reactions of bbrn+ (n = 0–2) at fluorinated and hydrocarbon self-assembled monolayer surfaces:...

9
JOURNAL OF MASS SPECTROMETRY J. Mass Spectrom. 2001; 36: 717–725 Reactions of BBr n + (n = 0 – 2) at fluorinated and hydrocarbon self-assembled monolayer surfaces: observations of chemical selectivity in ion–surface scattering Nathan Wade, Jianwei Shen, Jere Koskinen and R. Graham Cooks Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, USA Received 2 January 2001; Accepted 29 March 2001 Ion–surface reactions involving BBr n + (n = 0–2) with a fluorinated self-assembled monolayer (F-SAM) surface were investigated using a multi-sector scattering mass spectrometer. Collisions of the B + ion yield BF 2 + at threshold energy with the simpler product ion BF + ž appearing at higher collision energies and remaining of lower abundance than BF 2 + at all energies examined. In addition, the reactively sputtered ion CF + accompanies the formation of BF 2 + at low collision energies. These results stand in contrast with previous data on the ion – surface reactions of atomic ions with the F-SAM surface in that the threshold and most abundant reaction products in those cases involved the abstraction of a single fluorine atom. Gas-phase enthalpy data are consistent with BF 2 + being the thermodynamically favored product. The fact that the abundance of BF 2 + is relatively low and relatively insensitive to changes in collision energy suggests that this reaction proceeds through an entropically demanding intermediate at the vacuum–surface interface, one which involves interaction of the B + ion simultaneously with two fluorine atoms. By contrast with the reaction of B + , the odd-electron species BBr + ž reacts with the F-SAM surface to yield an abundant single-fluorine abstraction product, BBrF + . Corresponding gas-phase ion – molecule experiments involving B + and BBr + ž with C 6 F 14 also yield the products BF + ž and BF 2 + , but only in extremely low abundances and with no preference for double fluorine abstraction. Ion–surface reactions were also investigated for BBr n + (n = 0–2) with a hydrocarbon self-assembled monolayer (H-SAM) surface. Reaction of the B + ion and dissociative reactions of BBr + ž result in the formation of BH 2 + , while the thermodynamically less favorable product BH + ž is not observed. Collisions of BBr 2 + with the H-SAM surface yield the dissociative ion–surface reaction products, BBrH + and BBrCH 3 + . Substitution of bromine atoms on the projectile by hydrogen or alkyl radicals suggests that Br atoms may be transferred to the surface in a Br-for-H or Br-for-CH 3 transfer reaction in an analogous fashion to known transhalogenation reactions at the F-SAM surface. The results for the H-SAM surface stand in contrast to those for the F-SAM surface in that enhanced neutralization of the primary ions gives secondary ion signals one to two orders of magnitude smaller than those obtained when using the F-SAM surface, consistent with the relative ionization energies of the two materials. Copyright 2001 John Wiley & Sons, Ltd. KEYWORDS: ion– surface scattering; surface-induced dissociation; self-assembled monolayers; surface reaction; ion activation INTRODUCTION Low-energy (1–100 eV) ion–surface collisions are of grow- ing interest for (i) the preparation of chemically modified surfaces, 1,2 (ii) the elucidation of polyatomic ion structure, 3,4 (iii) the chemical analysis of surfaces 5,6 and (iv) as a means of uncovering novel chemical processes at interfaces. 7 Ł Correspondence to: R. G. Cooks, Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, USA. E-mail: [email protected] Contract/grant sponsor: National Science Foundation; Contract/grant number: CHE-9732670. Contract/grant sponsor: Foundation of Neste Corporation. Ion–surface collisions involve several different processes which can be briefly outlined as follows. Surface-induced dis- sociation (SID) occurs when the translational energy of the projectile ion is converted into internal energy upon collision with the surface, thus leading to the fragmentation of the projectile ion. A large, yet controllable, internal energy depo- sition into the ion is a characteristic of SID, 8 and attempts are being made to exploit this in the fragmentation of large or sta- ble molecules, especially biological ions. 9,10 Charge exchange involves electron transfer from the surface to the incoming projectile ion and can result in the chemical sputtering of ion- ized products from the surface; 11 thus it serves as a method DOI: 10.1002/jms.177 Copyright 2001 John Wiley & Sons, Ltd.

Upload: nathan-wade

Post on 15-Jun-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

JOURNAL OF MASS SPECTROMETRYJ. Mass Spectrom. 2001; 36: 717–725

Reactions of BBrn+ (n = 0–2) at fluorinated and

hydrocarbon self-assembled monolayer surfaces:observations of chemical selectivity in ion–surfacescattering

Nathan Wade, Jianwei Shen, Jere Koskinen and R. Graham Cooks∗

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

Received 2 January 2001; Accepted 29 March 2001

Ion–surface reactions involving BBrn+ (n = 0–2) with a fluorinated self-assembled monolayer (F-SAM)

surface were investigated using a multi-sector scattering mass spectrometer. Collisions of the B+ ion yieldBF2

+ at threshold energy with the simpler product ion BF+ž appearing at higher collision energies andremaining of lower abundance than BF2

+ at all energies examined. In addition, the reactively sputteredion CF+ accompanies the formation of BF2

+ at low collision energies. These results stand in contrast withprevious data on the ion–surface reactions of atomic ions with the F-SAM surface in that the threshold andmost abundant reaction products in those cases involved the abstraction of a single fluorine atom. Gas-phaseenthalpy data are consistent with BF2

