JOURNAL OF CHEMICAL PHYSICS VOLUME 112, NUMBER 18 8 MAY 2000

SPECIAL TOPIC: NEAR-FIELD MICROSCOPY AND SPECTROSCOPY

Scanning near-field optical microscopy with aperture probes: Fundamentals

and applications
Bert Hecht,a) Beate Sick, and Urs P. Wild
Physical Chemistry Laboratory, Swiss Federal Institute of Technology, ETH-Z, Universitätstr. 22,
CH-8092 Zurich, Switzerland
Volker Deckert and Renato Zenobi
Organic Chemistry Laboratory, Swiss Federal Institute of Technology, ETH-Z, Universitätstr. 16,
CH-8092 Zurich, Switzerland
Olivier J. F. Martin
Electromagnetic Fields and Microwave Electronics Laboratory, Swiss Federal Institute of Technology,
Gloriastrasse 35, ETZ ETH-Zentrum, CH-8092 Zurich, Switzerland
Dieter W. Pohl
Institute of Physics, University of Basel, Klingelbergstr. 82, CH-4056 Basel, Switzerland
共Received 13 October 1999; accepted 1 February 2000兲
In this review we describe fundamentals of scanning near-field optical microscopy with aperture
probes. After the discussion of instrumentation and probe fabrication, aspects of light propagation
in metal-coated, tapered optical fibers are considered. This includes transmission properties and field
distributions in the vicinity of subwavelength apertures. Furthermore, the near-field optical image
formation mechanism is analyzed with special emphasis on potential sources of artifacts. To
underline the prospects of the technique, selected applications including amplitude and phase
contrast imaging, fluorescence imaging, and Raman spectroscopy, as well as near-field optical
desorption, are presented. These examples demonstrate that scanning near-field optical microscopy
is no longer an exotic method but has matured into a valuable tool. © 2000 American Institute of
Physics. 关S0021-9606共00兲70316-3兴

I. INTRODUCTION relatively poor performers with respect to spectral and dy-

namic properties. Electron microscopes have to be operated
Scanning near-field optical microscopy 共SNOM or in vacuum, which limits their application in life sciences,
NSOM兲 is at the forefront of science and technology today requires special sample preparation, and complicates sample
because it combines the potentials of scanned probe technol- manipulation. SNOM combines the excellent spectroscopic
ogy with the power of optical microscopy. SNOM provides and temporal selectivity of classical optical microscopy with
us with eyes for the nanoworld. Among the main parameters a lateral resolution reaching well into the sub-100 nm re-
that might be of interest for a nano structure under investi- gime. However, for a long time technical problems jeopar-
gation are, besides shape and size, its chemical composition, dized the widespread application of SNOM. Near-field optics
molecular structure, as well as its dynamic properties. In or- 共NFO兲 therefore became a focus of research and develop-
der to investigate such properties, microscopes with high ment in the field of optical microscopy in recent years.1–9
spatial resolution as well as high spectral and temporal re- Today, we have reached the point where SNOM represents a
solving power are required. The classical optical microscope powerful tool for surface analysis that is technically and
excels with respect to spectroscopic and temporal selectivity, theoretically well understood. It is ready to be applied to a
although its resolution is restricted by diffraction to about large variety of problems in physics, chemistry, and biology.
half the wavelength, i.e., to 0.2–0.5 micrometer for visible In a simplified view of classical far-field optical micros-
light. Present-day science and technology, however, have an copy the object is illuminated by a monochromatic plane
increasing need for tools that allow to characterize, generate, wave. The transmitted or reflected light, scattered by the ob-
and manipulate structures as small as a few nanometers in ject in a characteristic way, is collected by a lens and imaged
size. Examples are readily found in biology, microelectron- onto a detector. For practical reasons, the lens is placed at
ics, and medical sciences. Electron microscopes as well as least several wavelengths ␭ of the illuminating light away
scanning tunneling and atomic force microscopes easily from the object surface, i.e., in the far field. Since high spa-
achieve 10 nm spatial resolution and beyond, but they are tial frequencies corresponding to the fine details of the object
generate Fourier components of the light field that decay
a兲
Electronic mail: [email protected] exponentially along the object normal,10 they cannot be col-

0021-9606/2000/112(18)/7761/14/$17.00 7761 © 2000 American Institute of Physics

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7762 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

lected by the lens. This effect leads to the well known Abbé
diffraction limit ⌬x⫽␭/(2 ␲ NA), 11 where NA is the nu-
merical aperture of the lens.
This resolution limit can be slightly pushed by scanning
confocal optical microscopy where a sharply focused spot of
light replaces the wide-field illumination.12,13 Using special
geometries of illumination and detection, sometimes in com-
bination with nonlinear effects such as multiphoton
excitation,13 further progress in this field is still being made; FIG. 1. Different types of scanning near-field optical microscopes: 共a兲 ap-
however, steps are small and no major breakthrough is to be erture SNOM with angular resolved detection, 共b兲 apertureless configura-
expected. tion, and 共c兲 scanning tunneling optical microscope.
In 1928, Synge, an Irish scientist, described an experi-
mental scheme that would allow optical resolution to extend
radiation emerging from the probe sample interaction zone
into the nanometer regime:14 He proposed to use a strong
into the far field.24,25
light source behind a thin, opaque metal film with a 100 nm
Figure 1共b兲 shows the intriguing ‘‘apertureless’’ NFO
diameter hole in it as a very small light source. The tiny spot
techniques.26–34 In this configuration a strongly confined op-
of light created this way should be used to locally illuminate
tical field is created at the apex of a sharply pointed probe tip
a thin biological section. In order to guarantee the local illu-
by external 共far-field兲 illumination. The relevant NFO signal
mination, he imposed the condition that the aperture in the
therefore often has to be extracted from a huge background
metal film be no further away from the section than the ap- of far-field scattered radiation. Resolutions ranging from
erture diameter, i.e., less than 100 nm. Images were to be 1–20 nm have been reported in laboratory experiments. The
recorded point by point detecting the light transmitted by the general applicability of the technique to a wider range of
biological section by means of a sensitive photo detector. samples is currently being investigated.
With his proposal, Synge was well ahead of his time. The third NFO microscopy configuration 关see Fig. 1共c兲兴
Unfortunately, he never tried to realize his idea. In 1932, he relies on the direct detection of localized fields in the near
actually proposed an alternative, discarding his original field above a sample by uncoated dielectric tips. The so-
scheme because of the difficult sample approach to a planar called scanning tunneling optical microscope 共STOM兲,35 also
screen.15 He therefore suggested to use the image of a point called photon scanning tunneling microscope 共PSTM兲,36 is
light source as an optical probe instead of an aperture. The schematically depicted in Fig. 1共c兲. In selected experiments
image was to be generated by an ellipsoidal mirror which the STOM led to results that could not have been achieved
would provide, in modern words, the largest possible nu- by other techniques.37–40
merical aperture. Ironically, this second scheme never would
have achieved SNOM-type resolution since it ignored the II. THE APERTURE SNOM
role of near fields which cannot be recovered by any conven-
tional imaging scheme, no matter what the numerical aper- A. General setup
ture is. In this article we will concentrate on the aperture
In 1984, shortly after the invention of the scanning tun- SNOM. Reaching a resolution of 50–100 nm on a routine
neling microscope,16 nanometer-scale positioning technology base and having a potential down to 10–30 nm, SNOM is at
was available and an optical microscope similar to Synge’s least a factor of 5 to 10 better in resolution than a standard
proposed and forgotten scheme was re-invented by Pohl and scanning confocal optical microscope with 1.4 NA. This rep-
demonstrated together with Denk and Duerig at the IBM resents an enormous progress in view of the relevance of the
Rüschlikon Research Laboratory.17–19 Independently, a simi- sub-100 nm regime today, and certainly, aperture SNOM is
lar scheme was proposed and developed by Lewis and his presently the most widely used and most developed NFO
group at Cornell University.20–22 The key innovation was the technique. This is also illustrated by the fact that all the
fabrication of a subwavelength optical aperture at the apex of commercial instruments currently available employ the aper-
a sharply pointed transparent probe tip that was coated with ture SNOM technique.
a metal. In addition, a feedback loop was implemented main- In our opinion the reasons for this situation are obvious.
taining a constant gapwidth of only a few nanometers while From a simplified point of view, imaging with SNOM is
raster scanning the sample in close proximity to the fixed similar to scanning confocal optical microscopy except that
probe. the illumination spot is smaller. Of course, one has to keep in
The promise to extend the power of optical microscopy mind that SNOM is a surface selective technique, whereas
beyond the diffraction limit triggered the development of scanning confocal optical microscopy can advantageously be
several experimental configurations that are able to generate applied to thick samples that require optical sectioning.12,13
optical images with nanometer resolution. Figure 1共a兲 shows Nevertheless, all the contrast mechanisms such as absorp-
the classical aperture SNOM configuration in which an ap- tion, phase, and fluorescence contrast, well-known in con-
erture probe is illuminating a small area of a sample ventional optical microscopy, can be transferred more or less
surface.4,5,17–23 In the most general configuration with re- directly to SNOM applications. An aperture represents a very
spect to light detection, the sample is placed on top of a confined light source without any background. This is in
hemispherical substrate which allows the capture of all the contrast to the apertureless techniques where a rather large,

