scanning electrochemical microscopy

The first five chapters of this book contain experimental and theoreticalbackground, which is essential for everyone working in this field: principles of SECM measurements Chapter 1, ins

This book is printed on acid-free paper.

Headquarters

Marcel Dekker, Inc

270 Madison Avenue, New York, NY 10016

Copyright 䉷 2001 by Marcel Dekker, Inc All Rights Reserved.

Neither this book nor any part may be reproduced or transmitted in any form or by anymeans, electronic or mechanical, including photocopying, microfilming, and recording,

or by any information storage and retrieval system, without permission in writing fromthe publisher

Current printing (last digit):

PRINTED IN THE UNITED STATES OF AMERICA

A little more than ten years have elapsed since publication of the first papersdescribing the fundamentals of scanning electrochemical microscopy(SECM) During this decade, the field of SECM has evolved substantially.The technique has been used in a variety of ways, for example, as an elec-trochemical tool to study heterogeneous and homogeneous reactions, forhigh-resolution imaging of the chemical reactivity and topography of variousinterfaces, and for microfabrication Quantitative theoretical models havebeen developed for different modes of the SECM operation The first com-mercial SECM instrument was introduced in 1999 The SECM technique isnow used by a number of research groups in many different countries Wethink the time has come to publish the first monograph, providing compre-hensive reviews of different aspects of SECM

The first five chapters of this book contain experimental and theoreticalbackground, which is essential for everyone working in this field: principles

of SECM measurements (Chapter 1), instrumentation (Chapter 2), tion of SECM ultramicroelectrodes (Chapter 3), imaging methodologies(Chapter 4), and theory (Chapter 5) Other chapters are dedicated to specificapplications and are self-contained Although some knowledge of electro-chemistry and physical chemistry is assumed, the key ideas are discussed atthe level suitable for beginning graduate students

prepara-Through the addition of submicrometer-scale spatial resolution, SECMgreatly increases the capacity of electrochemical techniques to characterizeinterfaces and measure local kinetics In this way, it has proved useful for

a broad range of interdisciplinary research Various applications of SECMare discussed in this book, from studies of biological systems, to sensors, toprobing reactions at the liquid/liquid interface Although we did not intend

to present even a brief survey of those diverse areas of research, each chapter

provides sufficient details to allow a specialist to evaluate the applicability

of the SECM methods for solving a specific problem We hope it will beuseful to all interested in learning about this technique and applying it

We would like to thank our students, co-workers, and colleagues whohave done so much to develop SECM The future for this technique, which

is unique among scanning probe methods in its quantitative rigor and itsability to study with ease samples in liquid environments, continues to be abright one

Allen J Bard Michael V Mirkin

I Background of Scanning Electrochemical Microscopy 1

IV Tip Position Modulation Instrumentation 44

V Constant-Current Mode Instrumentation 44

VI Experimental Difficulties in Data Acquisition 53VII Accessory Equipment for SECM 59

3 THE PREPARATION OF TIPS FOR SCANNING

Fu-Ren F Fan and Christophe Demaille

III Nondisk Tips and Tip Shape Characterization 104

Fu-Ren F Fan

II Principle and Methodology of SEM Imaging 111

V Conclusions and Future Projections 139

IV SECM of More Complicated Chemical Systems 170

V Numerical Solution of SECM Diffusion Problems

Using PDEase2 Program Package 182

6 HETEROGENEOUS ELECTRON TRANSFER REACTIONS 201

Kai Borgwarth and Ju¨rgen Heinze

7 KINETICS OF HOMOGENEOUS REACTIONS COUPLED

TO HETEROGENEOUS ELECTRON TRANSFER 241

8 CHARGE-TRANSFER AT THE LIQUID/LIQUID INTERFACE 299

Michael V Mirkin and Michael Tsionsky

IV Processes with Coupled Homogeneous Reactions 336

III Properties and Behavior of Ion-Selective Probes 417

IV Potentiometric Measurements in Scanning Probe

V Potentiometric Measurements in SECM 423

11 BIOLOGICAL SYSTEMS 445

Benjamin R Horrocks and Gunther Wittstock

I Approaches to Imaging Biological and Biochemical

IV Abbreviations, Acronyms, and Symbols 510

12 PROBING REACTIONS AT SOLID/LIQUID INTERFACES 521

Julie V Macpherson and Patrick R Unwin

II Measurement of Adsorption/Desorption Kinetics and

III Dissolution Kinetics of Ionic Single Crystals 536

13 MICRO- AND NANOPATTERNING USING THE

SCANNING ELECTROCHEMICAL MICROSCOPE 593

Daniel Mandler

I Patterning by the Direct Mode of the SECM 594

II Patterning by the Feedback Mode of the SECM 603

Allen J Bard

I Combining SECM with Other Techniques 629

III Instrumentation Improvements 636

ALLEN J BARD The University of Texas at Austin, Austin, Texas

BRADLEY D BATH ALZA Corporation, Mountain View, California

KAI BORGWARTH Institute for Physical Chemistry, Albert LudwigUniversity of Freiburg, Freiburg, Germany

CHRISTOPHE DEMAILLE The University of Texas at Austin, Austin,Texas

GUY DENUAULT University of Southampton, Southampton, England

FU-REN F FAN The University of Texas at Austin, Austin, Texas

JU¨ RGEN HEINZE Albert Ludwig University of Freiburg, Freiburg,Germany

BENJAMIN R HORROCKS University of Newcastle upon Tyne,Newcastle upon Tyne, United Kingdom

JULIE V MACPHERSON University of Warwick, Coventry, England

DANIEL MANDLER The Hebrew University of Jerusalem, Jerusalem,Israel

MICHAEL V MIRKIN Queens College–City University of New York,Flushing, New York

GE´ ZA NAGY Janus Pannonius University, Pe´cs, Hungary

ERIK R SCOTT Medtronic Corporation, Minneapolis, Minnesota

KLA´ RA TO´TH Institute of General and Analytical Chemistry, TechnicalUniversity of Budapest, Budapest, Hungary

MICHAEL TSIONSKY Gaithersburg, Maryland

PATRICK R UNWIN University of Warwick, Coventry, England

HENRY S WHITE University of Utah, Salt Lake City, Utah

DAVID O WIPF Mississippi State University, Mississippi State,

Mississippi

GUNTHER WITTSTOCK Wilhelm-Ostwald-Institute of Physical andTheoretical Chemistry, University of Leipzig, Leipzig, Germany

This volume is devoted to a complete and up-to-date treatment of scanningelectrochemical microscopy (SECM) In this introductory chapter, we coverthe historical background of the technique, the basic principles of SECM,and an overview of some of its applications (covered in more depth inlater chapters) A number of reviews of this field have also been published(1–6).

