Journal of Colloid and Interface Science234,410–417 (2001)doi:10.1006/jcis.2000.7328, available online at http://www.idealibrary.com on

Comparative Behavior of Aromatic Disulfide and Diselenide Monolayerson Polycrystalline Gold Films Using Cyclic Voltammetry, STM,

and Quartz Crystal Microbalance

Mohammed Aslam,∗ Krisanu Bandyopadhyay,∗ K. Vijayamohanan,∗,1 and V. Lakshminarayanan†∗Physical/Materials Chemistry Division, National Chemical Laboratory, Pune 411008, India; and†Raman Research Institute, Bangalore 560080, India

Received June 20, 2000; accepted November 6, 2000

A comparative investigation of the self-assembled monolayers ofdiphenyl disulfide (DDS), diphenyl diselenide (DDSe), and naph-thalene disulfide (NDS) on polycrystalline gold films using STM,QCM, and electrochemical techniques is presented. The geomet-ric constraint imposed by the rigid naphthalene ring for NDS in-hibits the cleavage of the S–S bond, thus adversely affecting themonolayer organization and stability relative to the monolayersformed with DDS and DDSe. A comparative analysis using tech-niques like cyclic voltammetry and quartz-crystal microbalance in-dicates that, for DDS, the facile cleavage of the S–S bond leads tostrong binding of the adsorbate molecules at the preferred surfacesites, resulting in a rather well-organized self-assembled structure.The STM pattern of NDS reveals a periodic domain (i.e., less than10 nm in size) while no such small domains are seen in the caseof DDS and DDSe due to the orientational flexibility of the rings.C© 2001 Academic Press

Key Words: self-assembled monolayers of DDS, DDSe and NDS;STM; QCM; cyclic voltammetry.

u ivc sft oo rS sss eo os fic eo re ip r

E

thelateoncon-ion

1)n oc-h auryl

enehes of

f geo-sedingstly,h at-ofrepa-tron

fidee the(8).n-

ularces.othuponundSe.a-andndupon

0CA

INTRODUCTION

During the past few years, preparation and characterizatioltrathin organic films obtained by self-assembly have receonsiderable attention for the modification of noble metalaces with a number of applications such as molecular recoion, nonlinear optics, tribology, and photopatterning methlogy (1–5). Among several functional groups known to foAMs on the surface of noble metal and semiconductingtrates, different types of the thiol/disulfide monolayers onurfaces have received attention due to their simplicity andf preparation (2–4). Although compact monolayers on gurfaces are generally formed by most organic thiol/disulompounds, the quality of the monolayer, such as the absf pinholes and other defects, critically depends on the pnce of a long hydrophobic tail. Structurally, the Au–S bondattern in both thiols and disulfides is the same except fo

1 To whom correspondence should be addressed. Fax: 0091-020-5893-mail: viji@ems.ncl.res.in.

eltion

41021-9797/01 $35.00opyright C© 2001 by Academic Pressll rights of reproduction in any form reserved.

n ofed

ur-gni-d-mub-Auaselddencees-ngan

044.

oxidative dissociation of the S–S bond for disulfides. Whilenature of surface attachment is believed to involve a Au–thiointeraction (1–4), the possibility of thiolate group dimerizatito form disulfides on gold surfaces has also been recentlysidered (6). Most of the data obtained from different diffractstudies are consistent with a (

√3×√3)R30◦ overlayer structure

formed during the chemisorption of thiol/disulfide on Au(11surfaces, and in such a model, each sulfur headgroup cacupy a threefold site on an underlying Au(111) surface witseparation of 5A. In this paper we present the results of ostudy on the ability of the monolayer formation of diphendisulfide (DDS), diphenyl diselenide (DDSe), and naphthaldisulfilde (NDS), which provide valuable information about trelative degree of organization and change in the interactionsubstrate headgroups (7). DDS was selected on the basis ometric similarity with respect of DDSe, whereas NDS was uto reflect the role of inherent rigidity of the naphthalene rso that S–S bond cleavage is sterically hindered (8, 9). Laselenol/diselenide monolayers have not yet received enougtention (10–12) despite their promising utility for a varietyapplications, such as photoresists and photocatalysts, in pration of semiconductor quantum dots, photon-induced electransfer systems, etc.

