silylation of an oh-terminated self-assembled monolayer surface through low-energy collisions of...

JOURNAL OF MASS SPECTROMETRYJ. Mass Spectrom. 2002; 37: 591–602Published online 16 April 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.317

Silylation of an OH-terminated self-assembledmonolayer surface through low-energy collisionsof ions: a novel route to synthesis and patterningof surfaces

Nathan Wade, Chris Evans, Sung-Chan Jo and R. Graham Cooks∗

Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, USA

Received 15 November 2001; Revised 24 February 2002; Accepted 28 February 2002

Using a multi-sector ion–surface scattering mass spectrometer, reagent ions of the general form SiR3+

were mass and energy selected and then made to collide with a hydroxy-terminated self-assembledmonolayer (HO-SAM) surface at energies of ∼15 eV. These ion–surface interactions result in covalenttransformation of the terminal hydroxy groups at the surface into the corresponding silyl ethers dueto Si — O bond formation. The modified surface was characterized in situ by chemical sputtering, alow-energy ion–surface scattering experiment. These data indicate that the ion–surface reactions havehigh yields (i.e. surface reactants converted to products). Surface reactions with Si(OCH3)3

+, followedby chemical sputtering using CF3

+, yielded the reagent ion, Si(OCH3)3+, and several of its fragments.

Other sputtered ions, namely SiH(OCH3/2OH2+ and SiH2(OCH3/OH2

+, contain the newly formed Si — Obond and provide direct evidence for the covalent modification reaction. Chemical sputtering of modifiedsurfaces, performed using CF3

+, was evaluated over a range of collision energies. The results showed thatthe energy transferred to the sputtered ions, as measured by their extent of fragmentation in the scatteredion mass spectra, was essentially independent of the collision energy of the projectile, thus pointing to theoccurrence of reactive sputtering.

A set of silyl cations, including SiBr3+, Si(C2H3)3

+ and Si(CH3)2F+, were similarly used to modify theHO-SAM surface at low collision energies. A reaction mechanism consisting of direct electrophilic attackby the cationic projectiles is supported by evidence of increased reactivity for these reagent ions withincreases in the calculated positive charge at the electron-deficient silicon atom of each of these cations.In a sequential set of reactions, 12 eV deuterated trimethylsilyl cations, Si(CD3)3

+, were used first as thereagent ions to modify covalently a HO-SAM surface. Subsequently, 70 eV SiCl3

+ ions were used tomodify the surface further. In addition to yielding sputtered ions of the modified surface, SiCl3

+ reactedwith both modified and unmodified groups on the surface, giving rise not only to such scattered productions as SiCl2OH+ and SiCl2H+, but also to SiCl2CD3

+ and SiCl2D+. This result demonstrates that selective,multi-step reactions can be performed at a surface through low-energy ionic collisions. Such processesare potentially useful for the construction of novel surfaces from a monolayer substrate and for chemicalpatterning of surfaces with functional groups. Copyright 2002 John Wiley & Sons, Ltd.

KEYWORDS: ion–surface scattering; surface modification; self-assembled monolayers; ion–surface reaction; chemicalsputtering

INTRODUCTION

It is desirable to be able to prepare functionalized sur-faces with controlled distributions of functional groups ofinterest. Applications of thin films in integrated optics,1 sen-sor design,2 microelectronic devices3 and various devicesthat mimic biological function4 depend on some means of

ŁCorrespondence to: R. Graham Cooks, Department of Chemistry,Purdue University, West Lafayette, Indiana 47907, USA.E-mail: [email protected]/grant sponsor: National Science Foundation;Contract/grant number: CHE-9732670.

chemical patterning. Substrates used for the fabrication of‘smart surfaces’ include conjugated polymers,5 ferromag-netic thin films,6 metal nanoparticle systems,7 assembledproteins8 and alternating layers of adsorbed polyionicspecies.9 The patterning required for the microfabrica-tion of such high-quality structures is usually carriedout with photolithography.10 Recently, a number of non-photolithographic techniques have been demonstrated,which include the use of x-rays,11 electron beams,12

scanning tunneling microscopy,13 conductive-tip atomicforce microscopy,14 micromachining,15 microwriting16 andmicrocontact printing.17

Copyright 2002 John Wiley & Sons, Ltd.

592 N. Wade et al.

Interactions of ions with surfaces are also being exploitedfor the chemical modification of materials. Among theseare ion implantation and surface modification plasmaprocessing.18,19 Plasma treatments have been applied tothe surfaces of polymers, metals and semiconductors, andmodifications may include the etching of material from thesurface, formation of thin films as a result of polymerizationof monomeric units from the plasma or surface reactionswhich lead to the introduction of new functional groups. Thechemical changes produced by plasma treatments, however,are the result of ions of unknown nature in addition toelectrons, photons and various high-energy neutral species.Ion beam techniques, on the other hand, can involve a specificanalyte ion through mass analysis.20 Although traditionallyused for the analysis of surfaces, ion beam techniques, suchas secondary ion mass spectrometry21 and ion scatteringspectroscopy,22 are now being modified to allow materialsprocessing.23,24

Chemical modification of surfaces is obviously best doneusing low-energy ion beams, and for this reason we havefocused our attention on collisions of ions with surfacesin the hyperthermal collision energy range (1–200 eV).Ion–surface collisions in this energy regime have been ofgrowing interest for (i) the elucidation of ion structure,25,26

(ii) the chemical analysis of surfaces27,28 and (iii) thepreparation of chemically modified surfaces.29,30 One ofthe main advantages of attempting surface modificationin this collision energy range is the ability to use theprojectile ion itself as a chemical reagent and in sodoing to modify selectively the outermost molecular layersof a surface. This is the case because the amount ofenergy transferred to the surface molecules (or to theprojectile ion) upon collision is in the range of tensof kcal mol�1 (1 kcal D 4.184 kJ), the energies needed todrive many chemical reactions involving covalent bondformation. This was initially made evident in low-energyion–surface collision studies in which scattered ions wereobserved which resulted from a transfer of an atomor group of atoms from the surface to the projectileion. These were properly recognized as the products ofion–surface reactions. An early notable example involvesH atom abstraction from hydrocarbon-covered surfacesusing odd-electron ions, such as the molecular cation ofbenzene or pyridine.31 It was further demonstrated thatthese ion–surface reactions chemically altered the surfacemolecular groups. For example, a transhalogenation reaction,performed with reagents such as PCl3

C or Si(NCO)C, hasbeen shown to transform fluorocarbon surfaces chemicallyinto terminal CF2X units, where X represents the halogenatom or pseudohalogen in the projectile ion.32 – 34 Thescattered ion beam then includes such reaction productsas PCl2FC.