SECM involves the measurement of the current through an

ultrami-croelectrode (UME) (an electrode with a radius, a, of the order of a few nm

to 25 ␮m) when it is held or moved in a solution in the vicinity of asubstrate Substrates, which can be solid surfaces of different types (e.g.,glass, metal, polymer, biological material) or liquids (e.g., mercury, immis-cible oil), perturb the electrochemical response of the tip, and this pertur-bation provides information about the nature and properties of the substrate.The development of SECM depended on previous work on the use of ul-tramicroelectrodes in electrochemistry and the application of piezoelectricelements to position a tip, as in scanning tunneling microscopy (STM) Cer-tain aspects of SECM behavior also have analogies in electrochemical thin-layer cells and arrays of interdigitated electrodes

The movement of the tip is usually carried out by drivers based onpiezoelectric elements, similar to those used in STM, as described in Chapter

2 Typically, inchworm drivers (Burleigh Instruments, Fishers, NY) are used,since they can move larger distances than simple piezoelectric tube scanners.However, where higher resolution is needed, piezoelectric pushers can beadded, so that the inchworms provide coarse drives and the pushers nm-resolution drives Generally the direction normal to the substrate is taken asthe z direction, while x and y are those in the plane of the substrate

There are several modes of operation of the SECM In the tip tion–substrate collection (TG/SC) mode, the tip is used to generate a reac-tant that is detected at a substrate electrode For example, the reaction O⫹

genera-ne→ R occurs at the tip, and the reverse reaction occurs at the substrate.This mode of operation is similar to that at the rotating ring-disk electrode(7) Similar behavior is observed for a pair of side-by-side microband elec-trodes (8,9) and in thin-layer cells (10) In the SECM, TG/SC is usuallyused in studies of homogeneous chemical reactions, where the reaction ofspecies R as it transits between tip and substrate causes a decrease in thesubstrate current (see Chapter 7) An alternative mode, where the substrate

is the generator and tip the collector (SG/TC mode), can also be employedand is used in studies of reactions at a substrate surface (Chapters 6, 9, 11,and 12) The SG/TC mode was first used to study concentration profiles near

an electrode surface without scanning and imaging (11–13)

The most frequent mode of operation of the SECM is the feedbackmode, where only the tip current is monitored As discussed in the nextsection, the tip current is perturbed by the presence of a substrate at closeproximity by blockage of the diffusion of solution species to the tip (negativefeedback) and by regeneration of O at the substrate (positive feedback) Thiseffect allows investigation of both electrically insulating and conducting sur-faces and makes possible imaging of surfaces and the reactions that occurthere This mode of operation with surface imaging was first described, alongwith the apparatus and theory, in a series of papers in 1989 (14–16)

II PRINCIPLES OF SECM

A Ultramicroelectrodes

An understanding of the operation of the SECM and an appreciation of thequantitative aspects of measurements with this instrument depends upon anunderstanding of electrochemistry at small electrodes The behavior of ul-tramicroelectrodes in bulk solution (far from a substrate) has been the subject

of a number of reviews (17–21) A simplified experimental setup for anelectrochemical experiment is shown in Figure 1 The solution contains a

species, O, at a concentration, c, and usually contains supporting electrolyte

to decrease the solution resistance and insure that transport of O to theelectrode occurs predominantly by diffusion The electrochemical cell alsocontains an auxiliary electrode that completes the circuit via the power sup-ply As the power supply voltage is increased, a reduction reaction, O⫹ ne

→ R, occurs at the tip, resulting in a current flow An oxidation reactionwill occur at the auxiliary electrode, but this reaction is usually not of in-terest in SECM, since this electrode is placed sufficiently far from the UME

FIG 1 Schematic diagram of a cell for ultramicroelectrode voltammetry.

FIG 2 Typical voltammogram for an ultramicroelectrde

that products formed at the auxiliary electrode do not reach the tip duringthe experiment The potential of the tip electrode is monitored against astable reference electrode, such as a silver/silver chloride electrode A plot

of the current flowing as a function of the potential of the UME is called avoltammogram; a typical one is shown in Figure 2 As shown, an S-shaped

curve is produced The current eventually limits to a value that is completelycontrolled by the rate of mass transfer by diffusion of O from the bulksolution to the electrode surface, where the electrochemical reaction hasdecreased its concentration to essentially zero For a conductive disk of

radius a in an insulating sheath, this steady-state diffusion-controlled current

when the tip is far from a surface is given by:

where D is the diffusion coefficient of species O, and F is the Faraday The

current at electrodes with other shapes, e.g., hemispheres or cones, can beexpressed in a similar way, as discussed in Chapter 3, but almost all SECMexperiments are carried out with disk-shaped electrodes, because they showthe best sensitivity The current is also relatively independent of the radius

of the insulating sheath, r g, often expressed in the SECM literature as

RG = r g /a Moreover, because the flux of O to a small disk by diffusion

(⬃Dc/a) is quite large, the current is relatively immune to convective effects

like stirring in the solution The current at a small disk also reaches steadystate in a relatively short time (⬃a2

/D) For example, a 10 ␮m radius diskwill attain steady state in a fraction of a second These characteristics implythat an ultramicroelectrode used as a scanning tip and moved in a solutioncan be treated as a steady-state system Finally, because of the small currentsthat characterize most experiments with ultramicroelectrode tips, generally

pA to nA, resistive drops in the solution during passage of current are erally negligible

gen-B Feedback Mode

The general principles of the feedback mode are shown in Figure 3 As

shown in Eq (1), the current, iT, ⬁, is measured at the ultramicroelectrode tipwhen it is far from any surface (A), the subscript, ⬁, implying this longdistance In fact, as we shall see, this distance only has to be a few tipdiameters The current under these conditions is driven by the hemisphericalflux of species O from the bulk solution to the tip (Fig 3A) When the tip

is brought near an electrically insulating substrate, like a piece of glass orplastic (Fig 3C), the substrate blocks some of the diffusion of O to the tip

and the current will decrease compared to iT, ⬁ The closer the tip gets to the

substrate, the smaller iT becomes At the limit when the distance between

tip and substrate, d, approaches zero, iTalso approaches zero This decrease

in current with distance is called negative feedback When the tip is brought

near an electrically conductive substrate, like a platinum electrode, whilethere is still blockage of diffusion of O to the tip by the substrate, there isalso the oxidation of the product R back to O This O generated at the

FIG 3 Basic principles of scanning electrochemical microscopy (SECM): (A) far

from the substrate, diffusion leads to a steady-state current, iT,⬁; (B) near a conductive

substrate, feedback diffusion leads to iT> iT, ⬁; (C) near an insulating substrate,

hin-dered diffusion leads to iT < iT, ⬁ (Reprinted with permission from A J Bard, G

Denuault, C Lee, D Mandler, and D O Wipf, Acc Chem Res 23, 357 (1990).

Copyright 1990 American Chemical Society.)

substrate diffuses to the tip and causes an increase in the flux of O compared

with iT, ⬁ Thus with a conductive substrate iT > iT, ⬁ In the limit as d

ap-proaches zero, the tip will move into a regime where electron tunneling canoccur and the tip current will get very large This increase of current with

distance is called positive feedback A plot of iTversus d, as a tip is moved

in the z direction, is called an approach curve.