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7763

contact,49 and electrostatic image forces.50 The phase signal

reacts faster to changes in the state of oscillation, because it
is independent of the dissipation of kinetic energy stored in
the resonance. Therefore, the use of pure phase feedback51 or
a combination of amplitude and phase feedback44,46,52 is ad-
vantageous in fast shear force feedback systems. Light
source, fiber, optical probe, shear force sensor, and mechani-
cal support often make up an independent unit in an SNOM
device as sketched in Fig. 2共a兲.
Light that is emitted by the aperture locally interacts
with the sample. It can be absorbed, phase shifted, or locally
excite fluorescence, depending on the sample and the con-
trast mechanism of interest. In any case, light emerging from
the interaction zone has to be collected with the highest pos-
sible efficiency. For this purpose, high NA 共oil immersion兲
microscope objectives or mirror systems are often used. A
distinct advantage of an oil immersion objective when using
a plane parallel substrate is that it can collect light that would
otherwise undergo total internal reflection and thus fail to
reach the detector. This is the so-called ‘‘forbidden
light.’’25,44,52,53 Forbidden light detection can lead to en-
FIG. 2. Standard SNOM setup consisting of 共a兲 an illumination unit, 共b兲 hanced contrast and resolution in phase and amplitude con-
collection and redistribution unit, and 共c兲 a detection module.
trast images.54 In the case of mirror-based detection, hemi-
spherical substrates providing a ‘‘solid immersion’’ are
intensive laser spot is focused onto a tiny tip apex. For fluo- sometimes used to capture the forbidden light.25,44,52,53
rescence studies, which are among the most relevant appli- The collected light is directed via a 共dichroic兲 mirror
cations of SNOM, a localized illumination is important be- either to a visual inspection port of the microscope or to a
cause excessive photobleaching of the sample has to be suitable detector 关see Fig. 2共c兲兴. Filters can be inserted to
avoided. remove unwanted spectral components. Light collection,
It is straightforward and tempting to extend existing con- redistribution, and filtering is advantageously done by means
ventional optical microscopes by a SNOM zooming stage. A of a standard 共inverted兲 optical microscope depicted in
typical SNOM setup, as it is realized in many laboratories Fig. 2共b兲.
and commercial instruments, is sketched in Fig. 2. Laser
B. Fabrication of near-field optical probes
light of a suitable wavelength is coupled into an optical fiber
that has an aperture probe at its far end. Control of the po- Great efforts have been devoted to the microfabrication
larization and spectral filtering of the light is necessary be- of NFO probes.55,56 However, few of the present designs are
fore coupling into the fiber. The tip is mounted onto a me- commercially available or have actually been used in NFO
chanical support that includes adjustment screws to facilitate experiments. Therefore, the NFO community still relies
a coarse approach of the tip to the sample 共into the range of heavily on aperture probes based on tapered optical fibers.
the z-piezo displacement兲. Furthermore, it is often necessary Since fiber probes of sufficient quality are still expensive or
to align the tip laterally onto the optical axis of the conven- even impossible to obtain commercially, tip fabrication re-
tional microscope’s objective or move it to specific areas of mains of major interest to scientists working in the area of
interest on the sample. near-field optics.
In order to measure and control the gapwidth between The fabrication process for fiber-based optical probes
tip and sample, a strongly gapwidth-dependent signal is can be divided in two main steps: 共a兲 the creation of a trans-
needed. Early SNOMs relied on electron tunneling parent taper with a sharp apex, and 共b兲 subsequent coating
feedback.17,18 Today, the majority of SNOMs utilizes a with aluminum to obtain an entirely opaque film on the cone
method similar to noncontact atomic force microscopy walls and to form a transmissive aperture at the apex. There
共AFM兲 which is called shear force feedback.41,42 The fiber are two methods for the preparation of tapered optical fibers
probe is vibrated at one of its mechanical resonances parallel with a sharp tip and reasonable cone angle: 共i兲 the ‘‘heating
to the sample surface, ideally at amplitudes below 1–5 nm. and pulling’’ method, and 共ii兲 chemical etching.
Amplitude and phase of this minute fiber oscillation are The heating and pulling method is based on local heating
monitored by a suitable displacement sensor.41–46 During the using a CO2 laser or a filament and subsequently pulling the
final approach (⬇0 – 20 nm兲 due to the action of shear fiber apart. The resulting tip shapes depend heavily on the
forces, the resonance frequency is detuned with respect to the temperature and the timing of the heating and pulling, as
driving oscillator leading to a decrease in amplitude and to a well as on the dimensions of the heated area.44,57,58 The pull-
phase shift. The origin of the shear force effect is not entirely ing method has the advantage that the glass surface on the
clear. There might be a range of mechanism involved, in- taper is very smooth, which positively influences the quality
cluding viscous damping in thin water films,47,48 intermittent of the evaporated metal layer. Furthermore, pulled tips often

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7764 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

FIG. 3. Evaporation geometry of the aluminum coating process: evaporation

takes place under an angle slightly from behind while the tip is rotating. The
deposition rate of metal at the apex is much smaller than on the side walls.

exhibit flat facets at the apex due to a final fracture during

pulling which facilitates the formation of an aperture during
evaporation. It is, however, difficult, though not impossible,
to obtain a large cone angle; correspondingly the transmis-
sion coefficient for a given aperture size is low, as to be FIG. 4. Aluminum-coated aperture probes prepared by pulling 共a兲,共b兲 and
discussed in Sec. III A 1. etching 共c兲,共d兲: 共a兲,共c兲 macroscopic shape, SEM and optical image. 共b兲,共d兲
Chemical etching of optical fibers in HF solutions be- SEM close-up of the aperture region, scale bar corresponds to 300 nm.
came popular since Turner’s etching technique59 was
introduced.60 It comprises dipping the fiber into an HF solu-
tion covered with an overlayer of an organic solvent. Tip
formation takes place at the meniscus that forms at the fiber respectively. A comparison of the different global shapes of
at the interface between HF and solvent. The mechanism is the probes in Figs. 4共a兲 and 4共c兲 before aluminum coating
based on the fact that the meniscus height is a function of the evidences the improved cone angle and overall shape of the
remaining fiber diameter. Chemical etching allows more re- tube etched tips 关Fig. 4共c兲兴. Figures 4共b兲 and 4共d兲 show scan-
producible production of larger quantities of probes in a ning electron microscopy 共SEM兲 images of the apex region
single step, thus opening the road towards a laboratory-scale of pulled and tube etched tips after aluminum coating. It is
‘‘mass production’’ of optical probes. A specific advantage obvious that both tips exhibit a comparably high quality of
of Turner’s method is that the taper angle can be tuned by the aluminum layer. The aperture can be observed for both
varying the organic solvent used as the overlayer. Accord- tips.
ingly, optical probes with correspondingly large transmission
coefficient can be produced.61,62 A major disadvantage of
Turner’s technique turned out to be the microscopic rough-
ness of the glass surface on the taper walls. It leads to pin- III. SQUEEZING LIGHT THROUGH A
holes and imperfections in the metal coating that interfere SUBWAVELENGTH APERTURE
with the true NFO signals.
This problem was solved only recently by introducing a To achieve the ultimate performance possible in aperture
method called tube etching.63,64 In tube etching the fiber is SNOM, the optical probe should combine two main proper-
not stripped before dipping it into HF. Tip formation takes ties: 共i兲 the spot size determined by the aperture diameter
place inside a small volume defined by the polymer coating should be as small as possible. At the same time, 共ii兲 the light
due to convective flow of HF. The delicate meniscus be- intensity at the aperture should be as high as possible. This
tween HF and the organic overlayer no longer plays a direct can be achieved either by optimizing the overall light
role in the tip formation process. The method is thus less throughput or by improving the damage threshold of the
prone to produce roughness on the taper while still providing metal coating thus increasing the maximum input power. Fi-
large cone angles. nally, for practical reasons, delivering the light into the vi-
In both techniques, the aperture is formed during the cinity of the aperture should be as convenient as possible.
evaporation of aluminum. Since the evaporation takes place This requirement is well met by aperture probes produced
under an angle slightly from behind, the deposition rate of from standard optical fibers. Optical probes with small spot
metal at the apex is much smaller than on the sides. This sizes and large transmission coefficients could open up fas-
geometrical shadowing effect leads to the self-aligned forma- cinating new areas of research such as nanoscale nonlinear
tion of an aperture at the apex, as is illustrated in Fig. 3. optics and optical surface modification on the nanometer
In Fig. 4 we compare typical fiber probes produced by scale. It is therefore of principal interest to understand the
heating and pulling 共left side兲 and tube etching 共right side兲, limitations and possibilities of aperture probes.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7765