In our earlier study it was shown that naphthalene disulforms a compact monolayer (96% surface coverage) despitfact that the hydrocarbon tail lacks the ordering influenceDDS and DDSe were found to form similar monolayers uder identical conditions with the molecular plane perpendicto the surface, although there are several important differenFor example, DDS dissociately chemisorbs on Au, while in bDDSe and NDS, the Se–Se and S–S bonds are preservedadsorption. Also, the number of pinholes and defects are foto be much less in DDS monolayers than in NDS and DDIn addition, we compare the relative ability of SAM formtion and the structure of these monolayers of NDS, DDS,DDSe formed on polycrystalline gold by using QCM, STM, acyclic voltammetry. These techniques are selected basedtheir proven ability to effectively unravel the molecular levdetails of the structure (13–15) and the degree of organizaof SAMs.

0

L

yCnicl

euthuen,resranotar ibila

e

aeall

suns-t

fttoe

oknre

peereop-

e ofe in-iteg-

ectedes

toue tohemassnt offace.illa-nergyad-e.di-ctionsron-ing

andmM

old-es inblanklower

AROMATIC DISULFIDE AND DISE

EXPERIMENTAL

Gold substrates for these experiments were prepared bthermal evaporation of 200-nm gold (purity 99.9%) on abuffer layer (20 nm) first deposited onto conventionally cleaglass plates, using reported procedures (8). This procedure ypolycrystalline Au films with strong (111) texture. The surfaroughness factor (R= actual/geometrical surface area) of gowas determined according to a standard procedure, whercharge required to cathodically reduce the oxide layer wasfor the calculation (9). To avoid surface contamination, allsubstrates were stored in polypropylene containers untilNDS was prepared by a reported procedure (16). All other chicals used in this work were purchased from Aldrich Co. awere used as received. Before the SAMs were depositedsubstrates were cleaned by the standard cleaning proceduscribed earlier (17). After repeated cleaning cycles usingfochromic acid, aqueous HF, and deionized water, the substwere blown dry with pure argon and were immediately traferred into a deaerated 1 mM solution of the appropriate cpound in acetonitrile. The substrates were kept for 24 h indeposition solution to obtain well-organized, crystalline film,though 10–12 h is sufficient for the chemisorption. The longecubation period generally allows the molecular film to assemin a crystalline-like solid phase and also provides the possibfor the desorption of physisorbed and chemisorbed contaminfrom the gold surface into the solution. The substrates weremoved from the solution, washed repeatedly with the solvand finally with absolute ethanol, and then dried in a streamAr prior to the measurements. All solvents were reagent grand were used without further purification. Double-distillwater was passed through a Millipore system (Milli-Q) to obtdeionized high-purity water (>18 MÄ-cm) and was used in athe experiments.

An Elchema elctrochemical nanobalance (with 0.1-Hz relution) interfaced with a computer was used to monitor the stle mass changes during SAM formation. All the experimewere performed at room temperature using vaccum-depo10-MHz Au-coated quartz crystals (0.2-cm2 exposed area). Initially, the solvent was introduced into the vessel to stabilizeQCM frequency, and then a small measured amount (25µl)of the specific compound was inserted with a syringe. Thequency change due to adsorption is directly proportional tomass, and after conversion the mass was plotted againstThe blank experiment conducted with solvent alone showeda minor mass change as compared to the mass change aftmonolayer formation.