A quantitative description of approach curves can be obtained by solvingthe diffusion equations for the situation of a disk electrode and a planarsubstrate (16), as discussed in Chapter 5 Typical approach curves for aconductive substrate (essentially infinite rate of regeneration of O from R)and an insulating substrate (zero rate of regeneration of O) are shown in

Figure 4 These curves are given in dimensionless form by plotting IT= iT/

iT, ⬁ (the tip current normalized by the current far from substrate) versus L = d/a (the tip-substrate separation normalized by the tip radius) Since this plot

involves only dimensionless variables, it does not depend upon the tration or diffusion coefficient of O From these curves one can readily find

concen-d from the measureconcen-d ITand a knowledge of a The approach curves for an insulator actually depend upon rg, since the sheath around the conductingportion of the electrode also blocks diffusion, but this effect is not usuallyimportant with most practical tips If the rate constant for electron transfer

at the substrate to species O is kb,s, the limiting curves repesent kb,s → 0

(insulator) and kb,s→ ⬁ (conductor) The approach curves for intermediate

values of kb,s can be found (Chapter 5) (Fig 5) These are very useful infinding the rate of heterogeneous charge transfer at an interface (see Chapters

6 and 8)

C Collection-Generation Modes

As discussed above, there are two modes of this type In the TG/SC mode,the tip is held at a potential where an electrode reaction occurs and thesubstrate is held at a different potential where a product of the tip reactionwill react and thus be collected In most cases the substrate is considerably

larger than the tip, so that the collection efficiency, given by iS/iT(where iS

is the substrate current), is essentially 1 (100%) for a stable tip-generated

species, R If R reacts on transit from tip to substrate, iS/iTbecomes smaller,

and its change with separation, d, allows determination of the rate constant

of the homogeneous reaction (Chapter 7)

The alternative mode is the substrate generation–tip collection (SG/TC)mode In this case the tip probes the reactions that are occurring on a sub-strate For example, a scan in the z direction can produce the concentrationprofile, while a scan over the surface can identify hot spots, where reactionsoccur at a higher rate

FIG 4 Diffusion-controlled steady-state tip current as a function of tip-substrateseparation (A) Substrate is a conductor; (B) substrate is an insulator (From Ref 2.)

A related method involves the use of the tip reaction to perturb a tion at a surface; an example of this approach is SECM-induced desorption(SECMID) (22) For example, the adsorption/desorption kinetics of protons

reac-on a hydrous metal oxide surface can be studied in an unbuffered solutireac-on

by bringing the tip near the surface and reducing proton (to hydrogen) atthe tip This causes a local change in pH that results in proton desorptionfrom the surface The tip current can be used to study the kinetics of protondesorption and diffusion on the surface (Chapter 12)

FIG 5 Approach curves as a function of the heterogeneous reaction rate constant

for electron transfer at the substrate, k, IT = iT/iT, ⬁ From top to bottom, k (cm/s) is

(a) 1, (b) 0.5, (c) 0.1, (d) 0.025, (e) 0.015, (f) 0.01, (g) 0.005, (h) 0.002, (i) 0.0001.Curve (a) is identical to that for mass transfer control and curve (i) for an insulatingsubstrate

D Transient Methods

Most SECM measurements involve steady-state current measurements Thiscan be a significant advantage in the measurement of kinetics, even for rapidprocesses, because factors like double-layer charging and adsorption do notcontribute to the observed currents However, one can also carry out transient

measurements, recording iT as a function of time This can be of use inmeasurements of homogeneous kinetics (Chapter 7) and for systems that arechanging with time It can also be used to determine the diffusion coefficient,

D, of a species without knowledge of the solution concentration or number

of electrons transferred in the electrode reaction (23)

E Fabrication

The SECM can also be used as a tool for modification of surfaces Forexample, metals or semiconductors can be etched or metals deposited on asurface by passing the tip close to the surface and carrying out an appropriate

electrochemical reaction Two different modes are possible In the direct

mode, the tip acts as the counterelectrode and the desired electrochemical

reaction occurs on the substrate For example, Cu can be etched from a Cusubstrate Spatial resolution is determined by the current density distribution

between tip and substrate In the feedback mode a reactant is generated at

the tip which promotes the reaction on the substrate For example, Cu can

be etched by bromine electrogenerated at the tip In this case resolution isdetermined by the lateral (x-y) diffusion of reactant as it diffuses from tip

to substrate Details of fabrication using SECM are covered in Chapter 13

III APPLICATIONS OF SECM

The chapters that follow illustrate a wide range of applications of SECMthat have appeared Given below is an overview and some examples thatmight help put the technique in perspective before the detailed treatments

A Imaging

By scanning the tip in the x-y plane and measuring current changes (the

constant height mode) (or, less frequently, by maintaining a constant current and measuring the changes in d in a constant current mode), one can obtain

topographic images of conducting and insulating substrates (Chapter 4) The

resolution of such images is governed by the tip radius, a, and d However,

by working in the thin film of water that condenses on a mica surface inhumid air, it is possible to obtain higher resolution with a conical tip that isonly slightly immersed in the water film Of particular interest is the use ofSECM to perform ‘‘chemical imaging,’’ observing differences in reactionrates at different locations on the surface This mode is useful in studyingbiological materials (e.g., enzyme sites) (Chapter 11) and surfaces that haveactive and passive sites

B Ultramicroelectrode Shape Characterization

It is frequently difficult to determine the actual shape of an trode by examination using an optical or scanning electron microscope Forexample, the conducting portion may be slightly recessed inside the glassmantle, or the shape may be that of a cone protruding from the insulator.Electrodes with radii of the order of 1␮m or less are particularly difficult

ultramicroelec-to characterize Simply determining a voltammogram with the tip in bulksolution is usually not useful in this regard, since almost all ultramicroelec-trodes will produce a steady-state wave-shaped voltammogram characteristic

of roughly hemispherical diffusion However, by recording an approach

curve, iTversus d, one can frequently identify recessed tips (where iTdoes

not increase at small d when the insulator hits the substrate) or tips with

shapes other than disks, which show different approach behavior ter 5).

(Chap-C Heterogeneous Kinetics Measurements

As suggested above, by recording an approach curve or voltammogram withthe tip close to a substrate, one can study the rates of electron transferreactions at electrode surfaces (Chapter 6) Because mass transfer rates atthe small tip electrodes are high, measurements of fast reactions withoutinterference of mass transfer are possible As a rule of thumb, one can mea-

sure k ⬚ values (cm/s) that are of the order of D/d, where D is the diffusion

coefficient (cm2

/s) For example, k⬚ for ferrocene oxidation at a Pt electrode

in acetonitrile solution was measured at a 1 ␮m radius tip at a d of about

0.1␮m yielded a value of 3.7 cm/s (24) The use of small tips and smallcurrents decreases any interference from uncompensated resistance effects

D Measurements of Homogeneous Kinetics

Rate constants for homogeneous reactions of tip-generated species as theytransit between tip and conducting substrate can be determined from steady-state feedback current or TG/SC experiments or by transient measurements(Chapter 7) Generally rate constants can be measured if the lifetime of thespecies of interest is of the order of the diffusion time between tip and

substrate, d2

/2D Thus first-order reaction rate constants up to about 105

s⫺1and second-order reaction rate constants up to about 108 M⫺1 s⫺1 are ac-cessible

E Biological Systems

There have been a number of applications of SECM to biological systems(Chapter 11) These include imaging of cells, studies of enzymatic reactions,and oxygen evolution on leaf surfaces SECM has also been applied ininvestigations of the transport of species through skin (Chapter 9) BecauseSECM is capable of monitoring a wide range of chemical species with goodspecificity and high spatial resolution, it should find wide application instudies of living organisms and isolated tissues and cells

F Liquid/Liquid Interfaces

There is considerable interest in ion and electron transfer processes at theinterface between two immiscible electrolyte solutions (ITIES), e.g., waterand 1,2-dichloroethane SECM can be used to monitor such processes(Chapter 8) It allows one to separate ion transport from electron transfer

and is relatively insensitive to the resistance effects often found with moreconventional (four-electrode) electrochemical measurements.