+ being the thermodynamically favored product. The fact that theabundance of BF2

+ is relatively low and relatively insensitive to changes in collision energy suggests thatthis reaction proceeds through an entropically demanding intermediate at the vacuum–surface interface,one which involves interaction of the B+ ion simultaneously with two fluorine atoms. By contrast withthe reaction of B+, the odd-electron species BBr+ž reacts with the F-SAM surface to yield an abundantsingle-fluorine abstraction product, BBrF+. Corresponding gas-phase ion–molecule experiments involvingB+ and BBr+ž with C6F14 also yield the products BF+ž and BF2

+, but only in extremely low abundancesand with no preference for double fluorine abstraction. Ion–surface reactions were also investigated forBBrn

+ (n = 0–2) with a hydrocarbon self-assembled monolayer (H-SAM) surface. Reaction of the B+ ionand dissociative reactions of BBr+ž result in the formation of BH2

+, while the thermodynamically lessfavorable product BH+ž is not observed. Collisions of BBr2

+ with the H-SAM surface yield the dissociativeion–surface reaction products, BBrH+ and BBrCH3

+. Substitution of bromine atoms on the projectile byhydrogen or alkyl radicals suggests that Br atoms may be transferred to the surface in a Br-for-H orBr-for-CH3 transfer reaction in an analogous fashion to known transhalogenation reactions at the F-SAMsurface. The results for the H-SAM surface stand in contrast to those for the F-SAM surface in that enhancedneutralization of the primary ions gives secondary ion signals one to two orders of magnitude smallerthan those obtained when using the F-SAM surface, consistent with the relative ionization energies of thetwo materials. Copyright 2001 John Wiley & Sons, Ltd.

KEYWORDS: ion–surface scattering; surface-induced dissociation; self-assembled monolayers; surface reaction; ionactivation

INTRODUCTION

Low-energy (1–100 eV) ion–surface collisions are of grow-ing interest for (i) the preparation of chemically modifiedsurfaces,1,2 (ii) the elucidation of polyatomic ion structure,3,4

(iii) the chemical analysis of surfaces5,6 and (iv) as a meansof uncovering novel chemical processes at interfaces.7

ŁCorrespondence to: R. G. 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.Contract/grant sponsor: Foundation of Neste Corporation.

Ion–surface collisions involve several different processeswhich can be briefly outlined as follows. Surface-induced dis-sociation (SID) occurs when the translational energy of theprojectile ion is converted into internal energy upon collisionwith the surface, thus leading to the fragmentation of theprojectile ion. A large, yet controllable, internal energy depo-sition into the ion is a characteristic of SID,8 and attempts arebeing made to exploit this in the fragmentation of large or sta-ble molecules, especially biological ions.9,10 Charge exchangeinvolves electron transfer from the surface to the incomingprojectile ion and can result in the chemical sputtering of ion-ized products from the surface;11 thus it serves as a method

DOI: 10.1002/jms.177 Copyright 2001 John Wiley & Sons, Ltd.

718 N. Wade et al.

for the analysis of the chemical components of a surface.12

Ion–surface reactions are events in which an atom or groupsof atoms are transferred between the projectile and surfacein the course of low-energy collisions. Transhalogenationreactions,13 for example, have been shown to transformchemically fluorocarbon surfaces into terminal CF2X units,where X may represent a halogen atom or pseudohalogengroup derived from the projectile ion. More recently, sur-face modification has been demonstrated with aromatic ions,such as the [M � H]C ion of chlorobenzene,1 which cova-lently bind to a carboxylic acid terminated self-assembledmonolayer surface through an ion–surface decarboxylationreaction which is reminiscent of the condensed-phase Kolbereaction. Soft landing is sorption of an intact projectile ion ona target surface.14 This physical modification may allow ionstorage at a surface.

One objective in low-energy ion–surface reaction studieshas been to widen the scope of this area of chemistrythrough the discovery of new reaction types and to seekan understanding of their reaction mechanisms, dynamicsand energetics. One of the best-studied ion–surface reactionsis the protonation or alkylation of incident polyatomicprojectiles, especially in the case of such radical cationsas the molecular ions of pyrazine and benzene,15,16 uponcollision at a surface bearing hydrocarbons. Ion–surfacereactions have also been extended to F-SAM target surfacesand the abstraction of fluorine atoms has proven to bean ubiquitous reaction.17 Multiple abstractions in a singlecollision event have been observed,18 for example, at 30 eVlaboratory collision energy tungsten ions, WCž, abstractup to five fluorine atoms to scatter reactively as WF5

C.The high ionization energy of the fluorinated moleculeswhich constitute these surfaces minimizes neutralizationof incoming projectile ions, allowing other ion–surfaceprocesses to compete with charge exchange. The F-SAMsurface is also fairly efficient as a target for translationalto internal energy conversion of projectiles,19 and remainsrelatively clean of hydrocarbon contaminants for longperiods under typical vacuum conditions (<10�8 Torr (1Torr D 133.3 Pa)).

