1. Introduction

Over the past ten years, Apertureless Scanning Near-field Optical Microscopy (ASNOM) has been developed for various applications such as molecular scale fluorescence and optical imaging with a resolution better than the diffraction limit [1–9]. While many configurations of ASNOM are used to accommodate different samples and optical contrast mechanisms, one of the most utilized techniques uses a large numerical aperture objective that focuses the light to a diffraction-limited region [6–9]. This approach produces a reduced excitation volume for pinpointing regions of interest, as well as producing a higher flux of photons if desired. Furthermore, by controlling the input polarization, different modes and polarizations can be produced. Optical near-field confinement is then achieved by placing a nanometric tip in the focal spot. The presence of the tip introduces a local perturbation whose nature depends on the distribution of polarization states within the spot. In particular, an electromagnetic field enhancement can occur at the tip apex where field components parallel to the tip axis are dominant [9]. This ASNOM configuration yields high-resolution imaging and spectroscopy including nanometer scale Raman spectroscopy [7,8]. While these enhanced optical interactions are far-field detected, little is known about the perturbation of the focal region induced by the presence of the tip. In this letter, we use a photochemical method that produces topographical changes in a polymer to reproduce the near-field and far-field modes produced with the diffraction limited illumination configuration. Gold tips are compared to dielectric tips for confirming the enhanced near-field produced with metal tips. Our approach relies on analysis of the topographical deformation of an azobenzene-containing photopolymer which self-develops upon illumination. The optically induced topography is characterized after exposure by atomic force microscopy. Studies of this type are important for understanding the electromagnetic field distribution surrounding nanometric metal probes, and should lead to the optimization of tip design and illumination conditions for controlling light-matter interactions.

2. Experimental

The experimental configuration is depicted in Fig. 1(a). It is based on an inverted optical microscope. The ~35 nm thick polymer films were exposed using an excitation wavelength λ of 532 nm located within the absorption band of our compound. A set of neutral density filters and a mechanical shutter were placed in the beam path to control the exposure time and the intensity. The beam was spatially filtered to give a clean Gaussian profile with an adjustable linear polarization. The light beam was focused onto the sample by a high-numerical objective lens (60 X, N.A.=1.4) to a spot size of approximately 250–300 nm. The self-developing photopolymer is a copolymer MMA-Co-Mdr1 made of polymethylmethacrylate (PMMA).[10] The photopolymer is composed of an inert matrix (PMMA) at the wavelength of the laser beam while the absorption is provided by Disperse Red One dye (DR1) which is grafted on the main matrix chain. The dye is characterized by a reversible trans to cis photoisomerization process. The conformations of the two states are depicted in Fig. 1(b). During isomerization cycles, the photopolymer is known to migrate according to the light intensity gradient, away from the high intensity regions and along the direction of polarization, if the polarization lies in the plane of the film [10]. Thus, the film decreases in thickness in these high intensity regions. Interestingly, some recent near-field studies showed that an increased thickness occurs when the local polarization is mainly composed of enhanced field vectors polarized perpendicularly to the plane of the film [11]. While the mechanism for this difference is not entirely characterized, it may be due to bleaching of the chromophores at high energy density as this polarization does not produce movement of the chromophores out of the illumination direction [10]. The subsequent absence of trans-cis isomerization processes produces different light-polymer interactions that lead to protrusions rather than depletions where the intensities are maximized [11].

 

Fig. 1. (a) Experimental arrangement for recording the local perturbation induced in a diffraction-limited spot by an ASNOM probe. Inset: SEM image of a gold tip (b) isomerization of the azobenzenic molecule.

Download Full Size | PDF

Gold tips were produced by electrochemical polishing in HCl solution [6]. A typical 30 nm radius gold tip is shown in the inset of Fig. 1(a). Additionally, glass tips were used as a control [12]. The tip was approached to the polymer by shear force microscopy [13] and positioned in the focal region. After illumination, the topographical deformations of the polymer were characterized by tapping-mode atomic force microscopy (TMAFM).