Cyclic voltammetry was performed in an oxygen-free atmsphere (bubbling Ar gas) by using a three-electrode cell: gcoated glass or SAM-modified gold substrates as the worelectrodes, a large area platinum flag counter electrode, asaturated calomel reference electrode (SCE). The suppoelectrolyte was 0.1 M aqueous KOH. All potentials in the t

are referred to as SCE. Cyclic voltammetry experiments wecarried out with an EG&G PAR Model 362 potentiostat and a R

ENIDE MONOLAYERS ON GOLD 411

ther-edeldsedthe

sedese.m-dthede-

ul-tes

s-m-hel-n-le

ityntsre-ntofdedin

o-b-tsited

he

re-heime.nlyr the

o-ld-ingd a

tingxt

0151 plotter. A home-built (18) scanning tunneling microsco(STM) was used to probe the SAM formation. STM studies wperformed at room temperature in air and the instrument waserated in constant current mode of 200 pA at a bias voltag+100 mV (substrate positive). Prior to these experiments, thstrument was calibrated with highly oriented pyrolytic graph(ZYA, Advanced Ceramic). An electrochemically etched tunsten tip was used as the probe. To ensure that the data collwere representative for the film morphology, multiple imagwere taken at different locations and scan ranges.

RESULTS AND DISCUSSIONS

Quartz Crystal Microbalance (QCM)

In situ quartz crystal microbalance (QCM) is consideredbe an ideal tool for investigating the adsorption processes dits high sensitivity (±0.1 ng) toward mass change (19, 20). TSauerbrey equation, which relates the frequency change toloading, can be used to determine the approximate amouadsorbed molecules on the gold-coated quartz crystal surHowever, several additional factors can also affect the osction frequency, such as viscous damping, surface stress, edisruption by nonshear coupling (20). The most importantvantage is its high sensitivity (±0.1 ng) toward mass changThe major limitation is that the frequency shift cannot berectly related to the absolute mass change and hence correby blank experiments (i.e., quartz crystal in the same enviment without the adsorbing molecule) is needed for obtainreliable kinetic information.

Figure 1 shows the mass change with time for NDS, DDS,DDSe when the gold-coated quartz crystal is exposed to 1

FIG. 1. The mass change with time for NDS, DDS, and DDSe when gcoated quartz crystals are exposed to 1 mM solution of these moleculacetonitrile; the mass values used are after correction by performing aexperiment. The error bars were determined on the basis of upper and

reEvalues of the fluctuation of the frequency signal. (a) Au/NDS; (b) Au/DDS;(c) Au/DDSe.

2

d

ra

ietflo.

sD

r

su

n

l

a

dhoe

f

eu

om-ma-yer

notrenty-umdichde., areentatre-b-

ntial

byongwas

odicdes,theE

de

S-rentbycan.

idero-rethedentstsoftiontedAuthehapet thatanenttivegidlowoneseono-nge

412 ASLAM

solution of these molecules in acetonitrile. The mass increseems to attain a plateau value at about 150, 260, and 34cm2 (corresponding to a time of 80 min) for monolayers of NDDDS, and DDSe, respectively. These values were obtainesubtracting the minor mass change (8 ng) from blank expments with the solvent only. Although the three curves shsimilar features, especially in the initial region, the one cosponding to DDS is steeper than the others, suggesting a fadsorption rate than that of NDS or DDSe. An approximatetimation of the surface coverage shows that DDS adsorbs wa few seconds and covers most of the surface (surface covclose to 0.9), while DDSe and NDS need about 1 h to atthe same coverage. In addition to the distinct kinetics of surcoverage, the difference in saturation coverage can be reto the molecular structure. The saturation coverage corresping to the final mass uptake is similar for DDS and DDSe (0and 0.99, respectively) while it is significantly lower for ND(0.8). Since the roughness of gold is identical in all casemore closely packed molecular assembly is suggested forand DDSe relative to a more defective type of monolayerNDS. This could be due to the easy relaxation appearing fbond cleavage for DDSe and DDS, while for NDS, the fusring may impose constraints for side movement and a leinteraction affinity of DDSe than for the corresponding disfides. These results underline the importance of intramolecconformational changes in controlling the monolayer packdensity and is in agreement with the results of surface-enhaRaman spectroscopy (SERS) investigations (17).