On the question of ion–surface reaction mechanisms, anumber of different pathways have been proposed to accountfor observed Ion–surface reactions in various systems. Inthe charge-transfer mechanism,16 the incoming projectile ionundergoes charge exchange with the surface functionalgroup producing a surface-bound radical cation. Fragmentions arising from this species interact with the neutralizedprojectile in a subsequent ion–molecule reaction at theinterface. This reaction pathway may be responsible forhydrogen atom and alkyl group abstraction by aromatic andheteroaromatic radical cations. An alternative mechanismis direct abstraction of surface atoms or groups by theprojectile ion, one example of which is a prompt Eley–Ridealprocess.20 Another ion–surface reaction mechanism thatdoes not involve charge exchange between the surfaceand projectile ion has been established. Using low-energyCsC ions to bombard various Si(111) surfaces,21 Yang et al.showed that ions such as CsSiC and CsH2OC were generatedvia a two-step process where collision-induced desorption

of the neutrals from the surface was followed by a gas-phase ion–molecule reaction. Such ion–neutral electrostaticrecombination reactions are facilitated by extensive energy lossby the projectile to the surface and the efficient secondaryneutral emission which occurs even at low impact energies.Concerted mechanisms have been proposed for ion–surfacereactions involving fluorine atom or fluorocarbon groupaddition to the projectile ion upon collision with an F-SAMsurface.17 For example, experimental observations22 andenthalpy calculations suggest that two reaction pathwayscontribute to XeFC formation from collisions involving XeCž.At low collision energies oxidative insertion of XeCž into aC—F bond is thought to occur, whereas at higher collisionenergies reaction appears to proceed via the formation of afluoronium ion intermediate.

We report here on reactions of boron cation and bro-moboron cations with an F-SAM surface and an H-SAMsurface, and also analogous reactions with perfluoro-n-hexane (C6F14) in parallel gas-phase ion–molecule exper-iments. The selectivity demonstrated by boron cations inthese ion–surface reactions is unique in comparison withpreviously studied ions, illustrating the dependence of theseprocesses on the chemical nature of the projectile ion. Theenergetics of these ion–surface reactions are analyzed andpossible mechanisms are discussed. Comparisons are alsomade between F and H atom abstraction, considering effectsdue to the nature of these two surfaces.

Boron cation and fluorocarbon chemistry is also ofintrinsic interest. Fluorocarbons are distinguished by theirchemical inertness23 and activation of the strong C—F bond(¾130 kcal mol�1 (1 kcal D 4.184 kJ)) has been an importantchemical objective,24 especially selective activation in multi-ply fluorinated compounds. In this study, formation of thestrong B—F bond (diatomic bond energy 138 kcal mol�1)along with the unique binding properties of boron assist inproviding the selectivity observed in these reactions. Boronchemistry is characterized by the electron deficiency of itscompounds. For example, BH3, the simplest of boranes, issuch a strong Lewis acid that it will coordinate with virtuallyany source of electron density, including itself.25 Boron isalso well known for the diversity of structural types in whichit can participate, a case in point being the metalloboranes.26

Borinium ions (R2BrC) have been proposed as transient inter-mediates in condensed-phase reactions,27 but owing to thehigh electrophilicity of these subvalent boron species theyare difficult to prepare in the condensed phase. The reac-tivity and binding affinities of dicoordinated boron cationswith organic substrates have recently been studied in the gasphase,28,29 and at a more fundamental level, potential energysurfaces for reactions of BC and BH2

C with H2 have beenstudied.30

EXPERIMENTAL

Ion–surface reactionsExperiments were performed using a custom-built hybridmass spectrometer of BEEQ (B D magnetic sector, E Delectrostatic analyzer, Q D quadrupole mass analyzer)configuration.31 Projectile ions (BC, BBrCž, BBr2

C) were gen-erated by 70 eV electron impact (EI) upon boron tribromide

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

Ion–surface collisions involving BBrnC (n D 0–2) 719

(Aldrich, Milwaukee, WI, USA). Ions were accelerated to2 keV translational energy, then mass selected and energyfocused using the B and E analyzers. Prior to collision, theion beam was decelerated to the desired translational energy,where the nominal laboratory collision energy was calculatedas the difference in potential between the ion source and tar-get. The surface was held in a UHV scattering chambermaintained at a nominal base pressure of 2 ð 10�9 Torr (typ-ical operating pressures were ¾5 ð 10�9 Torr). The samplewas rotated so that the primary ion beam was incident at55° to the normal, while scattered ions were collected over awide range of angles centered about the specular angle andextracted into the post-collision E and Q analyzer system formass analysis using the quadrupole mass analyzer.

Hydrocarbon and fluorocarbon thiolates were bound toa gold film prepared by thermal evaporation of 100 A ofchromium then 2000 A of gold on to silicon h100i wafers(International Wafer Service, Portola Valley, CA, USA). Themolecular assemblies were constructed by immersing thesubstrates in dilute (1 mM) ethanol solutions of dodecanethiolor the disulfide [CF3(CF2)7(CH2)2S]2 for a period of atleast 1 week at room temperature. The exposure period formonolayer formation was chosen based upon optimizingthe quality of experimental scattered ion mass spectra, usingXeCž chemical sputtering data as the test system. Detailedinformation concerning the preparation and properties ofthe surfaces has been provided elsewhere.32 The surfaceswere rinsed in ethanol and dried under argon before beingintroduced into the high-vacuum scattering chamber.