fiber leading to a reorganization of propagating modes ac-

companied by a rather strong back-reflection of light as a
first source of attenuation.
The mode structure in a metallic waveguide at optical
frequencies was calculated by Novotny and Hafner66 as a
function of the dielectric core diameter. The results can be
interpreted such that in a tapered waveguide, where the core
diameter decreases gradually, one mode after the other runs
into cutoff until only the final HE11 mode is still propagating.
An aluminum-coated ( ⑀ alu⫽⫺34.5⫹i8.5) dielectric ( ⑀ core
⫽2.16) waveguide supports the pure HE11 mode at a wave-
length ␭⫽488 nm for inner diameters between ⬇250 and
160 nm as sketched in Fig. 5. The transmission coefficient of
FIG. 5. Mode propagation in a tapered metal-coated optical fiber at a wave- light up to this point is mainly determined by the fraction of
length of 488 nm. Cutoff diameters taken from Ref. 66. power contained in cutoff modes as compared to the power
in the still-propagating HE11 mode. The magnitude of this
factor is likely to depend on the geometry of the taper. This
A. Transmission coefficient of aperture probes connection has been poorly understood up to now. Power not
comprised in the propagating modes is either reflected back
1. The taper region into the waveguide or absorbed in the metal coating, leading
The transmission coefficient of an aperture probe at a to a considerable heating of the metal.67–70
given wavelength is defined as the power of the light coupled Below an inner diameter of 160 nm, even the HE11 mode
into the taper region divided by the light power emitted by runs into cutoff. Cutoff means that the wave vector becomes
the aperture. In the case of an optical fiber, the power imaginary and thus the mode field decays exponentially. The
coupled into the taper region is easily determined by cleav- power that is actually delivered to the aperture depends on
ing the fiber behind the coupler and measuring the output. the distance between the HE11 cutoff diameter and the aper-
The measured value corresponds to the power in the fiber ture plane which is determined by the cone angle of the
minus ⬇4% reflection loss. taper. This fact was first noticed by Novotny et al.,71 who
The power emitted by the aperture is often measured in predicted that larger cone angles would drastically improve
the far field. In principle, this is correct since the nonpropa- the overall transmission coefficient of a tapered waveguide
gating field components around the aperture do not contrib- structure.
ute to the time-averaged Poynting vector.65 However, in al-
most all experimental situations the instantaneous electric 2. The aperture
fields dominate light-matter interaction. Thus, the far-field The transmission coefficient of a subwavelength hole in
transmission coefficient does not reflect the field enhance- an infinitely thin perfectly conducting screen was first calcu-
ment effects in the near field. However, in comparing tips lated rigorously by Bethe.72 Errors in his expression for the
with the same aperture size, the far-field transmission coef- near field were corrected in a paper by Bouwkamp.73 The
ficient can be valuable information. so-called Bethe/Bouwkamp model of a subwavelength aper-
The transmission coefficient of an aperture at the apex of ture is important because it provides closed analytic expres-
a conical tip is determined by the characteristics of its two sions for the resulting electric and magnetic field. Although
main structural components: 共i兲 the part that guides light into the model fails to accurately describe reality, the expressions
the vicinity of 共ii兲 the actual subwavelength aperture. In stan- still include the most characteristic features of the near- and
dard fiber probes, as depicted in Fig. 5, the guiding part far-field distribution of a realistic subwavelength aperture.
corresponds to a tapered, metal-coated dielectric waveguide. Within the Bethe/Bouwkamp model, it is expected that
The efficiency of guiding light to the aperture is determined the transmission coefficient of a subwavelength hole should
by the distribution of propagating modes in the tapered scale as a 4 , where a denotes the aperture diameter.72–74 An
waveguide. In the special case of a single-mode fiber with a increase of the aperture diameter from 20 to 100 nm hence
taper produced by heating and pulling, the initial effect of the results in a transmission coefficient increased by a factor of
taper is a spreading of the guided mode since the guiding 625. In order to illustrate this, we use transmission data cal-
core is getting thinner and thinner. When the mode starts to culated by Novotny et al.71 for a realistic tip structure with
‘‘feel’’ the vicinity of the metal coating, the waveguide different full taper angles ␣ and aperture diameters of 10 and
changes its character from a dielectric to a metallic hollow 20 nm. The calculations employed the multiple multipole
one filled with a dielectric. In the case of a taper produced by method75 to solve Maxwell’s equations.
chemical etching, the mode spreading effect is absent be- Scaling the numerical data according to the Bethe/
cause the core diameter is not affected by the chemical etch- Bouwkamp model provides a simple approximation of the
ing. As a result the influence of the metal coating starts to be transmission of a 100 nm aperture as a function of the taper
important at smaller waveguide diameters closer to the aper- angle 共see Fig. 6兲. For values up to ␣ ⫽30°, where the two
ture. In both cases, the mode structure in the metallic wave- curves run parallel, scaling the 10 nm curve by 104 and the
guide is completely different from the one in an unperturbed 20 nm curve by 5 4 leads to transmission coefficients that

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7766 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

ported in the literature and measured in our

laboratories.61,63,76 It should even be possible from such con-
siderations to estimate the effective aperture size from trans-
mission measurements if the taper angle at the apex is known
from an independent 共e.g., SEM兲 measurement.
Working with really small apertures (⬍40– 50 nm兲 is
often impossible since the transmission coefficient decreases
dramatically with decreasing aperture size. This cannot be
overcome by an increased input power because of the gen-
erally low damage threshold of the metal coating (⬇10
mW兲. The relevant NFO signals becomes so small that the
signal to noise ratio is no longer sufficient to provide reliable
results. This is the reason why most NFO studies today are
carried out with apertures of 80–100 nm.
FIG. 6. Transmission coefficient of an aperture probe as a function of the
full taper cone angle ␣ . Squares and rhombs: calculated values from Ref. 71
for a 20 and 10 nm diameter aperture, respectively. Circles and triangles:
B. Field distribution
interpolated values determined by scaling the numerical data according to Since we focus this article on aperture SNOM, it is im-
the Bethe/Bouwkamp a 4 law by a factor of 104 and 5 4 , respectively. This
provides a simple approximation for the transmission coefficient of a 100 portant to investigate the character of the electric field distri-
nm aperture. bution behind a small aperture. Let us consider a geometry
that is even simpler than the Bethe/Bouwkamp configuration,
namely a slit in a thin, good-conducting metallic screen 共Fig.
nearly coincide. At larger taper angles the interpolated data 7兲. The screen is illuminated at normal incidence by a plane
points start to diverge. This is likely due to the breakdown of wave with a wave number k 0 ⫽2 ␲ /␭, where ␭ is the wave-
the Bethe/Bouwkamp model because the finite penetration length 关Fig. 7共a兲兴. Right behind this aperture, the transmitted
depth of light into the aluminum (⬇7 nm兲 starts to wash out field will be more or less confined to its width.77
the difference in diameter between the two apertures and the The character of this field is best understood by consid-
transmission through the metal screen becomes dominant. ering its angular 共Fourier兲 spectrum.10 The electric field dis-
From Fig. 6 it can be deduced that for a 100 nm aperture tribution behind the aperture can be obtained by convoluting
and a taper angle of about 30°, a transmission between 10⫺6 the angular spectrum of the incident field with the Fourier
and 10⫺5 may be expected. For even larger taper angles transform of the aperture.10 The Fourier transformation must
around 42°, as they occur in etched fiber probes and in an- be performed in the plane of the screen and provides an
isotropically etched silicon structures, the transmission coef- angular spectrum of plane 共and evanescent兲 waves with field
ficient should reach values around 10⫺3 . These numbers are amplitudes ␥ (k 储 ,k z ). Here, k 储 and k z are respectively, the
in reasonable agreement with transmission coefficients re- transverse and longitudinal components of the k vector.