A number of examples of surface modification resultingfrom ion–surface collisions at a variety of collision ener-gies have been presented in the literature. For instance,OHC and NHC ions have been grafted into polystyrenefilms,35 as have fluorine-containing projectiles.36 Siliconnitride formation has been achieved by reactions of NC

and N2C ions with an Si(100) surface,37 and diamond-

like carbon films have been deposited on clean surfacesas a result of collisions of low-energy CC ion beams.38

In a different type of experiment, we have reportedion–surface modifications with polyatomic ions in whichthe entire projectile ion remains intact upon covalently bind-ing with the surface. One example involves aromatic ions,such as the [M � H]C ion of chlorobenzene,29 which cova-lently binds to a carboxylic acid terminated self-assembledmonolayer surface through an ion–surface decarboxyla-tion reaction, reminiscent of the condensed phase Kolbereaction.

A recent study from this laboratory showed that low-energy (<15 eV) collisions of the trimethylsilyl cation,Si(CH3�3

C, at an OH-terminated self-assembled monolayersurface resulted in the covalent modification of that surface.39

In situ chemical sputtering of the modified surface and x-rayphotoelectron spectroscopic analysis demonstrated that a 1 hmodification resulted in a yield of about 30%, i.e. silylationof 30% of the hydroxy monolayer. Furthermore, time-of-flight secondary ion mass spectrometric images illustratedthe fine spatial control of this chemical modification reactionachievable when a grid was placed in front of the surface.In this paper, we show that this ion–surface silylation reac-tion can be extended to include reagent ions of the generalform SiR3

C, where the corresponding silyl ether products,R3SiOCH2 —, are generated at the surface. Silylation of sur-face hydroxyl sites is well known as a very useful process forthe manipulation of the physicochemical properties of sur-faces. Applications include improvements or alterations inresistance to corrosion,40 electrical resistance,41 wettability,42

adhesion,43 adsorption of biomolecules44 and specializedapplications within separation science.45 These solution-phase reactions can be tailored to provide the desired surfaceproperties by appropriate selection of alkyl groups in theorganosilanes of general formula SiRX3, where X is typicallya halogen or methoxy group. Hydroxy-functionalized sur-faces have also served as templates for the fabrication ofmultilayer assemblies,46 where again silane chemistry hasfigured prominently.

In this paper, evidence of Si—O bond formation in thecourse of low-energy ion–surface collisions is provided,and the occurrence of the silylation reaction is shown tobe controlled by the chemical nature of the substituentgroups on the reagent ion, SiR3

C. These chemical effectsare discussed, and the reactions are examined with respectto both the collision energy and to competing ion–surfacecollision processes. In situ chemical sputtering was usedto monitor the changes in the surface produced by thesereactions. In addition, sequential ion–surface collisions (i.e.ion–surface collisions performed at the modified surface)show evidence of sequential modification of a surface. Inthese experiments, functional groups, previously introducedby the initial reagent ion, are further modified in the nextstage of reaction. These results suggest that multi-stepsynthetic procedures at surfaces with low-energy ion beamsshould be possible. The specificity of the silylation reaction,along with the spatial control provided by using a masked

Copyright 2002 John Wiley & Sons, Ltd. J. Mass Spectrom. 2002; 37: 591–602

Silylation of HO-SAM by ion–surface collisions 593

ion beam, makes this a promising method for chemicalpatterning and controlled modification of surfaces.

EXPERIMENTAL

Hydroxy-terminated self-assembled alkanethiol monolay-ers bound to a gold film through a sulfur linkage,HO(CH2)11 —S—Au, were used as the surfaces for theseexperiments. Substrates were prepared by thermal evap-oration of 100 A of chromium and then 2000 A of goldon to silicon wafers of orientation <100> (InternationalWafer Service, Portola Valley, CA, USA). The molecularassemblies were constructed by immersing the substrates indilute (1 mM) ethanol solutions of 11-mercaptoundecanol fora period of at least 1 week at room temperature. Detailedinformation concerning the preparation and properties of thesurfaces has been provided elsewhere.47 The surfaces wererinsed and sonicated in ethanol, then dried under argonbefore being introduced into the high-vacuum scatteringchamber.

All experiments were performed using a custom-built,hybrid mass spectrometer with geometry BEEQ (B Dmagnetic sector, E D electric sector, Q D quadrupolemass analyzer), a detailed description of which has beenprovided.48 Briefly, volatile samples were independentlyintroduced into the ion source (10�5 Torr nominal samplepressure (1 Torr D 133.3 Pa)) and ionized by electronionization (70 eV). Ions of interest were accelerated to 2keV translational energy and mass and energy selected,respectively, by the magnetic and electrostatic analyzersof the double-focusing mass spectrometer. Projectile ionsof interest were decelerated to low translational energies(<100 eV), then allowed to collide with the surface in a UHVscattering chamber maintained at a nominal base pressureof 2 ð 10�9 Torr (typical operating pressures were below5 ð 10�9 Torr).

For the ion–surface scattering experiments, the samplewas rotated so that the primary ion beam was incident at 55°

to the surface normal, while scattered ions were collected atthe specular angle. Collection angles were not varied, but iontrajectory simulations have shown that under the conditionsused, the analysis system accepts a wide range of scatteringangles, ¾30° on either side of the specular angle. Scatteredions were analyzed using a quadrupole mass analyzer pre-ceded by an electrostatic analyzer, used as a kinetic energyfilter. The kinetic energy analyzer was set in a low-resolutionmode so as to pass ions of a broad range of energies in therange of a few electronvolts. These conditions are knownto transfer efficiently the products of inelastic, reactive, andchemical sputtering processes. The selected conditions givemaximum scattered ion transmission without significantangular or velocity discrimination, although angular andvelocity resolved experiments have been performed pre-viously using the same instrument. Projectile ions usedin this study either as a reagent ion or for analysis ofthe surface included CF3

C, SiBr3C, Si(CH3)2BrC, SiCl3

C,Si(CH3)2(CH2Cl)C, Si(CH3)2ClC, Si(C2H5�3

C, Si(CH3)2C6H5C,

Si(CH3)2FC, Si(CH3)2NCOC, Si(CH3)2CNC, Si(OCH3�3C and

Si(CD3�3C, derived, respectively, from perfluorohexane,

tetrabromosilane, trimethylbromosilane, tetrachlorosi-lane, chloromethyl(trimethyl)silane, dimethyldichlorosi-lane, tetraethylsilane, trimethylphenylsilane, trimethylflu-orosilane, trimethylisocyanatosilane, trimethylcyanosilane,tetramethoxysilane (all from Aldrich, Milwaukee, WI,USA) and tetramethylsilane-d12 (C/D/N Isotopes, Quebec,Canada).

Standard conditions for performing surface modificationinvolved orienting the surface substrate perpendicular tothe primary ion beam. Collision energies for the projectileions were set at 12 eV, and the primary ion current densitywas maintained at 0.8 nA cm�2. The spot size was estimatedas about 25 mm2, and the reaction times, unless otherwiseindicated, were chosen to be 2 h. These conditions were usedfor the comparison of each of the reagent ions. Reaction timesand collision energies were varied for other purposes.

