawranek, J. P.; Mielke, Z. J. Mol. Struct. 2008, 887,180.

(71) Flinn, D. H.; Guzonas, D. A.; Yoon, R.-H. Colloids Surf. A 1994,87, 163.

(72) Adsorption on Silica Surfaces; Papirer, E., Ed.; Marcek Dekker:New York, 2000; Vol. 90.

(73) Choi, J.; Kato, T. IEEE Trans. Magn. 2005, 41, 599.(74) Wasserman, S. R.; Whitesides, G. M.; Tidswell, I. M.; Ocko, B. M.;

Pershan, P. S.; Axe, J. D. J. Am. Chem. Soc. 1989, 111, 5852.

JP901064R


Documents

AdsorptionIsotherm Orientation Alcohols SiO2 Kim