Ion–molecule reactionsDiscussions of the instrumentation and methodologies usedin triple quadrupole mass spectrometry can be found in theliterature.33 Ion–molecule reactions were carried out in aFinnigan TSQ 700 triple-quadrupole instrument (FinniganMAT, San Jose, CA, USA). The projectile ions, BC and BBrCž,were generated in a 70 eV EI ion source from BBr3, massselected using the first quadrupole (Q1), then allowed toreact with the neutral reagent molecules of C6F14 in ther.f.-only quadrupole (q2) used as a collision cell at nominallaboratory frame of reference collision energies of 0–20 eV.The neutral molecules were introduced as the vapor, andprior to allowing them to leak into the collision cell atnominal pressures (0.2–1.5 mTorr, corresponding to 15–85%attenuation of the primary ion beam) residual air wasremoved by repeating several freeze–thaw cycles whileallowing the sample vial to be evacuated. The productsformed by the ion–molecule reactions that took place in q2were mass analyzed by scanning the third quadrupole (Q3)over an appropriate m/z range. C6F14 was obtained fromLancaster (Windham, NH, USA).

RESULTS AND DISCUSSION

Ion–surface reactions with an F-SAM surfaceThe results from experiments on the ion–surface reactivitiesof BC, BBrCž and BBr2

C (Figs 1, 3 and 4) clearly show thevarious trends in fluorine abstraction reactions that occuras a function of collision energy. When BC was used as the

5 15 25 35 45 55 65 75 85

B+

B+

B+

B+

CF+

CF+

CF+

BF+•

BF+•

BF2+

BF2+

BF2+

CF3+

CF3+

CF3+

(a) 15 eV

(c) 35 eV

(d) 55 eV

(e) 75 eV

m/z (Thomson)

Ion

Abu

ndan

ce

(b) 25 eVB+

CF+BF2

+

Figure 1. Scattered ion mass spectra recorded upon collisionof BC with the F-SAM surface at energies of (a) 15, (b) 25,(c) 35, (d) 55 and (e) 75 eV.

projectile (Fig. 1), the BF2C ion was observed along with the

chemically sputtered product CFC at all collision energieshigher than 25 eV. It is interesting that the BFCž ion is absentat these low collision energies. This is contrary to what hasbeen observed in the reactions of most other non-metallicand metallic atomic ions,17,18 which undergo single fluorineabstraction preferentially as the lowest energy process, whilemultiple abstractions begin at higher collision energies. Inthe present case, the single fluorine containing product BFCž

appeared at a much higher energy, ¾50 eV, and formation ofBF2

C predominated over the formation of BFCž at all collisionenergies studied. It is also noteworthy that some BFCž wasformed as a dissociation product of the scattered BF2

C, andthis may account for the observed increase in the BFCž/BF2

C

ratio as a function of collision energy. The ion BF3Cž was not

detected in any of these experiments.Addressing the issue of the BF2

C production mechanism,the threshold for chemical sputtering from the F-SAMsurface using typical inorganic and metal-based cations(e.g. XeCž or W(CO)2

C) occurs between 30 and 40 eV. Here,however, CFC formation is observed at energies below 25 eV,while the usual chemical sputtering product, CF3

C, wasnot seen until higher collision energies. Formation of CFC

at such low collision energies is clearly not the result ofdissociative chemical sputtering, although similar resultshave previously been observed in isolated cases with theions SiCž, SiClC and AlC.17 This phenomenon was referred

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

720 N. Wade et al.

to as reactive sputtering since mechanistically a differentform of sputtering is involved, perhaps associated with anion–surface reaction rather than the normal charge exchangeprocess. In these cases, single fluorine abstraction productsaccompanied the reactively sputtered CFC ion, in contrastto what is observed with the boron cation. Consideringall the above, it is a reasonable proposal that fluorineabstraction by BC occurs at the terminal CF3 and the twofluorine atoms originate from the same terminal carbon. Itis known that the highly electron-deficient borane moleculetends to form the dimer B2H6 through a four-memberedbridged structure.34 Thus, it is proposed that at low energy,BC ions form preferentially a similar four-membered ringstructure, which serves as a reactive intermediate during theion–surface collision event.

Figures 2(a) and (b), illustrate reaction schemes whichdescribe two possible pathways leading to preferentialproduction of both BF2

C and CFC through two separatesurface intermediates, while Fig. 2(c) and (d) show analo-gous processes already reported in the literature. Figure 2(a)demonstrates the formation of the four-membered ring inter-mediate. This intermediate is analogous to the symmetricalfluoronium complex (F12C–FC –13CF) (Fig. 2(c)) which wassuggested to occur in a previous study involving the reactionof 13CCž with an F-SAM surface.35 This proposed fluoroniumion intermediate was useful for explaining the equal abun-dances of 13CFC and 12CFC at high collision energies in thescattered ion mass spectrum. An alternative route to the for-mation of BF2

C which must also be considered is oxidativeinsertion of the BC ion into two C—F bonds (Fig. 2(b)). A

+ BF2+

CF+ or C2F3+

+

++ BF2

B+

CRF

CFF F

F

CRF

CFF F

F

B+

C

C F

FR

F

F

FB

B+C

F

CF F

R

FF

CRF

CF

F

CRF

CF

F

:

CRF

CFF F

F

13C+

F12CFF13C

+

13CF+ + F12CF

F13CF + 12CF+

CRF

CF F

F

FC13

+

13

CRF

CF

F

FF C +

Xe+•

a

b

C

C FF

FF

F

RC

C FFXe

F

FF R

+

C

.C FF

FF R

c

d+ XeF+

.

Figure 2. Possible mechanisms for BC reactions with the terminal group of an F-SAM surface through (a) a cyclic four-memberedintermediate and (b) oxidative insertion into dual C—F bonds. (c) Reaction of 13CC with the terminal F-SAM surface through afluoronium ion intermediate and (d) reaction of XeCž with the terminal F-SAM surface through an oxidative insertion mechanism.

