30 The Electrochemical Society Interface • Fall 2003

he delivery of drugs into thebody, for therapeutic treat-ment, at a continuously con-trolled rate, is a general goalof the medical community.

One means of achieving controlledrelease is by molecular transport acrossthe skin, commonly referred to astransdermal drug delivery. Transdermaldrug delivery systems may be passive,in which a molecular species diffusesacross the skin, driven by a concentra-tion gradient (e.g, nicotine patch); oractive, in which an electric potentialgradient across the skin acceleratestransport via migration and/or elec-troosmosis. Active transport is referredto as iontophoresis, which has beenexplored by the medical community forover a century.1 Figure 1 shows aschematic diagram of an iontophoreticdevice in contact with the skin. Thedrug of interest is contained in a thinlayer of solution separating an electrodefrom the outer layer of skin, the stratumcorneum. An electrical current, typicallya few hundred microamperes persquare centimeter, is forced across theskin, driving the molecule into theunderlying tissues and the circulatorysystem.

Passive and active transdermal drugdelivery systems avoid difficulties asso-ciated with traditional methods of drugadministration. Many drugs cannot beadministered orally due to extensivemetabolism in the gastrointestinal tractor weak absorption into the blood-stream. Advances during the pastdecade in transdermal drug deliverysuggest that many drug molecules,including delicate peptides,2 proteins,3

and oligonucleotides4,5 can be deliv-ered across skin at a precisely controlledrate and in an essentially non-invasivefashion. Iontophoretic transport oflidocaine and fentanyl is currentlybeing employed for local and systemicanesthetic administration in clinical tri-als.6 Reverse iontophoretic transdermaltransport, in which molecules in theblood stream are transported across theskin for analytical detection, is beingdeveloped for the noninvasive moni-

Electrochemical Imagingof Molecular Transport in Skin

by Henry S. White

T

FIG. 2. Schematic diagram of the iontophoresis cell and scanning electrochemical microscope used for imag-ing fluxes across hairless mouse skin.24

FIG. 1. Schematic diagram of drug delivery from an iontophoresis cell placed on the skin.

toring of blood glucose levels by diabet-ic patients.7

The success of iontophoretic systemsdepends largely on a fundamentalunderstanding of transport pathwaysand mechanisms in skin, a highly com-plex biological membrane. Over the pastdecade, our laboratory has developedscanning electrochemical microscopy(SECM) to investigate and quantifytransport in artificial and biologicalmembranes, including both hairlessmouse8 and human skin.9 Spatially-resolved measurements of moleculartransport using SECM,10-12 with

micrometer and submicrometer resolu-tion,13 provides a means to identifymicroscopic structures in skin that areassociated with molecular transportpaths and mechanisms.14 SECM hasbeen used to quantify the individualcontributions of diffusion, migration,and convective to localized fluxes inskin15 and other membranes.16 Thiscapability has been especially useful incharacterizing transport mechanisms inskin, where the molecular flux varies sig-nificantly as a function of spatial posi-tion and as a function of the chemicalnature of the molecule being transport-

Electrochemical Imagingof Molecular Transport in Skin

The Electrochemical Society Interface • Fall 2003 31

trolyte, typically 0.1 M NaCl. In theconventional imaging mode of SECM,the SECM tip is placed in the receptorcompartment and is poised at a poten-tial such that the electroactive moleculeis oxidized or reduced as it emergesfrom pores in the membrane (Fig. 2).The magnitude of the tip current mea-sured by the SECM tip is proportionalto the local concentration of electroac-tive species above the pore opening. Inaddition, the local concentration of theredox species above the pore opening isproportional to the species flux withinthe membrane. Thus, a plot of theSECM tip current versus spatial positionprovides a direct means to visualizetransport paths in the membrane. Theresulting images are images of molecu-lar flux that directly reflect transportrates associated with physiologicalstructures in the skin. To study ion-tophoretic transdermal transport, aconstant electrical current, iapp, isapplied between two Ag/AgCl elec-trodes located on opposite sides of theskin sample (Fig. 2). This current can becontinuously varied, allowing transportrates to be studied over a wide range ofiontophoretic conditions.