3. Results

Figures 2(a) and 2(b) show the TMAFM images of the surface of the polymer after being exposed (laser power=50µW, exposure time=2s) by a focused Gaussian beam with two orthogonal linear polarizations in the plane of the polymer without the presence of any tip. The inset in Fig. 2(a) shows an array of different illuminated regions, illustrating the reproducibility of the experiment. The focal field of a tightly focused laser beam polarized along the x axis is composed of two transverse electric components Ex, Ey and one longitudinal one Ez. Ey and Ez are typically one order of magnitude weaker than Ex for a Gaussian excitation [14]. Therefore, it is reasonable to assume that the modifications of the polymer surface is mostly influenced by focal fields oriented along the main polarization. The topography is in agreement with our current knowledge of the polymer photo-deformation, i.e. that the molecules migrate away from the regions of maximum intensity along the incident polarization in the plane of the film [10]. As a result, and as observed in Figs. 2(a) and (b), the reorganization of the copolymer surface leads to a depletion in the central focal region surrounded by two lateral protuberances along the direction of the polarized light. Additionally, the longitudinal fields present in the focal region could also weakly contribute to the origin of the maxima. The polymer does not respond linearly even at these low optical intensities (<1mW/µm²), creating a significantly wider depression than the actual focal spot size. The images of Figs. 2(a) and (b) have similar features but show significant differences in their details. The degree of polarization of the incident beam was checked after each optical element without revealing any asymmetries between the two orthogonal polarizations. Anisotropy of the polymer is also a very unlikely origin considering the fabrication process and the random distribution of the dye within the matrix. Further experiments are currently under way to understand the origin of the fine details of the photo-induced surface changes. The reorganization of the copolymer without the perturbative presence of a probe for the two polarizarization are considered as a reference for the following experiments involving gold and glass tips.

 

Fig. 2. AFM images of the polymer surface taken after the exposure. The white arrows correspond to the incident polarization (a), (b): exposure without any tip. Note that there is no change in images (a) or (b) in the presence of a glass tip. Inset in (a): array of dots written under the same condition as for (a). (c)-(f): exposure with a gold tip placed in the focal spot. (e) and (f) are smaller scan areas of the focal regions (c) and (d). The dashed circle in Figs. 2(c)-(f) highlight the near-field response of the polymer when the tip is present.

Download Full Size | PDF

Figures 2(c) to (d) show the surface modifications of two areas illuminated with two different orthogonal polarizations with a gold tip located in the center of the focal spot. Figures 2(e) to (f) are detailed scans over the same regions as for Figs. 2(c) and (d). The interaction between the gold tip and the surface produces a local sub-wavelength elongated protuberance parallel to the incident polarization. The dimensions of this structure are, on average, 230 nm in the direction of the polarization and 130 nm in the direction perpendicular to it. Note that the convolution over the AFM tip diameter mean that this feature is smaller. The AFM tip diameter of ~10 nm indicates that the protuberance is at least 10% smaller than these dimensions, and the optical near-field has dimensions well below the diffraction limit.

4. Discussion

To understand the origin of the deformation caused by the presence of the tip, the spatial distribution of the electric field surrounding the tip extremity has been evaluated by electromagnetic calculations based on a three dimensional Finite Difference Time Domain (FDTD) method.[15] The results of the calculation are presented in Fig. 3. Figure 3(a) presents the geometry of the problem. The tip was placed at a 5 nm distance from the polymer surface (n=1.7) and was illuminated by an incident plane wave linearly polarized along the x axis and propagating along the z-axis. A plane wave model is justified because the depolarization of a Gaussian beam is weak at the focus plane of a high-numerical aperture [14], so that the center region is Gaussian-type and polarized along the incident polarization.