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

Ion–surface collisions involving BBrnC (n D 0–2) 721

similar reaction intermediate, involving oxidative insertionof XeCž into a singular C—F bond, has been suggested to beresponsible for fluorine abstraction from the F-SAM surface(Fig. 2(d)).22

Tabulated gas-phase thermochemical values can be usedto compare the relative energy demands of the variousion–surface reactions. This commonly used approach indi-cates that formation of BF2

C is more exothermic comparedwith single fluorine abstraction. This is shown by the follow-ing reaction energetics:

BC C C3F8 ���! BF2C C C3F6 H D �97 kcal mol�1 �1�

BC C C3F8 ���! BFCž C C3F7ž H D 16 kcal mol�1 �2�

where the product C3F6 is the olefin and C3Fž7 is the primary

radical.Corresponding values for the previously studied atomic

ions, SiCž and AlC, for the same reactions, are as follows:

SiCž C C3F8 ���! SiF2Cž C C3F6

H D �36 kcal mol�1 �3�

SiCž C C3F8 ���! SiFC C C3F7ž

H D �13 kcal mol�1 �4�

AlC C C3F8 ���! AlF2C C C3F6

H D �25 kcal mol�1 �5�

AlC C C3F8 ���! AlFCž C C3F7ž

H D 57 kcal mol�1 �6�

In the case involving SiCž, the dominant scattered ionproduct was SiFC, while no SiF2

C was observed. For collisionsinvolving AlC, AlF2

C is observed but not at collision energiesbelow the AlFCž onset and not in greater abundance.We assume that the more favorable thermochemistry forabstraction of two fluorines in the case of boron, relative tothe other two ions, leads to its unique behavior.

The efficiency of the BC reaction is also noteworthy. Thecollision energy dependence of the BC reaction indicates thatthe efficiency of the fluorine abstraction reaction is low below25 eV. The relative abundance of BF2

C increases slowly above35 eV, but the BF2

C abundance does not increase above 30%of the base peak (BC ion), even at 75 eV. This is in contrastto previous studies,17 e.g. SiCž reactions with the F-SAMsurface at energies as low as 20 eV to yield SiFC at 150% theabundance of the projectile ion. The low abundance of theBF2

C product and its insensitivity to collision energy changeswould not be expected for such an exothermic reaction. Apossible reason is that the reaction has a high entropy barrier.There is no information available on the potential energysurface for the BC –CF3 system, but theoretical calculationsindicate that a barrier in the exothermic BC –H2 reaction toform HBHC originates from the insertion step that requiresthe interplay of the s and p electrons of BC in the excitedstates.30 The ground state of the BC ion has a 1S(2s2) electronicconfiguration, that of the first excited state being 3P(2s2p).The energy gap between these two states is 4.6 eV. Henceit is possible that the formation of BF2

C involves an energy

barrier and occurs via an electronic excited state of the BC

ion. However, another study showed that only 5% of theBC formed during 70 eV electron ionization of BBr3 existsin an excited state.36 Thus, the favored explanation for theinefficiency of the BC reaction is an entropically demandingintermediate.

The ion–surface reactions of BBrCž (Fig. 3) clearly showsignificantly higher reactivity towards the F-SAM surfacethan the bare BC ion. It is reasonable to suggest that the highreactivity of BBrCž in C—F bond cleavage is probably dueto the existence of the unpaired electron. Even at only 15 eVcollision energy, the C—F bond cleavage product BBrFC isthe most abundant peak in the spectrum. Unlike the casewith BC, the fluorine pick-up reaction with BBrCž increasesrapidly with collision energy. Moreover, on increasing thecollision energy, a number of other product ions are seen,including BrCF2

C, BF2C and BFCž. The products, BF2

C andBFCž, are representative of a transhalogenation reaction,involving abstraction of fluorine atoms by the projectile andtransfer of the Br atom to the surface. Evidence of the brominetransfer is observed in the product BrCF2

C, which is a resultof bonding to the surface through BrC with associated C—Ccleavage. The ion BF2

C begins to be observed at or below25 eV collision energy and again is more abundant than BFCž

in the experimental range of collision energies covered.For BBr2

C (Fig. 4), Br-for-F transhalogenation appears tobe the major reactive scattering process and, not surprisingly,no BBr2FCž is formed. The peak intensity of the majorproduct BBrFC rises rapidly with the collision energy,as does the abundance of the SID product BBrCž. It is

5 25 45 65 85 105 125 145 165

B+CF+

BBr+•

BBrF+

BBrF+

BBrF+

C2F5+

C2F4+•

C3F3+

CF3+

CF3+

CF3+

CF2+•

BF2+

BF2+

BF2+

BF+• BrCF2+

BrCF2+

BrCF2+

BBrF+

B+ CF+

CF+

BBr+•

BBr+•

BBr+•

m/z (Thomson)

Ion

Abu

ndan

ce

(a) 15 eV

(b) 25 eV

(c) 45 eV

(d) 65 eV

Figure 3. Scattered ion mass spectra recorded upon collisionof BBrCž with the F-SAM surface at energies of (a) 15, (b) 25,(c) 45 and (d) 65 eV.