G Membranes and Thin Films

Different types of films on solid surfaces (e.g., polymers, AgBr) and branes separating solutions have been examined by SECM (Chapters 6 and9) SECM is a powerful technique for examining transport through mem-branes, with the ability to scan the surface to locate positions of differentpermeability It has also been used with polymer films, e.g., polyelectrolytes

mem-or electronically conductive polymers, to probe the counterion (dopant) fluxduring redox processes SECM can be particularly useful in probing filmthickness as a film is grown on a surface (25) SECM is unique in its ability

to probe inside some thin films and study species and electrochemical cesses within the films (26,27) For example, the tip current versus z-dis-placement curve as a conical tip (30 nm radius, 30 nm height) was movedfrom a solution of 40 mM NaClO4 into a nominally 2000 A˚ thick Nafionfilm containing Os(bpy)2 3 ⫹ on a glass/ITO substrate (Fig 6) (26) The tipwas held at 0.80 V versus SCE, whereOs(bpy)2 3 ⫹is oxidized to the 3⫹ form

pro-at a diffusion controlled rpro-ate The different stages of penetrpro-ation of the tipinto the film, from initial contact to tunneling at the ITO can clearly be seenand the film thickness established Moreover, with the tip at position c, avoltammogram can be recorded (Fig 7) From such a voltammogram, onecan determine the diffusion coefficient of 2 ⫹and information about

Os(bpy)3the kinetics and thermodynamics of the reaction occurring in the film

H Surface Reactions

Measurements of the rates of surface reactions on insulator surfaces, such

as dissolution, adsorption, and surface diffusion, are possible (Chapter 12).For example, proton adsorption on an oxide surface can be studied usingthe tip to reduce proton and induce a pH increase near the surface (22).Then, by following the tip current with time, information about proton de-sorption kinetics is obtained Studies of corrosion reactions are also possible.Indeed, work has been reported where a tip-generated species has initiatedlocalized corrosion and then SECM feedback imaging has been used to study

it (28) In these types of studies, the tip is used both to perturb a surfaceand then to follow changes with time

I Semiconductor Surfaces

SECM has been used to probe heterogeneous electron transfer reaction netics on semiconductor electrodes, such as WSe (29) In these studies, as

FIG 6 (Top) A scheme representing five stages of the SECM current-distanceexperiment (A) The tip is positioned in the solution close to the Nafion coating.(B) The tip has penetrated partially into Nafion and the oxidation of OS(bpy)2 3 ⫹occurs The effective tip surface grows with penetration (C) The entire tip electrode

is in the film but is not close to the ITO substrate (D) The tip is sufficiently close

to the substrate to observe position SECM feedback (E) The tunneling region tom) Dependence of the tip current versus distance The letters a–e correspond tothe five stages A–E described above The displacement values are given with respect

(Bot-to an arbitrary zero point The current observed during the stages a–d is muchsmaller than the tunneling current and therefore cannot be seen on the scale of curve

1 (the left-hand scale) Curve 2 is at higher current sensitivity to show the distant curve corresponding to stages a–d (the right-hand current scale) The solid

current-line is computed for a conically shaped electrode with a height, h = 30 nm, and a radius, r0= 30 nm for zones a–c, and SECM theory for zone d The tip was biased

at 0.80 V vs SCE, and the substrate at 0.20 V vs SCE The tip moved at a rate of

30 A˚ /s (From Ref 26.)

in those at the liquid/liquid interface, the use of a separate metal probeelectrode is useful in freeing the measured response from resistance effects

It also allows one to examine differences in behavior at different points on

a surface As discussed in Chapter 13 on applications to fabrication, SECMhas also been used to etch semiconductor surfaces and study the nature ofthe etching reactions

J Electrochemistry in Small Volumes of Solution

Because of its ability to position an electrode tip with high spatial resolution

in three dimensions, SECM can be used to probe electrochemistry in a smallvolume of liquid (e.g., on a conductive substrate that serves as a counter/reference electrode) For example, a solution volume of 3–20␮L was used

to probe the adsorption isotherms on a mineral surface (30) Probing evensmaller volumes, e.g., of liquids contained in pores, should be possible.Since electrochemical generation is an ideal method for producing small,controlled amounts of reactants, studies in which one wants to probe chem-istry with very limited amounts of sample appear to be a good application

In such studies, means to maintain the sample volume and prevent ration, for example, by close control of the humidity or using an overlayer

evapo-of an immiscible liquid, will be required

K Thin Liquid Layers

The SECM has been used to form thin liquid layers and probe ical reactions in them When the tip is pushed through the interface between

electrochem-FIG 7 Voltammogram at a microtip electrode partially penetrating a Nafion filmcontaining 0.57 MOS(bpy)2 3 ⫹ Scan rate, v = 5 mV/s The substrate was biased at

0.2 V vs SCE The solid line is computed with a heterogeneous rate constant, k⬚ =1.6⫻ 10⫺4cm/s and D = 1.2⫻ 10⫺9cm2

/s (From Ref 26.)

two immiscible liquids, for example, through an aqueous layer above a layer

of mercury or a layer of benzene above an aqueous layer, a thin film (severalhundred nm to a few␮m) of the top liquid layer is trapped on the surface

of the tip (31–33) Electrochemical measurements can be used to probereactions in this layer Another type of thin layer that has been studied isthe one that forms on a surface when exposed to humid air In this case, awater layer that can be as thin as a few nanometers forms on a hydrophilicsurface (e.g., mica) The SECM tip can probe into this layer, although studieshave mainly been aimed so far at imaging rather than investigating theproperties of the layer (Chapter 4)

L Potentiometry

While most SECM studies are carried out with amperometric tips that drivefaradaic (electron transfer) reactions, it is also possible to use potentiometrictips that produce a potential change in response to concentration changes ofspecies These are usually typical ion selective electrode tips, although other

types, such as Sb tips for pH detection, have been described (34) Probes ofthis type and their applications are discussed in Chapter 10 They are par-ticularly useful for studies of species that do not show electroactivity, like

Na⫹, K⫹, and Ca2 ⫹ Note, however, that ions of this type can be determined

in an amperometric mode by the use of micropipet electrodes that respond

to the transport of ions across an interface between two immiscible liquids(35,36)

M Fabrication

A variety of studies have now been done that demonstrate that the SECMcan carry out metal deposition, metal and semiconductor etching, polymerformation, and other surface modifications with high resolution Such pro-cesses are discussed in Chapter 13 These SECM approaches have the ad-vantage over analogous STM procedures in that the conditions of deposition

or etching are usually known and well defined, based on electrochemicalstudies at larger electrodes

REFERENCES

1 A J Bard, F.-R F Fan, D T Pierce, P R Unwin, D O Wipf, and F Zhou,

Science 254:68–74, 1991.

2 A J Bard, F.-R F Fan, and M V Mirkin, in Electroanalytical Chemistry, Vol.

18, A J Bard, ed., Marcel Dekker, New York, 1994, pp 243–373

3 M Arca, A J Bard, B R Horrocks, T C Richards, and D A Treichel, Analyst

119:719–726, 1994.

4 M V Mirkin, Mikrochim Acta 130:127–153, 1999.

5 A J Bard, F.-R F Fan, and M V Mirkin, in The Handbook of Surface

Imaging and Visualization, A T Hubbard, ed., CRC, Boca Raton, Fl, 1995,

pp 667–679

6 A J Bard, F.-R F Fan, and M V Mirkin, in Physical Electrochemistry:

Principles, Methods and Applications, I Rubinstein, ed., Marcel Dekker, New

York, 1995, pp 209–242

7 A J Bard and L R Faulkner, Electrochemical Methods, Wiley, New York,

1980, p 298

8 C Amatore, in Physical Electrochemistry: Principles, Methods and

Applica-tions, I Rubinstein, ed., Marcel Dekker, New York, 1995, pp 131–208.