Localized TransportPathways in Skin

A key issue in transdermal drugdelivery is the identification of trans-port pathways that allow molecules topass through the skin. To address thisquestion, we use SECM to image mole-cules emerging from skin as they aretransported across the sample by eitherpassive or active transport. For exam-ple, Fig. 3 shows the steady-statevoltammetric response of the tip whenit is positioned directly above a hair fol-licle opening (curve 1) and at a lateraldistance of ~150 µm from the opening(curve 2).17 In this experiment, theredox-active molecule hydroquinone(HQ) is transported across HMS by dif-fusion alone and is detected by oxida-tion at the SECM tip. The sigmoidalshaped voltammetric wave recordedabove the hair follicle (curve 1) corre-sponds to oxidation of HQ moleculesthat have diffused across the skin sam-ple through the hair follicle. The mag-nitude of the tip voltammetric currentis proportional to the local rate atwhich HQ permeates the skin.(Transport rates can be readily quanti-fied from SECM tip approach curves).16

The SECM tip current decreases to back-ground levels (curve 2) when the tip ismoved away from the pore opening,

FIG. 3. (a) Schematic drawing illustrating the SECM tip positioned directly above the opening of a hair folli-cle. (b) Voltammetric response of a 2.7 µm-radius SECM tip positioned directly above a hair follicle openingin hairless mouse skin (curve 1) and 150 µm away from the opening (curve 2). The voltammetric current cor-responds to oxidation of HQ that has diffused through the hair follicle.17

ed. A brief summary of the use of SECMfor imaging biological membranes, andsome results obtained in our studies ofiontophoretic transdermal transport, arepresented in the following sections.

Electrochemical Visualizationof Molecular Fluxes inBiological Membranes

Figure 2 shows the iontophoresis celland key components of the SECM. Ininvestigations of transdermal transport,the SECM tip (a small carbon fiber) istypically rastered across the surface of askin sample at a tip-surface separation ofa few micrometers using three piezoelec-tric (x, y, z) inchworm motors. In manyof our experiments, the skin sample isexcised from 7 week-old male hairless

mice immediately after euthanasia byCO2 asphyxiation. Hairless mouse skin(HMS) is a useful model for understand-ing transport in human skin. The factthat HMS contains very few hairs isadvantageous in SECM studies, as hairsprotruding from the skin would interferewith the motion of the SECM tip.However, in understanding the resultspresented below, it is important to notethat the hair follicles in HMS are intact,offering low resistance pathways fortransport across the stratum corneum.

In SECM studies of membrane trans-port, the sample (e.g., HMS) separates adonor solution, which contains an elec-troactive species (Az+) at millimolar con-centrations, from a receptor solution inwhich the redox species is absent. Bothsolutions contain a supporting elec-

(a)

(b)

32 The Electrochemical Society Interface • Fall 2003

demonstrating that the diffusive flux ofHQ is localized to the hair follicle. Thesites of high diffusive flux are indepen-dently identified as hair follicles using adye staining technique in which col-loidal Prussian blue is precipitated atthe opening of the follicle. Similarexperiments using small organic andinorganic redox species, with differentcharges (z = +1, 0, and –1) indicate thathair follicles in HMS act as the primaryroute for transdermal transport.

Titration of ElectroosmoticFluxes within Individual

Hair Follicles

A number of studies during the pastdecade have demonstrated that anapplied electrical current can alsoenhance the rate of transport of electri-cally neutral molecules.18-20 This resultsuggests that iontophoresis can inducean electroosmotic flow of solutionacross skin tissues. Analogous to elec-troosmotic flow in capillary elec-trophoresis and synthetic membranes(e.g., Nafion), electroosmosis in skinresults from a fixed ionic charge in theepidermal layers.

The three SECM images presented inFig. 4 demonstrate that electroosmotictransport of HQ is operative inside thehair follicle when a constant ion-tophoretic current, iapp, is appliedacross the skin sample. The SECMimages were recorded by rastering theSECM tip at a fixed height above theopening of a hair follicle while detect-ing HQ at the tip. The middle imagewas obtained at iapp = 0, correspondingto diffusion of HQ. The bottom image,obtained at iapp = 50 µA, clearly demon-strates that a positive iontophoretic cur-rent enhances the rate of moleculartransport through the hair follicle.Because HQ is electrically neutral at pH= 6.0, its transport is not directly influ-enced by the applied current. Rather,the enhancement in molecular trans-port must result from electroosmoticflow of solution inside the hair follicle.

In our experiments, the applicationof a positive iapp corresponds to theupward migration of electrolyte cations(Na+) from the donor to the receptorsolution and/or the downward migra-tion of anions (Cl-), i.e., the anode is inthe donor compartment. The directionof electroosmotic flow that is inducedby the current depends upon thepermselective properties of the skinsample. HMS, which exhibits a net neg-ative charge at pH = 6.0,21 displayscation-selective membrane properties.Thus, the majority charge carrier in skin

at pH = 6.0 is Na+. The cation permselec-tivity of HMS may arise from proteinchemistry similar to that of human skin;it is reported that the observed permse-lectivity results from a larger number ofcarboxylate (-COO-) than ammonium (-NH3

+) groups (associated with proteinamino acid residues) that reside in epi-dermal and dermal tissues.22 Theincrease in HQ flux at positive iapp (Fig.4) is consistent with the reported cation-selectivity of HMS.