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

722 N. Wade et al.

B79Br81Br+

B79Br81Br+

B79Br81Br+

B79Br81Br+

BBrF+

BBrF+

BBrF+

{

BBr+•

BBr+•

BBr+•

{ BrCF2+

{

CF3+

CF3+

CF3+

CF+

CF+

(d) 65 eV

(c) 45 eV

(b) 35 eV

(a) 15 eV

{{

{{

C2F4+•

5 25 45 65 85 105 125 145 165 185 205

m/z (Thomson)

Ion

Abu

ndan

ce

BF2+

BF2+

Figure 4. Scattered ion mass spectra recorded upon collisionof B79Br81BrC with the F-SAM surface at energies of (a) 15,(b) 35, (c) 45 and (d) 65 eV.

possible that dissociation at the surface generates reactivespecies BBrCž which further reacts to form BBrFC, or SIDand ion–surface reactions could occur at the surface ina concerted fashion. Minor products, such as BrCF2

C andBF2

C, were also observed above 45 eV collision energies. It

is obvious that the fluorine abstraction reactivity of bothBBrCž and BBr2

C is even more pronounced than that of BC.Previous work has shown the increased reactivity associatedwith an unpaired electron using a distonic ion reagent.7 Itmust also be taken into consideration that the high reactivityof BBrCž to induce C—F bond cleavage could also be affectedby excited-state formation of reactant ions in the 70 eV EIion source. It is known that 11B79Br neutral molecule hasseveral relatively low-lying excited states. Relative to theground state X(1C) these are approximately 2.32, 2.34 and4.2 eV higher in energy, the highest being a readily accessibleexcited state A(1), and the transition corresponding toapproximately 97.0 kcal mol�1.37

Ion–molecule reactionsWhen BC ion was reacted with C6F14 in the gas phase,both BFCž and BF2

C were observed at low collision energies(Fig. 5). In the case of BBrCž (Fig. 6), little or no BF2

C

is observed above ¾8 eV nominal laboratory frame ofreference collision energies, whereas BFC is observed inmodest abundance up to ¾20 eV. For BBr2

C, no processesother than direct collision-induced dissociation (CID) wereobserved under the conditions used. The fluorine abstractionproducts are much less abundant in these gas-phasecollisions than in the ion–surface collisions. In a previousstudy involving fluorocarbon molecules, it was shown thation–molecule reactions could be used as a predictor forion–surface reactions.38 Not only were similar fluorineabstraction products observed, but the results suggestedthat similar complex intermediates were formed in the gasphase to those at the ion–surface interface, each resultingin halogen exchange or alkene formation as the chemistrydictated, the major difference being the larger abundanceof reaction products produced from ion–surface reactions.

Figure 5. Mass spectrum of product ions formed upon collision of mass-selected 11BC ion at 0.3 eV nominal collision energy(laboratory frame of reference) with C6F14 in q2 at 0.4 mTorr nominal pressure.

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

Ion–surface collisions involving BBrnC (n D 0–2) 723

Figure 6. Mass spectrum of product ions formed upon collision of mass-selected 11B79BrC ion at 2 eV nominal collision energy(laboratory frame of reference) with C6F14 in q2 at 0.4 mTorr nominal pressure.

These differences in C—F activation can be attributed to thehigher T–V conversion at the surface, which is a result ofthe highly efficient energy dissipation in the solid matrixas compared with the isolated molecule environment. Thephenomenon observed in the ion–surface reactions, namelyBF2

C formation at the reaction onset instead of the BFCž

product, is not reproduced in the gas-phase collisions. Otherwork has demonstrated that differences between reaction atthe surface and reaction in the gas phase can also be attributedto the organization of the molecules in the monolayer.6 In thecase of a gaseous molecule, no preferential orientation existsand the ion has a finite probability of undergoing reactionwith resulting cleavage of any bond, whereas reaction at thesurface monolayer is limited to the terminal portion of themolecule and is dependent on the geometry of the moleculeat the surface.6

Ion–surface reactions with H-SAM surfaceFor comparison with the reactions at the F-SAM surface,BBrn

C (n D 0–2) ions were reacted with a hydrocarbonsurface. The reaction of BBr2

C with the C12 H-SAM surfaceyields both hydrogen and alkyl group abstraction products.Figure 7 shows the product ion spectra due to BBr2

C reactivecollisions upon the H-SAM surface at collision energies of 25,45, 65 and 85 eV. It should be noted that ion neutralizationoccurred more readily at the H-SAM than the F-SAM surface.The secondary ion signals were found to be one to twoorders of magnitude lower than those obtained when usingthe F-SAM surface. Note the low abundance of BBr2

C at65 eV [Fig. 7(c)], in comparison with that with the F-SAMsurface [Fig. 4(d)]. The presence of BBrHC and BBrCH3

C is

0 10080604020 120 140 160 180 200

Ion

Abu

ndan

ce

{ {

{ {

x 5

{ {

{

m/z (Thomson)

BH2+

BH2+

B79BrH+

B81BrH+

B79BrCH3+

B81BrCH3+

(a) 25 eVBBr2+ (171)

BBr2+ (171)

B79BrH+

B81BrH+

B79BrH+

B79BrH+

B81BrH+

B79BrCH3+

B81BrCH3+

B79BrCH3+

B81BrCH3+

B79BrCH3+

(b) 45 eV

(d) 85 eV

BBr2+ (171)

(c) 65 eV

B79BrC2H5+

B81BrC2H5+

C2Hx+

C3Hy+

C4Hz+

BBr2+ (169)

Figure 7. Scattered ion mass spectra recorded upon collisionof BBr2Cž with the H-SAM surface at energies of (a) 25, (b) 45,(c) 65 and (d) 85 eV.

clear evidence for the occurrence of ion–surface reactionsinvolving H and alkyl radical transfer. The presence of theseproduct ions was confirmed using the bromine isotopic data