FIG. 7. Character of the electric fields

behind an aperture: 共a兲 A small aper-
ture of size a in an infinitely thin and
perfectly conducting screen is illumi-
nated at normal incidence with a plane
wave propagating in the z direction
with a vector k z ⫽k 0 ⫽2 ␲ /␭. 共b兲–共d兲
The spectrum of the transmitted field
is given for three different aperture
sizes. In each case the vertical axis
represents the field amplitude ␥ of the
k vector with components (k 储 ,k z ),
where the transverse component k 储 is
real and the longitudinal component k z
can be either real or imaginary. The
latter corresponds to an evanescent
field strongly localized at the vicinity
of the aperture. The ratio of propagat-
ing and evanescent fields strongly de-
pends on the aperture size a relative to
the wavelength ␭.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7767

Since our aim here is not to give an exact treatment of the direction becomes dominated by the evanescent components;
scattering by the slit but rather to illustrate the physical phe- i.e., for the vast majority of the transverse components, k 储 is
nomena relevant to aperture SNOM, we will assume that the larger than k 0 关Fig. 7共d兲兴. This means that the transmitted
transverse component k 储 is real. However, the longitudinal field is strongly localized at the vicinity of the aperture and
component k z can be either real or imaginary. It is related to decreases rapidly away from the aperture. As a result, the
k 储 via far-field power emitted by the aperture decreases, while
strongly confined and enhanced fields appear at the vicinity
k 20 ⫽k 2储 ⫹k z2 . 共1兲 of the aperture. Experimental evidence for the existence of
Together, k 储 and k z describe the radiation properties of the nonradiative fields close to an aperture are the ‘‘forbidden
plane and evanescent waves that constitute the field behind light’’ emission,24,44,54 surface plasmon excitation,79 as well
the aperture. The angular spectra for different aperture sizes as an increased fluorescence from single molecules close to
are depicted in Figs. 7共b兲–7共d兲. Field amplitudes ␥ (k 储 ,k z ) an aperture.80
concentrated at small k 储 correspond to large apertures,
whereas a distribution of ␥ (k 储 ,k z ) over a broad range of k 储 IV. TIP–SAMPLE INTERACTION AND IMAGE
values corresponds to small apertures. FORMATION
From Eq. 共1兲 we see that, depending on the magnitude of We previously emphasized the importance of a localized
k 储 , two different types of solutions are obtained for k z light source for SNOM. However, in most SNOM experi-

再 冑k 20 ⫺k 2 ments the field is detected at a large distance of the tip, in the

储 k 储 ⭐k 0
k z⫽ 共2兲 far field. The mechanisms by which some components of the
i 冑k 2 ⫺k 20
储 k 储 ⬎k 0 . evanescent illuminating field can be transformed into propa-
gating field components that carry information about the
As an example, let us consider an aperture of width a. The
sample are at the core of image formation in SNOM.
angular spectrum of this aperture will contain significant
Accurate and versatile computational techniques are
field amplitudes ␥ (k 储 ,k z ) at k 储 ⫽2 ␲ /a. For this particular
mandatory to understand these intricate contrast mecha-
k 储 , the expression for the longitudinal wave vector compo-
nisms. One of the intrinsic difficulties in the simulation of
nent reads
near-field optical images is related to the fact that a practical

k z⫽ 冑冉 冊 冉 冊
2␲
␭
2
⫺
2␲
a
2
. 共3兲
experiment contains elements that have very different sizes:
The sample is usually quite small, but it is deposited on a
substrate with a quasi-infinite extension. The tip is very large
It is obvious that for a⬍␭, k z becomes imaginary. too, but its interaction with the sample quite subtle. Finally,
The two sets of solutions described by Eq. 共2兲 are de- as already mentioned, while the relevant light–matter inter-
picted as a dashed line in Figs. 7共b兲–7共d兲: for 兩 k 储 兩 ⭐k 0 the action takes place in the near field close to the sample, the
extremity of the k vector describes a circle in the (k 储 ,k z ) image is formed by detecting intensity changes in the far
plane, the k z component being real. On the other hand, for field, at very large distances from the sample.
兩 k 储 兩 ⬎k 0 the solutions follow a square-root curve, with an A great effort has been devoted, over the last 10 years, to
imaginary k z component. A field component with an imagi- the theoretical understanding of image formation in near-
nary wave vector in the propagation direction corresponds to field optical microscopy, and two excellent review articles
an evanescent field. It hence remains strongly bound to the have been published by Girard and Dereux81 and Greffet and
aperture and does not propagate into the far field. The strong Carminati.82 However, it is only recently that self-consistent
localization is responsible for the resolution that can be calculations taking into account these different length scales,
achieved in scanning near-field optical microscopy.78 and in particular the interaction between the tip and the
For a given wavelength, the ratio of propagating (k z sample in a fully three-dimensional manner, have been
real兲 and evanescent (k z imaginary兲 field amplitudes in the presented.83,84 These results were obtained with the Green’s
angular spectrum strongly depends on the size a of the aper- tensor technique.85 The decisive advantage of this approach
ture as compared to the wavelength ␭ 关see Figs. 7共b兲–7共d兲兴. to simulate SNOM experiments lies in the fact that it can
When aⰇ␭, the angular spectrum of the transmitted field is accommodate complex systems that include both infinite
very similar to that of the incident field, i.e., most of the substrates and strongly localized samples. Furthermore, it al-
transmitted field propagates in the forward direction with a lows the computation of the measured far field while taking
propagation vector k z ⬇k 0 , k 储 ⬇0 关Fig. 7共b兲兴. Few field com- into account the near-field interaction in a fully self-
ponents with a small k 储 can appear, leading to a slight diver- consistent manner. Unfortunately, this approach is not very
gence of the beam transmitted through the aperture. accurate when high permittivity materials are in play. This
When the aperture size becomes comparable to the can, however, be superseded by using the filtered Green’s
wavelength, the lateral confinement increases and the spec- tensor technique.86
trum broadens. As a result the transmitted field diverges Figure 8 presents results obtained with this approach.
strongly 关Fig. 7共c兲兴. The angular spectrum of the transmitted The geometry under study is depicted in Fig. 8共a兲, which
field now also includes a small range of finite amplitudes shows a relatively large tip that is used to illuminate a small
with 兩 k 储 兩 ⬎k 0 . glass cube on a glass surface. In Fig. 8共b兲 we show an image
Finally, when the aperture is much smaller than the calculated when the tip is raster-scanned over the cube at a
wavelength, the spectrum of the transmitted field in the z constant height above the surface. The field transmitted

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7768 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

FIG. 8. Simulation of image formation in near-field microscopy. 共a兲 An accurate model for an SNOM experiment should include an aperture tip with a
reasonably large extension, a sample, and an infinite substrate supporting the sample. 共b兲 Total transmitted electric field intensity computed for the geometry
depicted in 共a兲. The sample is made of glass (␧⫽2) and deposited on a glass substrate. The tip has a 50 nm opening with a 70 nm thick aluminum coating.
The sample size is 40⫻40⫻40 nm3 . The incident field is polarized in the x direction. To reproduce an experimental image, each pixel on this figure
corresponds to a different calculation with a particular tip–sample position, the tip apex being kept at a constant height 共2 nm兲 above the top sample surface.
共c兲 The strong signals observed in 共b兲 are related to depolarization fields created on the sample sides that are normal to the incident field. 共d兲 Comparison of
different scanning modes: constant height 共dashed line兲 and constant gap 共solid line兲 measured along the dashed line in 共b兲. The corresponding tip motion is
depicted at the bottom of the figure. When the sample is scanned in constant gap mode, i.e., when the tip motion follows the sample outline, the depolarization
signal is much stronger as it is in constant height mode.

through the system is computed, as in a realistic experiment tion signal that is observed in the near-field image 关Fig.
共see Fig. 2兲. We observe in Fig. 8共b兲 a strong asymmetry in 8共b兲兴.
the measured signal. In order to understand this effect, one If instead of scanning at a constant height above the
must recall that when a system like the glass cube on the surface, a constant gapwidth is kept between the tip and the
surface is illuminated with some external field E0 , it gener- sample, one measures a very different signal 关Fig. 8共d兲兴. In
ates a depolarization field Ed such that the total field Et this scanning mode, the interaction between the incident field
⫽E0 ⫹Ed fulfills the boundary conditions imposed by Max- and the sample sides takes place during the entire upward
well’s equations.87 In our calculation, the illumination field motion of the tip necessary to follow the contour of the
E0 was chosen to be polarized in the x direction. Such an sample. This leads to the prominent depolarization peaks that
incident field is therefore parallel to the top and bottom sides are observed in Fig. 8共d兲 when the incident field is normal to
of the sample 关Fig. 8共b兲兴. Since an electric field parallel to an
the sample side. These peaks are located far away from the
interface must be continuous across this interface, the inci-
sample because they include the tip radius in addition to the
dent field already satisfies the electromagnetic boundary con-
sample side. The strong signal is hence merely related to the
ditions and no depolarization field is created along these
tip motion and constitutes a good illustration of a topo-
sides 关Fig. 8共c兲兴.
On the other hand, this x-polarized incident field does graphic artifact.88 Such an artifact is very difficult to distin-
not comply with the boundary conditions along both vertical guish from the real near-field signal. Scanning at constant
sides, since it is normal to the sample interfaces. As a matter height therefore is the preferred mode of operation when
of fact, the total electric field Et should now have a discon- working with slightly corrugated samples. On the other hand,
tinuity proportional to the sample permittivity at this inter- our results also show that a careful control of the illumina-
face 关when the electric field is normal to an interface, it is the tion polarization can provide another handle in order to keep
product of the field with the permittivity that must be con- these artifacts under control. In particular, a series of mea-
tinuous across the interface, Fig. 8共c兲兴. Therefore, the sample surement of the same system with different illumination po-
generates a strong depolarization field so that the total field larizations should allow discrimination between topography
fulfills the boundary conditions. It is this strong depolariza- artifact and real near-field signal.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7769