Ab initio calculations were carried out using standardprocedures in the Gaussian 98 suite of programs.49 Single-point energies and atomic charge densities were determinedfor each of the silylium cations and model reaction productsusing the Hartree–Fock level of theory by employing thepolarization 6–31G(d,p) basis set.50

RESULTS AND DISCUSSIONEvidence for surface modification: reaction withSi(OCH3/3

+ and Si(C2H5/3+

The mass spectrum of ions scattered during 70 eV collisionsof CF3

C ions at an OH-terminated self-assembled monolayer(HO-SAM) surface is represented in Fig. 1(a). Because ofthe relatively high ionization energy of CF3

ž (9.8 eV) andthe strength of its C—F bond,51 charge exchange betweenthis projectile ion and an organic surface, not projectileion fragmentation, is the dominant ion–surface collisionevent. Charge exchange ionizes surface-bound molecules,and momentum transfer due to the impinging projectilereleases the product ions into the gas phase where they maysubsequently fragment. This process, known as chemicalsputtering, yields ions diagnostic of the chemical functional-ity present at the surface. Previous studies have examined thechemical sputtering phenomena and its capabilities for char-acterizing surfaces.52 Evidence of the presence of terminalhydroxy functional groups on this self-assembled monolayersurface is provided by the sputtered ions, H3OC (m/z 19),CH2OHC (m/z 31) and C2H4OHC (m/z 45), in the massspectrum. Other ions sputtered from the surface occur atnominal masses that correspond to CnH2n�1

C and CnH2nC1C.

These ions represent either fragmentation products from thehydroxy-terminated monolayer or fragments from hydro-carbon molecules adsorbed to the surface. Adventitioushydrocarbons and water are typically observed in moderateamounts even under high-vacuum conditions.

In an attempt at surface modification, the HO-SAMsurface was bombarded with projectile ions of formulaSi(OCH3�3

C at ¾12 eV collision energy for a period of 2 h. Theprimary ion current density was maintained at 0.8 nA cm�2.The number of cations colliding with the surface over thisperiod corresponds to ¾25% of the surface monolayer, usingestimates that were used for coverage determination in thecase of ion–surface reaction with the trimethylsilyl cation.39


594 N. Wade et al.

m/z (Thomson)

Rel

ativ

e Io

n A

bund

ance

CH2OH+

H3O+

Si(OCH3)3+

SiH(OCH3)2+

SiH2(OCH3)+

SiH(OCH3)(OH)+

Si(OCH3)2OH+

H3O+

CH2OH+, SiH3+

C2H4OH+

SiH(OCH3)2OH2+

SiH2(OCH3)OH2+

10 20 30 40 50 60 70 80 90 100 110 120 130

Si(OCH3)3+

SiH(OCH3)2+

CH2OH+, SiH3+

SiH2(OCH3)+

Si(OCH3)2OH+

SiH(OCH3)2OH2+

SiH(OCH3)(OH)+

SiH2(OCH3)OH2+

(c)

(b)

(a)

Figure 1. Scattered ion mass spectra recorded upon 70 eV collisions of CF3C with (a) HO-SAM surface, (b) HO-SAM surface after

reaction with Si(OCH3�3C ion and (c) HO-SAM surface after condensed-phase reaction with Si(OCH3�4.

Following this reaction, chemical sputtering with 70 eVCF3

C was again used to interrogate the surface. Verificationof surface modification is provided in the scattered ionmass spectrum (Fig. 1(b)). Several abundant chemicallysputtered product ions that appear in this spectrum arenot observed in the scattering of CF3

C from the original HO-SAM surface. These are characteristic of the modificationreaction and include the original projectile ion, Si(OCH3�3

C

(m/z 121), and several of its fragments, Si(OCH3�2OHC (m/z107), Si(OCH3)2HC (m/z 91), Si(OCH3)(OH)HC (m/z 77),Si(OCH3�H2

C (m/z 61), and Si(OCH3�C (m/z 59); however,the possibility that these products involve reaction withwater in addition to fragmentation is not excluded. Thesputtered ion at m/z 31 is judged to include both CH2OHC

from unmodified portions of the surface and SiH3C, another

common fragment ion of Si(OCH3�3C. Previous experiments

have shown that an abundant m/z 31 ion occurs fromhydroxyl-terminated SAMs when examined with othercharge exchange reagent ions.63 The observation of these

sputtered ions in relatively large abundance is indicative ofthe fact that Si(OCH3�3

C has been transferred to the surface.More importantly, evidence that Si—O bond formation hasoccurred is found in the two peaks at m/z 79 and 109.Ions with these m/z values are not observed in an EImass spectrum of Si(OCH3�4 or in the gas-phase collision-induced dissociation mass spectrum of Si(OCH3�3

C. Rather,these ions are likely to result from fragmentation of thenewly formed silyl ether. Possible chemical formulas forthese ions are SiH2�OCH3�OH2

C and SiH(OCH3�2OH2C,

respectively, where one oxygen atom originates from theoriginal monolayer surface and two additional H atomshave been abstracted by this oxygen atom. The hydrogentransfer process observed in these fragmentation productsmay be similar to that observed from the sputtering of H3OC

from the terminal HO-SAM surface (Fig. 1(a)).The above evidence for surface modification and cova-

lent bond formation is not definitive. In fact, previousexperiments involving both fluorinated and hydrocarbon



self-assembled monolayer surfaces53,54 showed that colli-sions of ions at energies below 15 eV resulted in physicalentrapment (soft-landing) of some projectile ions withinthe monolayer framework, not chemical bonding. The soft-landed ions were released and observed in subsequent chem-ical sputtering experiments. In a previous study involving thetrimethylsilyl cation,39 two experiments were performed inan effort to distinguish the proposed silylation reaction fromearlier soft-landing experiments. These experiments wererepeated here for the trimethoxysilyl cation. After bombard-ment of the HO-SAM surface with Si(OCH3�3

C, the surfacewas removed from the vacuum chamber, washed with acidicanhydrous methanol and reintroduced to vacuum. The argu-ment is that a covalent modification would persist after thetreatment, whereas physically adsorbed molecules would bereleased from the surface. Chemical sputtering with CF3

C

yielded a spectrum identical with that observed in Fig. 1(b),while similar treatment has been shown to erase the modi-fication peaks from the chemical sputtering mass spectrumfor a soft-landing experiment.