9 A J Bard, J A Crayston, G P Kittlesen, T V Shea, and M S Wrighton,

Anal Chem 58:2321, 1986.

10 A T Hubbard and F C Anson, in Electroanalytical Chemistry, Vol 4, A J.

Bard, ed., Marcel Dekker, New York, 1970, pp 129–214

11 R C Engstrom, M Weber, D J Wunder, R Burgess, and S Winquist, Anal.

Chem 58:844, 1986.

12 R C Engstrom, T Meaney, R Tople, and R M Wightman, Anal Chem 59:

2005, 1987

13 R C Engstrom, R M Wightman, and E W Kristensen, Anal Chem 60:652,

1988

14 A J Bard, F.-R F Fan, J Kwak, and O Lev, Anal Chem 61:132, 1989.

15 J Kwak and A J Bard, Anal Chem 61:1221, 1989.

16 J Kwak and A J Bard, Anal Chem 61:1794, 1989.

17 R M Wightman and D O Wipf, in Electroanalytical Chemistry, Vol 15, A J.

Bard, ed., Marcel Dekker, New York, 1989, pp 267–353

18 M I Montenegro, M A Queiro´s, and J L Daschbach, eds., Microelectrodes:

Theory and Applications, Kluwer Academic Publishers, Dordrecht, 1991.

19 J Heinze, Angew Chem Int Ed 32:1268–1288, 1993.

20 R J Forster, Chem Soc Rev., 289–297, 1994.

21 C G Zoski, in Modern Techniques in Electroanalysis, P Vanysek, ed.,

Wiley-Interscience, New York, 1996, pp 241–312

22 P R Unwin and A J Bard, J Phys Chem 96:5035, 1992.

23 A J Bard, G Denuault, R A Friesner, B C Dornblaser, L S Tuckerman,

Anal Chem 63:1282, 1991.

24 M V Mirkin, T C Richards, and A J Bard, J Phys Chem 97:7672, 1993.

25 C Wei and A J Bard, J Electrochem Soc 142:2523, 1995.

26 M V Mirkin, F.-R F Fan, and A J Bard, Science 257:364, 1992.

27 M Pyo and A J Bard, Electrochim Acta 42:3077, 1997.

28 D Wipf, Colloids Surfaces A: Physicochem Eng Aspects 93:251, 1994.

29 B R Horrocks, M V Mirkin, and A J Bard, J Phys Chem 98:9106, 1994.

30 P R Unwin and A J Bard, Anal Chem 64:113, 1992.

31 M V Mirkin and A J Bard, J Electrochem Soc 139:3535, 1992.

32 C Wei, A J Bard, and M V Mirkin, J Phys Chem 99:10633, 1995.

33 M Tsionsky, A J Bard, and M V Mirkin, J Phys Chem 100:17881, 1996.

34 B R Horrocks, M V Mirkin, D T Pierce, A J Bard, G Nagy, and K Toth,

Anal Chem 65:1213, 1993.

35 T Solomon and A J Bard, Anal Chem 67:2787, 1995.

36 Y Shao and M V Mirkin, Anal Chem 70:3155, 1998.

INSTRUMENTATIONDavid O Wipf

Mississippi State UniversityMississippi State, Mississippi

I INTRODUCTION

A scanning electrochemical microscope is a scanning probe microscope(SPM) The scanning electrochemical microscopy (SECM) instrument nec-essarily resembles other SPM instruments, but differences in the probe tipand signal lead to differences in design and capabilities Differences alsoarise from the larger amount of research and engineering development inthe commercially successful atomic force and scanning tunneling micro-scopes The majority of the SECM instruments in use today are custom-built by the investigator Although a commercial instrument dedicated toSECM has appeared on the market, both the commercial and ‘‘home-made’’SECM instruments are less highly engineered than their other SPM cousins.Thus, there is still much opportunity for individual investigators to appre-ciate the design of SECM instruments and to make significant progress inSECM development

This chapter discusses the components of the SECM instrument ning with an overview of the major components, the discussion considersdifferent choices in instrument construction and their effect on performance.The design of a commercial instrument is discussed in light of the range ofchoices presented in the overview Several instrumental approaches to theimportant problem of ‘‘constant-current’’ imaging are discussed and evalu-ated Further improvement in SECM instrumentation will likely involve use

Begin-of ever-smaller tips in order to improve image resolution, and some practicalproblems related to the use of small imaging tips are discussed Finally,some designs for construction of useful auxiliary equipment for SECM arepresented

FIG 1 An illustration of the SECM instrument.

II OVERVIEW OF THE SECM APPARATUS

The illustration of an SECM instrument shown in Figure 1 outlines thediscussion in this section An important aspect of the SECM is the position-ing system, which includes the positioning elements, translator stages, andmotor controllers Equally important is the data acquisition system, whichbegins with use of a potentiostat or electrometer to amplify the probe signal.After amplification, the signal is digitized with an analog-to-digital converter(ADC) and stored on a computer Computer software is required to controlthe positioning and data acquisition system as well as to display and analyzethe SECM data Other important parts of an SECM are a probe mountsystem, video microscope, and vibration isolation

A Positioners and Translators

Accurate and reproducible positioning of the probe in three dimensions is

an important design element in SECM An SECM will typically allow

move-ment in three orthogonal directions: x, y, and z Ideally, the positioning

elements for the SECM will allow a probe to move at desired scan rate (e.g.,

␮m/s) over a given range In addition, the positioner axes of motion are

ideally decoupled, and movement of one axis will not produce movement

in the other axes In practice, the positioning system used will only achievethese goals over a limited range of movement and scan speeds

The smallest practical scan range is set by the SECM tip size Sinceonly image regions of tip size or larger will contain unique surface infor-mation, at least one image dimension should be significantly larger than the

tip diameter Thus, a minimum scan range is about 2 d (d = tip electrode

diameter) A maximum scan range is set by physical limitations of the sitioning device and, perhaps, by time or computer memory limitations Themaximum scan size is also limited by the maximum scan rate of the posi-tioner, which again will depend on physical limitations of the positioner.The stability of the sample, tip, and solution as well as the patience of theoperator will set the minimum scan rate One hour is required to complete

po-a 100 d ⫻ 100 d image at a scan rate of 8.3 d/s, assuming that data points are collected at intervals of one-third d to avoid aliasing artifacts Positioning

accuracy and precision should also be considered For most imaging

exper-iments, a lateral position error of 0.1 d or less is sufficient Vertical accuracy

is more important than lateral accuracy, and errors of less than 0.01 d are

desirable

Most SECM experiments use tips with 1–25 ␮m diameters Thus, apositioner for these probes should be able to scan regions of 100–1000␮msquare at scan rates up to 50 ␮m/s For these conditions, motorized posi-tioners are suitable With smaller probes, the greater accuracy of piezoelec-tric tube or tripod positioners of the type found in STM and AFM instru-ments is required