The calculated distance between the patches is in agree-

ment with the shear force image 关Fig. 9共b兲兴 of the sample.
The shear force image represents a convolution of a rather
blunt protrusion on the aperture rim 关cf. Figs. 4共c兲 and 4共d兲兴
with the real sample topography. Therefore, the triangular
shape of the metal patches, which was confirmed indepen-
dently by SEM, is not visible.
Constant-height-mode optical images 关Figs. 9共c兲 and
9共d兲兴 were obtained with the same tip as used for Fig. 9共b兲
and at the same location on the aluminum island film, col-
lecting the forbidden light emitted from the probe–sample
interaction zone, only. The light emitted by the optical probe
was linearly polarized 关arrows in Figs. 9共c兲 and 9共d兲兴 in two
orthogonal directions. The extinction ratio measured was
about 15. Figure 9共e兲 presents a cut through two neighboring
dots at the position indicated by the white bar in Fig. 9共c兲. In
Fig. 9共f兲 a high-pass-filtered representation of Fig. 9共c兲 is
displayed achieving a better visibility of the small structures
by removing long-ranged 共far-field兲 modulations.
The optical scan images 关Figs. 9共c兲 and 9共d兲 show a
distinct fine structure. The hexagonal pattern, corresponding
to the metal patches, is well visible. The structures in the
optical images show systematic distortions. In particular, the
patches are elongated in the direction of polarization. This
might be attributed to the polar concentration of the electric
energy on opposite sides of the rim of a small aperture.71,73,91
A few nanometers behind the aperture plane this rather sharp
double-peak structure is washed out, forming an approxi-
mately elliptic spot. The direction of strongest confinement
hence corresponds to the direction perpendicular to the po-
larization in agreement with the observed elongation. It has
FIG. 9. Metal island film: 共a兲 Sketch of metal patches arranged in hexagons
in the interstices of densely packed 220 nm latex spheres. 共b兲 Shear force to be pointed out that the metal patches strongly influence
image. 共c兲, 共d兲 Constant height forbidden light images taken at the same the field distribution at the aperture,92 therefore, our discus-
position as 共b兲. The light emitted by the optical probe was linearly polarized sion of spot shapes must remain qualitative if no realistic
in 共c兲 and 共d兲 in two orthogonal directions symbolized by the white arrows. simulations are available.
共e兲 Intensity profile along the white line in 共c兲. 共f兲 High-pass-filtered version
of 共c兲 suppressing long-ranged intensity changes. The wavelength used for The optical resolution in optical images 关Figs. 9共c兲, 9共d兲,
illumination was 633 nm. and 9共f兲, respectively兴 can be estimated from the line cut Fig.
9共e兲 to be at least 50 nm.

V. SNOM APPLICATIONS
2. Dielectric grating
A. Amplitude and phase contrast
Phase contrast is obtained when imaging samples induce
In order to obtain a phase or amplitude contrast image of differences in the optical path. A typical example for such a
a sample as discussed in Sec. IV, light emitted from the sample is a dielectric grating. An AFM topograph of such a
probe–sample interaction zone is collected and recorded di- grating is shown in Fig. 10共a兲, accompanied by a height pro-
rectly by means of a suitable photodetector. file in 10共b兲. The grating structure has a period of 383 nm
1. Metal island film and a step height of 8 nm which was etched into a thin glass
substrate.
To demonstrate amplitude contrast, we chose a 15 nm Figure 10共c兲 shows a very high resolution constant
thick aluminum island film produced by means of the latex height mode optical image recorded at a gapwidth below 1
sphere shadow mask technique.89,90 The structure is sketched nm. Figures 10共d兲 and 10共e兲 show constant height mode op-
in Fig. 9共a兲. The objects of interest are the metal patches at tical images recorded gap of a few nm and ⬇100 nm, re-
the interstices between the spheres. They are formed by spectively. This series of images clearly evidences the strong
evaporation and subsequent dissolution of the spheres. To dependence of the resolution on the gapwidth.
characterize their size, we propose to choose the altitude of
the inscribable equilateral triangle as a standard. The length
of this altitude is h⫽( 冑3⫺3/2)d⬇0.23d, where d is the
B. Fluorescence imaging
sphere radius. For the d⫽220 nm latex spheres used for the High resolution optical imaging of biological samples
present test sample, the altitude hence is 50 nm. The nearest- with fluorescent labels is among the most promising fields of
neighbor distance between metal patches is d/ 冑3⫽127 nm. application for SNOM. SNOM has the potential to image the

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7770 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

FIG. 10. Dielectric grating: 共a兲 AFM topography. 共b兲 Line cut through 共b兲. 共c兲 Very high-resolution constant height mode image close to contact. 共d兲,共e兲
Constant height mode allowed images recorded at a gapwidth of a few, and ⬇100 nm, respectively.

distribution of such labels down to the level of single fluo- todiode. Figure 11共a兲 shows a near-field fluorescence image
rophores 共see, for example, Refs. 4,5,93, and 94 and refer- of two rectangular patches of proteins, nominally 500
ences therein兲. ⫻3000 nm2 in size. Some areas of the sample show small
Controllable deposition of well-defined species of bio- irregularities. This feature makes the sample suitable for an
logically active molecules on surfaces is of interest in biol- SNOM resolution test. Figure 11共b兲 shows the simulta-
ogy and biochemistry, since it opens up the possibility for neously recorded shear force topographs of the structure. To
nanometer-scale patterning and to carry out chemical assays estimate the optical resolution, intensity profiles were taken
on a nm scale with minute quantities of reagents.95–97 in the optical image 关Fig. 11共a兲兴 at two positions marked by
We investigated microcontact-printed patches of protein arrows 1 and 2. The profiles are depicted in Figs. 11共c兲 and
molecules, specifically chicken immunoglobulin labeled with 11共d兲 as gray lines. The width of the small feature in Fig.
tetramethyl-rhodamine-isothiocyanate 共TRITC兲,95 using 11共d兲 suggests a resolution of about 80 nm. To exclude the
SNOM. The patches consist of a monolayer of close-packed possibility of a z-motion artifact,88 line cuts were taken at the
protein molecules, ⬇4⫻10⫻14 nm3 in size. Each protein same positions in the topographic image. The results are
molecule is labeled by about 4 dye molecules. The fluoro- shown in Figs. 11共c兲 and 11共d兲 as black lines and prove that
phores were excited at 514 nm and detection was performed the feature in the optical image is not induced by a topo-
above 570 nm, as described in Sec. II A using an oil immer- graphic crosstalk. A comparison of the line profiles taken in
sion objective and a single photon counting avalanche pho- the optical and topographic image in 11共c兲 suggests a shift of

FIG. 11. 共a兲 High-resolution (512

⫻512 pixel兲 scanning near-field fluo-
rescence image of a contact-printed
pattern of TRITC-labeled chicken im-
munoglobulin molecules immobilized
on a glass surface acquired simulta-
neously to the shear force image 共b兲.
The white arrows 共1 and 2兲 mark the
positions where the profiles shown in
共c兲 and 共d兲 were taken. 共c兲 Fluores-
cence and topography profile along 1,
and 共d兲 Fluorescence and topography
profile along 2.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7771

FIG. 12. SNOM–SERS imaging of brilliant cresyl blue labeled DNA fragments on an SERS substrate. Lateral dependence of 共a兲 topography measured in
shear force mode, 共b兲 Raman intensity at 800 cm⫺1 corresponding to Raman scattering of the tip material 共glass兲, 共c兲 Raman intensity at 1641 cm⫺1 共BCB兲
before normalization, and 共d兲 near-field surface enhanced Raman intensity at 1641 cm⫺1 after normalization with the glass signal. In 共b兲, 共c兲, and 共d兲, the pixel
size was only 100⫻100 nm, and for each graph the intensity of one line of the complete Raman spectrum was plotted. In 共c兲 and 共d兲, the contrast between
the lowest and the highest Raman intensity was a factor of 3. The arrows in the spectral images correspond to the lateral positions of the Raman spectra shown
in 共e兲. The square marks a BCB band and the circle denotes a glass band. Note that the BCB signal never vanishes completely.

