ABSTRACT: The validity of reflection-back-to-the-fiber SNOM (scanning near-field optical microscopy) has been unduely questioned by an erratic approach curve that disputed the enhancement of near-field reflectance. It is shown now that only truncated (broken) tips without metal coating do not experience the enhancement when approached close to shear-force distance. However, sharp uncoated tips continue to show up the near-field enhancement, and chemical contrast on rough surfaces continues to be of basic value at submicroscopic resolution. It is pointed out how a good tip may be immediately differentiated from a broken one.
KEYWORDS: near-field reflectance; sharp tips; broken tip artefacts; SNOM; submicroscopic resolution; chemcal contrast
Scanning near-field optical microscopy (SNOM) has been shown to provide submicroscopic resolution with tapered uncoated tips that pick up reflected light with increased intensity when approached to constant shear-force distance above organic1,2 and inorganic3 surfaces. The phenomenon of increased near-field reflectivity has first been described by Courjon et al.4 However, in contrast to the well established facts, an intriguing paper of Sandoghdar et al.5 claims decrease of that reflectivities upon shear-force approach due to interference effects and disputed the sudden discontinuity in the reflection signal as one approaches the tip to the sample. Unfortunately, that paper5 confused the issue without providing relevant experimental details. In view of numerous approach curves that clearly validate (i) the near-field enhancement on various surfaces6,7 and (ii) the repeatable chemical contrast determinations on rough organic surfaces we address the issue with some of our common approach records and evaluate the possible errors at the failure of Sandoghdar et al.5 to observe "the sudden increase".
EXPERIMENTAL, TIP CHARACTERIZATION
Multi-mode and single-mode (SMC-A0515B) fibers of 125 µm from Spectran were pulled as sharp as possible using a Sutter Instruments P-2000 laser-based fiber puller for our experiments at 488 nm. At optimized pulling parameter settings (Heat 235, Filament 0, Velocity 5, Delay 120, Pull 160) the pulling time was about 120 ms. Microscopic inspection indicated a total probe length of about 0.5 mm. SEM micrographs gave radii of curvature of the probe tips of about 15 nm and taper angles of typically 18 - 20°. Resonance frequencies were about 150 kHz and Q-factors ranged from 200 to 300 (see Ref. 7 for further details). The broken tips were judged under a microscope at 400-fold magnification. The force setting for automatic coarse and close approach at a DME RASTERSCOPE 4000 SNOM was 0.017 nN for a force constant of 1.024 N m-1 in these experiments. The level of parasitic light was measured to be 14 % of the background intensity by disconnecting the unengaged illuminated fiber.
When an illuminated multi-mode tip (488 nm, total exit intensity 200 µW) was coarse-approached to a polyvinylpyrrolidone layer on glass at an initial recorder reading of 180 mV, a considerable increase of reflected light intensity was obtained upon the short touch of the setpoint for the shear-force damping. From there the tip is forced to move to the fixed close approach start position (Fig. 1). The scan of the chemically uniform surface after close approach gave a recorder reading of 860 mV, more than 4.7 times of the total background (Fig. 1). With the single-mode tip under equal conditions the increase was 4.8 fold. Thus, without necessity for assessment of the precise tip to sample distance we do observe Courjon´s enhanced near-field reflectivity4 in the approach record (Fig. 1). Similar near-field increases were obtained when organic crystal surfaces1, for example (110) of anthracene etc., were approached with single-mode or multi-mode sharp tips.
Further, we took approach records of tips which had suffered the loss of their sharp cone end, in order to evaluate the conditions for the failure that was claimed by Sandoghdar et al.5 As expected, we got a decrease (from 520 mV to 365 mV recorder reading) in the reflected light intensity that came back through the truncated "tip" (2 µm wide, Fig. 2). Hence only broken or truncated tips reproduce Sandoghdar´s claims, and such tips do produce artefacts both in the topographic and in the optical image. Only sharp tips are suitable for artefact-free SNOM measurements with submicroscopic resolution.1
As Sandoghdar et al. did not report any details of their tip5 we also checked the response of still more heavily truncated tips. Extremely blunt truncated cones (40 µm wide) did not give a change in reflected light intensity when approached under shear-force control, at least in 10 µm scans.
Our results indicate, that Sandoghdar et al.5 have either used uncoated tapered tips that were truncated or broken, or that their tips broke during their approach procedure. Broken tips are useless for collection of enhanced near-field reflected light and do give topographic artefacts but no SNOM-response: any chemically and physically imposed correlations between AFM and SNOM contrasts are lost when running broken tips under shear-force control. Hence a warning is necessary, because every tip could be broken or eventually break during approach and/or scanning:
Approach records similar to Fig.1 must be documented in order to ensure the near-field enhancement prior to every SNOM measurement!
Furthermore, some of the reported changes in contrast behavior during scans may be due to tip breakage. Experimentally, tip breakage is immediately recognized by the considerable decrease in the response signal. Good tips do, of course, not show up a taper end under a light microscope, whereas unsuitably truncated (broken) tips do. Properly performed reflection-back-to-the-fiber SNOM continues to provide valuable and reliable chemical contrast on technically important rough surfaces at submicroscopic resolution.1
1. G. Kaupp and A. Herrmann, J. Phys. Org. Chem. 10, 675 (1997); G. Kaupp, A. Herrmann and M. Haak, J. Vac. Sci. Technol. B 15, 1521 (1997); G. Kaupp, Lecture at the NFO-4 Jerusalem, 1997, book of abstracts, p. 92: G. Kaupp and A. Herrmann, Ultramicroscopy (1998), in press; G. Kaupp, Supermicroscopy in Supramolecular Chemistry: AFM, SNOM, and SXM, in Comprehensive Supramolecular Chemistry, Vol. 8, 381-423 + 21 color plates, ed. J. E. D. Davies, Elsevier, Oxford,1996; G. Kaupp, Chemie in unserer Zeit 31, 129 (1997).
2. C. K. Meyer, A. K. Kirsch, G. Vereb, D. Arndt-Jovin and T. M. Jovin, Lecture at the NFO-4 Jerusalem, 1997, book of abstracts, p. 39; N. F. van Hulst, M. H. P. Moers and B Bölger, J. Microsc. 171, 95 (1993); A. Jalocha and N. F. van Hulst, Opt. Commun. 119, 17 (1995).
3. S. I. Bozhevolnyi, M. Xiao and O. Keller, Appl. Opt. 33, 876 (1994); H. Bielefeld, I. Hörsch, G. Krausch, M. Lux-Steiner, J. Mlynek and O. Marti, Appl. Phys. A 59, 103 (1994), a paper which was retreated in Ref. 5; P. J. Moyer and M. A. Paesler, SPIE Scanning Probe Microscopy II 1855, 59 (1993).
4. D. Courjon, J.-M. Vigoureux, M. Spajer, K. Sarayeddine and S. Leblanc, Appl. Opt. 29, 3734 (1990); C. Girard and M. Spajer, ibid. 29, 3726 (1990).
5. V. Sandoghdar, S. Wegscheider, G. Krausch and J. Mlynek, J. Appl. Phys. 81, 2499 (1997).
6. M. Spajer, D. Courjon, K. Sarayeddine, A. Jalocha and J.-M. Vigoureux, J. Phys. (Paris) III 1, 1 (1991); F. Zenhausern,M. P. O'Boyle and H. K. Wickramasinghe, Appl. Phys. Lett. 65, 1623 (1994).
7. S. Madsen, Ph.D.-thesis, Microelectronic Centret at the Technical University of Denmark, Lyngby, 1997.