1 Motorized Positioners

Many SECM designs employ Burleigh Instruments ‘‘Inchworm’’ motors toprovide the large lateral scan range required for 1␮m and larger tips (1–3) The Inchworm positioner is a linear motor in which three piezoelectricelements act to move a central shaft (4) The two end elements are alternatelyclamped and disengaged from the shaft, while a center element expands andcontracts At the start of a movement cycle, the center element is fullycontracted, the right element is clamped to the shaft, and the left element isunclamped The center element is expanded by application of a staircasevoltage ramp to propel the shaft to the right When the center element isfully extended, the left element is clamped and the right element is un-clamped The center element is now contracted by a staircase ramp of op-posite slope At full contraction, the cycle repeats Because of the staircaseramp, the Inchworm moves in discrete steps of about 4 nm The centerelement can expand by 2␮m and so the clamping occurs at 2␮m intervals.The principal advantage of the Inchworm positioner is the lack of traditional

magnetic motors, which eliminates rotating parts and gears Other tages include backlash elimination on reversal of motion and no power dis-sipation when the motor is stationary Once stopped, the Inchworm shaft isfirmly clamped so no creep or vibration is transmitted to the probe Therange of motion is 2.5 cm or more and is limited only by the shaft length,thus the Inchworm can act as both a ‘‘coarse’’ and ‘‘fine’’ positioning ele-ment Disadvantages of the Inchworm are the higher cost due to a proprietarysupplier and the specialized motor controllers The Inchworm motors andcontroller also require some maintenance The author’s experience is thatactively used motors will require factory service at 1- to 2-year intervals Aspare motor is recommended Another quirk of the Inchworm motor is thatclamping (called a click in some reports) produces two objectionable effects.The probe does not move smoothly at the clamp position, which produces

advan-an artifact in the probe signal (3) This is especially noticeable in distance curves In addition, the end of the probe may move as much as⫾1

current-␮m in a direction perpendicular to the Inchworm shaft at the clamp Thus,directly mounting the probe to the Inchworm motor may cause problemsand even tip crashes with the smaller probes A factory modification is avail-able to minimize this motion

Stepper-motor positioners can be less expensive (but not always) thanthe Inchworm motors and are available in almost an infinite variety ofchoices Nearly every optics supplier (such as Newport, Melles Griot, etc.)supplies positioning systems that are suitably precise for use with SECM.Several groups have used stepper motors successfully for SECM positioning(5–8) Stepper motor systems rely on gears to transfer rotary to linear mo-tion With any gear system, backlash limits the positioning accuracy, partic-ularly for imaging applications, which requires many scan reversals Themotors consume power while stationary and a concern is the possibility ofelectrical interference at the tip

2 Continuous Positioners

Continuous positioners are capable of high-resolution continuous motion.Most positioners of these types are based on lead zirconium titanate (PZT)piezoelectric ceramic (4) Two common configurations for piezoelectric SPMpositioners are tripods or tube scanners Tripod scanners are simply threeorthogonal ‘‘sticks’’ of piezoelectric material The probe is attached to thevertex of the tripod (9) Application of voltage to any one of the sticks movesthe probe in that axis direction It also moves the probe a small, but pre-dictable, amount in the other two axes Tube scanners use a piezoceramictube The tube has five electrodes, one coating the entire inner wall and four

90⬚ segments on the outer wall (10) The probe is attached parallel to thetube axis The tube bends by applying a voltage to opposing electrode seg-

ments: each of the opposing two electrodes are polarized either positively

or negatively with respect to the inner electrode As the tube bends, theprobe tip is moved laterally Vertical motion is produced by polarizing allouter segments simultaneously Greater vertical motion is often achieved byusing a composite scanner with separate piezoelectric elements for verticaland lateral motion (11,12) A disadvantage of the tube scanner is that lateralmotion also generates some vertical movement The vertical component ofthe motion is a complicated function of the lateral motion, but it is roughlyarc shaped The vertical motion can be minimized with careful design but

is unavoidable Because of their lower cost and greater vibration immunity,tube scanners are often chosen over tripod scanner for SPM positioners.The distance resolution of both the tube and tripod scanners is limited

by the precision of the driving voltage A typical piezoelectric positionermay move 10 nm per applied volt; thus, subangstrom motions are feasible

In contrast, micrometer-scale movements will require hundreds of volts Thismeans that the practical maximum lateral movements are about 100 ␮m.Vertical movements of tube scanners are constrained to a few micrometers.The limited movement range is a handicap for using tips larger than a fewmicrometers in SECM because current-distance curves cannot be collectedand the scan range is too small Note that the scan range can be increased

by use of levers to provide a mechanical advantage but resolution is portionately degraded due to increased position noise

pro-Tube and tripod scanners have several nonideal properties The behavior

of piezoelectric scanners is well covered in the literature (4,13,14), and only

a brief outline of these problems are given here Hysteresis, linearity, and

‘‘creep’’ are the most critical Hysteresis causes the piezo extension versusvoltage curve to differ depending on whether the voltage is increased ordecreased In SPM applications, nonlinearity and hysteresis are compensated

by calibration with a well-characterized sample (such as a diffraction ing) and by applying corrections during scanning or during data analysis(15,16) Because of hysteresis, data are only calibrated when acquired in thecalibrated direction and scan rate Because vertical movements cannot beconfined to single directions and scan rates, they are uncalibrated and datainterpretation is complicated Creep is the slow relaxation to a steady-stateextension following a rapid voltage change Creep is apparent when a largechange in driving voltage occurs For example, when zooming from a largescan area to a small feature, creep causes a slow drift of the area scanned,and several minutes may be required to equilibrate the scanners

grat-Several SPM instrument manufacturers have corrected scanner ideality by incorporating sensing devices in the piezoelectric elements Thismodification allows real-time feedback control of the extension and elimi-nates many of the above problems A number of different sensor types are

non-available Strain-gauge sensors indicate extensions by a resistance change,capacitance sensors detect the capacitance change between a fixed and mov-ing element, and optical sensors use interferometry to detect a positionchange The devices become desirable when large scan areas are used andwhen high linearity is required (as in metrology applications) Because thesensors ultimately limit the resolution of the piezoelectric positioner and canadd noise, they are not often used in atomic resolution images in SPM.However, in SECM applications the larger distances usually used are notaffected by the decreased resolution.

Micrometers are also useful positioning elements Although manuallyoperated micrometers are not suitable for most imaging applications, theyprovide an inexpensive way to acquire current-distance curves and performsubstrate modifications (17,18) Submicron positioning is available with dif-ferential micrometers Motorized micrometers are widely available, and mi-crometers with DC servo motor drives in closed-loop operation (see below)can give excellent results

3 Motor Controllers

A motor controller contains the necessary electronics to drive a motor andprovides a simplified electrical interface between the operator and the po-sitioning elements Piezoelectric and motorized positioners require largepower levels and complicated drive signals that cannot be readily accom-modated by the computer running the experiment Motor controllers provideall the necessary electronics for driving the positioning elements, and thecomputer may only be required to supply signals to start and stop thepositioning

Controllers provide ‘‘closed-loop’’ or ‘‘open-loop’’ positioning Anopen-loop controller does not verify that a movement has occurred at thespeed or distance desired Typically, a calibration scheme is used to deter-mine the amount of movement produced by the positioner As an example,four TTL signals are used to control each axis with the Burleigh Inchwormopen-loop controller: axis select, forward/reverse, halt/run, and clock Tomove a motor, the computer generates a set of TTL levels to select the axis

and direction (i.e., z forward) and then provides a set of pulses to the clock

input The number and frequency of the pulses determines the amount ofmovement Each axis must be individually calibrated to determine the av-erage amount of movement per pulse as a function of both scan directionand speed Calibration is accomplished either by imaging a well-character-ized feature or by generating enough pulses to allow measurement of themovement with a caliper Position is calculated with a computer programusing the number of clock pulses and the calibration curve More sophisti-cated open-loop controllers incorporate microprocessors to maintain position

and calibration information These controllers might accept higher levelcommands such as ‘‘move 100 micrometers’’ or ‘‘return to origin,’’ but theystill rely on a calibration procedure.

