Figure 3. SEM image of an integrated singlecrystal silicon cantilever and tip which has anend radius of 2 to 10nm. Tips for AFM aretypically made of silicon or silicon nitride. Bar = 100µm.

(STEM).1, 2 In 1942, Zworkin et. al.developed the first SEM for bulksamples. This configuration containsmany of the basic principles of today’s SEMs.2, 3 Cambridge ScientificInstruments produced the firstcommercial instrument in 1965. A number of improvements haveoccurred since this time, resulting in an increase in resolution from 50nm in 1942 to ~0.7nm today. Besides the development of morphologicalimaging, the SEM has beendeveloped to detect signals which are used to determine compositionalinformation, such as X-rays, back-scattered electrons, cathodoluminesce,Auger electrons, and specimen current.

The development of the AFM waspreceded by the development of theScanning Tunneling Microscope (STM)in 1981 at IBM Zurich ResearchLaboratory by Binnig and Rohrer.4

Its ability to view the atomic lattice of a sample surface earned theinventors the Nobel Prize in Physics in 1986. Although the STM providessubangstrom resolution in all threedimensions, it is limited to conductiveand semiconductive samples. To imageinsulators as well as conductors, theAtomic Force Microscope (AFM) wasdeveloped in 1986,5 and the firstcommercial AFMs were produced in1989 by Digital Instruments, nowVeeco Instruments.

AFM provides three-dimensionalsurface topography at nanometerlateral and subangstrom verticalresolution on insulators and conductors.From this beginning, the field ofScanning Probe Microscopy (SPM)was born which consists of a family of techniques that involves scanning asharp tip across the sample surfacewhile monitoring the tip-sampleinteraction to form a high resolutionimage. Although the AFM has becomethe most commonly used form of SPM,

many other SPM techniques have beendeveloped which provide informationon differences in friction, adhesion,elasticity, hardness, electric fields,magnetic fields, carrier concentration,temperature distribution, spreadingresistance, and conductivity.

Imaging Mechanisms

Scanning Electron Microscopy

The operation of the SEM consists of applying a voltage between aconductive sample and filament,resulting in electron emission from thefilament to the sample. This occurs in a vacuum environment ranging from10-4 to 10 -10 Torr. The electrons areguided to the sample by a series ofelectromagnetic lenses in the electroncolumn. A schematic of a typical SEMis shown in Figure 2. The resolutionand depth of field of the image aredetermined by the beam current andthe final spot size, which are adjustedwith one or more condenser lensesand the final, probe-forming objectivelenses. The lenses are also used toshape the beam to minimize the effects of spherical aberration,chromatic aberration, diffraction, and astigmatism.

The electrons interact with the samplewithin a few nanometers to severalmicrons of the surface, depending onbeam parameters and sample type.Electrons are emitted from the sampleprimarily as either backscatteredelectrons or secondary electrons.Secondary electrons are the mostcommon signal used for investigationsof surface morphology. They areproduced as a result of interactionsbetween the beam electrons andweakly bound electrons in theconduction band of the sample. Someenergy from the beam electrons istransferred to the conduction bandelectrons in the sample, providing

Figure 4. TappingMode AFM image of 1.4Åmonoatomic steps on epitaxial silicondeposited on (100) Si. 1µm scan.

enough energy for their escape fromthe sample surface as secondaryelectrons. Secondary electrons are low energy electrons (<50eV), so only those formed within the first fewnanometers of the sample surface haveenough energy to escape and bedetected. High energy beam electronswhich are scattered back out of thesample (backscattered electrons) canalso form secondary electrons whenthey leave the surface. Since theseelectrons travel farther into the samplethan the secondary electrons, they canemerge from the sample at a muchlarger distance away from the impactof the incident beam which makes theirspatial distribution larger. Once theseelectrons escape from the samplesurface, they are typically detected by an Everhart-Thornley scintillator-photomultiplier detector. The SEMimage formed is the result of theintensity of the secondary electronemission from the sample at each x,ydata point during the rastering of theelectron beam across the surface.