about 100 nm between the intensity and the topographic images 共b兲, 共c兲, 共d兲 and spectra from selected sample posi-
maximum. This is an effect that often occurs in SNOM im- tions, of stained DNA strands on a specially prepared silver
ages. It can be explained by assuming that a protrusion is sphere substrate. Since the substrate has a rough topography,
located on the rim of the metal coating 关see Figs. 4共c兲 and it is important to carefully distinguish between topographi-
4共d兲兴 that actually is responsible for the shear force interac- cally induced contrast and intensity changes due to absorp-
tion. Thus, the topographic features are expected to be dis- tion of the substrate, and finally the Raman signal of the
placed from the center of the aperture. sample itself. Fortunately reflectivity and Raman intensity
can be separated, at least to a first approximation.
C. Near-field Raman spectroscopy The two images, Figs. 12共b兲 and 12共c兲, which were re-
corded sitting on different Raman bands, show quite differ-
In combination with vibrational spectroscopies, near-
ent features. Figure 12共b兲 was recorded at a glass Raman
field optics can yield molecular contrast on a lateral scale
band of the near-field probe, whereas Fig. 12共c兲 correspond
determined by the aperture size. The main difficulties that
to a Raman signal of the labeled DNA. The signals of 12共b兲
are encountered applying this combination of techniques are
represent the reflectivity 共absorption兲 of the silver sphere
the low efficiency of Raman scattering and the lack of suit-
substrate. Of course, a higher reflectivity of the sample re-
able fiber material for infrared spectroscopy. Nevertheless,
sults in an increase in the detected Raman intensity as well;
there are examples of both near-field Raman,98–101 and near-
therefore, we use the reflectivity signal to correct for the
field infrared spectroscopy.33,102–104 A method to increase the
sample topography107 in all our spectra. The result is dis-
small Raman scattering cross section to make it signals de-
played in Fig. 12共d兲. It demonstrates that one of the promi-
tectable in SNOM experiments is the use of rough noble
nent spots of Fig. 12共c兲 is actually due to a locally increased
metal substrates. Surface enhancement effects observed in
the signals from samples deposited on such substrates can
reach several orders of magnitude depending strongly on the
morphology and shape of the metal features. The reason for
the strong enhancement is still subject to some
controversy.105,106 Two different mechanisms seem to be in-
volved. The so-called pure electromagnetic effect is due to
plasmon resonances of the rough metal films or particles.
The electromagnetic 共near兲 field of a metal particle within
this model can exceed the applied field by orders of magni-
tude. An additional enhancement can be caused by charge
transfer or bond formation between the sample and the me-
tallic substrate, which can strongly enhance the polarizability
of the molecule. FIG. 13. Far-field 共a兲, 共b兲 and near-field Raman spectra 共c兲, 共d兲 of p-xylene.
Figure 12 shows topography 共a兲 and near-field Raman 共a兲,共c兲 Parallel polarization; 共b兲,共d兲 Vertical polarization geometry.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7772 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

FIG. 14. Near-field optical ablation: 共a兲 Absorption spectrum of rhodamine B at a concentration of 10⫺5 M in methanol solution. Arrows indicate the
wavelengths used for the ablation experiments before recording the images on the right. 共b兲,共c兲 Topography of a rhodamine B film after coupling 2.1 ␮ J into
the SNOM tip at ␭⫽650 nm 共b兲, and after coupling 1.4 ␮ J into the SNOM tip at ␭⫽532 nm 共c兲, a wavelength near the maximum absorption of rhodamine
B. The total topographic contrast in the z direction is 10 nm 共b兲, and 26 nm 共c兲, respectively.

reflectivity. The relation between reflectivity and Raman sig- one also has to take into account transient thermal
nal can be seen more clearly by comparing the selected un- expansion69,115,70 as a possible mechanism for the creation of
corrected Raman spectra in Fig. 12共e兲. The arrows in the surface indentations, but it is likely to be negligible.114
optical images correspond to the lateral positions of the cho- We now discuss an experiment that gave very clear re-
sen spectra. Obviously, a substantial change in the ratio of sults about the ablation mechanism from a rhodamine B film.
the BCB–DNA band 共square兲 and the glass scattering This sample was irradiated either on resonance 共532 nm兲 or
共circle兲 occurs only from 1 to 2–3, whereas the absolute off resonance 共650 nm兲 with respect to its optical absorption
intensities of BCB–DNA and glass intensity vary in each
maximum 关Fig. 14共a兲兴. SNOM tips with apertures of ⬍100
spectrum.
nm diameter were used. No significant drop in optical output
In the case of surface enhanced Raman scattering the
of the SNOM tip was measured for similar input energies at
signal intensities can be very high. In fact, single-molecule
sensitivity in experiments using conventional optical micros- 400 and 650 nm input wavelength, although some decrease
copy has already been claimed by several groups.108,109 For is expected due to the increasing difference between wave-
NFO experiments the detection of a few hundred molecules length and aperture size. Figures 14共b兲 and 14共c兲 show shear-
was shown.110 The method thus seems to have the potential force topographic images recorded after several laser pulses
for molecular recognition on a nm scale. had been fired onto the sample. No ablation was observed for
Near-field Raman spectroscopy is not limited only to the off-resonance wavelength 关Fig. 14共b兲, ␭⫽650nm, 2.1 ␮ J
solid surfaces, but can be used as well in liquids.111 Figure pulse energy兴. In contrast, well defined, ⬇70 nm diameter
13 compares the near-field and far-field Raman spectra of 共FWHM兲, 5 nm deep holes were created by a wavelength
xylene. The data demonstrate that the distinct polarization that is strongly absorbed by the sample 关Fig. 14共c兲, ␭⫽532
properties in the near field of an optical probe are useful to nm, 1.4 ␮ J pulse energy兴. This clearly points to an optical
investigate, for instance, the conformation and the degree of ablation mechanism, either photochemical or photothermal.
crystallinity of a sample. Energy transfer by absorption of photons appears to be more
efficient than by a ballistic process. The photon energy used
D. Pulsed laser ablation through SNOM tips here is not sufficient for inducing chemical bond dissociation
Optical probes produced by tube etching are stable in the rhodamine film, in contrast to ablation processes with
enough to withstand several tenths of mJ of pulsed laser ultraviolet laser radiation, where a photochemical ablation
radiation coupled into them.112 The intensity at the aperture mechanism is often used to rationalize the observed
can be in excess of 108 W/cm2 , which is enough to cause phenomena.116 We therefore suggest that a photothermal
ablation of a sample close to the tip. Several mechanisms are mechanism is responsible for the ablation of the rhodamine
possible for laser-induced ablation through SNOM tips.113 film: we believe that the absorbed photon energy is con-
The photochemical model assumes that the energy is depos- verted to vibrational excitation, causing a rapid temperature
ited in the sample via optical absorption, and ablation fol- increase that leads to thermal desorption of the molecules.
lows as a consequence of chemical bond breaking. Similarly,
Another interesting observation in Fig. 14共c兲 is that the
in a photothermal scenario, the absorbed optical energy
ablated material is redeposited on the sample surface close to
would be converted into heat, causing ablation from the lo-
the ablation crater. The distribution of redeposited material
cally heated spot. Third, a ballistic mechanism is conceiv-
able, where material from the tip is sputtered onto the was not symmetric in this experiment, but was always to the
sample, causing ablation of material from its surface. A bal- left side of the crater, independent of the scan direction. In
listic mechanism was the most likely explanation for an ex- other SNOM ablation experiments, the distribution was more
periment carried out on an anthracene sample using a com- symmetric. The origin of the directionality appears to be the
pletely metallized SNOM tip.114 In the absence of any tip itself: frequently, the apex of an SNOM tip exhibits some
photons reaching the sample, sputtering by either metal at- uncontrolled asymmetry, causing the ablated material to be
oms from the tip, or by material adsorbed on its surface, was blown off in a specific direction. This has interesting appli-
believed to be the reason for the ablation process. Fourth, cations for the analytical nanosampling of material surfaces.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Scanning near-field optical microscopy 7773