Other evidence of covalent modification was providedwhen ethanolic tetramethoxysilane was reacted with anHO-SAM surface in the presence of a catalytic amountof triethylamine. This condensed-phase reaction occursby nucleophilic attack of the surface terminal hydroxylfunctionalities on the silicon atom with concomitant lossof methanol and results in the formation of the trimethoxysi-lyl ether-terminated monolayer surface. This syntheticsurface was likewise analyzed by ion–surface scatteringusing the projectile ion CF3

C. Chemically sputtered ionsagain dominate the spectrum (Fig. 1(c)), and the sameions which were characteristic of the ion–surface modifi-cation, namely Si(OCH3�3

C, Si(OCH3�2OHC, Si(OCH3�2HC,Si(OCH3�(OH)HC, Si(OCH3�H2

C and Si(OCH3)C, are alsoobserved. The only significant difference between the spectraof the gas-phase and condensed-phase surfaces is a decreasedpeak at m/z 31 for the solution-phase prepared surface. Thesmaller amount of the sputtered ion CH2OHC seems to indi-cate that the condensed phase reaction had a higher yieldthan the ion–surface modification, i.e. more of the terminalhydroxyl groups were converted into silyl ethers.

Another example of the silylation reaction is providedby an experiment in which ions of the form Si(C2H5�3

C

were reacted at the HO-SAM surface at a collision energyof 12 eV for a period of about 2 h. The primary ioncurrent density was maintained at 0.8 nA cm�2. Figure 2(a)shows the scattered ion mass spectrum resulting from CF3

C

chemical sputtering of the modified surface at 80 eV collisionenergy. Characteristic ions sputtered from the modifiedsurface include Si(C2H5�3

C, Si(C2H5�2HC, Si(C2H5�H2C and

Si(C2H5�C. As is the case with the modification involvingSi(OCH3�3

C, the nature of some of the chemically sputteredions yields direct evidence of Si—O bond formation for themodification reaction involving Si(C2H5�3

C. Ions at m/z 49and 77 probably represent the sputtered ions SiH3OH2

C andSiH2(C2H5�OH2

C. Again, ions containing these m/z valuesare not observed upon dissociation of the Si(C2H5�3

C ion orin the EI mass spectrum of Si(C2H5�4

Cž.

Reaction with Si(C2H5/3+: effects of changing

collision energy of sputtering agentIn the study of ion–surface collisions, much attentionhas been given to the transfer of kinetic energy of theprojectile ion into its internal modes, often leading to itsfragmentation.55 This fragmentation process is referred toas surface-induced dissociation (SID). It is well understoodthat with proper selection of both the target surface andthe collision energy, the internal energy uptake by theprimary ion can be controlled.56 Therefore, the resultingfragmentation pattern observed in the scattered ion massspectrum is indicative of this energy transfer. Determinationsof the transfer of the kinetic energy of the projectile ion toenergy of the surface have been made. In these studies,the kinetic energy and the internal energy of the scatteredions were measured, and it was assumed that the remainingenergy was imparted to the surface. For ethanol ions collidedwith a hydrocarbon monolayer on a stainless-steel target,57 itwas reported that ¾60% of the kinetic energy of the projectileion was transferred into the surface, whereas moleculardynamic simulations58 of Si(CH3�3

C colliding with an alkyl-terminated SAM showed that 30% of the kinetic energy ofthe projectile ion was transferred to the surface (note thatthese data refer to different materials and different impactangles).

It is possible to consider the energy transfer associatedwith chemical sputtering in a similar way. Figure 2(a)shows a fragmentation pattern observed as a result ofchemical sputtering of the Si(C2H5�3O-terminated surface.This fragmentation is indicative of the amount of energytransferred to the surface molecules as a result of theion–surface collision involving CF3

C. It has been shownexperimentally that the appearance energy of Si(C2H5�C

from tetraethylsilane is 19.4 eV and that of SiH2Cž is 25.7 eV.59

From Fig. 2(a), Si(C2H5�C (m/z 57) is clearly present (note thatm/z 57 could also represent the fragment ion, Si(C2H3�H2

C�,while SiH2

Cž (m/z 30) is not observed, indicating that its onsetenergy has not been reached. This allows one to estimate thatthe energy transferred to the surface is between 20 and25 eV. For an 80 eV collision, this indicates that 25–30% ofthe translational energy of the primary ion is transferred intothe ionization and fragmentation of the surface molecule.This leads to the question of whether, if the collision energyof CF3

C were varied, a change in the fragmentation patternwould be observed.

Figure 2(b) shows the result of interrogation of themodified surface with CF3

C at 40 eV collision energy.Noticeable differences from the 80 eV spectrum are observed.The elastically scattered ions and the ion–surface reactionproduct, CF2HC, are observed in high abundance. Moreimportantly, it is observed that the fragmentation pattern ofthe sputtered ions resulting from the surface silyl ethermolecules at 40 eV collision energy is the same as thatobserved at 80 eV collision energy. The same fragmentationpattern is observed for all collision energies in the range40–100 eV (below 40 eV collision energy, reasonable spectracould not be obtained owing to the low secondary ioncurrent). It has been shown previously that onset energies forthe chemical sputtering of surface groups are dependent on


596 N. Wade et al.

m/z (Thomson)

10 30 50 70 90 110 130

Rel

ativ

e Io

n A

bund

ance Si(C2H5)3

+

SiH(C2H5)2+

SiH2(C2H5)OH2+

SiH2(C2H5)+

(a) CF3+ // HO-SAM Surface

@ 80 eV Collision EnergyAfter Si(C2H5)3

+ Deposition

Si(C2H5)3+

SiH(C2H5)2+

SiH2(C2H5)OH2+

CF3+

SiH2(C2H5)+

SiH3OH2+ CF2H+

SiH3OH2+

(b) CF3+ // HO-SAM Surface

@ 40 eV Collision EnergyAfter Si(C2H5)3

+ Deposition

Si(C2H5)+CH2OH+

CH2OH+

Figure 2. Scattered ion mass spectra recorded upon (a) 80 and (b) 40 eV collisions of CF3C with an HO-SAM surface after reaction

with Si(C2H5�3C ion.

the exothermicity of the charge-exchange reaction and henceon the ionization energy of the sputtering reagent.60 Ionswith high ionization energies can sputter particular surfacegroups at lower energies than lower ionization energy ions.The additional energy needed to drive the endothermiccharge-exchange processes is provided by the kinetic energyof the projectile ion. Using known gas-phase thermochemicaldeterminations,59 the charge-exchange process is estimatedto be ¾0.1 eV endothermic in this experiment, i.e. it isapproximately thermoneutral. The results demonstrate thatthe internal energy obtained by the surface molecules as aresult of chemical sputtering has at most a weak dependenceon changes of the collision energy of the sputtering agent,at least at the energies examined. This result suggests thatthe process involved is exothermic reactive sputtering, inwhich the CF3

C ion is involved in C—O bond formation.These results are consistent with those observed in chemicalsputtering of an F-SAM surface with the projectile ionsArC, XeCž and BrC (C. Evans, T. Pradeep, D. Denault andR. G. Cooks, work in progress).