Atomic Force Microscopy

AFM consists of scanning a sharp tipon the end of a flexible cantileveracross a sample surface whilemaintaining a small, constant force. An integrated silicon tip and cantilevercan be seen in Figure 3. The tipstypically have an end radius of 2nm to 20nm, depending on tip type. Thescanning motion is conducted by apiezoelectric scanner which scans thetip in a raster pattern with respect tothe sample (or scans the sample withrespect to the tip). The tip-sampleinteraction is monitored by reflecting a laser beam off the back of thecantilever into a split photodiodedetector. By detecting the difference in the photodetector output voltages,changes in the cantilever deflection oroscillation amplitude are determined. A schematic can be seen in Figure 1.

The two most commonly used modesof operation are contact mode AFMand TappingMode AFM, which are conducted in air or liquidenvironments. Contact mode AFMconsists of raster-scanning the probe(or sample) while monitoring thechange in cantilever deflection with thesplit photodiode detector. A feedback

Figure 5. (a) SEM image of rugged polysilicon thin film. 100,000x, Bar = 0.1µm; (b)TappingMode AFM image of the same with roughness measurement. 1µm scan.

a. b.

loop maintains a constant cantileverdeflection by vertically moving thescanner to maintain a constantphotodetector difference signal. Thedistance the scanner moves verticallyat each x,y data point is stored by the computer to form the topographicimage of the sample surface. Thisfeedback loop maintains a constantforce during imaging.

TappingMode AFM consists ofoscillating the cantilever at itsresonance frequency (typically~300kHz) and lightly “tapping” the tip on the surface during scanning. The laser deflection method is used to detect the root-mean-square (RMS)amplitude of cantilever oscillation. Afeedback loop maintains a constantoscillation amplitude by moving thescanner vertically at every x,y datapoint. Recording this movement forms the topographical image. The advantage of TappingMode overcontact mode is that it eliminates thelateral, shear forces present in contactmode. This enables TappingMode to image soft, fragile, and adhesivesurfaces without damaging them,which can be a drawback of contactmode AFM.

Comparison of Techniques

There are a number of different waysto compare and contrast these twotechniques. Although investigations that use both SEM and AFM tocharacterize a material are common,there are just a few studies that directly discuss the complementar ynature of the techniques. 6-13 Acomparison of these techniques will be conducted with respect to 3factors: (1) Surface Structure, (2)Composition, and (3) Environment. The comparisons are presented fortypical equipment configurations andoperating procedures.

Surface Structure

Although both SEM and AFM aresimilar in lateral resolution, there aresituations in which one technique can provide a more completerepresentation of the sample surface,depending on the information desired.One principle difference is in how the two techniques process verticalchanges in topography. Below we will discuss measurements of differentvertical scales of topography,beginning with very smooth surfacesand working up to very rough surfaces to determine how the surfacetopography affects the ability of eachtechnique to perform the measurement.

Atomically Smooth SurfacesAtomically smooth surfaces can occureither naturally, such as on mineralsurfaces, or by processing, such aspolishing and epitaxial growth onsemiconductor, data storage, andoptical surfaces. A TappingMode AFMimage of an epitaxial silicon surface isshown in Figure 4. Note that, unlikeSEM, the AFM can measure in allthree dimensions (x, y, and z) with asingle scan. Since the AFM has avertical resolution of <0.5Å, it canresolve the 1.4Å monoatomic silicon

steps on the surface as well ascalculate an RMS roughness of 0.7Å.14

On a sample this smooth, the SEM hasdifficulty resolving these features due tothe subtle variations in height..

Thin FilmsOn most thin films, the SEM and AFMproduce a similar representation of thesample surface. A common applicationof surface investigations of thin filmsconsists of determining changes inmorphology with variations ofdeposition parameters, such astemperature, pressure, time, etc. Figure

Figure 6. (a) SEM image of partially GaP-covered Si after chemical beam epitaxy deposition for10 minutes. 30,000x, Bar = 1µm; (b) AFM image of the same sample as in figure 6a showing thepresence of nodules during the growth of GaP by chemical beam epitaxy. 10µm scan; (c) Cross-sectional measurement with AFM across the image in Figure 6b showing 3 nodules which have aheight of approximately 70nm.16

b.

c.

a.