VI. PERSPECTIVES sample, and detection scheme兲 and conversely to experimen-

tal difficulties to precisely control all the relevant parameters.
With this review, by focusing on fundamentals and se-
As a matter of fact, our simulations illustrated, for example,
lected applications of aperture SNOM, we cover only a small
in Fig. 8, the importance of the illumination polarization for
part of the research currently pursued in the area of NFO. For
image interpretation; still, the exact polarization state of the
a more complete picture the reader is referred to other NFO
field emerging from a real SNOM tip is difficult to control.
review articles in this issue and to recent reviews.2–5,9
One might wish, for the future, that experimentalists try to
The rapid development of NFO techniques is far from
design their experiments such that they can be compared to
coming to an end. Many problems remain to be solved and
theoretical models. On the other hand, theoreticians have to
many careful experiments are still to be done.
extend their numerical capabilities. Both sides would profit
With respect to the technical problem of tip fabrication,
from such collaborations.
there is still plenty of room for improvements, in particular
regarding tip stability, damage threshold, and transmission
ACKNOWLEDGMENTS
coefficients. Microfabricated near-field optical probes55,56
will help to solve these problems and improve reproducibil- The authors wish to thank A. Bernard for the kind gift of
ity by providing better defined experimental conditions. the contact printed sample and L. Novotny for helpful dis-
Also, new tip concepts which deviated from the simple ap- cussions and for providing the tip transmission data. We fur-
erture scheme may be of advantage. ther acknowledge financial support from the NFP36 and the
Another possibility to improve the performance of aper- ETH-Zürich; O.J.F.M. gratefully acknowledges support from
ture probes is the use of ultraviolet 共UV兲 light.117–122 This the Swiss National Science Foundation.
will shift the cutoff region in the taper towards the apex and
1
therefore strongly increase the transmission coefficient. This D. Pohl, in Scanning Near-field Optical Microscopy (SNOM), Advances in
can be exploited by using smaller apertures. In addition, di- Optical and Electron Microscopy Vol. 12 共Academic, New York, 1991兲,
pp. 243–312.
rect excitation of certain biomolecules might become pos- 2
H. Heinzelmann and D. Pohl, Appl. Phys. A: Mater. Sci. Process. 59, 89
sible instead of using fluorescent markers. Furthermore, in 共1994兲.
Raman spectroscopy it is advantageous to use UV excitation 3
C. Girard and A. Dereux, Rep. Prog. Phys. 59, 657 共1996兲.
because of the ␯ 4 dependence of the scattering efficiency and
4
R. Dunn, Chem. Rev. 99, 2891 共1999兲.
5
P. Barbara, D. Adams, and D. O’Connor, Annu. Rev. Mater. Sci. 29, 433
because of the potential of resonant field enhancement at 共1999兲.
small silver particles in this spectral region 共for a review, see 6
Near Field Optics, NATO ASI Series E: Applied Sciences, Vol. 242, edited
Ref. 123兲. by D. Pohl and D. Courjon 共Kluwer Academic, Dordrecht, 1993兲, Pro-
Apertureless techniques are becoming more and more ceedings of NFO I: Besançon/Arc & Senans, France, 26–28 October
1992.
important.26–34,124 The reported resolutions are exciting, but 7
Near Field Optics, Ultramicroscopy, Vol. 57, edited by P. Kruit and M.
their utilization for relevant applications such as fluorescence Isaacson 共North-Holland, Elsevier, Amsterdam, 1995兲, Proceedings of
imaging seems to be less straightforward than for aperture NFO II, Raleigh, NC, 20–22 October 1993.
8
SNOM. This is mainly related to the difficulties of discrimi- Near Field Optics and Related Techniques, Ultramicroscopy, Vol. 61,
edited by P. Kruit, M. Paesler, and N. van Hulst 共North Holland, Elsevier,
nating between effects generated in the enhanced field at the Amsterdam, 1995兲, Proceeding of NFO III, Brno, Czech Republic, 9–11
tip apex and the background related to the diffraction limited May 1995.
9
spot illuminating the tip. For fluorescence imaging, fluores- Proceedings NFO-5, J. Microscopy, Vol. 194, edited by S. Kawata
共Blackwell Science Ltd., Oxford, 1999兲, Proceeding of NFO 5, Shira-
cence resonant energy transfer 共FRET兲 microscopy,125–128 hama, Japan, 6–10 December 1998.
another apertureless contraption, is one of the most promis- 10
J. W. Goodman, Introduction to Fourier Optics, Physical and Quantum
ing candidates for a significant improvement of resolution. Electronics Series 共McGraw-Hill, New York, 1968兲.
We also did not mention all of the numerous new con-
11
M. Abbé, Archiv. Mikroscop. Anat. Entwicklungsmech. 9, 413 共1873兲.
12
T. Wilson and C. J. R. Sheppard, Theory and Practice of Scanning Optical
trast mechanisms that are constantly evolving. We did not Microscopy 共Academic, London, 1984兲.
mention, e.g., time-resolved SNOM, which combines na- 13
Handbook of Biological Confocal Microscopy, edited by J. Pawley 共Ple-
nometer scale optical resolution with femtosecond time reso- num, New York, 1995兲.
lution. A number of groups are already working in this
14
E. Synge, Philos. Mag. 6, 356 共1928兲.
15
E. Synge, Philos. Mag. 11, 65 共1931兲.
area.129–132 Pulsed laser techniques in principle provide the 16
G. Binnig and H. Rohrer, Helv. Phys. Acta 55, 726 共1982兲.
possibility to utilize nonlinear optical techniques like second 17
D. Pohl, W. Denk, and M. Lanz, Appl. Phys. Lett. 44, 651 共1984兲.
harmonic generation, hyper-Raman, etc. Such techniques can 18
U. Dürig, D. Pohl, and F. Rohner, J. Appl. Phys. 59, 3318 共1986兲.
help to avoid stray light problems in SNOM34,124 and also
19
D. Pohl, US Patent US4,604,520 共1986兲.
20
A. Lewis, M. Isaacson, A. Harootunian, and A. Muray, Ultramicroscopy
can improve the signal-to-noise ratio. 13, 227 共1984兲.
As another example, the combination of local desorption 21
A. Harootunian, E. Betzig, M. Isaacson, and A. Lewis, Appl. Phys. Lett.
using SNOM with mass spectrometric detection provides a 49, 674 共1986兲.
22
E. Betzig, M. Isaacson, and A. Lewis, Appl. Phys. Lett. 51, 2088 共1987兲.
promising new analytic method.133 23
E. Betzig and J. Trautman, Science 257, 189 共1992兲.
From our point of view, SNOM is a field where the 24
H. Heinzelmann, B. Hecht, L. Novotny, and D. Pohl, J. Microsc. 177, 115
interaction between modeling and experiment can be ex- 共1994兲.
tremely fruitful. This interaction, however, has not yet been
25
B. Hecht, H. Heinzelmann, and D. Pohl, Ultramicroscopy 57, 228 共1995兲.
26
U. Fischer and D. Pohl, Phys. Rev. Lett. 62, 458 共1989兲.
as successful as one might wish. This is probably related to 27
F. Zenhausern, Y. Martin, and H. Wickramasinghe, Science 269, 1083
the intrinsic limitations of numerical simulations to account 共1995兲.
for a realistic experimental configuration 共including tip, 28
Y. Inouye and S. Kawata, J. Microsc. 178, 14 共1995兲.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.
7774 J. Chem. Phys., Vol. 112, No. 18, 8 May 2000 Hecht et al.