Closed-loop controllers use a feedback process that relies on informationabout the actual probe motion As discussed above, some piezoelectric po-sitioners use strain gauges for determining the actual extension of the pie-zoelectric element Motorized positioners use other methods A commonapproach is to incorporate a linear distance encoder in the translation stage.Usually the encoder consists of a pair of ruled patterns printed on a trans-parent strip With the transparent strip attached to positioner, a pattern oflight and dark is generated as the ruled pattern moves past a stationaryphotodetector A microcomputer counts the number of transitions and gen-erates an absolute movement By arranging the patterns so that the marks

of the paired patterns are 90⬚ out of phase with each other (i.e., ‘‘in rature’’), motion direction is also determined The resolution of the encoder

quad-is determined by the precquad-ision in positioning the marks on the strip phisticated closed-loop controllers used in micropositioning applications alsouse the phase differences of the photodetector pairs to increase the encoderresolution Since the encoder ‘‘knows’’ the absolute position of the transla-tion stage, any error between the programmed movement and the actualmovement can be corrected by the controller

So-Rotating motors, such as a stepper motor, use a similar encoding schemebut use rules marked on the rotating shaft or pairs of slotted disks (mostcomputer mice use slotted disks) There are other methods of monitoringmovement, such as optical interferometry, but all rely on instantly correctingthe difference between the programmed and measured movements The mi-croprocessor in closed-loop controllers is able to understand high-level com-mands, thus, closed-loop controllers are usually much simpler to use forprecision positioning Some drawbacks to closed-loop controllers includeexpense, which may double or triple the cost of the open-loop system Inaddition, the resolution of the controller is limited by the encoding method.For example, the Burleigh Inchworm system is available with resolutionsfrom 0.02 to 0.1␮m, with cost greatly increased at higher resolution Ac-curacy in closed-loop systems is related to encoder resolution with accuracybeing at least 10–100 times worse, that is, an encoder with resolution of0.05 ␮m may only have a guaranteed accuracy of ⫾1 ␮m Finally, theclosed-loop system is a feedback system, and, like all feedback systems, itmay become unstable and cause the positioner to oscillate or overshoot thedesired position While not particularly likely with higher quality controllers,imaging difficulty or tip crashes may result

Controllers for piezoelectric tube or tripod scanners require ity, low-noise voltage amplifiers Since piezoelectric materials have a large

high-stabil-electrical capacitance, they also require amplifiers that produce the highpower required for wide-bandwidth control of reactive loads Closed-loopcontrollers mate an amplifier with a feedback loop and accept input fromstrain gauge or capacitive positioning sensors For simple low-frequencyopen-loop positioning, commercial amplifiers may be more costly than nec-essary An amplifier can be made using high-voltage operational amplifiersfrom companies such as Apex or Burr-Brown Including the cost of a high-voltage power supply, a home-built amplifier can cost less than one tenth of

a commercial unit Apex, in particular, has helpful product literature ondesigning amplifiers for piezoelectric actuators (19)

B Translation Stages

Most of the positioners mentioned above are coupled to a translation stage.Stages are primarily used to constrain motion from a positioner to a singleaxis and to provide a mount for the probe or sample In considering a trans-lation stage, the following sources of positioning errors must be considered.Runout error refers to the amount of ‘‘off-axis’’ linear motion; for example,

an x-axis translator has out-of-plane motion (straightness) in the z-axis and in-plane motion (flatness) in the y-axis Tilt and wobble are angular measures

of the off-axis motion, and these are specified by the three orthogonal ponents: roll, pitch, and yaw Abbe error is a consequence of tilt and wobble

com-It is produced by amplifying any translator angular error by the distancebetween the plane of travel and point of measurement For example, theangular error in the stages is greatly amplified at a probe mounted several

tens of centimeters above the x and y translation stages Another significant

error results from ‘‘play,’’ which is a consequence of looseness of badlymade or worn translator parts Backlash is often the result of play in thetranslator Friction and stiction (static friction) are also important since fric-tion varies with motion velocity, leading to velocity-dependent errors Stic-tion limits the translators ability to make small, incremental motions.Several translation stage designs are available Crossed-roller bearingstages provide high-quality, long-range positioning (at a high cost) and havelow friction and stiction qualities In comparison to the less expensive ball-bearing stages, crossed-roller stages have lower angular errors (100 ␮radcompared to 150␮rad) and are less susceptible to contamination For smallerpositioning ranges, flexure stages are ideal The stage is attached to stiff flatsprings, and motion is produced by elastic deformation of the spring Sincethere is no friction or stiction, very small and rapid motions are possible Inaddition, a mechanical advantage can be built into the stage, amplifying thepositioner motion Note that the two axes do not move independently in aflexure stage, and coupled motion between axes becomes objectionable at

FIG 2 Examples of two types of tip mounts: (A) home-built; (B) commercialmicropipet holder.

large excursions For this reason, flexure stages are not suitable for ments much greater than 2000␮m

move-C Probe Mounting

Although many types of probes are used in SECM experiments, as a rule,vibration is minimized by making probes and probe mounts as small andrigid as possible Careful design of the probe and mount is necessary forSPM experiments at nanometer resolution (14,20) Since most SECM ex-periments use probes with diameters of several micrometers, larger vibra-tions can be tolerated and probe mounting becomes less critical The twomounts shown in Figure 2 are suitable for larger SECM tips Figure 2Ashows a probe holder composed of a section of aluminum angle bracketthreaded to accept a Kel-F insert Two attachment holes in the bracket permitthe mount to be bolted to the vertical translation element The Kel-F insert

is a threaded section of rod that screws into the bracket The insert contains

a central hole that accepts the SECM probe and an intersecting horizontalhole tapped to accept a nylon screw, which is a clamp for the probe Asidefrom its excellent chemical stability, the Kel-F insert is used to electricallyisolate the probe and the translation stages A thin electrolyte film on theprobe body can easily wreak havoc if allowed to contact a nonisolated con-

ductive element A nylon screw is preferred over a metal screw as the ing element since it is more difficult to break an electrode if the clamp isovertightened The clamp also allows the probe to slip up in case of a tipcrash.

clamp-Commercially available probe holders are also useful These mounts aresold as holders for micro-pipets and are available from electrophysiologysuppliers A pipet holder is illustrated in Figure 2B The probe is laid in agroove and held by a screw and washer This mount is not as rigid as theprevious example, but it is convenient for use with fragile electrodes Forexample, the holder can be used for probe preparation steps, such as polish-ing or etching, without removing the probe from the holder This can min-imize electrode breakage caused by manual handling