Cyclic Voltammetry (CV)

Considering the simple parallel plate capacitance modethe electrode–electrolyte interface, which has been appliedcessfully for electrodes derived from long chain alkanethmonolayers (21–23), double-layer capacitance studies canvide additional insights into the average structure of the sorganized assemblies. Figure 2a shows cyclic voltammogrof bare gold and the monolayer-coated electrodes (DDS, DDand NDS), respectively, recorded at the scan rate of 0.5 V s−1 inthe potential range from 0.0 to+0.8 V vs SCE. An approximatecalculation of the double-layer capacitance from the nonfaracurrent measured at a 0.25-V constant potential indicates tuniform and compact monolayer is formed due to the strchemisorption of DDS, DDSe, and NDS on the gold surfacshown by the decrease in double-layer capacitance compthat of to bare gold. The differential capacitance decreases29 µF cm−2 for bare gold to 7.3, 8.0, and 11.6µF cm−2 forDDS-, DDSe-, and NDS-functionalized substrates, respectivAlthough these capacitance values are different from the rtively better values measured in 0.1 M aqueous NaF electrofor well-defined oriented surfaces (17), the decreasing orderrelative change suggest that the capacitance values reflect mlayer compactness. For example, the monolayer formed for Nshows disordered/defective organization compared to thosDDS and DDSe, pointing toward the importance of struct

in controlling the degree of ordering. In addition, the avera

ET AL.

aseng/

S,by

eri-owre-ster

es-thinrageainaceatednd-

92S, aDS

foromedserl-

ularingced

forsuc-iolpro-elf-msSe,

aicat angas

aredorm

ely.ela-lyteandono-DSforre

structure and quality of the monolayer assemblies could be cpared by this method despite the limitation due to the estition of capacitance, which is not very sensitive to monoladefects.

The voltammogram of a bare gold electrode, althoughwell defined, clearly indicates a marked increase in the curaround+0.7 V vs SCE, corresponding to the oxidation of polcrystalline Au as reported earlier (24) in an alkaline medi(a pH≥14). Polycrystalline Au is known to give two flattenepeaks at 1.15 to 1.2 V vs SCE in an alkaline medium, whcorrespond to the oxidation/reduction reactions of Au oxiSometimes, organic impurities, especially redox active onesalso known to give certain anomalous peaks at slightly differvalues. Also, holding the potential for a long period of time+1.0 V vs SCE is known to give a sharp cathodic peak corsponding to the reduction of the oxide film. We have not oserved any of these features as the upper limit of the poterange during the anodic scan was+0.8 V vs SCE. However,the fact that the increase of the current around+0.7 V vs SCE(Fig. 2a) could be attributed to gold oxidation was verifiedtaking voltammograms beyond 1.2 V vs SCE, where a strcathodic peak was observed. In addition, when the potentialheld at 1.2 V vs SCE, an increase in the height of the cathpeak was found. However, for the monolayer-coated electrothe onset of oxygen evolution was found to adversely affectfilm quality and hence the anodic limit was fixed to 0.8 V vs SCfor all further voltammograms. This is expected for Au oxiand subsequent oxygen evolution.

Compared to the voltammogram of bare gold, the DDmodified electrode (Au/DDS) shows a drastic decrease in cur(almost 100 times) due to the monolayer formation followeda plateau region (above 0.6 V vs SCE) during the anodic sA reverse scan gives a distinct cathodic peak at+0.43 V vsSCE, which might be attributed to the reduction of the oxfilm formed in the forward scan, as known from the electchemistry of gold in alkaline media (24). Interestingly, theis no sharp cathodic peak for DDSe (Au/DDSe), althoughi–E response of a sharp current rises to a potential indepenlimiting current (i.e., sigmoid-type behavior), which suggediffusion through pinholes (25, 26). The origin of this typethe voltammogram can be attributed to the oxidation/reducof an Au substrate, perhaps through more uniformly distribupinholes (similar to a microarray electrode). NDS-modified(Au/NDS) shows behavior close to that of DDS, except thatcathodic reduction peak is not as pronounced. These peak schanges along with an increased peak separation suggesthe reduction of the oxide film is more irreversible for NDS thfor DDS. Nevertheless, the magnitude of the oxidation curris significantly greater, presumably due to the large defecnature of the monolayer. This could be attributed to the rinature of fused rings present in NDS while DDS and DDSe alconformational flexibility during organization. This comparisof the voltammetric behavior clearly indicates that, among thmolecules, NDS shows poor compactness and defective mlayers. The same trend is also confirmed by the order of cha