Reaction with Si(CD3/3+: energetics and

ion–surface processesReaction proceeding through covalent attachment of theprimary ion to the surface is only one of the events that occurduring ion–surface collisions involving silylium cations.Figure 3 shows the scattered ion mass spectrum resultingfrom 55 eV collisions of Si(CD3�3

C with the HO-SAM surface.From this mass spectrum, at least three different typesof ion–surface processes occur, elastic scattering, inelasticscattering and reactive scattering. The elastically scatteredprimary ion (m/z 82) is observed in low abundance, whereasthe inelastic (SID) products are observed with much greaterintensity and include SiCž (m/z 28), SiDC (m/z 30), SiD3

C

(m/z 34), SiCD3C (m/z 46), Si(CD3�DCž (m/z 48), Si(CD3�D2

C

(m/z 50) and Si(CD3�2Cž (m/z 64). Other observed product

ions are believed to be reactively scattered ions and includeSiD2HC (m/z 33), Si(CD3�2HC (m/z 65) and Si(CD3�2OHC

(m/z 81). Note that none of the ions observed in this massspectrum are the result of chemical sputtering, which is a verydifferent situation to that observed in the scattered ion mass



Rel

ativ

e Io

n A

bund

ance

m/z (Thomson)

10 20 30 40 50 60 70 80 90

Si(CD3)3+

Si(CD3)2OH+

Si(CD3)2+•

Si(CD3)2H+

Si(CD3)D2+Si(CD3)+

Si(CD3)D+•

SiD+

SiD3+

SiD2H+

Si+•

Figure 3. Scattered ion mass spectrum recorded upon 55 eV collisions of Si(CD3�3C with the HO-SAM surface.

Collision Energy (eV)

10 20 30 40 50 60 70 80 90

Perc

ent I

on A

bund

ance

0

10

20

30

40

50

60

70

80

90

100

Elastically ScatteredDissociatively Scattered

Reactively Scattered

Figure 4. Energy-resolved mass spectra (ERMS) plot recorded from Si(CD3�3C scattering from an HO-SAM surface at collision

energies ranging from 20 to 90 eV. (ž) Elastically scattered ions; (�) dissociatively scattered ions; (�) reactively scattered ions.

spectrum of CF3C, and is ascribed simply to the difference

in ionization energies of the corresponding radicals. Figure 4summarizes how the production of each of these types ofions is affected by the collision energy. This energy-resolvedmass spectra (ERMS) plot expresses each of the product ionabundances grouped into the three categories mentionedabove. These results are normalized over the range ofcollision energies, 20–90 eV. At low collision energy, i.e.<30 eV, elastic scattering dominates the scattered ion massspectrum, since not enough energy is provided to dissociatethe ion. With increasing collision energy, fragmentationof the primary ion increases, resulting in an increase indissociatively scattered ions. Reactively scattered ions forthis system begin to appear at ¾50 eV collision energyand continue to increase as the collision energy increases.The dissociative ion–surface reaction products observedpresumably do not occur at lower collision energies because

of the activation barrier associated with abstraction of atomsor group of atoms from the surface. Note that one of theproduct ions, Si(CD3�2OHC, involves formation of the Si—Obond, presumably as a result of the silylation reaction.

To understand how the various reactive processes maybe related, the efficiency of the silylation reaction was alsoevaluated as a function of collision energy. The ERMS plotprovides information about the scattered ions resulting fromcollisions of the Si(CD3�3

C ion with the HO-SAM surface,but no information is provided concerning changes whichmay be incurring at the surface. It has been shown abovethat this information can be acquired by interrogating thesurface with CF3

C after Si(CD3�3C bombardment of the

surface for some period of time. This was done for a widerange of collision energies and the result is illustrated inFig. 5. The bombardment time was 2 h at each collisionenergy examined and the primary ion current density was


598 N. Wade et al.

00 10 20 30 40 50 60 70

10

20

30

40

50

60

70

80

90

100

Collision Energy (eV)

% S

putte

red

Ions

Ori

gina

ting

from

Tri

met

hyls

ilyl M

odif

ied

Surf

ace

Figure 5. Effect of collision energy on silylation of an HO-SAM surface using deuterated trimethylsilyl cations. Chemical sputteringusing CF3

C was used for analysis of the modified surface. The ordinate represents the fraction of chemically sputtered ionsoriginating from the modified surface groups vs the total number of chemically sputtered ions arising from the modified andunmodified surface.

held constant at 0.8 nA cm�2 for each. The ordinate in theplot is the fraction of sputtered ions ascribed to the newlyformed silyl ether functionality compared with the totalchemically sputtered ion abundance (i.e. those resulting fromboth the HO-terminated substrate and the modified surface).The reaction proved to be most efficient at about 15 eVcollision energy. The reactivity decreases with increasingcollision energy until about 60 eV, where it no longer occurs.At collision energies <15 eV the reactivity also decreasesrapidly. The important correlation to note is the loss ofthe elastically scattered ion in the ERMS plot and thesimultaneous decrease in the reaction efficiency as observedin Fig. 5. Clearly, the decrease in the reaction efficiency is aresult of this decrease in the survival of the intact primaryion. Formation of the Si—O bond still occurs at highercollision energies, e.g 60 eV, as evidenced by the production, Si(CD3�2OHC; however, the increased energy of theinteraction produces ion and neutral fragments other thanthe intact silyl ether product on the surface.

Substituent effects on silylation reactionsNucleophilic substitution occurs rapidly at silicon andtypically proceeds in the condensed phase through afive-coordinated intermediate. Reactions involving trivalentpositive silicon have been difficult to study owing to thedifficulty of preparing these cations in the condensed phase.60

These cations have been studied in the gas phase, whereefforts have involved determination of thermodynamicstabilities and the effects resulting from changing substituentgroups.61 In our experiments, nucleophilic addition totrivalent silicon was achieved through low-energy collisions

of these silylium cations (SiR3C) with a surface-attached

alcohol. Formation of the silyl ether product on the surfacehas been established, yet the mechanism by which the protonattached to the alcohol leaves is unknown. Proton loss issuggested to occur after Si—O bond formation, but it mayalso happen through a concerted process. To understand thethermodynamics of the situation, the model reaction belowmay be considered, in which Si(OCH3�3

C interacts with a1-butanol molecule.

Si(OCH3�3C C CH3CH2CH2CH2OH

! [CH3CH2CH2CH2OHSi�OCH3�3]C �1�

[CH3CH2CH2CH2OHSi�OCH3�3]C

! CH3CH2CH2CH2OSi�OCH3�3 C HC �2�

Step (1), which describes electrophilic attack by the silylcation at the lone pair electrons of the oxygen atom, resultsin the formation of a positively charged intermediate. ThisSi—O bond formation step should be facile, and is calculatedto be exothermic by 2.20 eV. However, the loss of a protonfrom this intermediate, as described in step (2), is highlyendothermic. The heat of reaction for the entire process iscalculated to be 7.15 eV endothermic. The driving force forthis reaction must be the kinetic energy of the projectile.These energy values also agree qualitatively with the datacollected in Fig. 5, which shows how the reaction efficiencydecreases dramatically as collision energies fall below 10 eV.