5 shows SEM and AFM images of apolysilicon thin film at approximatelythe same lateral magnification. Thetwo images show similar surfacestructure, however, they differ in theother types of information that can beacquired on this sample. The three-dimensional nature of the AFM can beused to calculate changes in roughnessand surface area variations due todifferences in deposition parameters.For the SEM, a large area view of thevariations in surface structure can beacquired all at once (such as severalmm’s), whereas a 100µm x 100µm

Figure 7. (a) Cross-sectional SEM image of polysilicon lines which shows undercutting due toreactive ion etching. Scale bar = 100nm; (b) Cross-sectional measurement of developed andincompletely developed vias in photoresist acquired by TappingMode AFM. In order to image thehigh aspect ratio structures on the sample, a silicon tip machined with a focused ion beam wasused to scan the vias. 6.2µm scan.

a.

b.

area is typically the largest areaviewed by an AFM. These images arean example of “rugged” polysiliconfilms which are used as capacitors inmemory devices. By making these filmsrough, the surface area is increasedwhich makes it possible to hold morecharge without increasing the lateraldimensions of the capacitors on thechips. By adjusting the depositionparameters and using the AFM toanalyze the surface area of the films,the deposition parameters needed toproduce a film with the maximumsurface area were determined. 15

Another example of the differencebetween the two techniques is ininterpreting subtle differences in height.In the SEM image, changes in slopecan result in an increase in electronemission from the sample surface,producing a higher intensity in theimage. However, it can sometimes be difficult to determine whether thefeature is sloping up or down. Forinstance, in the SEM image in Figure6a it is very difficult to determinewhether the small round structures arebumps or pits, even when tilting thesample stage in the SEM. The onlyother option would be to cleave thesample through one of these featuresand look at the sample in cross-section, which would be tedious and time consuming. Since the AFMdata contains the height information,determining whether a feature is abump or pit is straightforward. As canbe seen in Figures 6b and 6c, thefeatures on this sample are bumps.This information was used in the studyof the growth mechanisms of GaP on Si during chemical beam epitaxydeposition.16 Determination of whetherthese features were small bumps ordepressions would have changed howthe deposition process was altered toproduce an epitaxial GaP film.

High Aspect Ratio StructuresSemiconductor processing commonlyrequires measurements of high aspectratio structures such as trenches andvia holes. In a SEM, these structuresare typically measured in cross sectionby cleaving the wafer and imaging the sample on end to obtain thedimensions of the structure. A commonexample of this is seen in Figure 7a. In contrast, the AFM image of a trenchor via is made by scanning over thesample surface. The ability of the AFM to measure these structuresnondestructively makes it possible for the wafer to be returned to theproduction line after the measurement

is acquired. An AFM image of vias in photoresist is shown in Figure 7b. To image some higher aspect ratiostructures, the proper tip shape isneeded for the AFM to scan narrowopenings and steep sidewalls.Although the SEM measurement isdestructive to the sample, the ability to image the undercuts of these lines is a useful application that AFMs arenot typically designed to perform withthe exception of the Dimension X3DAutomated AFM (please see VeecoInstruments Application Notes on AFMMetrology of undercuts and near-vertical side walls with modelDimension X3D).

Rough SurfacesOne of the key advantages of theSEM with respect to other types ofmicroscopy is its large depth of field.This ability makes it possible to imagevery rough surfaces with millimeters of vertical information within a singleimage. A SEM image of non-wovenpolyethylene oxide fibers can be seenin Figure 8a. The depth of field andsmall beam size makes it possible toimage the fibers far below the toplayer. This ability also makes it possibleto measure very rough surfaces overlarger lateral areas as well. Althoughthe AFM can measure vertical surfacevariations below 0.5Å, its ability tomeasure a tall structure comes fromhow far the scanner can movevertically. Standard scanners typicallyhave 5 to 6µm of vertical range,however, in some configurations thevertical range approaches 10µm orlarger. For scanning areas that haveheights of greater than 5 to 10µm’s ofvariation, the SEM would be bettersuited for the analysis.