When the direction of the current isreversed, i.e., when iapp is negative, thedirections of Na+ migration and elec-troosmotic flow are also reversed. In thiscase, electroosmotic flow should opposethe diffusion of HQ through the hair fol-licle, resulting in a decrease in the totalflux of HQ. SECM images recorded atnegative iapp (top image, Fig. 4) showthat the HQ transport rate is indeedreduced when electroosmotic flowopposes the diffusion of HQ.

The direction of electroosmotic flowis determined by the acid-base propertiesof the epidermal and dermal tissues. Theisoelectric point, pI, of skin, i.e., the pHat which skin is electrically neutral, hasbeen measured for bulk samples to bebetween 4.5 and 4.6.23 HMS is negative-

ly charged at pHs above the pI and, thus,exhibits cation selectivity at pH = 6.0, asdemonstrated by the transport studiesdescribed above. At pHs below the pI,skin is expected to have a net positiveelectrical charge, and should exhibitanion permselectivity. Thus, it should bepossible to reverse the direction of elec-troosmotic flow by lowering the pH ofthe contacting solutions from a valueabove the pI to a value below the pI.

We repeated the iontophoretic exper-iments shown in Fig. 4 at pHs above,below, and near the pI of HMS to deter-mine if the electroosmotic transportcould be reversed in the hair follicle bysimply adjusting the pH of the contact-ing solution. The dependence of elec-troosmotic flow on pH may be quanti-fied using the enhancement factor, E,which is defined as the ratio of the flux-es of HQ under iontophoretic (Niont)and diffusive (Ndiff) conditions, i.e., E =Niont/Ndiff. Figure 5 shows E as a func-tion of pH for iapp = -50 and 50 µA.Figure 5 demonstrates that E is equal tounity at pH ~ 3.5, indicating that elec-troosmotic flow is not operative near thereported pI of HMS. This finding, inaddition to the sigmoidal shape of the Evs. pH curves, suggests that the direction

FIG. 4. SECM images of HQ emerging from the opening of a hair follicle as a function of the applied ion-tophoretic current at pH = 6.0. The center image corresponds to diffusion of HQ from the donor to the recep-tor solution. The lower image corresponds to positive current (iapp = 50 µA), inducing electroosmotic flow inthe same direction as diffusion. The upper image corresponds to a negative current (iapp = -50 µA), resultingin electroosmotic flow opposing diffusion.17

The Electrochemical Society Interface • Fall 2003 33

entrance, resulting in a decrease in thetip current as the tip is rastered acrossabove pore. Analogous to the increasein tip current measured in convention-al SECM images (Fig. 4), the decrease incurrent in RIM signifies a local region inthe membrane where the flux is large.In principle, SECM-RIM has an inher-ently lower sensitivity than conven-tional imaging, because the signal ismeasured relative to a large back-ground. However, SECM-RIM providesa means to image molecular transportinto biological membranes. This capa-bility is necessary for monitoring theuptake of molecular species into singlecells and biological tissues.

We have succeeded in obtainingSECM-RIM images of both syntheticmembranes and HMS samples. Forexample, Fig. 6 shows an SECM-RIMimage of a 16 µm radius hair follicle inexcised HMS. The image corresponds tothe diffusion of acetaminophen intothe hair follicle (from a 5 mM solution),demonstrating the first step toward invivo imaging of transdermal trans-port.24 A quantitative theory of SECM-RIM imaging has also been recentlydeveloped, which allows electroosmot-ic velocities and molecular fluxes to bequantified directly from the images.25 �

Acknowledgments

The success of this project reflectsthe creativity of three of the author’sformer PhD students: Bradley D. Bath,Eric R. Scott, and Olivia D. Uitto, as wellas the generous support provided byALZA, Inc.

References

1. S. Leduc, Ann d’Electrobiol., 3, 545 (1900).2. K. S. Bhatia and J. Singh. Pharm. Res., 15,

1857 (1998).3. S. Mitragotri, D. Blankschtein, and R.

Langer, Science, 269, 850 (1995).4. R. H. Brand, A. Wahl, and P. L. Iversen. J.