29
P. Gleyzes, A. Boccara, and R. Bachelot, Ultramicroscopy 57, 318 共1995兲. 80
W. Moerner et al., Phys. Rev. Lett. 73, 2764 共1994兲.
30
C. Girard and O. M. A. Dereux, Phys. Rev. Lett. 75, 3098 共1995兲. 81
See Ref. 3.
31
J. Koglin, U. Fischer, and H. Fuchs, J. Biomed. Opt. 1, 75 共1996兲. 82
J.-J. Greffet and R. Carminati, Prog. Surf. Sci. 56, 133 共1997兲.
32
T. Sugiura, S. Kawata, and T. Okada, J. Microsc. 194, 291 共1999兲. 83
O. J. F. Martin, J. Microsc. 194, 235 共1999兲.
33
B. Knoll and F. Keilmann, Nature 共London兲 399, 134 共1999兲. 84
O. J. F. Martin, in Near-field Optics: Physics, Devices, and Information
34
E. Sanchez, L. Novotny, and X. Xie, Phys. Rev. Lett. 82, 4014 共1999兲. Processing, Proceedings of SPIE Vol. 3791 共SPIE, Bellingham, 1999兲, pp.
35
D. Courjon, K. Sarayeddine, and M. Spajer, Opt. Commun. 71, 23 共1989兲. 1–8.
36
R. Reddick, R. Warmack, and T. Ferrell, Phys. Rev. B 39, 767 共1989兲. 85
O. J. F. Martin and N. B. Piller, Phys. Rev. E 58, 3909 共1998兲.
37
O. Marti et al., Opt. Commun. 96, 225 共1993兲. 86
N. B. Piller and O. J. F. Martin, IEEE Trans. Antennas Propag. 46, 1126
38
J. Krenn et al., Phys. Rev. Lett. 82, 2590 共1999兲. 共1998兲.
39
A. Choo et al., Ultramicroscopy 57, 124 共1995兲. 87
J. D. Jackson, Classical Electrodynamics, 3rd ed. 共Wiley, New York,
40
N. van Hulst, M. Moers, and E. Borgonjen, NATO Series E: ‘‘Photons and 1999兲.
Local Probes,’’ edited by O. Marti and R. Möller 共Kluwer Academic, 88
B. Hecht et al., J. Appl. Phys. 81, 2492 共1997兲.
Dordrecht, 1995兲, Vol. E 300, p. 165. 89
U. C. Fischer and H. P. Zingsheim, J. Vac. Sci. Technol. 19, 881 共1981兲.
41
R. Toledo-Crow, P. Yang, Y. Chen, and M. Vaez-Iravani, Appl. Phys. 90
H. W. Deckman and J. H. Dunsmuir, Appl. Phys. Lett. 41, 377 共1982兲.
Lett. 60, 2957 共1992兲. 91
L. Novotny and D. Pohl, in Ref. 40, Vol. 300, p. 21.
42
E. Betzig, P. Finn, and S. Weiner, Appl. Phys. Lett. 60, 2484 共1992兲. 92
L. Novotny, D. W. Pohl, and B. Hecht, Ultramicroscopy 61, 1 共1995兲.
43
K. Karrai and R. Grober, Appl. Phys. Lett. 66, 1842 共1995兲. 93
M. Garcia-Parajo et al., Bioimaging 6, 43 共1998兲.
44
B. Hecht, Forbidden Light Scanning Near-Field Optical Microscopy, 94
N. van Hulst, J. Chem. Phys. 112, 7799 共2000兲, this issue.
Ph.D. thesis, University of Basel 共Hartung–Gorre, ISBN 3-89649-072-9, 95
A. Bernard et al., Langmuir 14, 2225 共1998兲.
96
Konstanz, 1996兲. T. Schmidt, G. Schütz, H. Gruber, and H. Schindler, Anal. Chem. 68,
45
G. Tarrach, M. Bopp, D. Zeisel, and A. Meixner, Rev. Sci. Instrum. 66, 4397 共1996兲.
3569 共1995兲. 97
W. Trabesinger et al., Anal. Chem. 71, 279 共1999兲.
46
M. Pfeffer, P. Lambelet, and F. Marquis-Weible, Rev. Sci. Instrum. 68, 98
D. A. Smith et al., Ultramicroscopy 61, 247 共1995兲.
4478 共1997兲. 99
C. L. Jahncke, H. D. Hallen, and M. A. Paesler, J. Raman Spectrosc. 27,
47
T. Okajima and S. Hirotsu, Appl. Phys. Lett. 71, 545 共1997兲. 579 共1996兲.
48
S. Davy, M. Spajer, and D. Coujon, Appl. Phys. Lett. 73, 2594 共1998兲. 100
D. Zeisel et al., Anal. Chem. 69, 749 共1997兲.
49 101
M. Gregor, P. Blome, J. Schöfer, and R. Ulbrich, Appl. Phys. Lett. 68, 307 S. Webster, D. N. Batchelder, and D. A. Smith, Appl. Phys. Lett. 72,
共1996兲. 1478 共1998兲.
50
H. Bielefeldt, Ph.D. thesis, University of Konstanz, 1994. 102
M. K. Hong et al., SPIE 2863, 54 共1996兲.
51
A. Ruiter, J. Veerman, K. van der Werf, and N. van Hulst, Appl. Phys. 103
A. Piednoir, C. Licoppe, and F. Creuzet, Opt. Commun. 129, 414 共1996兲.
Lett. 71, 28 共1997兲. 104
A. Lahrech, R. Bachelot, P. Gleyzes, and A. C. Boccara, Appl. Phys.
52
B. Hecht, D. Pohl, H. Heinzelmann, and L. Novotny, in Ref. 40, Vol. 300, Lett. 71, 575 共1997兲.
p. 93. 105
M. Moskovits, Rev. Mod. Phys. 57, 783 共1985兲.
53
T. Huser et al., J. Opt. Soc. Am. A 16, 141 共1999兲. 106
A. Otto, I. Mrozek, H. Grabhorn, and W. Akemann, J. Phys.: Condens.
54
B. Hecht et al., J. Appl. Phys. 84, 5873 共1998兲. Matter 4, 1143 共1992兲.
55
S. Münster et al., J. Microsc. 186, 17 共1997兲. 107
V. Deckert, D. Zeisel, and R. Zenobi, Anal. Chem. 70, 2646 共1998兲.
56
W. Noell et al., Appl. Phys. Lett. 70, 1236 共1997兲. 108
K. Kneipp et al., Phys. Rev. Lett. 78, 1667 共1997兲.
57
E. Betzig et al., Science 251, 1468 共1991兲. 109
S. Nie and S. R. Emory, Science 275, 1102 共1997兲.
58
G. Valaskovic, M. Holton, and G. Morrison, Appl. Opt. 34, 1215 共1995兲. 110
D. Zeisel, V. Deckert, R. Zenobi, and T. Vo-Dinh, Chem. Phys. Lett.
59
D. Turner, US Patent 4,469,554 共1984兲. 283, 381 共1998兲.
60 111
P. Hoffmann, B. Dutoit, and R.-P. Salathé, Ultramicroscopy 61, 165 J. Grausem, B. Humbert, and A. Burneau, Appl. Phys. Lett. 70, 1671
共1995兲. 共1997兲.
61 112
D. Zeisel, S. Nettesheim, B. Dutoit, and R. Zenobi, Appl. Phys. Lett. 68, See Ref. 63.
2491 共1996兲. 113
B. Dutoit, D. Zeisel, V. Deckert, and R. Zenobi, J. Phys. Chem. B 101,
62
T. Yatsui, M. Kourogi, and M. Ohtsu, Appl. Phys. Lett. 73, 2090 共1998兲. 6955 共1997兲.
63
R. Stöckle et al., Appl. Phys. Lett. 75, 160 共1999兲. 114
D. Zeisel, S. Nettesheim, B. Dutoit, and R. Zenobi, Appl. Phys. Lett. 68,
64
P. Lambelet et al., Appl. Opt. 37, 7289 共1998兲. 2491 共1996兲.
65
M. Born and E. Wolf, Principles of Optics, 6th ed. 共Cambridge University 115
H. J. Mamin, Appl. Phys. Lett. 69, 433 共1996兲.
116
Press, Cambridge, 1980兲. D. Singleton, G. Paraskevopoulos, and R. S. Taylor, Chem. Phys. 144,
66
L. Novotny and C. Hafner, Phys. Rev. E 50, 4094 共1994兲. 415 共1990兲.
67 117
A. L. Rosa, B. Yakobson, and H. Hallen, Appl. Phys. Lett. 67, 2597 I. Smolyaninov, D. Mazzoni, and C. Davis, Appl. Phys. Lett. 67, 3859
共1995兲. 共1995兲.
68
D. Kavaldjiev, R. Toledo-Crow, and M. Vaez-Iravani, Appl. Phys. Lett. 118
S. Monobe et al., Opt. Commun. 146, 45 共1998兲.
67, 2771 共1995兲. 119
A. Kirsch et al., Biophys. J. 75, 1513 共1998兲.
69
M. Stähelin et al., Appl. Phys. Lett. 68, 2603 共1996兲. 120
A. Jenei et al., Biophys. J. 76, 1092 共1999兲.
70
C. Lienau, A. Richter, and T. Elsaesser, Appl. Phys. Lett. 69, 325 共1996兲. 121
S. Nishikawa and T. Isu, J. Microsc. 194, 15 共1999兲.
71
L. Novotny, D. Pohl, and B. Hecht, Opt. Lett. 20, 970 共1995兲. 122
A. Kramer et al., Biophys. J. 78, 458 共2000兲.
72
H. Bethe, Phys. Rev. 66, 163 共1944兲. 123
H. Metiu, Prog. Surf. Sci. 17, 153 共1984兲.
73
C. J. Bouwkamp, Philips Res. Rep. 5, 321 共1950兲. 124
Y. Kawata, C. Xu, and W. Denk, J. Appl. Phys. 85, 1294 共1999兲.
74
D. V. Labeke, D. Barchiesi, and F. Baida, J. Opt. Soc. Am. A 12, 695 125
V. L. S. K. Sekatskii, Appl. Phys. B: Lasers Opt. 63, 525 共1996兲.
共1995兲. 126
V. L. S. K. Sekatskii, JETP Lett. 65, 465 共1997兲.
75
C. Hafner, The Generalized Multiple Multipole Technique for Computa- 127
S. Vickery and R. Dunn, Biophys. J. 76, 1812 共1999兲.
tional Electromagnetics 共Artech, Boston, 1990兲. 128
G. Shubeita, S. Sekatskii, M. Chergui, and G. Dietler, Appl. Phys. Lett.
76
J. A. Veerman, A. M. Otter, L. Kuipers, and N. F. van Hulst, Appl. Phys. 74, 3453 共1999兲.
Lett. 72, 3115 共1998兲. 129
J. Levy et al., Phys. Rev. Lett. 76, 1948 共1996兲.
77
H. Levine and J. Schwinger, Commun. Pure Appl. Math. 3, 355 共1950兲. 130
A. Richter et al., Appl. Phys. Lett. 73, 2176 共1998兲.
78
O. J. F. Martin, C. Girard, and A. Dereux, J. Opt. Soc. Am. A 13, 1801 131
B. A. Nechay et al., Appl. Phys. Lett. 74, 61 共1999兲.
共1996兲. 132
T. Guenther et al., Appl. Phys. Lett. 75, 3500 共1999兲.
79
B. Hecht et al., Phys. Rev. Lett. 77, 1889 共1996兲. 133
D. A. Kossakovski et al., Ultramicroscopy 71, 111 共1998兲.

Downloaded 08 Jan 2001 to 129.132.41.15. Redistribution subject to AIP copyright, see http://ojps.aip.org/jcpo/jcpcpyrts.html.