Another example of a complex three-dimensional surface structure whichshows how the SEM and AFM can

complement each other can be seen in Figure 8b. The convoluted three-dimensional Y2O3 oxide crystal showngrowing out of a relatively flat Y2O3

thin film on a Si substrate is easilyimaged in the SEM (Figure 8b).Although the AFM would haveproblems imaging the obtuse anglesand enclosed areas of this surface, theroughness of the Y2O3 film can bemeasured whereas in the SEM imagethe surface roughness is not evident.Therefore, the two techniques togethergive a more complete “picture“ of the sample.

Composition

SEM is the only one of the twotechniques which provides elementalanalysis, however, both SEM and AFMare associated with techniques whichcan provide compositional informationthrough analyzing materials andphysical properties of the sample.Some of the most common of thesemethods are described next.

SEMAlong with the secondary electronemission which is used to form amorphological image of the surface in

the SEM, a number of other signalsare emitted as a result of the electronbeam impinging on the surface, asshown in Figure 9. Each of thesesignals carries information about the sample which provide clues to its composition.

Two of the most commonly usedsignals for investigating compositionare x-rays and backscattered electrons.X-ray signals are commonly used to

Figure 9. Signals emitted from a samplesurface after interaction with an electron beam.

Figure 8. (a) SEM image of a non-woven textile sample of polyethylene oxide fibers. The large depth of field of the SEM makes it possibleto image fibers which are 10’s of µm’s below the upper layer of fibers. Bar = 10µm; (b) SEM image of Y2O3 crystal. Bar = 1µm.

b.a.

provide elemental analysis by theattachment of an Energy-DispersiveSpectrometer (EDS) or Wavelength-Dispersive Spectrometer (WDS) to theSEM system. X-ray emission results frominelastic scattering between the beamelectrons and the electrons of thesample atoms. This interaction results inthe ejection of an inner shell electronfrom the atom, creating a vacancy thatis filled by an outer shell electron. Thisjump from an outer to inner shell resultsin a change in energy that produceseither a x-ray or Auger electron. Theemitted x-ray has energy equal to thischange. The x-rays are then detectedby either a lithium-drifted silicondetector for an EDS system, or a gasproportional counter detector for aWDS system. A typical x-ray spectrumcollected with an EDS system is shownin Figure 10.

Backscattered electrons are the resultof beam electrons being scatteredback out of the sample. In this case,the incident beam electrons undergo anumber of scattering events within thespecimen in which very little energy islost, allowing these electrons to gomuch deeper into the sample thansecondary electrons and still emerge

from the sample surface to bedetected. The percentage of beamelectrons that become backscatteredelectrons has been found to bedependent on the atomic number ofthe material, which makes it a usefulsignal for analyzing the materialcomposition. Once these electronsescape from the surface they aredetected by either the Everhart-Thornleydetector or a solid state detector. Anexample of a backscattered image ofa PbSn alloy is shown in Figure 11.

AFMAlthough an AFM does not provideelemental analysis, it can supplycompositional information bydifferentiating materials based onphysical properties, such as stiffness,elasticity, compliance, friction,adhesion, magnetic and electrostaticfields, carrier concentration,temperature distribution, spreadingresistance, and conductivity. Many of these techniques consist of lookingsimultaneously at another signal whileperforming standard AFM imaging.One of the most common techniquesfor mapping differences in materialsproperties is PhaseImaging.PhaseImaging is conducted during

Figure 10. EDS X-ray spectrum of an AlGaN thin film on SiCsubstrate showing the presence of N, Ga, and Al.

Figure 11. Backscattered SEM image of an PbSn alloy showingcontrast based on the atomic number of the two components.The brighter areas are Pb-rich. 5,000x, Scale bar = 1µm.

TappingMode AFM operation bymonitoring the phase lag between theoscillating drive signal used to drivethe cantilever and the oscillatingdetection signal from the photodiodedetector. This signal will indicatedifferences in viscoelasticity and/oradhesion across the imaged area. This technique is commonly applied tomapping the distribution of polymers ina heterogeneous system, or mappingthe distribution of filler, such as silica or carbon black, in a polymer matrix.An example of PhaseImaging on apolyethylene film is shown in Figure12. Other ways to get similarinformation include Force ModulationAFM, which maps differences inelasticity across the sample surface,and Lateral Force Microscopy (LFM),which maps differences in frictionacross the sample surface.