The ion–surface silylation reaction was examined furtherby altering the substituent groups on the silicon atom ofthe reagent ion. All attempted silylation reactions were



Table 1. Effect of chemical structure and local atomiccharge on silylation efficiency

Reagent iona

Atomic chargedensity on

silicon(HF/6–31G)b

Ratio ofchemicallysputteredions(%)c

SiCl3C 0.953 0

SiBr3C 0.977 0

Si(CH3)2ClC 1.088 0Si(CH3)2BrC 1.091 0Si(CH3)2CNC 1.129 9Si(CH3)2(C6H5)C 1.133 46Si(CH3)3

C 1.152 84Si(C2H5)3

C 1.164 85Si(CH3)2NCOC 1.182 42Si(CH3)2NC2H6

C 1.182 75Si(CH3)2CH2ClC 1.183 53Si(CH3)2FC 1.401 72Si(OCH3)3

C 1.603 56

a Silylium ions used in attempted ion–surface modificationexperiments.b Local atomic charge associated with silicon atom of thereagent ion as calculated by Hartree–Fock ab initio theorywith a basis set of 6–31 G(d,p).c Percentage of chemically sputtered ions observed in the70 eV scattered ion mass spectra of CF3

C that result from amodification reaction.

interrogated by subsequent chemical sputtering using CF3C.

Table 1 lists in the first column those silylium ions examined.Note that each experiment was performed under identicalconditions: 12 eV collision energy for a period of 2 h, witha primary ion current density of 0.8 nA cm�2. The thirdcolumn expresses the fraction of sputtered ions originatingfrom the modified surface compared with the total numberof sputtered ions (i.e. those originating from both themodified surface and the OH-terminated substrate). Thiscolumn provides only semi-quantitative information aboutthe reactivities of these ions, since it is recognized that theCF3

C sputtering efficiency of some surface groups may bedifferent from those of others. The second column displaysthe local charge associated with the silicon atom of thereagent ion as calculated by ab initio methods. Silyliumcations bear a strong electropositive charge on the siliconatom, and the trend here shows that those ions which containthe most positive charge on the silicon atom are the mostlikely to react. Carbon is slightly more electronegative thansilicon, so attachment of alkyl groups to the silicon atomwithdraws electron density from the silicon atom. Consistentwith their high net charge, those silyl cations containing onlyalkyl substituents reacted readily with the HO-SAM surface.In the case of another successful reactant, Si(CH3�2FC, thefluorine group is an even more electronegative group, andit withdraws enough electron density such that the siliconatom has a formal local positive charge of 1.401. On the otherhand, the trichlorosilyl ion, SiCl3

C, which has a calculatedpositive charge on its silicon atom of 0.953, did not react.Although the Cl atom is more electronegative than Si, the

atom is much larger and �-orbital overlap between the Siand Cl atoms may exist. Such an interaction would allowelectron density from the lone pair electrons on the Cl atomto be shared with the silicon atom, decreasing the positivecharge on the silicon atom.

Clearly, each of the above reactions is endothermic andrequires the energy provided by the kinetic energy of theprimary ion. However, the electrophilicity of the reagent ionis also a determining factor for the success of the silylationreaction. Other chemical factors such as steric hindrance tothe ion–surface interaction or the stability of the ion itselfare also expected to affect the reactivity of these ions.

Sequential ion–surface reactionsUp to this point, the silylation reaction has been examinedthrough in situ analysis with CF3

C. The characteristicof this ion, as mentioned, is that it produces primarilychemically sputtered ions in the secondary ion massspectrum. In the course of low-energy ion–surface collisions,other processes, namely ion–surface reactions and surface-induced dissociation have also been used to identify ordifferentiate surface types. For these reasons there is interestin surface analysis with the more reactive trichlorosilyl cationSiCl3

C. This ion was used to compare a fresh HO-SAM surfacewith an HO-SAM surface modified with the deuteratedreagent ion Si(CD3�3

C. Figure 6(a) displays the scattered ionmass spectrum resulting from 70 eV collisions of SiCl3

C withthe HO-SAM surface. As in the collisions of CF3

C with theunmodified HO-SAM surface, the chemically sputtered ions,H3OC, CH2OHC and C2H4OHC appear in the scattered ionmass spectrum as a result of collisions with SiCl3

C. Collisionswith SiCl3

C, however, produce several scattered ions that donot result from chemical sputtering. Elastic scattering ofthe projectile ion is indicated by a small peak at m/z 133,and inelastic scattering of the projectile ion leads to theSID product ions, SiClC and SiCl2

Cž. The remaining peaksare the result of reactively scattered ions that incorporateatoms or groups of atoms abstracted from the surface.These reactively scattered product ions are indicative ofthe chemical functionality present at the surface and includethe ions SiCl2OHC and SiCl2HC. In general, the dissociativeion–surface reaction product, SiCl2XC, which results fromscattering of SiCl3

C is useful for analysis of surfaces. Whenfluorocarbon and hydrocarbon monolayer surfaces wereexamined in previous studies with the SiCl3

C ion,62 SiCl2FC

and SiCl2CH3C were scattered from these respective surfaces.

The spectrum recorded as a result of 70 eV collisions of SiCl3C

with the Si(CD3�3C-modified HO-SAM surface (Fig. 6(b))

shows numerous differences from the unmodified HO-SAM surface. First, the elastically scattered ion, SiCl3

C, isabsent from the spectrum in Fig. 6(b). Loss of the elasticallyscattered ion can be attributed either to increased dissociationof the primary ion due to greater energy transfer to thation, or to increased neutralization of that ion. The latter ismost likely because of the increase in chemically sputteredions observed in Fig. 6(b), which are known to result fromneutralization of the primary ion. The sputtered ions fromthe modified surface dominate the spectrum and includeSi(CD3�3

C and common dissociation products of that ion,


600 N. Wade et al.

m/z (Thomson)10 30 50 70 90 110 130 150

Rel

ativ

e Io

n A

bund

ance

CH2OH+

C2H4OH+

SiCl+

SiCl2+•

SiCl2H+

SiCl2OH+ SiCl3+

SiD2(CD3)+

Si(CD3)3+

SiCl2CD3+

SiCl2OH+

SiCl2H+

SiCl2D+

SiCl2+•

SiD+

Si(CD3)+

SiCl+

Si(CD3)2+•

SiD(CD3)2+

SiCl3+ // HO-SAM Surface

After Si(CD3)3+ Deposition

@ 70 eV Collision Energy

CH2OH+

(a)

(b) × 4

SiCl3+ // HO-SAM Surface

@ 70 eV Collision Energy

Figure 6. Scattered ion mass spectra recorded upon 70 eV collisions of SiCl3C with (a) the HO-SAM surface and (b) the HO-SAM

surface after modification with Si(CD3�3C ion.