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

724 N. Wade et al.

shown in Fig. 7. In contrast to observations made at theF-SAM surface, no SID products, e.g. BBrCž or BC, wereformed, a result that is consistent with the softer natureof this surface.39 At 85 eV collision energy, BBrC2HC

5 andBH2

C were also observed. Reactions of BC and BBrCž withthe H-SAM surface [Fig. 8(a) and (b)] gave only 11BH2

C

(m/z 13) (10BH2C (m/z 12) when 10BC was used). No BBrHC

was observed when BBrCž was collided with an H-SAMsurface. In both cases, chemical sputtering occurs extensivelyto produce very abundant sputtering peaks, C2Hx

C (x D 3 or5) and C3Hy

C (y D 3, 5 or 7), in addition to the ion–surfacereaction products. The reactions of BBrn

C with the H-SAMsurface are one of the few examples of hydrogen or/alkylgroup abstraction by non-organic ions. These results suggestthat the Br-for-H or Br-for-CH3 transfer reaction could occur,in a similar fashion to the halogen-for-F transfer reaction inthe F-SAM surface mentioned above. However, the observedabundance of charge exchange products, e.g. CxHy

C, suggeststhat the exact reaction mechanism may be different from thatwith the F-SAM surface, particularly in terms of electrontransfer. As in many cases involving hydrogen/hydrocarbongroup abstraction, it is possible that charge exchange betweenthe projectile ions and the hydrocarbon adsorbates may beinvolved in the formation of BH2

C and BCxHyC.

When the projectile ion BC was collided with the H-SAMsurface only the pick-up of two H atoms was observed;the abstraction of a single H atom to form the respectiveodd electron species BHCž was not observed at all. Reactionenergetics, when carefully considered, support the evidence

5 10 15 20 25 30 35

m/z (Thomson)

BBr+ (m/z 90) / H-SAM Surface@ 80 eV Collision Energy

B+

B+

BH2+

BH2+

CH3+

CH3+

C2H3+

C2H3+

C2H5+

C2H5+

Ion

Abu

ndan

ce

B+ (m/z 11) / H-SAM Surface@ 70 eV Collision Energy

Figure 8. Scattered ion mass spectra recorded upon collisionof (a) BC with the H-SAM surface an energy of 70 eV and(b) BBrCž with the H-SAM surface at 80 eV collision energy.

that dual H-atom pick-up would be favored over single-atomabstraction. The pick-up of a single H by BC is estimated tobe endothermic whereas that of two hydrogen atoms isexothermic, as judged using values for the following modelsystem:

BC C C3H8 ���! BHCž C C3Hž7 Hrxn D 54 kcal/mol�1

�7�

BC C C3H8 ���! BH2C C C3H6 Hrxn D �23 kcal/mol�1

�8�The trend is similar to the corresponding reactions (1) and (2)at the F-SAM surface. In addition to the absence of BHCž, itis an important observation that no odd-electron ions werescattered as a result of collisions of BC, BBrCž or BBr2

C withthe H-SAM surface. This was not the case for experimentsinvolving the F-SAM surface. Three important propertiesseem to cause the contrasts between the H-SAM surfaceand the F-SAM surface in terms of their effect as targetsin ion–surface collision processes. First, the H-SAM surfacehas a lower ionization energy (IE) than the F-SAM surface(IE ¾ 13.38 eV for C3F8 and 10.94 eV for C3H8), second, theC—H bond energy (98 kcal mol�1 in C2H6) is much lowerthan the C—F bond energy (130 kcal mol�1 in C2F6), andthird, the hydrocarbon chains are expected to be more flexiblethan the fluorocarbon chains. These effects result in a smalleramount of energy transfer to a projectile ion upon collisionwith a surface bearing the hydrocarbon monolayer and tomore extensive neutralization. We have already observed theincreased reactivity of odd-electron ions over even-electronions, and it would seem less likely for a less stable odd-electron ion to survive the scattering process from an H-SAMsurface since this surface has a greater ability to lose electronsthrough a charge-exchange process or to have a C—H bondcleave through some ion–surface reaction.

CONCLUSION

Bare BC ion reacts with the F-SAM surface at low energy togive rise to a unique product ion distribution. The appearanceof the product ion BF2

C at energies below that needed togenerate BFCž implies that the multiple abstraction takesplace in a single step, not in a sequential fashion. A four-membered intermediate is suggested, involving BC and theterminal CF3 group and leading to the formation of BF2

C andCFC. The preference for double fluorine atom abstraction bythe boron cation contrasts with previously studied reactionsof other atomic ions colliding with the F-SAM surface, whichprefer single fluorine abstraction at the reaction threshold.This reaction is a demonstration of the selective chemicalnature by which fluorine abstraction and other ion–surfacereactions may occur. Note, also, that although the chemicallysimilar AlC ion did not exhibit double F atom abstraction atreaction threshold, AlF2

C was a competitive product ion withAlFCž at higher energies, and the thermodynamic preferencesof the AlC ion are similar to those of BC.

Although gas-phase experiments have been able toreproduce many of the same reaction products as observedin ion–surface reactions, it has been shown that energytransfer for bond activation is much less efficient. By fixing

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725

Ion–surface collisions involving BBrnC (n D 0–2) 725

the molecular reagent on the surface, a reaction which is notobserved in the gas-phase, BBr2

C formation, can be madeprominent. The high reactivity of BBrCž emphasizes the roleof the unpaired free electron of the projectile in C—F bondcleavage. BBrC

n ions also undergo ion–surface reactions withthe H-SAM surface, yielding hydrogen and alkyl groupattachment products. These halogen-for-hydrogen reactionsat the H-SAM surface suggest that the transhalogenationreaction established at the F-SAM surface could be extendedto the H-SAM surface to achieve halogen-for-hydrogenmodification.