Pharm. Sci., 87, 49 (1998).5. K. Oldenburg, K. T. Vo, G. A. Smith, and

H. E. Selick. J. Pharm. Sci., 84, 91 (1995).6. S. K. Gupta, M. Southam, G. Sathyan, and

M. Klausner. J. Pharm. Sci., 87, 976 (1998).7. J. A. Tamada, N. J. V. Bohannon, and R. O.

Potts, Nature Med., 1, 1198 (1995).8. E. R. Scott, H. S. White, and J. B. Phipps,

Solid State Ionics, 53-56, 176 (1992).9. O. D. Uitto and Henry S. White, Pharm.

Res., 20, 646-652 (2003).10. E. R. Scott, H. S. White, and J. B. Phipps,

Anal. Chem., 65, 1537 (1993).11. E. R. Scott, A. I. Laplaza, H. S. White, J. B.

Phipps, Pharm. Res., 10, 1699 (1993).12. E. R. Scott, J. B. Phipps, H. S. White, J.

Invest. Dermatology, 104, 142 (1995).13. J. V. Macpherson, P. R. Unwin, Anal.

Chem., 72, 276 (2000).

FIG. 5. Flux enhancement factor, E, at positive and negative iontophoretic currents as a function of pH. E isequal to unity at pH ~ 3.5, corresponding to the pH of the epithelial cell layers comprising the hair folliclestructure.17

FIG. 6. (Top) Schematic of the reverse-imaging mode (RIM) used to image molecule entering the skin. Thearrows represent the direction of transport of an electroactive molecule, Az+, into the skin tissue. Qualitativesketch of the tip current as it is scanned above the pore. (Bottom) SECM image of acetaminophen above a16.2 mm radius hair follicle in hairless mouse skin. The donor compartment contained 5 mM aceta-minophen.24

and magnitude of electroosmotic flow inthe hair follicle is determined by theacid-base equilibria of the protein aminoacid residues located in the epithelialcells comprising the hair follicle struc-ture.

Toward In Vivo Imaging ofDrug Transport

We are currently developing newSECM methodologies for investigatingtransdermal transport in vivo.24,25 In this

experiment, it is necessary to have boththe solute molecule and SECM tip posi-tioned in the solution that is contactingthe exterior surface of the membrane ortissue being imaged. We refer to thisoperation of SECM as reverse imagingmode (RIM)—the tip is scanned on thedonor side of the membrane in order toobserve molecules entering into themembrane (Fig. 6). In SECM-RIM, thediffusion of molecules into the poredepletes solute in the donor solutionimmediately adjacent to the pore

34 The Electrochemical Society Interface • Fall 2003

14. B. D. Bath, H. S. White, E. R. Scott,Scanning Electrochemical Microscopy; A. J.Bard, M. Mirkin, Eds.; John Wiley: NewYork, 2002.

15. B. D. Bath, E. R. Scott, J. Bradley Phipps,and H. S. White, J. Pharm Sci., 18, 1537(2000).

16. B. D. Bath, H. S. White, and E. R. Scott,Anal. Chem., 72, 433 (2000).

17. B. D. Bath, H. S. White, and E. R. Scott,Pharm. Res., 17, 471 (2000).

18. M. J. Pikal, Adv. Drug Deliv. Rev., 9, 201(1992).

19. M. B. Delgado-Charro and R. H. Guy,Pharm. Res., 11, 929 (1994).

20. S. M. Sims and W. I. Higuchi, J. MembraneSci., 49, 305 (1990).

21. M. J. Pikal and S. Shah, Pharm. Res., 7, 213(1990).

22. R. R. Burnette and B. Ongpipattanakul, J.Pharm. Sci., 76, 765 (1987).

23. A. Luzardo-Alverez, M. Rodríguez-Fernández, J. Blanco-Méndez, R. H. Guy,and M. B. Delgado-Charro, Pharm. Res.,15, 984 (1998).

24. O. D. Uitto and H. S. White, Anal. Chem.,73, 533 (2001).

25. O. D. Uitto, H. S. White, and K. Aoki, Anal.Chem., 74, 4577 (2002).

About the Author

HENRY S. WHITE is a Professor inthe Department of Chemistry at theUniversity of Utah. In addition tostudies of iontophoresis across skin,his research interests include electro-analytical chemistry, magnetic field-induced transport, oxide films andcorrosion, and electrochemical phe-nomena at electrodes of nanometerdimensions. He received the 2002Faraday Medal from theElectrochemical Group of the RoyalSociety of Chemistry (London) andis an Associate Editor of the Journalof the American ChemicalSociety. He may be reached atwhite@chem. utah.edu.