Si(CD3�2DC, Si(CD3�2Cž, Si(CD3�D2

C, SiCD3C and SiDC. Note

that these are the same fragment ions as observed in thescattered ion mass spectrum of Si(CD3�3

C impacting uponthe HO-SAM surface in Fig. 3. More important in this massspectrum is the array of reactively scattered ions of the formSiCl2XC, where X represents chemical entities picked upfrom the surface. The scattered ions SiCl2HC and SiCl2OHC

are still present and represent functional groups from theunmodified surface, but two new ions are observed, SiCl2DC

and SiCl2CD3C. These are clearly diagnostic of the modified

surface. Beyond the qualitative aspects of this result, it isimportant to note that the ratio of SiCl2OHC to SiCl2CD3

C is1 : 3 in this mass spectrum. Further evaluation of this massspectrum shows that ¾90% of the chemically sputtered ionsresult from the silylation modification.

The quantitative aspects of this silylation reaction wereaddressed in a previous paper,39 and the results proved thatthe trimethylsilylation reaction results in ¾30% conversionof the HO-terminated monolayer. Therefore, the enhancedobservation of the modified portions of the surface is the

result of a greater susceptibility of these molecular groupsto chemical sputtering and reaction. More than likely,this increased reactivity stems from the fact that thesecovalently bound reagents on atomic dimensions are moreaccessible than the OH-terminal substrate after reaction.Ion–surface interactions, including chemical sputtering andreactive scattering, have been shown to be very sensitiveto the depth placement of chemical groups within amonolayer.27

The ion–surface reaction products SiCl2DC andSiCl2CD3

C are the first examples in which a surface thathas been modified through low-energy ion reaction hasundergone a subsequent ion–surface reaction involvingatoms or groups of atoms that were produced in theprevious ion–surface reaction. As mentioned, the ability toperform sequential, multi-step reactions at surfaces shouldbe a significant capability in chemical modification ofsurfaces. The other important implication of this result isthe demonstration that low-energy collisions of ions can beused to characterize modified surfaces. Scheme 1 displays



Au

S

CH2

CH2 CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2CH2

CH2CH2 CH2 CH2

(CH2)8 (CH2)8 (CH2)8 (CH2)8 (CH2)8

OH

S

OH

S

OH

S

OH

S

OH

Si(CD3)3+ (12 eV)

(a) Silylation Reaction

SiCl3 + (70 eV)

Au

S

OH

S

CH2

CH2

CH2

CH2 CH2 CH2 CH2

CH2

CH2CH2

CH2CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

CH2

(CH2)8 (CH2)8 (CH2)8 (CH2)8 (CH2)8

(CH2)8 (CH2)8 (CH2)8 (CH2)8 (CH2)8

O

SiD3C CD3

CD3 CD3 CD3

D3C CD3 D3C CD3

S

O

Si

S

OH

S

O

Si

Au

S

OH

S

O

SiCl

CD3 CD3 CD3

CD3 CD3 CD3

S

OH

S

O

SiCl

S

O

SiCl

(b) Halogen-Alkyl Exchange Reaction

Scheme 1. (a) Collisions of the Si(CD3�3C ion at the HO-SAM

surface at 12 eV energy results in the transformation of the sur-face to the corresponding silyl ether. (b) The Si(CD3�3

C-modifiedsurface from step (a) is treated with SiCl3

C at 70 eV collisionenergy and results in the incorporation of a Cl atom into thesurface monolayer by Si—Cl bond formation.

these events pictorially. Step (a) describes the silylationreaction involving the Si(CD3�3

C ion, which transforms the

surface into the corresponding silyl ether. Step (b) describesa second reaction event, collision of 70 eV SiCl3

C ions.Figure 6(b) illustrates the scattered ions that result from thislatter collision process. The ion–surface reaction productsSiCl2DC and SiCl2CD3

C result from a Cl-for-alkyl exchangereaction. Evidence for this reaction type has been providedpreviously,33 where selective modification through covalentsubstitution of the halogen atom into the surface moleculeshas been shown. The transfer reaction was verified by anexperiment in which an HO-terminated SAM surface wasmodified using Si(CH3�3

C, then treated with SiCl3C ion

a collision energy of 20 eV. The resultant surface uponinterrogation using CF3

C gave SiCl(CH3�2C ion at m/z 93

in addition to the other ions that come from HO-SAMsurface after modification by Si(CH3�3

C. This demonstratesthe formation of an Si—Cl bond at the surface to complementthe evidence from the reactively scattered ions. Owing to theselective chemical nature of ion–surface reactions, the abilityto perform multiple ion–surface reactions in sequence hasimportant implications for routes to new synthetic surfaces.

CONCLUSION

Evidence of an ion–surface silylation reaction has beendemonstrated using low-energy collisions of mass- andenergy-selected silyl cations with an OH-terminated surface,resulting in the formation of the corresponding silyl etherproduct on the surface. Direct quantitative measurementof surface conversion due to silylation reactions was notperformed in this study, but the ion abundances observedin the scattered ion mass spectra were comparable to thoseobserved in the case of trimethylsilylation39 where surfaceconversion was ¾30% of a monolayer. These endothermicreactions are driven by the kinetic energy of the reagention. The reaction efficiency was greatest for the silyliumcations at ¾15 eV and decreased with increasing collisionenergy as fragmentation of the reagent ion became morelikely. Furthermore, the chemical dependence of this reactionseemed to be controlled by the electrophilicity of the reagention. In general, silylium cations, SiR3

C, with R groups able towithdraw electron density away from the Si atom are morelikely to react.

Observation of Si—O bond formation was accomplishedin these experiments through subsequent in situ ion–surfacescattering. Although chemically sputtered products havebeen utilized before to identify surface modification, theproduct ions, SiCl2DC and SiCl2CD3

C, observed duringSiCl3

C scattering are the first cases in which surfacemodification as a result of low-energy ion–surface collision isdemonstrated in the form of ion–surface reaction products.These products also are evidence for a halogen-for-alkylsubstitution reaction performed on the surface previouslymodified through silylation. This is a clear demonstration ofthe use of low-energy ion–surface reactions to perform multi-step synthesis at a surface. These techniques are promisingfor performing selective chemical writing on surfaces.

AcknowledgementsThis work was supported by the National Science Foundation,CHE-9732670.


602 N. Wade et al.

REFERENCES1. Rogers J, Prybyla J, Meier M, Dodabalapur A. Polym. Mater. Sci.

Eng. 1999; 81: 22.2. Sailor M, Lee E. Adv. Mater. 1997; 9: 783.3. Beh W, Kim I, Qin D, Xia Y, Whitesides G. Adv. Mater. 1999; 11:

1038.4. Rieke P, Tarasevich B, Wood L, Marsh B, Song L, Fryxell G,

Engelhard M, Baer D, John C. Mater. Res. Soc. Symp. Proc. 1994;330: 139.