Figures 3(b) to (d) and Figs. 3(e) to (g) show the amplitude of the components of the electric field (Ex, Ey, Ez) on the polymer surface for a gold tip and a glass tip, respectively. We note, via the differing y-axis scales for Figs. 3(b)-(d) and those of Fig. 3(e)-(g), that the field amplitude for Ex and Ez in the vicinity of the gold tip is higher than for the glass tip. The calculations also show that Ey can be neglected for both the dielectric and the metallic probes. Importantly, the calculation shows that due to a significant depolarization, the tip’s near field appears to be mainly composed of longitudinal field Ez as well as a lateral field Ex. The latter gives rise to an almost homogenous central spot beneath the tip. Figures 3(d) and 3(g) show a strong depolarization from Ex to Ez, resulting in two lobes polarized along the tip axis. Figure 3 shows that the two components (Ex, Ez) of the gold tip near-field have to be taken into account for interpretation of the local polymer perturbation. Figure 3 shows that the amplitude of Ez is enhanced by a factor of 2 to 4 with respect to Ex in the region of significant depolarization (i.e., where Ez has the largest amplitude). The light that is depolarized and oriented along the metal tip z-axis is enhanced through a mechanism analogous to the lightning rod effect [16]. However, the magnitude of the enhancement is weaker than for the case of proper illumination polarization along z [6], but nonetheless a significant component of the field is produced on the polymer surface by the tip that is polarized perpendicular to the film plane. The magnitudes of the fields do not suggest a significant plasmon resonance in this illumination region. This removes possible heating concerns of the gold tip near the high-extinction plasmon resonance [17]. Furthermore, recent calculations show that the steady-state temperature increase of a gold tip caused by laser heating is miniscule for the laser intensities and illumination geometry used in our experiment [16]. In order to discriminate possible mechanical interactions from the pure optically induced deformations, a glass tip was placed in the focal spot. In the case of a glass tip, the field amplitude Ex is reduced in a region located directly under the tip (see Fig. 3(e)). Consequently, the photopolymer should migrate toward this local field minimum and produces a protuberance directly facing the tip extremity. However, a comparison of the magnitude of the local fields Ex for the two tip materials indicates that modifications of the surface introduced by a glass tip must be small compared to reorganization initiated by a gold tip. This conclusion is supported by the fact that no topographical perturbation was observed inside the depletion of the diffraction pattern. TMAFM images identical to Figs. 2(a) and 2(b) were obtained after exposure. In other words, neither the mechanical interactions between the probe and the copolymer, nor the local field distribution are sufficient to modify significantly the chromophores reorganization; unlike for recent experiments using contact mode AFM [18].

 

Fig. 3. Calculated amplitude of the different field components on the polymer surface beneath a 40 nm radius tip. (a) geometry of the problem. (b) to (d): case of the gold tip. (e) to (g): case of the glass tip. The size of the calculated images is 300×250 nm².

Download Full Size | PDF

Based on both the experimental results and the numerical calculations, we propose the following mechanism to explain the subwavelength patterns. The process is illustrated by Fig. 4, where the red solid single and doubled arrows represent the components of the electric field Ex and Ez, respectively. The dotted blue arrows represent the resulting displacement of the photopolymer, again with single and double lines corresponding to the associated optical fields. As described above, the far-field response of the polymer along the direction of polarization results in a depletion of polymer surrounded by two protrusions, as shown in Fig. 4(a) in conjunction with a possible weak effect due to the focal longitudinal component. In the presence of the tip, the near-field components of Ex will also laterally displace the polymer following the same mechanism. Furthermore, based on previous observations, [10,11] one might expect that the field component oriented along the z direction should produce protrusions. The different spatial profiles of the confined Ex and Ez fields give us an opportunity to discern the competing forces on the polymer. A cross-section of Fig. 2(f), shown in Fig. 4(b), illustrates the net effect. The matter in the near-field zone of the tip first shows a distinct increase in height of approximately 40nm above the floor. This is entirely consistent with prior studies where strong Ez components produce protrusions in the polymer [10,11]. In the center of the elevated region near the tip, a dip is also observed that is consistent with the confined Ex field pushing matter laterally away from the tip. Thus, the spatial features of the exposed photopolymer support both the presence and predicted modes of Ex and Ez components in the confined fields near the tip. The TMAFM images are affected by the imaging tip diameter, so that protusions appear larger than they are and depressions will appear smaller. This topography and the small peanut-like shape parallel to the incident polarization visible in Fig. 2(e) are in agreement with the proposed displacement mechanism.

 

Fig. 4. (a) Experimental topography profile along the pattern of Fig. 2(d) without the presence of the tip. (b) Experimental topography profile in the presence of the metal tip extracted from Fig. 2(f). The red single and double solid arrows represent the different orientation of the electric field components Ex and Ez, respectively. The dotted blue arrows indicate the corresponding response of the polymer to the electric field components.

Download Full Size | PDF

5. Conclusions

We have used topographic changes in a polymer film to reproduce the local perturbation of a diffraction-limited spot by a metallic ASNOM probe. Our method relies on the observation of the photoinduced migration of azobenzene molecules grafted to a polymer. We have observed the consequence of lateral charge density at the edges of the extremity of the metal tip, associated with a strong near-field depolarization. A possible mechanical origin of the observed pattern has been ruled out by the use of glass tips as a reference. Our approach is believed to represent a powerful tool for understanding optical confinement near metal nanoprobes used in near-field optics as well as opening the door to optically induced manipulation of molecules by metal tips. Further efforts to realize this goal could center on an improved degree of confinement of the enhanced field through illumination with a higher-order laser beam [9].