gein differential capacitance values estimated from Fig. 2a.

AROMATIC DISULFIDE AND DISELENIDE MONOLAYERS ON GOLD 413

FIG. 2. Cyclic voltammograms of bare gold and DDS-, DDSe-, and NDS-modified gold substrates (a) in the range from 0 to+0.8 V and (b) in the potentialrange of 0 to−1.0 V, in 0.1 M KOH (reference: SCE; counter electrode: platinum disk; scan rate: 0.5 V s−1).

E

aut

bDedeec

te

eoln.c

bif

Dee

hoifisDaua

nioe3tic

on

ri

oe

tes

414 ASLAM

To compare the electrochemical stability of these monolers with respect to reductive desorption in an alkaline medicyclic voltammograms were recorded in the negative potenrange. Figure 2b shows cyclic voltammetric response of thegold surface and of three monolayer-coated electrodes (DDDSe, and NDS), at a scan rate of 200 mV/s, in the pottial range from 0.0 to−1.0 V vs SCE. The electrodes modifiewith DDS, DDSe, and NDS show a reductive desorption pat−0.65,−0.73, and−0.49 V vs SCE. This behavior can bexplained on the basis of the earlier work by Porter andworker (27) on electrode reactions of disulfide monolayersgold in aqueous solution. The difference in the reducitve potial values also supports the relative compactness of the Dmonolayer since it is more difficult to desorb, perhaps duethe fact that it has been formed by a more facile cleavagS–S, in contrast to the easier desorption of NDS. DDSe shintermediate behavior and the low degree of the structurategrity of the NDS monolayer may be attributed to the preseof pinholes, grain boundaries, trapped solvent molecules, etcalso supplemented by the increased broadness of the redudesorption peak.

To determine if the above change in reductive desorptionhavior is only related to molecular structure and packing oris a function of time, scan-rate-dependent voltammograms wtaken (Fig. 3) in the same electrolyte for DDS, DDSe, and Nmonolayer-modified electrodes. The voltammetric responsthe current with an increase in scan rate for all molecules wfound to be similar as all peak potentials were found to smore cathodic with higher scan rates. For example, Fig. 3a shsuch a scan-rate-dependent voltammogram for a DDS-modAu electrode, where the peak seems to broaden with therate. DDSe also shows similar behavior (Fig. 3b), but for N(Fig. 3c), the shift in potential with an increase in scan ris substantially more confirming of the poor degree of strtural integrity. In addition, at higher scan rates the peakssharper and all these observations, including the pronounasymmetric shape at 500 mV s−1 scan rate, are in agreemewith the more irreversible nature of the reductive desorptof the NDS molecule. For all the cases, a plot of the currpeakslp, with scan rate, gives a straight line (insets of Fig.but does not pass through the origin, perhaps due to deviafrom the ideal surface confinement caused by the appearanpinholes, grain boundaries, trapped solvent, etc. or due toirreversible formation of the species involved in the voltammgrams. This can be compared with the voltammetric respofor diffusing species, where the peak current normally vawith the square root of the scan rate. The higher slope oflp vsthe square root of the scan rate plot for DDSe is also in gagreement with the less defective monolayers due to the bcoverage.