5. Jager E, Smela E, Inganas O. Science 2000; 290: 1540.6. Peeters F, De Boeck J. Handb. Nanostruct. Mater. Nanotechnol.

2000; 3: 346.7. Yang M, Gracias D, Jacobs P, Somorjai G. Langmuir 1998; 14:

1458.8. Fang A, Ng H, Su X, Li S. Langmuir 2000; 16: 5221.9. Xueping J, Hammond P. Langmuir 2000; 16: 8501.

10. Seebohm G, Craighead H. Electron. Mater. Ser. 2000; 6: 97.11. Brandow S, Chen M, Shirey L, Dressick W. Langmuir 2001; 17:

228.12. Eck W, Stadler V, Geyer W, Zharnikov M, Golzhauser A,

Grunze M. Adv. Mater. 2000; 12: 805.13. Gorman C, Carroll R, He Y, Tian F, Fuierer R. Langmuir 2000; 16:

6312.14. Maoz R, Cohen D, Sidney R, Sagiv J. Adv. Mater. 1999; 11: 55.15. Abbott N, Rolison D, Whitesides G. Langmuir 1994; 10: 2672.16. Lopez G, Biebuyck L, Hans A, Harter R, Kumar A, White-

sides G. J. Am. Chem. Soc. 1993; 115: 10774.17. He H, Li Q, Zhou Z, Zhang H, Li S, Liu Z. Langmuir 2000; 16:

9683.18. Rie K. Surf. Coat. Technol. 1999; 112: 56.19. Pfender E. Plasma Chem. Plasma Process. 1999; 19: 1.20. Kuettel O, Groening P, Agostino R, Schlapbach L. J. Vac. Sci.

Technol. A 1995; 13: 2848.21. Vickerman J. Adv. Spectrosc. 1998; 26: 71.22. Moon D, Kim H, Kang H. J. Surf. Anal. 1999; 5: 144.23. Brown I, Anders A, Dickinson M, MacGill R, Monteiro O. Surf.

Coat. Technol. 1999; 112: 271.24. Lau W. Nucl. Instrum. Methods Phys. Res. B 1997; 131: 341.25. Cooks RG, Mabud M, Horning S, Jiang X, Paradisi C, Traldi P. J.

Am. Chem. Soc. 1989; 111: 859.26. Dongre A, Somogyi A, Wysocki V. J. Mass Spectrom. 1996; 31:

339.27. Gu C, Wysocki V. J. Am. Chem. Soc. 1997; 119: 12010.28. Pradeep T, Shen J, Evans C, Cooks RG. Anal. Chem. 1999; 71:

3311.29. Shen J, Evans C, Wade N, Cooks RG. J. Am. Chem. Soc. 1999; 121:

9762.30. Ada E, Kornienko O, Hanley L. J. Phys. Chem. B. 1998; 102: 3959.31. Cooks RG, Ast T, Mabud M. Int. J. Mass Spectrom. Ion Processes

1990; 100: 209.

32. Pradeep T, Feng B, Ast T, Patrick J, Cooks RG, Pachuta S. J. Am.Soc. Mass Spectrom. 1995; 6: 187.

33. Shen J, Grill V, Evans C, Cooks RG. J. Mass Spectrom. 1999; 34:354.

34. Wade N, Pradeep T, Shen J, Cooks RG. Rapid Commun. MassSpectrom. 1999; 13: 986.

35. Nowak P, McIntyre N, Hunter D, Bello I, Lau W. Surf. InterfaceAnal. 1995; 23: 873.

36. Ada E, Kornienko O, Hanley L. J. Phys. Chem. B 1998; 102: 3959.37. Park K, Kim B, Kang H. J. Chem. Phys. 1992; 97: 2742.38. Kasi S, Kang H, Rabalais J. J. Chem. Phys. 1988; 88: 5914.39. Evans C, Wade N, Pepi F, Cooks RG. Anal. Chem. in press.40. Grundmeier G, Matheisen E, Stratmann M. J. Adhes. Sci Technol.

1996; 10: 573.41. Vuillaume D, Boulas C, Collet J, Allan G, Delerue C. Phys. Rev.

B: Condens. Matter. 1998; 58: 16491.42. DeRosa L, Monetta T, Mitton D, Bellucci F. J. Electrochem. Soc.

1998; 145: 3830.43. Brandow S, Chen M, Aggarwal R, Dulcey C, Calvert J,

Dressick W. Langmuir 1999; 15: 5429.44. Fang J, Knobler C, Gingery F. J. Phys. Chem. B 1997; 101: 8692.45. Bereznitski Y, Jaroniec M. J. Chromatogr. A 1998; 828: 51.46. Takahagi T, Nagasawa Y, Ishitani A. Jpn. J. Appl. Phys., Part 1

1996; 35: 3542.47. Chidsey C, Liu G, Rowntree P, Scoles G. J. Chem. Phys. 1989; 91:

4421.48. Winger B, Laue H, Horning S, Julian R, Lammert S, Riederer D,

Cooks RG. Rev. Sci. Instrum. 1992; 63: 5613.49. Frisch MJ, et al. Gaussian 98, Revision A.7. Gaussian: Pittsburgh,

PA, 1998.50. Hehre W, Ditchfield R, Pople J. J. Phys. Chem. 1972; 56: 2257.51. Ehlert T. J. Phys. Chem. 1969; 73: 949.52. Pradeep T, Evans C, Shen J, Cooks RG. J. Phys. Chem. B 1999; 103:

5304.53. Miller S, Luo H, Pachuta S, Cooks RG. Science 1997; 275: 1447.54. Shen J, Yim Y, Feng B, Grill V, Evans C, Cooks RG. Int. J. Mass

Spectrom. 1999; 182: 423.55. Mabud M, Dekrey M, Cooks RG. Int. J. Mass Spectrom. Ion

Processes 1985; 67: 285.56. Laskin J, Denisov E, Futrell J. J. Phys. Chem. B 2001; 105: 1895.57. Kubista J, Dolejsek Z, Herman Z. Eur. Mass Spectrom. 1998; 4:

311.58. Schultz D, Wainhaus S, Hanley L, deSaint-Claire P, Hase W. J.

Chem. Phys. 1997; 106: 10337.59. deRidder J, Dijkstra G. Recl. Trav. Chim. Pays. Bas 1967; 86: 737.60. Lambert J, Zhao Y. Angew. Chem. 1997; 36: 400.61. Patai S. The Chemistry of Organic Silicon Compounds. Wiley:

Chichester, 1989; 187.62. Wade N, Evans C, Pepi F, Cooks RG. J. Phys. Chem. B 2000; 104:

11230.


silylation of an oh-terminated self-assembled monolayer surface through low-energy collisions of...

Documents