There are also techniques that can beused to investigate long range forcesacross the imaged area. MagneticForce Microscopy (MFM) and ElectricForce Microscopy (EFM) map themagnetic and electrostatic fieldgradients, respectively, which extendfrom the sample surface. Thesetechniques are performed by using

either a magnetic or conductive probeto map the attractive and repulsiveforces between the tip and the sample.MFM is commonly used to detect thedomain structure of magnetic bitswritten on magnetic media, to evaluatethe performance of magnetic heads,and to investigate the magneticstructure of experimental materials. This is conducted by a routine calledLiftMode in which a TappingModetopographic image and a magneticimage are acquired over the samearea. LiftMode consists of first collectinga line scan in TappingMode of thesurface topography. The tip is thenlifted above the surface and a secondscan is made over the same line usingthe saved topographic scan to maintaina constant tip-sample separation. Thelong-range magnetic forces shift theresonance frequency of the oscillatingcantilever, which is detected to producethe magnetic image. An example ofbits written on a textured hard disk isshown in Figure 13.

Although the AFM is applied so that itis nondestructive to the sample surface,it can be used to study differences inmechanical properties by performingnanoindention to investigate hardnessdifferences between materials. This

hydrated samples. These twotechniques compensate for this need by different means: an environmentalchamber for a SEM, and a fluid cell for the AFM. Second, the SEM isrequired to work in a vacuumenvironment due to the nature of thetechnique which brings up the issues of vacuum compatibility of the sample,the conductivity of the surface. Toimage poorly conductive surfaceswithout sample charging may requireconductive coatings or staining, which may alter or obscure the features of interest; or it may require low-voltage operation or anenvironmental chamber, which maysacrifice resolution.

For SEM, hydrated samples areaddressed by placing a specimen inan environmental chamber with eitheran electron transparent window or asmall aperture for the beam to enter the chamber. The chamber is typicallyflushed with an inert gas saturated withwater vapor. Common applications areto either investigate hydrated surfacesto preserve their surface structure whenhydrated, or to reduce charging oninsulating samples. An example ofimaging of a pesticide film on skin canbe seen in Figure 15. For the electron

technique uses a diamond tip mountedon a stiff, stainless steel cantilever. ATappingMode AFM image is collectedwith the probe to determine the area ofinterest for indentation, the nano-indention is then made at a specifiedforce, and an image is then collectedof the indented area. An example ofcomparing the difference betweendiamond like carbon films on a harddisk is shown in Figure 14. In thisexample, the two films demonstrate a difference in hardness fromindentations made at the same forces,producing different sized indents.Scratching and wear testing may alsobe conducted with this configuration toinvestigate adhesion and delaminationof films under a small applied force.

Environment

One of the primary differencesbetween these two types of microscopyis the environment in which they areperformed, i.e., SEM is only conductedin a vacuum environment. In addition to vacuum, AFM is conducted in anambient, gas or liquid environment.There are several issues which makeenvironment an important issue. First,there is a frequent need in fields suchas biology and biomaterials to study

Figure 12. Phase image of two components which are used to form apolyethelene (PE) film. The phase image (right) clearly shows the distributionof the two polymers due to differences in stiffness which is not evident fromthe topographic image (left). 2µm scan.

Figure 13. Magnetic Force Microscopy (MFM) image of overwrittentracks on a textured hard disk. The topography (left) was imaged usingTappingMode; the magnetic force image of the same area (right) wascaptured with LiftMode (lift height 35 nm) by mapping shifts in cantileverresonant frequency. 25µm scan. (17)

beam to interact with the surface in this configuration, it must go though anenvironment of gas and water vapor.One drawback of this configuration is that it will result in an increase inscattering of the electron beam on theway to and from the surface, whichmay result in the sacrifice of imagequality and resolution.