Scanning Tunneling Microscopy (STM)

STM is an important tool for studying nanoscopic structuresSAMs to a high degree of accuracy and the relationship betwe

T AL.

y-m,ialareS,

n-

ak

o-onn-DStoofwsin-ce, astive

e-itereSinre

iftwsed

canStec-re

cedtnnt),onse ofthe-se

es

odtter

FIG. 3. The cyclic voltammogram taken in 0.1 M KOH at various scan ra−1

of

en(0.05 to 0.5 V s ) for DDS-, DDSe-, and NDS-modified gold electrodes; insetshows variation of peak current with scan rate.

AROMATIC DISULFIDE AND DISELENIDE MONOLAYERS ON GOLD 415

FIG. 4. STM images of (a) bare gold and (b) DDSe-, (c) DDS-, and (d constant

o.ie

wb

a

r

b

a

r

andha-hein

re-

current mode.

molecular structure, compactness, and nature of the adsmolecules can be well understood using this technique (28)example, several STM studies of SAMs using long chain thon Au(111) surfaces clearly illustrate the surface structurwell as subtle changes upon annealing (29). STM images shin Figs. 4a–4d also indicate the difference in the nature ofmonolayer for these three molecules on an Au surface. Incase of gold-coated glass (Fig. 4a), the surface does not shoperiodicity or terraces, while the crystalline structure is visiwith granular size variation from 30 to 60 nm, having degorges and hills (no single-crystal surfaces were used). Relto bare gold, DDSe-functionalized gold surfaces show (Fig.uniform size domains bundled together with a size ranging f25 to 30 nm and nearly similar patterns have been obsefor DDS-modified gold surfaces (Fig. 4c). For NDS, no superiodic domains have been observed (Fig. 4d), the surfacecovered with larger domains, showing a continuous coveragthe surface with NDS. The streaks are probably from the smdomains and pinholes are clearly visible in all images. ThSTM images are qualitatively similar to those reported eafor other organoselenium and organosulfur monolayers (11

Figure 5 displays a high-resolution STM pattern of NDwhich reveals a periodic domain (i.e., less than 10 nm in siz

) NDS-modified SAMs on gold substrates; all images were obtained in the

rbedForolsas

ownthetheanyle

eptive

4b)omrvedcheinge ofller

eselier).

Significantly, no such small domains are seen for DDSDDSe monolayers. This is due to the rigidity of the naphtlene ring in NDS, which hinders the monolayer formation. Torientational flexibility of rings in DDS and DDSe is missing

FIG. 5. Magnified STM image of a NDS molecule after scanning a rep

S,e).sentative domain of a small area in constant current mode when the tunnelingcurrent was kept at 200 pA and the bias voltage was 100 mV.

E

eb

Sfi

n

ltsso-, in

m-u

ate–no-DSthe

Se.eredwo

low-ment

r–CA,

om-

and

and

416 ASLAM

FIG. 6. STM images of DDSe-modified gold at different scan ranges, keing the tunneling current (200 pA) in constant current mode by applying avoltage of 100 mV.

NDS. This illustrates the importance of geometric constraintsthe formation of a compact monolayer. STM images of DDmonolayers on gold, at various scan ranges (Fig. 6), conthat the entire surface has uniform coverage. Domains of silar size are present, despite the presence of pinholes. Simpatterns are also observed for DDS monolayers as showFig. 7.

FIG. 7. STM images of DDS-modified gold at different scan ranges, keepithe tunneling current (200 pA) in constant current mode by applying a bvoltage of 100 mV.

T AL.

p-ias

fore

rmim-ilarin

ng

Thus, voltammetric data in combination with the QCM resusuggest that while monolayer formation by adsorption fromlution is fast for DDS, reductive desorption is faster for NDSan alkaline medium.

CONCLUSIONS

A comparison of the results of QCM, STM, and electrocheical investigations of SAMs of NDS, DDS, and DDSe on Afilms indicates that molecular geometry as well as substrheadgroup interaction play a key role in determining the molayer organization and stability. The facile S–S cleavage in Dis one of the main reasons for the enhanced stability ofmonolayer, as also illustrated by the structurally similar DDBy contrast, NDS forms a thermally unstable and less ordmonolayer, presumably due to the structural rigidity of the tsulfurs.