AcknowledgementsThis work was supported by the National Science Foundation,CHE-9732670, and by the Foundation of Neste Corporation (J.K.).

REFERENCES1. Shen J, Evans C, Wade N, Cooks RG. J. Am. Chem. Soc. 1999; 121:

9762.2. Ada E, Kornienko O, Hanley L. J. Phys. Chem. B 1998; 102: 3959.3. Cooks RG, Mabud M, Horning S, Jiang X, Paradisi C, Traldi P. J.

Am. Chem. Soc. 1989; 111: 859.4. Dongre A, Somogyi A, Wysocki V. J. Mass Spectrom. 1996; 31:

339.5. Gu C, Wysocki V. J. Am. Chem. Soc. 1997; 119: 12 010.6. Pradeep T, Shen J, Evans C, Cooks RG. Anal. Chem. 1999; 71:

3311.7. Shen J, Grill V, Cooks RG. J. Am. Chem. Soc. 1998; 120: 4254.8. Mabud M, Dekrey M, Cooks RG. Int. J. Mass Spectrom. Ion

Processes 1985; 67: 285.9. Gu C, Somogyi A, Wysocki V, Medzihradszky K. Anal. Chim.

Acta 1999; 397: 247.10. Fuhrer K, Schultz A, Stone E, Gillig K, Ruotolo B, Park Z, Rus-

sell D. In Proceedings of the 48th ASMS Conference on MassSpectrometry and Allied Topics, Long Beach: CA, 2000.

11. Vincenti M, Cooks RG. Org. Mass Spectrom. 1988; 23: 317.12. Pradeep T, Evans C, Shen J, Cooks RG. J. Phys. Chem. B 1999; 103:

5304.13. Wade N, Pradeep T, Shen J, Cooks RG. Rapid Commun. Mass

Spectrom. 1999; 13: 986.

14. Miller S, Luo H, Pachuta S, Cooks RG. Science 1997; 275: 1447.15. Ast T, Mabud M, Cooks RG. Int. J. Mass Spectrom. Ion Processes

1988; 82: 131.16. Phelan L, Hayward M, Flynn J, Bernasek S. J. Phys. Chem. B 1998;

102: 5667.17. Pradeep T, Ast T, Cooks RG, Feng B. J. Phys. Chem. 1994; 98:

9301.18. Pradeep T, Riederer D, Hoke S, Ast T, Cooks RG, Linford M. J.

Am. Chem. Soc. 1994; 116: 8658.19. Morris M, Riederer D, Winger B, Cooks RG, Ast T, Chidsey C.

Int. J. Mass Spectrom. Ion Processes 1992; 122: 181.20. Williams E, Jones G, Fang L, Zare R, Garrison B, Brenner D. J.

Am. Chem. Soc. 1992; 114: 3207.21. Yang M, Hwang C, Kang H. J. Chem. Phys. 1997; 107: 2611.22. Feng B, Shen J, Grill V, Evans C, Cooks RG. J. Am. Chem. Soc.

1998; 120: 8189.23. Banks E, Barlow M. Fluorocarbon and Related Chemistry. Thanet

Press: Margate; 1971.24. Banks R, Sharp D, Tatlow J. (eds). Fluorine: the First Hundred

Years. Elsevier Sequoia: New York; 1986.25. Housecraft C. Boranes and Metalloboranes: Structure, Bonding, and

Reactivity. John Wiley & Sons: New York; 1990.26. Kennedy J. Prog. Inorg. Chem. 1986; 34: 211.27. Noth H, Rasthofer B, Narula C, Konstantinov A. Pure Appl.

Chem. 1983; 55: 1453.28. Ranatunga T, Kenttamaa H. J. Am. Chem. Soc. 1992; 114: 8600.29. Ma S, Wong P, Cooks RG. J. Mass Spectrom. 1997; 32: 159.30. Nichols J, Gutowski M, Cole S, Simmons J. J. Phys. Chem. 1992;

96: 644.31. Winger B, Laue H, Horning S, Julian R, Lammert S, Riederer D,

Cooks RG. Rev. Sci. Instrum. 1992; 63: 5613.32. Chidsey C, Liu G, Rowntree P, Scoles G. Chem. Phys. 1989; 91:

4421.33. Yost R, Boyd R. Methods Enzymol. 1990; 193: 154.34. Liebman J, Greenberg A, Williams R. Advances in Boron and the

Boranes. VCH: New York; 1988.35. Ast T, Pradeep T, Feng B, Cooks RG. J. Mass Spectrom. 1996; 31:

791.36. Lin K, Cotter R, Koski W. J. Chem. Phys. 1974; 60: 3412.37. Destoky C, Dubois I, Bredohl H. J. Mol. Spectrosc. 1989; 136: 216.38. Patrick J, Pradeep T, Luo H, Ma S, Cooks RG. J. Am. Soc. Mass

Spectrom. 1998; 9: 1158.39. Pradeep T, Miller S, Cooks RG. J. Am. Soc. Mass Spectrom. 1993;

4: 769.

Copyright 2001 John Wiley & Sons, Ltd. J. Mass Spectrom. 2001; 36: 717–725