One of the primary attractions to theAFM is its ability to image insulatingsurfaces at high resolution in liquid.Imaging samples in a hydrated statewith an AFM is commonly performedby enclosing the sample and probe in a liquid environment, as shown inFigure 16. Since AFM does not rely on conductivity, the image andscanning mechanism are not disturbedby the presence of the liquid. Commonapplications for AFM investigations in liquid are in the biological sciences,biomaterials, crystal growth, forceinteraction studies, and forinvestigating processes in-situ (Figures17, 18). The resolution of the imagewill be determined by the radius of thetip, the applied force, and the noisefloor of the instrument. Because of

Figure 15. Environmental SEM image of a pesticide film onskin. A hydrated environment was needed in order tomaintain the integrity of the pesticide layer and to reducecharging. Bar = 1mm.

these factors, this configuration allowsthe study of hydrated specimens at alateral resolution of 1 to 5nm and avertical resolution down to 0.5Åwithout sample damage, as seen inthe image of the GroES chaperon(Figure 17). With the appropriateaccessories, AFM can also be used invaried gaseous environments and atelevated temperatures. The latter isparticularly important for research anddevelopment of polymers.

Further Discussion

One thing to keep in mind whencomparing these two techniques is thatalthough SEM and AFM appear verydifferent, they actually share a numberof similarities. Both techniques raster a probe across the surface to detectsome interaction with the surface toform an image. Both have a lateralresolution which is similar in scale(although under certain conditionsAFM is superior). And both techniqueshave image artifacts that the operatoris trained to identify. The SEM has hada much longer time to mature as a

Figure 14. Indentations on two different diamond-like carbon thin filmsusing three different forces (23, 34, and 45µN) with four indents madeat each force to compare differences in hardness. 500nm scans.

technique and we have developedgood understanding of how to identifyand avoid artifacts, but the rapidadoption and implementation of AFMhas resulted in a similar understandingof artifacts. This article has avoideddiscussing such artifacts unless they are relevant to the comparison.Furthermore, by using two techniqueswhich are complementary, onetechnique will often compensate for theimaging artifact of the other technique.

However, one should be wary ofcombined systems in which an AFM isplaced inside the SEM chamber. Oneof the true advantages of the AFM is its ability to perform high resolutionmeasurements outside of a vacuumenvironment. Placing it inside a vacuumenvironment reduces its flexibility andincreases its operating time. CombinedAFM/SEM systems often have reducedcapabilities and typically compromisethe performance of both instruments.

Figure 16. Fluid cell for an AFMwhich allows imaging in anenclosed, liquid environment.

Figure 17. Image of two GroES moleculespositioned side-by-side in physiologic fluid,demonstrating 10Å lateral resolution and1Å vertical resolution. The entire moleculemeasures 84Å across, and a distinct 45Åheptameric “crown” structure protrudes 8Åabove the remaining GroES surface andsurrounds a central depression. 18nm scan. Image courtesy of Z. Shao,University of Virginia.18

Figure 18. Living human vascular endothelial cells imaged in culture media by TappingMode AFM. These images, collected at 30 minuteintervals, reveal the movement of living cells which were incubated with 200ng/ml vascular endothelia growth factor (VEGF). WithTappingMode, the nucleus as well as other submembraneous structures are visible. These cells appear flatter and more elongated comparedto control (untreated) cells. 65µm scans. Sample courtesy of Georges Primbs, Miravant Inc.

Summary

SEM and AFM are complementarytechniques that provide a morecomplete representation of a surfacewhen used together than if each werethe only technique available. Thesetechniques overlap in their capabilitiesto provide nanometer scale lateralinformation. However, they deviate in the fact that the AFM can providemeasurements in all three dimensions,including height information with avertical resolution of <0.5Å, whereasthe SEM has the ability to image veryrough samples due to its large depthof field and large lateral field of view.

The SEM can provide elementalanalysis using X-ray detection, whereasthe AFM can provide compositionalinformation based on physicalproperties. The fact that the twotechniques operate in differentenvironments can be a strength when used together since the AFMdoes not encounter vacuum issues(difficult sample preparation, samplemodification, etc.) and may imagesamples in an enclosed fluid or otherenvironment. The vacuum environmentof the SEM makes it possible toconduct a number of techniques thatrequire vacuum, such as X-ray analysis.

By having both techniques side-by-sidein an analytical facility, the overallscope of analytical capabilities isbroadened, adding to the flexibility of the facility.