ACKNOWLEDGMENTS

Mohammed Aslam thanks CSIR, New Delhi, for a Senior Research Felship. K. Vijayamohanan acknowledges financial support from the Departof Science and Technology, Government of India.

REFERENCES

1. Ulman, A., “An Introduction to Ultrathin Organic Films from LangmuiBlodgett to Self-assembly,” p. 237. Academic Press, San Diego,1991.

2. Nuzzo, R. G., and Allara, D. L.,J. Am. Chem. Soc.105,4481 (1983).3. Hickman, J. J., Ofer, D., Laibinis, P. E., and Whitesides, G. M.,Science

252,688 (1991).4. Nuzzo, R. G., Zegarski, B. R., and Dubois, L. H.,J. Am. Chem. Soc.109,

733 (1987).5. Ulman, A.,Chem. Rev.96,1533 (1996).6. Fenter, P., Eberhardt, A., and Eisenberger, P.,Science266,1216 (1994).7. Patai, S., and Rappoport, Z., “Organic Selenium and Tellurium C

pounds,” Vol. 1, p. 44. Wiley, New York, 1986.8. Bandyopadhyay, K., Sastry, M., Paul, V., and Vijayamohanan, K.,Langmuir

13,866 (1997).9. Oesch, U., and Janata, U.,Electrochim. Acta28,1237 (1983).

10. Samant, M. G., Brown, C. A., and Gordon, J. G., II,Langmuir8, 1615(1992).

11. Dishner, M. H., Hemminger, J. G., and Feher, F. J.,Langmuir13, 4788(1997).

12. Bandyopadhyay, K., and Vijayamohanan, K.,Langmuir14,625 (1998).13. Bryant, M. A., and Pemberton, J. E.,J. Am. Chem. Soc.113, 3629

(1991).14. Murty, K. V. G., Venkataramanan, M., and Pradeep, T.,Langmuir14,5446

(1998).15. Bain, C. D., Troughton, E. B., Tao, Y. T., Evall, J., Whitesides, G. M.,

Nuzzo, R. G.,J. Am. Chem. Soc.111,321 (1989).16. Gamage, S. A., and Smith, R. A. J.,Tetrahedron46,2111 (1990).17. Bandyopadhyay, K., Vijayamohanan, K., Venkataramanan, M.,

Pradeep, T.,Langmuir15,5314 (1998).18. Lakshminarayanan, V.,Curr. Sci.74,413 (1998).

ias19. Sauerbrey, G.,Z. Phys.155,206 (1959).20. Buttry, D. A., and Ward, M. D.,Chem. Rev.92,1355 (1992).

L

.ski,

AROMATIC DISULFIDE AND DISE

21. Porter, M. D., Thomas, B. B., Allara, D. L., and Chidsey, C. E. D.,J. Am.Chem. Soc.109,3559 (1987).

22. Miller, C., Cudent, P., and Gr¨atzel, M., J. Phys. Chem.95, 877(1991).

23. Sabatini, E., and Rubinstein, I.,J. Phys. Chem.91,6663 (1987).24. Icenhower, D. E., Urbach, H. B., and Harisson, J. H.,J. Electrochem. Soc

117,1500 (1970).25. Whiteman, R. M.,Anal. Chem.53,1127A (1981).

ENIDE MONOLAYERS ON GOLD 417

26. Amatore, C., Saveant, J. M., and Tessier, D.,J. Electroanal. Chem.147,39(1983).

27. Chuan-Jian, Z., and Porter, M. D.,J. Am. Chem. Soc.116, 11616(1994).

28. Burgess, I., Jeffrey, C. A., Cai, X., Szymanski, G., Galus, Z., and LipkowJ.,Langmuir15,2607 (1999).

29. Dishner, M. H., Peter, T., Hemminger, J. C., and Feher, F. J.,Langmuir14,6676 (1998).