The two mounts described above are too large for probes with meter-sized tips In addition to vibration problems, the mounts would not fitonto piezoelectric tube or tripod scanners Small SECM probes are verysimilar to electrochemical STM probes These probes are often mounted byplugging the probe into a small electrical pin-socket glued to the tube scan-ner The socket thus serves as a mechanical and electrical contact to theprobe

nano-D Vibration Isolation

Vibration increases noise and decreases lateral and vertical resolution Themost effective vibration isolation takes place in the instrument design Agoal is to keep vibrations resonances out of the band of frequencies usedfor data collection and feedback control, i.e., about 1–1000 Hz Minimiza-tion of size and maximization of rigidity shifts vibration resonances to highfrequencies (14,20) In addition, locating the SECM away from vibrationsources such as air vents, pumps, or heavy machinery is advised Note thatbypassers are a source of vibrational and electrical interference and locationswith foot traffic are inadvisable

Additional vibration isolation may not be required for SECM ForSECM imaging with probes of 10 ␮m and larger, isolation is not critical.For smaller tips, some attempt should be made to determine if vibrationisolation is needed One simple check is to observe an increase in noise asthe probe approaches the surface in a feedback experiment Another check

is to determine the frequency distribution of vibrations Use a lock-in plifier to examine frequencies from 10 mHz to 10 kHz with the probe po-sitioned as close as possible to an electrode surface in a positive feedbackconfiguration Alternatively, perform a Fourier transform of the probe noisesignal Any peaks in the signal versus frequency plot are candidates forvibrational resonances Discriminating vibrational and electrical noise is pos-

am-sible by moving the probe slightly away from the electrode surface tional peaks will decrease at larger probe-substrate separations.

Vibra-Isolating the apparatus from the external environment is an effectivemeans of reducing vibration, since energy from the environment is not avail-able to excite mechanical resonances For vibrations transmitted through air

by acoustic or convective motion, the instrument is isolated by placing it in

a closed box lined with acoustical foam or lead sheets Energy from thebuilding floor is transmitted to the instrument through the support Trans-mission of frequencies larger than 10 Hz is effectively minimized by usingpneumatic or spring supports Elastic bands (‘‘bungees’’) are often used as

an inexpensive spring device The instrument is placed on a metal platesupported by three or four bungee lines hanging from the ceiling or othersupport About 10 kg of mass is added to the plate to extend the bungees

to about one half of their maximum extension Adjust the load and bungeelength until the natural oscillation frequency of the suspended plate is about

1 Hz

Air-tables are also effective isolation devices These are available mercially and minimally consist of a rigid top supported by pneumatic cush-ions in the legs The exceptional rigidity found in optical tables is unnec-essary and less expensive workstation-type tables can be used A simplehomemade table often works as well as a commercial table This consists

com-of a motorcycle or bicycle inner tube supporting a 1 cm thick aluminumplate The plate should be large enough to completely cover the inner tubewith additional space on the edges for wooden support blocks, which supportthe plate during experiment set-up Add mass for balance and adjust the tubepressure to give a 1–10 Hz resonant frequency

High-frequency resonances can be excited by the movements of thetranslators These can be damped by using rubber pads or foam under theinstruments Natural rubber or Sorbothane are often recommended for thispurpose

E Signal Transduction and Amplification

The probe in SECM produces a signal that must be transduced and amplifiedprior to recording At a voltammetric tip, electrolysis of either a mediator

or a substrate-produced substance produces a faradaic current signal At apotentiometric tip, the activity of a solution phase species generates a voltagesignal

The requirements for accurate recording of these two types of signalsare quite different and lead to different equipment requirements A sensitivepotentiostat is required for a voltammetric probe and an electrometer isneeded for a potentiometric probe

FIG 3 Electrical schematic of a potentiostat circuit.

1 Potentiostat and Current Transducer

A potentiostat is used for most voltammetric measurements The potentiostatuses an electrical feedback loop to control the potential of the workingelectrode (the electrode at which the reaction of interest occurs) with respect

to a reference electrode, even in the presence of ohmic drop A third trode, the auxiliary, is used to supply the current flowing at the workingelectrode The use of a third electrode eliminates current flow through thereference electrode, permitting smaller reference electrodes and more accu-rate potential control A simplified diagram of a potentiostatic circuit is given

elec-in Figure 3 (21) Most modern electrochemical equipment employ versatilehigh-input impedance differential amplifiers known as operational amplifiers(OAs) as circuit elements In the potentiostat circuit, OA1 provides the aux-iliary electrode with the required voltage and current to maintain the desiredpotential difference between the reference and working electrodes OA2 is

a buffer amplifier (voltage follower) that prevents significant current drawthrough the reference electrode and outputs a low-impedance measure of thereference electrode potential In operation, the amplifier OA1 adjusts thecurrent and voltage at the auxiliary electrode to whatever is necessary tominimize the voltage difference at the inputs The negative feedback loopthat exists between the output and input of OA1 includes OA2, the referenceand auxiliary electrode, and the solution resistance between reference andauxiliary Thus, the voltage at the auxiliary electrode will be the invertedsum of the reference electrode voltage (ER), any externally applied voltage(EW), and the ohmic voltage arising from current flow through the solutionresistance This means that, if the reference electrode is close to the workingelectrode, the effect of the ohmic drop is minimized and the potential dif-ference between the working electrode and reference electrode is maintained

or ‘‘clamped’’ close to EW Clamping the voltage at EWis the basis for the

name ‘‘voltage clamp’’; a term used outside the electrochemical literature

as a more descriptive name for the potentiostat circuit

The working electrode is maintained at the circuit common potential(which may not be earth ground) in the potentiostat circuit of Figure 3 Thecurrent-transducer amplifier, OA3, maintains the electrode potential at com-mon potential and provides a voltage output proportional to the current input

As before, the amplifier uses negative feedback to minimize the potentialdifference between the inputs by adjusting the output In the process, theinverting input (⫺) is set to be virtually the same as the potential of thenoninverting input (⫹), and thus the inverting input is at ‘‘virtual ground.’’Note that the output voltage across the feedback resistor, RF, produces acurrent to oppose the current flowing into the input and changing RFchangesthe signal gain An alternate method of maintaining the working electrode

at the common potential is to connect a small-valued measuring resistorbetween the working electrode and common By measuring the potentialdrop that develops across the resistor, the current flow can be calculatedfrom Ohm’s law

Although the circuit in Figure 3 could be used in voltammetric ments, most potentiostat circuits are considerably more sophisticated Ad-ditional circuitry is added to provide damage protection from excessive sig-nal inputs or outputs, additional current or voltage range (i.e., compliance)

experi-at the auxiliary electrode, and for variable gain and filtering of the inputcurrent Some circuit designs maintain the reference electrode rather thanthe working electrode at the circuit common A floating working electrodecomplicates the design of the working electrode amplifier slightly but allows

a number of devices to share a reference electrode For example, both apotentiometric and an amperometric working electrode can use the samereference electrode if the amperometric electrode is not set at ground poten-tial Some ‘‘potentiostats’’ eliminate the auxiliary electrode and apply EWdirectly to the reference electrode OA1 and OA2 are eliminated, but thisrequires that the total current flow through the cell is sufficiently small toavoid polarizing the reference electrode Experiments with disk ultramicro-electrodes of 10␮m or smaller diameter can often use these ‘‘two-electrode’’potentiostats

2 Bipotentiostat

A bipotentiostat is simply a potentiostatic circuit designed to allow taneous potential control of two working electrodes in an electrochemicalcell A simplified schematic of a bipotentiostat circuit is illustrated in Figure

simul-4 Note that connections in this schematic are always indicated by a dot at

the intersection, and crossing lines without a dot are not connected Part of

this circuit is identical to the potentiostat circuit, with the addition of