References

1. Van Ardenne, M. (1938) Z. Phys. 109, 407

2. Goldstein, J. I., Newbur y, D.E., Echlin, P., Joy, D.C., Fiori, C., Lifshin, E., Scanning Electron Microscopy and X-ray Microanalysis, 1981, Plenum Publishing Corp., New York, p. 3.

3. Zworykin, V.K., Hillier, J., Snyder, R.L., (1942), ASTM Bulletin 117, p. 15.

4. Binning G, Roher H, Gerber C, Weibel E, “Surface Studies by Scanning Tunneling Microscopy,” Phys. Rev.Lett. 49 (1982) 57.

5. Binnig G B, Quate C F, and Gerber Ch., “Atomic Force Microscope,” Phys, Rev. Lett., 12 (1986) 930

6. Neves, B.R.A., Salmon, M.E., Russell, P.E., Troughton, E.B. “Comparitive Study of Field Emission-ScanningElectron Microscopy and Atomic Force Microscopy to Access Self-Assembled Monolayer Coverage on AnyType of Substrate,” Microscopy and Microanalysis 5 (1999) 413.

7. Castle, J.E., Zhdan, P.A., “Characterization of Surface Topography by SEM and SFM: Problems and Solutions,”J. Phys. D: App Phys, 30 (1997) 722.

8. Lemoine, P., Lamberton, R.W., Ogwu, A.A., “Complementary Analysis Techniques for the Morphological Studyof Ultrathin Amorphous Carbon Films,” J. App. Phys., 86 (1999) 6564.

9. Van Helleputte, H.R.J.R., Haddeman, T.B.J., Verheijen, M.J., Baalbergen, J-J., “Comparative Study of 3DMeasurement Techniques (SPM, SEM, TEM) for Submicron Structures,” Microelectronic Engineering, 27 (1995) 547.

10. Nessler, R., “Scanning Microscopy Technologies: Scanning Electron Microscopy and Scanning ProbeMicroscopy,” Scanning 21 (1999) 137.

11. Fischer, A., Jentoft, F.C., Weinberg, G., Schlogl, R., Niesen, T.P., Bill, J., Aldinger, F., De Guire, M.R., Rühle,M., “Characterization of Thin Films Containing Zirconium, Oxygen, and Sulfer by Scanning Electron andAtomic Force Microscopy,” J. Mat. Res. 14 (1999) 3725.

12. Campbell, A.L, Bunning, T.J., Stone, M.O., Church, D., Grace, M.S., “Surface Ultrastructure of Pit Organ,Spectacle, and Non Pit Organ Epidermis of Infrared Imaging Boid Snakes: A Scanning Probe and ScanningElectron Microscopy Study” J. Struct. Biol. 126 (1999) 105.

13. Reynolds, G.W., Taylor, J.W., “Correlation of Atomic Force Microscopy Sidewall Roughness Measurementswith Scanning Electron Microscopy Line-Edge Roughness Measurements on Chemically Amplified ResistsExposed by X-ray Lithography,” J. Vac. Sci. Technol. B 17 (1999) 2723.

14. Strausser, Y.E., Doris, B., Diebold, A.C., Huff, H.R., “Measurement of Silicon Surface Microroughness withAFM,” Extended Abstracts of the 185th Meeting of the ECS, vol. 94-1, Pennington, NJ (1994) 461.

15. Strausser, Y.E., Schroth, M., Sweeney, J.J., Characterization of the low-pressure chemical vapor depositiongrown rugged polysilicon surface using atomic force microscopy,” J. Vas. Sci. Technol. A 15, 1997, 1007.

16. Kelliher, J.T., Thornton, J., Dietz, N., Lucovsky, G., Bachmann, K.J., “Low temperature chemical beam epitaxy of gallium phosphide/silicon heterostrucutres, Materials Science and Engineering, B22 (1993) 97.

17. Babcock, K., “Magnetic Force Microscopy: High resolution imaging for data storage,” Data Storage,September 1994.

18. Mou, J., Czajkowsky, D.M., Sheng, S., Ho, R., Shao, Z., “High resolution surface structure of E. coli GroeSOligomer by Atomic Force Microscopy,” FEBS Letters 381 (1996) 161.